Clay Mineralogy: Ralph Grim

CLAY MINERALOGY
RALPH E. GRIM
Research Professor of Geology
U,E:iver~itL_Qf_]llinois
, I L L-,1' A R Y
1334
Date:
New York Toronto London

McGRAW-HILL BOOK COMPANY, INC.
1953
CLAY MINERALOGY
Copyright, 1953, by the McGraw-Hill Book Company, Inc.
Printed in the United States of America. All rights reserved.
This book, or parts thereof, may not be reproduced in any
form without permission of the publishers.
Library of Congress Catalog Card Number: 52-13807
I IX
24835
THE MAPLE PRESS COMPANY, YORK, PA.

PREFACE
The present volume attempts to summarize available data on the
structure, composition, properties, occurrence, and mode of origin of the
various clay minerals whose identities have been reasonably well estab-
lished. The distribution of the clay minerals in rocks of various litho-
logic types, geologic age, and condition of formation is considered, and
an attempt is made to analyze the errvironmental conditions under which
the individual clay-mineral groups are formed and are stable. Because
of the scantiness of available data, the general conclusions resulting from
such analyses must be considered as tentative and preliminary. Future
researches may show that some major revisions are necessary.
The aluminum and ferric iron hydrate minerals found in some clay
materials are not considered herein. Generally such minerals have not
been classed with the clay minerals, and their structure, properties,
occurrence, and modes of origin have already been reported in detail in
monographic form in the literature.
The specific properties of the individual clay minerals are discussed,
but no attempt is made to consider the larger rock properties of clay
materials except incidentally as they are related to specific mineral
properties. For example, the changes taking place when the individual
clay minerals are heated are discussed at some length, and the relation
to the refractoriness of clays is indicated. A detailed discussion of the
refractoriness of clays, which is influenced by other factors than the clay-
mineral composition, is not, however, included.
The clay minerals are the major factor controlling the larger rock
properties of clay materials, such as plasticity, strength, sensitivity, etc.
Other factors also influence these properties, and an adequate discussion
of them would require the presentation of much additional fundamental
data. Further, an analysis of such rock properties, even if restricted to
the clay-mineral viewpoint, would require a volume at least the size of
the present one.
A brief discussion is presented of the various concepts of the composi-
tion of clay materials that have been advanced in the past, as well as a
brief history of the development of the present clay-mineral concept.
Such discussion is rcecessary as a background for the consideration of the
clay minerals themselves.
Clay materials have been studied by scientists for a great many years,
but within the last 30 years there has been a tremendous expansion of
v
vi Preface
clay investigations. Many investigators approaching the subject from
different disciplines have devoted all or most of their attention to clay
materials. As a consequence of this enlarged effort, fundamental infor-
mation on all as'pects of clay materials has been greatly extended since
about 1920. The reason for this great expansion of interest in clays has
been twofold. In the first place, new research tools, such as X-ray-
diffraction analysis, became available for studying extremely small
particles. In the second place, the economic importance of clay mate-
rials was more generally appreciated. A long list of important commer-
cial applications of clay-mineral studies could be given, but the following
will suffice to illustrate the application of clay mineralogy in diverse fields.
In the ceramic industry, only certain clays with peculiar properties
can be used to manufacture certain products. Thus, only certain clays
of particular clay-mineral composition will withstand high temperatures
and, therefore, can be used for the making of refractory brick. Studies
of the changes taking place when clay minerals are heated to elevated
temperatures have greatly enhanced the understanding of exactly what
happens when clay products are burned. A consequence of such clay-
mineral data has been to improve the quality of some ceramic products
and to reduce the time necessary to fire them.
In the oil industry certain types of bentonite clay are essential for the
preparation of the muds required for the drilling of many oil wells, and
other types of bentonite clay form the basis for many of the catalysts
used in the refining of petroleum products. Detailed clay-mineral studies
have been and are being made of the particular bentonites needed for
each of these uses. The result of this work has vastly aided drilling-mud
practice and the refining operation. It has also been of great value in
the search for adequate supplies of the particular type of bentonite needed
for each of these uses.
The most'important outlet for certain types of kaolinite clays is in the
paper industry, where they are used for fillers and coating materials.
Researches into the structure and properties of kaolinite have permitted
improvements in the clays produced for the paper trade; these have
resulted in improvements of such paper properties as acceptance to ink,
rate of drying, etc.
Construction engineers are frequently faced with the problem of build-
ing a structure through a clay material, as in the case of a tunnel, on
clay material, as in the foundation of a building, or with a clay material,
as in the subgrade of a highway or an earth-filled dam. The method of
procedure, of necessity, is to obtain samples of the material which must
be used or is available, and to test them in the laboratory under condi-
tions representing field conditions as nearly as possible. On the basis of
the empirical laboratory data, the engineer arrives at the structural design
Preface vii
and, in so doing, frequently must predict how the soil material will act
when it is placed under different conditions, as, for example, when the
water table is changed, after a base-exchange reaction has taken place
in the clay, etc. Obviously the likelihood of the accuracy of such predic-
tions is increased if the fundamental factors controlling the properties in
question are understood. Much progress remains to be made on this
subject, but currently it is possible to warn the engineer on the basis of
simple clay-mineral determinations when he is facing materials that are
likely to give some misleading empirical test data.
In the field of agriculture the tilth of a soil, its content of plant
nutrients, and its treatment possibilities with fertilizers are all to a very
large extent contingent on the clay-mineral composition of the soil. It is
not difficult to understand, therefore, why soil investigators have been
in the forefront of those carrying on clay-mineral investigations.
Geologists have been interested in clay-mineral researches for a variety
of reasons, but two of them are particularly important from an economic
standpoint. It seems likely that the clay-mineral composition of a sedi-
ment will ultimately prove to be an important clue in unraveling the
conditions under which it was deposited. Also it seems likely that the
clay minerals may have played an important role in the origin of petro-
leum by acting as catalysts in the alteration of the original buried organic
material to hydrocarbon compounds. Clay-mineral researches should
therefore provide significant information on the origin of petroleum and
important criteria for the location of source beds of petroleum. In short
such researches should be of great aid in the search for petroleum by
geologists.
The author wishes to express his thanks to the persons who have per-
mitted the reproduction of data from their published reports. Professor
Thomas A. Bates of Pennsylvania State College kindly obtained special
electron micrographs for this volume. Professor Paul F. Kerr of Colum-
bia University, and the American Petroleum Institute have graciously
allowed the reproduction of some electron micrographs and other data
from '~heir reports of work done on their Project 49. Professor G. W.
Brindley and the Mineralogical Society of Great Britain have permitted
the use of data from the recent monograph "X-ray Identification and
Structure of the Clay Minerals." Doctor S. B. Hendricks has very
willingly allowed the author to follow his plan of illustrating the clay-
mineral structures. All these and other sources of information in the
volume are, of course, acknowledged specifically at the proper place.
The author wishes to express his great indebtedness to Dr. W. F.
Bradley of the Illinois State Geological Survey, with whom he has been
associated in clay-mineral researches for 20 years. Many of the subjects
in the present volume have been discussed many times through the years
viii Preface
with Dr. Bradley, and his thoughts and work have contributed immensely
to many of the conclusions presented herein. Doctor Bradley has read
much of the :manuscript and offered many significant comments. Any
omissions or errors in the data presented are, however, solely those of the
author.
The author began his studies of the clay minerals as a member of the
Illinois State Geological Survey. These studies were continued in that
organization for twenty years always with the enthusiastic support of
Dr. M. M. Leighton, Chief of the Survey. The author wishes to acknowl-
edge this support and the encouragement which came from Dr. Leighton's
appreciation of the importance of clay mineralogy.
Great difficulty was experienced in presenting the X-ray-diffraction
data, since it is recorded in both angstrom and kX units, and it was not
feasible to translate all the essential data into either one of these units.
The volume, therefore, contains both kX and angstrom {[nits. In
general, X-ray-diffraction data are given in kX units and dimensional
data of lattices, ions, etc., in angstrom units. This point is discussed in
some detail on page 84, and the factor for translating one unit to the
other is given there. As shown in that discussion, the difference between
the two units is very small, and for all intents and purposes the values are
substantially equivalent for much of the clay-mineral data.
RALPH E. GRIM
URBANA, ILL.
April, 1953
CONTENTS
PREFACE. V
1. INTRODUCTION 1
Definitions
Factors Controlling the Properties of Clay Materials 3
Clay-mineral composition. Nonclay-mineral composition. Or-
ganic material. Exchangeable ions and soluble salts. Texture
2. CONCEPTS OF THE COMPOSITION OF CLAY MATERIALS 11

Old Concepts 11
Clay-mineral Concept . 16
3. CLASSIFICATION AND NOMENCLATURE OF THE CLAY MINERALS 27

Classification of the Clay Minerals 27
Nomenclature of the Clay Minerals. 2\J
Questionable and Discredited Clay Minerals 38
4. STRUCTURE OF THE CLAY MINERALS. 43

General Statement . 43
Allophane Minerals 45
Kaolinite Minerals . 46
Halloysite Minerals 52
Montmorillonite Minerals 55
Illite Minerals 65
Chlorite Minerals 69
Vermiculite . 72
Sepiolite, Palygorskite, Attapulgite Minerals 77
Mixed-layer Minerals . 79
5. X-RAY DIFFRACTION DATA 84
Kaolinite and Halloysite Minerals 87
Montmorillonite Minerals 91
Illite Minerals 93
Chlorite Minerals 97
Vermiculite Minerals 99
Sepiolite, Palygorskite, Attapulgite Minerals 99
Mixed-layer Structures 102
6. SHAPE AND SIZE-ELECTRON MICROGRAPHS 106

Data for the Clay Minerals 108
Allophane. Kaolinite. Dickite. Nacrite. Halloysite. Mont-
morillonite. Illite. Vermiculite and chlorite. Attapulgite-se-
piolite-palygorskite. Mixed-layer minerals
ix
x Contents
7. ION EXCHANGE. 126
Importance of Ion Exchange. 126
Cation Exchange 128
History. Cation-exchange capacity. Other minerals with cation-
exchange capacity. Causes of cation exchange. Position of ex-
changeable cations. Rate of the exchange reaction. Variations
due to particle size. Effect of grinding. Relation of tempera-
ture. Environment of the exchange reaction. Hydrogen clays.
Clogging of cation-exchange positions. Replaceability of exchange-
able cations. Fixation of cations. Theory of cation exchange.
Determination of cation-exchange capacity and exchangeable
cations
Anion Exchange. 156
Phosphate fixation
8. CLAY-WATER SYSTEM . 161
Nature of Adsorbed Water 162
Density of Initially Adsorbed Water. 171
Evidence for the Crystalline State of the Initially Adsorbed Water 171
Thickness of Adsorbed Nonliquid Water. 174
Time Factor 175
Influence of Cations and Anions . 175
Stepwise Hydration of Montmorillonite 181
Stability of Montmorillonite Hydration 182
Influence of Adsorbed Organic Molecules 183
Heat of Wetting. 183
Values for heat of wetting. Causes of heat of wetting. Heat of
wetting in solutions of electrolytes. Effect of firing
9. DEHYDRATION, REHYDRATION, AND CHANGES TAKING PLACE ON HEATING 190
Methods of Study . 190
Vapor pressure-water content determinations. Dehydration
curves. Differential thermal analyses. Identification of high
temperature phases
Allophane 211
Kaolinite 212
Dehydration and phase changes on heating. Rehydration
Halloysite 217
Montmorillonite. 220
Vermiculite . 231
Illite . 233
Chlorite . 238
Sepiolite, Palygorskite, Attapulgite 241
Clay-mineral Mixtures 245
10. CI,AY-MINERAL-ORGANIC REACTIONS 250
Introduction 250
Contents Xl
Reactions with Montmorillonite and Halloysite 252

Ionic reactions. Adsorption of polar molecules. Comparison of
ionic and polar complexes
Reactions with Clay Minerals Other Than Montmorillonite and Halloy-
site 262
Kaolinite. Illite. Chlorite. Vermiculite. Attapulgite
Resistance of Adsorbed Organic Molecules to Biological Decomposition. 264
Organophilic Clay-mineral Complexes 265
Structural Implications of Montmorillonite-Organic Complexes . 269
Analytical Tec!:mique Based on Clay-mineral-Organic Reactions. 271
X-ray techniques. Differential thermal techniques. Optical
methods. Cation-exchange capacity. Surface-area determina-
tion. Geometry and properties of organic molecules
Staining Tests for Clay Minerals 274
11. OPTICAL PROPERTIES 278

Kaolinite 278
Halloysite 280
Montmorillonite. 282
Illite . 284
Chlorite . 284
Vermiculite 285
Sepiolite, Attapulgite, Palygorskite 285
Influence of Immersion Media on Optical Properties 286
Oriented-aggregate Technique. 288
Interlayer Mixtures. 289
Orientation in an Electrical Field. 290
Form Birefringence 291
Discussion of the Application of Optical Methods in Clay-mineral Studies 292
12. MISCELLANEOUS PROPERTIES 295

Solubility of the Clay Minerals 295
General statement. Solubility of clay minerals in acids. Nature
of the acid reaction. Decomposition by electrodialysis. Cation
liberation by neutral salts. Rational analysis. Solubility of clay
minerals in alkalies
Infrared Spectra of the Clay Minerals 303
General statement. Data for the clay minerals
Surface Area 308
.Elensity . 312
Kaolinite. Halloysite. Illite. Montmorillonite. Vermiculite.
Chlorite. Sepiolite, attapulgite, palygorskite
13. ORIGIN AND OCCURRENCE OF THE CLAY MINERALS 316

Synthesis of the Clay Minerals 316
Introduction. Syntheses from mixtures of oxides and hydroxides
at elevated temperatures and pressures. Syntheses from mix-
tures of crystalline minerals ~nd chemical reagents at elevated
temperatures and pressures. Syntheses from mixtures of oxides
and hydroxides at ordinary temperatures and pressures. Trans-
formations of clay minerals at ordinary temperatures and pres-
sures. General conclusions from synthesis data
xii Contents
Clay Minerals of Hydrothermal Origin . 323
Introduction. Types of clay minerals in hydrothermal deposits.
Mode of formation and occurrence. Nature of the hydrothermal
solutions. Relation to parent materials. Relation to mineraliza-
tion. Clay minerals associated with hot springs, fumaroles, etc.
Soils and Weathering . 330
Factors controlling weathering processes. Soil-profile develop-
ment. Classification of great soil groups of the world. Descrip-
tion of great soil groups. Tundra soils. Podsolic soils. Laterite
soils. Zonal aridic soils. Clay-mineral composition of soils.
Discussion of weathering products formed from various types of
rock under varying conditions. Reversion of weathering cycle.
Occurrence of halloysite and allophane in soils. Nature of the
al tera tion process
14. ORIGIN AND OCCURRENCE OF THE CLAY MIl\fERALS (CONT.) 348

Recent Sediments 348
Marine environment. Xonmarine conditions
Ancient Sediments . 355
Clay-mineral composition in relation to mode of origin. Clay
minerals in relation to geologic age. Clay minerals in relation
to lithology. Miscellaneous clay materials
ApPENDIX 369
Chemical Analyses 369
INDEX 375
,.
CHAPTER 1
Introduction
DEFINITIONS
.
Clay is used as a rock term and also as a particle-size term in the
mechanical analysis of sedimentary rocks, soils, etc. As a rock term it is
difficult to define precisely, because of the wide variety of materials that
have been called clays. In general the term clay implies a natural,
earthy, fine-grained material which develops plasticity when mixed with a
limited amount of water. By plasticity is meant the property of the
moistened material to be deformed under the application of pressure,
with the deformed shape being retained when the deforming pressure is
removed. Chemical analyses of clays show them to be essentially
silica, alumina, and water, frequently with appreciable quantities of iron,
alkalies, and alkaline earths.
The difficulty is that some material called clay does not meet all the
above specifications. Thus, so-called flint clay has. substantially no
plasticity when mixed with water. It does, however, have the other
attributes of clay.
The term clay has no genetic significance. It is used for material
that is the product of weathering, has formed by hydrothermal action,
or has been deposited as a sediment.
As a particle-size term, the clay fraction is that size fraction composed
of the smallest particles. The maximum size of particles in the clay
size grade is defined differently in different disciplines. In geology the
tendency has been to follow the Wentworth 1 scale and to define the clay
grade as material finer than about 4 microns. In soil investigations, the
tendency is to use 2 microns as the upper limit of the clay size grade.
Although there is no sharp universal boundary between the particle
size of the clay minerals and non clay minerals in argillaceous sediments,
a large number of analyses have shown that there is a general tendency
for the clay minerals to be concentrated in a size less than about 2 microns,
or that naturally occurring larger clay-mineral particles break down easily
to this size when the clay is slaked in water. Also such analyses have
1 Wentworth, C. K., A Scale of Grade and Class Terms for Clastic Sediments, J.
Geol., 30, 377-392 (1922).
1
2 Clay Mineralogy
shown that the nonclay minerals usually are not present in particles
much smaller than about 1 to 2 microns. A separation at 2 microns is
frequently about the optimum size fo_r the best split of the clay-mineral
and non clay-mineral components of natural materials. There is, there-
fore, a fundamental reason for placing the upp0r limit of the clay size
l;rade at 2 microns.
Clays contain varying percentages of clay-grade material and there-
fore, varying relative amounts of non clay-mineral and clay-mineral
components. The writer knows of no clay which does.not contain some
nonclay-mineral material coarser than the clay grade, although the
amount in some hydrothermal clays is extremely small (less than 5 per
cent). Many materials are called clays in which the clay-grade and
clay-mineral component make up considerably less than half the total
rock. In such materials the non clay is frequently not much coarser
than the maximum for the clay grade, and the clay-mineral fraction may
be particularly potent in causing plasticity. In general fine-grained
materials have been called clay so long as they had distinct plasticity
and insufficient amounts of coarser material to warrant the appellations
silt or sand. If particle-size analyses are made, the term clay would
be reserved for a material in which the clay grade dominates. However,
names have been and are applied most frequently solely _on the basis of
the appearance and bulk properties (e.g., plasticity) of the sample.
Shale is a fine-grained, earthy, sedimentary rock with a distinct
laminated, or layered, character. The layering may be due to a general
. parallel arrangement of flake-shaped or elongate particles or to an
alternation of beds of somewhat different com'position. The lamination
is parallel to the bedding and has not been developed by postdepositional
metamorphic action. The requirements of composition are substantially
the same for a shale as for a clay. Occasionally, however, natural
materials are called shale with little regard to composition. Thus,
thinly layered rocks composed essentially of quartz and/or carbonate
with little clay-mineral component have been called shale. Sometimes,
although by no means always, shales are more indurated and harder than
clays. The term shale is sometimes used by engineers for any hard,
indurated, argillaceous rock regardless of any lamination.
Argillite is a fine-grained argillaceous material that is massive and some-
what indurated and hard. It differs from shale in being massive rather
than laminated and from clay by being harder.
The term soil is likely to have a considerably different meaning when
used by a geologist, by an agronomist, and by a civil engineer. Soil to a
geologist is the weathered regolith at the earth's surface that supports
vegetation. It is thought of generally as being loose, argillaceous, and
with some organic content. To the agronomist it is the loose regolith
Introduction
at the earth's surface. It need not be weathered nor contain any vege-
tation; it may be gravel, for example. Also according to agronomists,
a soil is likely to be composed of a series of horizons and have properties
quite independent of the underlying parent bedrock. The civil engineer
tends to divide the material at the earth's crust into two categories (1)
rock and (2) soils. Rock is defined as something that is hard and con-
solidated. Soil, according to Terzaghi and Peck,2 "is a natural aggregate
of mineral grains that can be separated by such gentle means as agitation
in water." Substantially any loose material at the earth's crust, regard-
less of particle-size distribution, composition, or organic content, is soil
to the engineer. It mayor may not be weathered. Similarly soil to the
engineer can extend to any depth below the surface so long as the material
is not indurated substantially. Shale to the engineer is similar to soil
except that the term is applied to material that is slightly harder and is
definitely argillaceous. The term clay is primarily a particle-size term
to the engineer.
The author has found it convenient to use the expression clay material
for any fine-grained, natural, earthy, argillaceous material. Clay
material includes clays, shales, and argillites of the geologist. It would
also include soils, if such materials were argillaceous and had appreciable
contents of clay-size-grade material.
No attempt will be made herein to consider the definitions of rela-
tively minor types of argillaceous materials with somewhat specific
properties, such as loam, gumbo, etc. Description of such materials
can be obtained from standard textbooks on soils and sedimentary
rocks.
FACTORS CONTROLLING THE PROPERTIES OF CLAY MATERIALS
The factors which control the properties of clay materials or the

attributes which must be known to characterize completely a clay
material may be classified as follows:
a. Clay-mineral Composition. This refers to the identity and rela-
tive abundance of all the clay-mineral components. Since certain clay
minerals which may be present in very small amounts may exert a
tremendous influence on the attributes of a clay material, it is not ade-
quate to determine only the major clay-mineral components. Thus a
small amount (5 % ) of montmorillonite in a clay is likely to provide a
material very different from another clay with the same composition in
all ways except for the absence of montmorillonite. In order to make
complete clay-mineral determinations, it is frequently necessary to
2 Terzaghi, R., and R. Peck, "Soil Mechanics in Engineering Practice," Wiley,
New York (1948).
4 Clay Mineralogy
fractionate the clay grade to concentrate minor constituents so that
adequate analytical data can be obtained. Fortunately such a concen-
tration can often be attained, because the various clay minerals fre-
quently occur in particles of different sizes or break down easily in water
to particles of different size. Also the clay minerals must be determined
in their natural state. For example, care must be taken that the analysis
will reveal the natural hydration state of the minerals and their ion-
exchange composition. Clays composed of halloysite have very different
physical properties depending on whether the mineral is in the 4H 20
form, the 2H 20 form, or an inte~mediate state. Montmorillonite clays
have very different properties when Na+ is the exchange cation and when
Ca++ is the cation.
In clay materials containing a considerable amount of nonclay-mineral
material, it is frequently necessary to remove the non clay-mineral
material before the clay minerals can be identified completely. 3 Fre-
quently this involves merely a particle-size separation. Sometimes, as
in the presence of pigmentary iron oxide or hydroxides, extremely fine
carbonate, and pigmentary organic material, other methods must be
attempted. Considerable caution is necessary to avoid significant
change in the clay-mineral components in such separations. For
example, the use of acids to remove the iron or carbonate, even if very
dilute, may dissolve certain of the clay minerals if they are present (see
Chap. 12). In the case of iron oxide or hydroxide, recent biological
methods 4 of removal appear to be quite satisfactory. Strong oxidizing
. agents to eliminate the organic material are likely to alter the clay
minerals significantly.
b. Nonclay-mineral Composition. This refers to the identity of the
nonclay minerals, their relative abundance, and the particle-size distri-
bution of the individual species. Calcite, dolomite, large flakes of mica,
pyrite, feldspar, gibbsite, and other minerals are very abundant in some
clay materials.
Obviously, it is impossible or unjustifiably time~consuming to get all
the data concerning the non clay minerals in most investigations of clay
materials. The lengths to which one can and must go depend largely on
the problem at hand and the purpose of the investigation. It is fre-
quently adequate to determine the identity of only the more abundant
nonclay minerals and their sorting and particle-size distribution in a
general way. Thus the "heavy minerals" may be of no significance in
3 Grim, R. E., P. F. Kerr, and R. H. Bray, Application of Clay Mineral Technique

to 1llinois Clay and Shale, Bull. Geol. Soc. Am., 46, 1909-1926 (1935).
4 Allison, L. E., and G. D. Scarseth, A Biological Reduction Method for Removing
Free Iron Oxides from Soils and Colloidal Clays, J. Am. Soc. Agron., 34, 616-623
(1942).
,
Introduction 5
relation to the physical properties of a clay but may be of important
diagnostic value in determining whether or not a clay has formed by the
alteration of volcanic ash. As another example, the study of a soil from
the point of view of soil mechanics demands that sorting within the silt
size range be studied in considerable detail, since the presence of some
silt materials may yield a material of unique physical properties of great
importance to the construction engineer. The analysis of a clay material
must be tailor-made to the material being studied and to the purpose of
the investigation and must provide comparable results from one sample
to another. One cannot blindly use a set analytical procedure for all
materials and all problems and still get adequate data without a tre-
mendous waste of time and effort.
The nonclay minerals in clay materials tend generally to be concen-
trated in particles coarser than about 2 microns. There are, however,
materials in which they are much finer grained. In some Wyoming
bentonites, for example,5,6 a considerable amount of cristobalite is
present in particles considerably less than 1 micron in diameter intimately
mixed with the clay mineral montmorillonite. Many clay materials con-
tain extremely fine iron oxide or hydroxide, which acts as a pigment.
The identification of the coarser nonclay minerals can be made with the
p'etrographic microscope. The determination of those occurring in
extremely fine particles requires X-ray techniques. Neither of these
methods permits very precise quantitative determinations. In the case
of extremely fine silica, the maximum accuracy by X-ray diffraction is
about 2 % if the quantity is small and somewhat less if the quantity is
large (4%). Numerous attempts 7 ,8 have been made to determine the
amount of nonclay minerals chemically, e.g., the amount of free and
combined silica in a clay material. Such methods are based on a differ-
ence in solubility of the constituents. Unfortunately variations of
solubility with particle size cause the results to have questionable value.
The absence of accurate quantitative methods for determining the non-
clay-mineral components of clay materials frequently makes it impossible
to obtain exact data on the chemical composition of the clay minerals
themselves in such materials. 9
5 Gruner, J. W., Abundance and Significance of Cristobalite in Bentonites and
Fuller's Earths, Econ. Geol., 35, 867-875 (1940).
6 Roth, R. S., The Structure of Montmorillonite in Relation to the Occurrence and
Properties of Certain Bentonites, Ph.D. thesis, University of Illinois (1951).

7 Trostel, L ..J., and D. J. Wynne, Determination of Quartz (Free Silica) in Refrac-
tory Clays, J. Am. Ceram. Soc., 23, 18-22 (1940).
8 Sauzeat, H., Can Free Quartz Be Determined in a Rock, Rev. indo minerale, 529,
114-117 (1948).
9 Kelley, W. P., Calculating Formulas for Fine Grained Minerals on the Basis of
Chemical Analysis, Am. Mineral., 30, 1-26 (1945).

Clay Mineralogy
c. Organic Material. This refers to the kind and amount of organic
material contained in the clay material. In general the organic material
occurs in clay materials in two ways: it may be present as discrete particles
of wood, leaf matter, spores, etc., or it may be present as organic mole-
cules adsorbed on the surface of the clay-mineral particles (see Chap. 10).
The discrete particles may be present in any size from large chunks easily
visible to the naked eye to particles of colloidal size which act as a pig-
ment in the clay-mineral material.
The total amount of organic material can be determined simply by
readily available standard analytical procedures. Values may be
obtained from the difference between total loss on ignition and determina-
tion of loss of water, sulfur, and other inorganic volatiles. Such values
are not precise but are usually adequate. Differential thermal analyses
provide a crude determination of amount of organic material. Fine
pigmentary organic material gives a dark gray or black color to a clay
material, but there is no direct relationship,between the color and organic
content. A very small amount of organic material may have a very
large pigmenting effect.
Determination of the kind of organic material is a more difficult
problem. 10 Sometimes, if the discrete particles are relatively large,
they can be identified visually or microscopically. By means of X-ray-
diffraction technique, the presence of adsorbed organic molecules may
usually be determined. At the present stage of our knowledge it is
usually impossible to go further and identify the organic components
present in small amounts and of extremely fine size, either discrete or
adsorbed. A large amount of fundamental work, applying modern
methods, such as infrared absorption, must be done on the organic
material in clay materials to provide basic data before any more or less
routine procedures can be devised for such analyses.
The study of the organic content of clay minerals is a problem worthy
of intensive research for a variety of reasons. For example, the organic
content often is important in determining the properties of a clay material,
and also a knowledge of clay-mineraI-organic relations might throw much
light on some important geologic processes. It might improve our
understanding of the origin of petroleum, since the clay minerals may
well have acted as catalysts in the transformation of the parent organic
matter into hydrocarbons, 11
d. Exchangeable Ions and Soluble Salts. Some clay materials con-
tain water-soluble salts which may have been entrained in the clay at the
10 Francis, M" Sur la matiere organique dans les argiles, Verre silicates ind., 14, 155-
158 (1949).
11 Grim, R. E., Relation of Clay Mineralogy to the Origin and Recovery of Petro-
leum, Bull. Am. Assoc. Petroleum Geol., 31, 1491-1499 (1947).

Introduction 7
time of accumulation or may have developed subsequently as a conse-
quence of weathering or alteration processes, as in the oxidation of pyrite
to produce sulfates. It is frequently neces$;ary to wash out the soluble
salts before other attributes of the material are studied. Some salts may
act to flocculate the clay, so that it cannot be dispersed for particle-size
analysis or for fractionation preliminary to clay-mineral analysis until
the salts are washed out. Common water-soluble salts found in clay
materials are chlorides, sulfates, and carbonates of alkalies, alkaline
earths, aluminum, find iron.
The clay minerals and some of the organic material found in clay
materials have significant ion-exchange capacity. The ion-exchange
capacity of the clay minerals and the organic components, as well as the
identity and relative abundance of the exchangeable ions which are
present, are extremely important attributes of clay materials. It is
difficult to distinguish sometimes between ()xchangeable ions and those
present in a moderately soluble compound, So that determinations of ion-
exchange characteristics are difficult in a material containing appreciable
water-soluble salts. This whole matter, t()gether with analytical pro-
cedures, is discussed in detail in Chap. 7.
e. Texture. The textural factor refers to the particle-size distribution
of the constituent particles, the shape of the particles, the orientation of
the particles in space and with respect to each other, and the forces tend-
ing to bind the particles together.
Some knowledge of the particle-size distribution of the coarser grains
can be obtained quickly by microscopic examinations, and detailed
determinations can be made by sieving and/or wet sedimentation meth-
ods. Fine-grained particles require wet methods, and this applies to
the clay-mineral fraction. It must be remembered that wet methods
are likely to reflect only the degree to which clay-mineral units or aggre-
gates have been cleaved or broken down in the process of making the
analysis rather than any inherent attribute of the natural material. In
dispersing clays in water for analysis, the material is usually agitated,
which splits and cleaves natural particles. The particle-size distribution
records the amount of agitation applied. There are clay-mineral
materials from which literally any particle-size distribution can be
obtained by relatively slight variations of the pteparation procedure.
In general the particle-size distribution of clay materials composed of
montmorillonite, vermiculite, and the attaplligite-sepiolite clay minerals
would be more affected by analytical prQ(~edures than clay materials
composed of the other clay minerals.
The use of chemical dispersing agents almost certainly will alter the
base-exchange composition of the material, and consequently such
agents must not be used or at least used only with great caution, if
8 Clay Mineralogy
exchangeable ions are to be determined. It is generally essential to
determine the exchangeable ions on the" as received" material, since any
mixing in water or washing is likely to cause a significant change. Also
such chemicals are likely to yield salts in the resulting fractions which
complicate the identification of any clay minerals therein.
The dewatering of the fine clay grades may well yield a material in a
different hydration state than the original material, and such dehydra-
tion, if it is complete, may tend to conceal some of the clay-mineral
components. Thus, completely collapsed montmorillonite from which
all adsorbed water has been removed is easily mistakable for illite. If
clay-mineral determinations are to be made on clay fractionations, it is
essential they be only air-dried, not oven-dried.
It is obvious that the particle-size-grade analysis of clay materials is
difficult, and care must be taken to devise a tailor-made procedure best
suite.d to the material at hand and to the objectives of the investigation
if pertinent, reproducible, and comparable data are to be obtained.
The shape of the finest particles is revealed best by electron-microscope
studies. Such investigations have shown the hexagonal outline of the
flake-shaped units of kaolinite, the elongate tubular shape of the halloy-
site minerals, the irregular flake shape of the illite, chlorite, vermiculite,
and most montmorillonite mineral particles, and the elongate lath or
fiber shape of some of the montmorillonite minerals and of attapulgite-
sepiolite-palygorskite. Information on the thickness as well as areal
dimensions can frequently be obtained from electron micrographs;
kaolinite particles that have been studied show a ratio of areal diameter
to thickness of 2-25: 1, whereas for montmorillonite it is 100-300: 1.
In the application of the electron beam in electron microscopy, consider-
able heat is developed in the specimen so that some concern has been felt
as to whether or not some of the observed results are due to this heat and
the possible resulting dehydration rather than to the natural mineral.
The microscope using ordinary light can, of course, be used to study
the coarser particles. The lower limit for the study of the shape of
particles by ordinary microscopic methods is about 5 microns.
Some information regarding the orientation of extremely fine particles
can sometimes be obtained from the study of thin sections. In the
absence of appreciable amounts of nonclay-mineral components, aggre-
gate parallel orientation of the anisotropic clay-mineral particles ~s
compared to random orientation is shown by uniform extinction and bire-
fringence characteristics. Thin-section studies appear to have distinct
limitations. The thickness of the sections is many times that of the
individual clay-mineral components, so that many individuals lie on top
of each other. The presence of even small amounts of organic material
or free ferric iron oxide or hydroxide will mask the individual components
Introduction 9
and distort the optical values. Also the clay material must be dried in
preparation for the cutting of the section, so that the texture observed
may be not quite that of the original material. Even with these deficien-
cies of thin-section study, it is usually worth while to cut sections and
study them in any clay-material investigation. Such studies,12 for
example, have revealed particularly pertinent data on the paragenesis
of hydrothermal clay minerals in wall-rock alteration associated with
ore bodies.
Some few attempts have been made to devise new methods of studying
the texture of clays in their natural state, such as the replica electron-
micrograph methods, the relation of preferred aggregate orientation to
certain physical properties, and the cutting and study of thin sections
of frozen clay materials. It appears that work with such methods has
not reached the point where general results are available. The texture
of clay materials is an important and promising field for research that
should attract able investigators.
So little is known in detail about the forces binding the particles
together in clay materials that the possible types of binding forces can
merely be enumerated.
1. Forces due to the attraction of the mass of one clay-mineral particle
for the mass of another particle.
2. Intermolecular forces resulting from the nearness of one particle
to another with the overlap of fields of force of molecules in the surface
layers of adjacent particles.
3. Electrostatic forces due to charges on the lattice resulting from
unbalanced substitution within the lattice, broken bonds on edges of the
lattice, and the attractive force of certain ions adsorbed on clay-mineral
surfaces. Examples are to be found in the bonding action of K+ between
mica layers and of multivalent ions with one valence tied to one particle
and another valence tied to a secona particle.
4. The bonding action of adsorbed polar molecules. Oriented water
molecules (see Chap. 8) between two clay-mineral surfaces may form a
bridge of considerable strength if only a few molecules thick and of no
strength if more than a few molecules thick. Similarly, adsorbed polar
organic molecules could serve as a bond between clay-mineral particles.
In any given clay material all the bond forces probably are at work,
and they are interrelated. Thus the nature of the adsorbed ion will
itself influence bonding and also affect the development of oriented
adsorbed water, which in turn is related to bonding.
The matter of the bonding force in clay materials is of particular
importance to soil-mechanics investigations and construction engineers,
12 Sales, R. H., and C. Meyer, Wall Rock Alteration at Butte, Montana, Am. Inst.
Mining Met. Engrs. Tech. Pub. 2400 (1948).

10 Clay Mineralogy
since it largely determines the sensitivity and strength of soil materials.
Construction failures have occurred because the strength properties of a
soil that developed during construction could not be predicted ade-
quately from empirical laboratory testing data. Without fundamental
data on how and why clay materials are held together, it is impossible
always to predict safely from any empirical data how a clay material
will act when load is applied, when the water table is altered, or when
other conditions are changed.
ADDITIONAL REFERENCES
Atterberg, A., Die Plastizitiit der Tone, Intern. Mitt. Bodenk., pp. 10':"'43 (1911).
Baver, L. D., "Soil Physics," Wiley, New York (1940).
Casagrande, A., Classification and Identification of Soils, Proc. Am. Soc. Civil Engrs.,
pp. 783-810 (1947).
Glossop, R., and A. W. Skempton, Particle Size in ,silts and Sands, J. Inst. Civil
Engrs. (London), no. 5492, 81-105 (1945).
Grim, R. E., Modern Concepts of Clay Materials, J. Geol., 50, 225-275 (1950).
Jenny, H., "Factors of Soil Formation," McGraw-Hill, New York (1941).
Joffe, H., "Pedology," Rut!!;ers University Press, New Brunswick, N. J. (1949).
Knight, H. G., New Size Limit of Clay-Silt, Soil Sci. Soc. Am. Proc., 2, 592 (1937).
Krumbein, W. C., and F. J. Pettijohn, "Manual of Sedimentary Petrography,"
Appleton-Century-Crofts, New York (1938).
Oden, S., General Introduction to the Chemistry and Physical Chemistry of Clays,
Bull. Geol. Inst. Univ. Upsala, 15, 175-194 (1916).
Ries, H., "Clays, Occurrence, Properties and Uses," 3d ed., Wiley, New York (1927).
Twenhofel, W. H., "Principles of Sedimentation," McGraw-Hill, New York (1950).
Von Moos, A., and F. de Quervain, "Technische Gesteinkunde," Birkhauser, Basel
(1948).
- CHAPTER 2
Concepts of the Composition of Clay Materials
OLD CONCEPTS
Because of the importance of clay materials in ceramics and other

industries, in agriculture, in geology, and elsewhere, their investigation
goes back far into antiquity. Many people have devoted most of their
lives to the study of clay materials. From the first, investigators learned
that clays and soils had widely varying properties. Even soils and clays
which had the same color and general appearance and the same texture
were found to differ widely in other characteristics. As soon as pro-
cedures for the ultimate analysis of clay materials were worked out, it was
learned that clay materials differed widely in their chemical composition.
The finest fractions of clay materials, which were thought to be the
essence of the material, showed wide variations in the amounts of alumina,
silica, alkalies, and alkali earths that they contained. It was also found
that clays of the same ultimate chemical composition frequently had very
different physical attributes, and that clays with substantially the same
physical properties might have very different chemical compositions.
It is obvious from the foregoing statements that there must be varia-
tions not only in the amounts of the ultimate chemical constituents, but
also in the way in which they are combined, or in the manner in which
they are present in various clay materials. A review of the older litera-
ture shows that a considerable number of concepts were suggested to
portray the fundamental and essential components of all clay materials
and to explain their variation in properties. These concepts essentially
present ideas of the nature of the way in which the alumina, silica, etc.,
are made up into the fundamental building blocks of clay materials.
Until very recent years there have been no adequate analytical tools to
determine with any degree of certainty the exact nature of the funda-
mental building blocks of most clay materials. It is understandable,
therefore, that many different concepts were suggested and that there
was no general agreement among the workers in this field.
It is desired to present here very brief statements of the older concepts
to serve as a background for the later consideration of the development of
the present, generally accepted clay-mineral concept. For a more
detailed discussion of these early ideas of the nature of clays and soils,
11
12 Clay Mineralogy
reference should be made to the works of Blanck,l Strerome,2,3 Oden,4
Bradfield,5 Marshall,6 and Kelley.7 There is no definite sequence in the
development of these older concepts; in general they existed contem-
poraneously in the minds of various investigators.
A very old idea is that there is a single pure clay substance and that
this pure clay substance is the mineral kaolinite or something substan-
tially similar to it. Clay materials, according to this concept, are com-
posed of kaolinite, frequently with varying amounts of other materials
considered to be impurities. Differences in the chemical composition
between kaolinite and natural clays are explained by the presence of these
impurities. Kaolinite, however, was thought to be the essence of clays.
This concept of the general prevalence of kaolinite was held widely by
geologists and, indeed, unfortunately persists in the thinking and writings
of some present-day members of this profession. Also the concept has
persisted to some extent in many other quarters. The Webster's New
International Dictionary, 1934 edition, in the definition of clay states
"the essential constituent of pure clay, or kaolin, is the mineral kaolinite
. . . Most clays, however, contain other hydrous aluminous minerals
with more or less finely comminuted quartz, feldspar, mica, etc."
Some clays are composed almost wholly of kaolinite, and in some of
such clays, the particles of kaolinite are large enough to be seen and
identified positively by the microscope with relatively low magnification.
Such clays are the rare exceptions in which the fundamental building
blocks could be identified definitely in the years prior to the development
of modern research tools for studying extremely small particles. Also
such kaolinite clays are of particular importance in the ceramic art and
to geologists; hence they were among the first to be studied in detail. It
was a simple matter to extend the findings of the study of these kaolinite
clays to all clay materials. It is now established, of course, that there are
many clay materials in which there is no kaolinite present. Merrill 8 and
1 Blanck, E., Anorganische Bestandteile des Bodens, "Handbuch der Bodenlehre,"
vol. 7, pp. 1-60, Springer, Berlin (1933).

2 Stremme, H., Die Chemie des KaOlins, Fortschr. Mineral. Krist. Petrog., 2,87-128,
Dresden (1912).
3 Stremme, H., Allgemeines tiber die wasserhaltigen Aluminumsilicate, "Handbuch
der Mineralchemie," vol. 2, part 2, pp. 30-94, 130-134, Doelter, Ed., Steinkopff,
Dresden (1917).
, Oden, S., General Introduction to the Chemistry and Physical Chemistry of Clays,
Bull. Geol. Inst. Univ. Upsala, 15, pp. 175-194 (1916).
fi Bradfield, R., The Chemical Nature of Colloidal Clay, Missouri Agr. Expt. Sta.
Res. Bull. 60 (1923).

Marshall, C. E., "The Colloid Chemistry of the Silicate Minerals," Academic
Press, New York, (1949).
7 Kelley, W. P., "Cation Exchange in Soils," Reinhold, New York (1948).
8 Merrill, G. P., "Non-metallic Minerals," Wiley, New York (1904).
Concepts of the Composition of Clay Materials 13
Ries 9 early emphasized the error of considering that kaolinite was the
base of all clays, but the error tended to persist.
Another concept very widely held, particularly by soil investigators,
was that the essential component of all clay materials was a colloid com-
plex. Particularly in early days, all colloidal material was thought to
be amorphous, and the colloidal complex in clays was thought to be
amorphous. The complex was thought to be partly inorganic and partly
organic when the clay material contained some organic material.
In ~eral there were two more or less clearly defined ideas concerning
the character of the colloidal complex. One of the ideas, with which the
names of Van Bemmelen 10 and Stremme l l are particularly associated,
regarded the complex not as a definite compound but as a loose mixture
of the oxides of silicon, aluminum, and iron. In later years, the researches
of Thugutt,12 Bradfield,13 and many others showed that clay materials
generally did not contain a colloidal mixture of oxides. The other idea
regarded the complex as a compound or a mixture of compounds. The
compounds were generally thought of as salts of weak ferroalumino-
siliceous acids. In some cases these compounds were considered defi-
nitely to be amorphous, but mostly there was no real concept of the struc-
ture of the colloidal complex.
Way14 in his early work on the exchange reaction in soils concluded
that the exchange complex in soils was a hydrous aluminum silicate quite
similar to artificial precipitates produced in the laboratory.
Van Bemmelen 10 and later Stremme l l divided their colloidal fraction
into two parts. One part, which was soluble in hydrochloric acid, they
called allophaneton, and a second part not soluble in hydrochloric acid
but soluble in hot concentrated sulfuric acid came to be called kaolinton.
The allophaneton was thought to be highly colloidal of very varying
composition and largely responsible for the plastic and adsorptive proper-
ties of clay materials. Kaoli_nton was thought to be largely amorphous
but at times containing also some crystalline material. Its composition
showed relatively little variation and usually approached that of the
gRies, H., "Clays, Occurrence, Properties and Uses," 3d ed., Wiley, New York
(1927).
10 Van Bemmelen, J. M., Die Absorptionverbindung und das Absorptionvermogens
der Ackererde, Landw. Verso Sta. 35, 69-136 (1888).

11 Stremme, H., On Allophane, Halloysite and Montmorillonite, Centro Mineral.
Geol. pp. 205-211 (1911).

12 Thugutt, H., Are Allophane, Montmorillonite, and Halloysite Units or Are They
Mixtures of Alumina and Silica Gels?, Centro Mineral Geol., pp. 97-103 (1911).
13 Bradfield, R., The Colloidal Chemistry of the Soil, "Colloid Chemistry," J.
Alexander, ed., vol. III, pp. 569-590, Reinhold, New York (1928).
14 Way, J. T., On the Power of Soils to Absorb Manure, J. Roy. Agr. Soc. (Engl.) 13,
123-143 (1852).
14 Clay Mineralogy
mineral kaolinite. Attempts were made to classify clay materials on the
basis of their kaolinton and allophaneton content.
Mellor 15 and Searle 16 also developed the idea that there were two
essential components of clay. One, which was called "clayite," was
thought to be the true clay substance in kaolins and was considered to
be an amorphous substance with about the same chemical composition
as the mineral kaolinite. The other, for which the name" pelinite" was
suggested, was the true clay substance in clay materials other than the
kaolins. The latter was thought of as an amorphous material of varying
composition but of generally higher silica content than "clayite" and
also with appreciable alkalies and/or alkaline earths.
Wiegner 17 in his extensive studies of cation exchange viewed the
exchange material as made up of three parts: (1) a kernel, (2) a layer of
adsorbed anions external to the kernel but lying in contact with it, (3)
exchangeable cations attracted to the particle by the adsorbed anions.
The kernel was considered to be a hydrous compound chiefly of alumina
and silica of variable composition and of unknown structural attributes.
In the extensive studies of cation exchange in soils by Gedroiz,18 this
investigator considered the complex as zeolitic material, but not as zeolitic
in the mineralogical sense. In other words the complexes had cert~in
of the properties of zeolites but were not considered to have their precise
composition or structure. The nature of their structure was not known.
Another slight variation of this same concept, which has been carried
down to the present in the work of Mattson,19.20 is that the colloid com-
plex is made up of a relatively inert framework of silica, iron, and alumi-
nous materials encased in an active amorphous envelope of a varying
compound of silica, alumina, and iron with alkalies and alkaline earths.
Mattson considers this latter compound to be an amorphous isoelectric
precipitate of hydrated sesquioxides and silicic acid. In the light of
advances in clay mineralogy and the finding of the general crystalline
nature of the components of clay materials, Mattson 21 has somewhat
,. Mellor, J. W., Nomenclature of Clays, Trans. Ceram. Soc. (Engl.), 8, 23 (1908).
16 Searle, A. B., Clay and Clay Products, Brit. Assoc. Advancement Sci. Rept., pp.
113-154 (1920).
17 Wiegner, G., Ionenumtausch und Struktur, Trans. Intern. Congr. Soil Sci. 3rd
Congr., Oxford, 3, 5-28 (1936).
18 Gedroiz, K. K., On the Absorptive Power of Soil, Commissariat of Agriculture
U.S.S.R., Petrograd (1922). Translated by S. A. Waksman and distributed by U.S.
Department of Agriculture.
19 Mattson, S., The Laws of Soil Colloidal Behavior, III, Isoelectric Precipitates,
Soil Sci., 30, 459-495 (1930).

20 Mattson, S., The Laws of Soil Colloidal Behavior, IX, Amphoteric Reactions and
Isoelectrie Weathering, Soil Sci., 34, 209-239 (1932).

21 Mattson, S., The Constitution of the Pedosphere, Ann. Roy. Agr. Coll. Sweden,
5, 261-276 (1938).
modified his concept by postulating the colloidal complex as a crystalline
kernel covered with an amorphous heterogeneous coating which lacks a
definite composition and is not identical with the nucleus. According to
Mattson, X-ray-diffraction analysis would reveal only the character of
the crystalline nucleus and not of the heterogeneous coating which is the
essence of the complex. Mattson concepts have been criticized by
Kelley22 and Marshall,6 and there is no doubt that in many clay mate-
rials, X-ray analyses have shown that substantially all the components
are definit3 crystalline compounds.
In a very recent work, Puri 23 has considered soils to be composed essen-
tially of ferroaluminosilicates of varying composition but all composed
of the same framework. According to him, when soils from different
localities,are subject to treatment by mild acids, a framework residue is
obtaineawhich in every case behaves in the same manner. Studies of
the structures of the clay minerals have, of course, shown that there are
important and significant differences in the structure of the various
components of the finest fractions of soils.
Asch 24 and Byers 26 and his colleagues in the U.S. Department of Agri-
culture considered that the essential components of soils were a number
of substances rather than a single compound. They viewed these sub-
stances as aluminosilicic acids or salts of such acids with definite composi-
tions and with definite structures. This concept approaches the present
clay-mineral concept, and, in fact, Byers et al. suggested clay-mineral
names, e.g., montmorillonitic, halloysitic, for their postulated acids.
It has long been known by mineralogists that the zeolite minerals are
silicate compounds that possess the property of cation exchange. When
W ay 14 and his successors showed that soil materials had cation-exchange
capacity and that it resided in the silicate complex, an understandable
step was to postulate that soil materials contained zeolites. Lemberg 26
in 1876 particularly developed the concept of the presence of zeolites in
soil materials. Later, when the general idea was that the colloidal com-
plex was amorphous, it was postulated, notably by Gans,27 that the
complex was zeolitic. That is, the complex was an amorphous counter-
22 Kelley, W. P., Mattson's Papers "The Laws of Soil Colloidal Behavior," Review
and Comments, Soil Sci., 56, 443-456 (1943).

23 Puri, A. N., "Soils: Their Physics and Chemistry," Reinhold, New York (1949).
24 Asch, W., and D. Asch, "The Silicates of Chemistry and Commerce," Constable,
London (1914).
25 Byers, H. G., L. T. Alexander, and R. S. Holmes, The Composition and Constitu-
tion of the Colloids of Certain of the Great Soil Groups, U.S. Dept. Agr. Tech. Bull.
484 (1935).
26 Lemberg, J., Ueber Silicatumwandlungen, Z. deut. geol. Ges., 28, 519-621 (1876).
27 Gans, R., Ueber die chemische oder physikalische Natur der kolloidalen wasser-
haltigen Tonerdesilikate, II, Centro Mineral. Geol., pp. 728-741 (1913).
16 Clay Mineralogy
part of the crystalline zeolite minerals. Even in relatively recent work
the colloidal complex is sometimes referred to as zeolitic, although the
work of Gedroiz 28 and many others has shown wide differences between
the properties of mineral zeolites and the finest fractions of clay materials
and indicated that the exchange complex can be considered zeolitic only
in the sense that it possesses cation-exchange capacity.. Modern X-ray
analyses have revealed one or two instances when zeolites do occur in
bentonite clays, but such minerals are not general and significant com-
ponents of clay materials.
It is generally recognized that the small size of the particles in clay
materials is one of the reasons for their special attributes. It was sug-
gested, notably by Oden,4 that particle size is the major factor and that,
in fact, clays can be composed of almost any minerals if they are fine
enough-about 1 micron was considered the upper size limit. According
to Oden, clays are composed of a heterogeneous array of extremely small
particles of crystalline and amorphous components. Some clays, espe-
cially those of glacial origin, may contain an unusually large variety of
minerals in extremely small particle sizes. Present data indieate that
certain minerals, i.e., the clay minerals, must be present in appreciable
amounts if the clays are to have the plastic properties associated with the
term clay. The shape of such particles, their adsorptive and surface
properties, in addition to their small size, are essential if a material is to
have the characteristics of clay.
CLAY -MINERAL CONCEPT
For many years some students of clay materials have suggested that
such materials are composed of extremely small particles of a limited
number of crystalline minerals. For example, Le Chatelier29 and Lowen-
stein 30 arrived at this conclusion in 1887 and 1909, respectively. This is
the clay-mineral concept, but prior to about 1920 to 1925 there were no
adequate research tools to provide positive evidence for it. The clay-
mineral concept, therefore, is not new; rather it has been well established
and generally accepted in recent years.
In 1923, Hadding 31 in Sweden and in 1924 Rinne 32 in Germany, working
28 Gedroiz, K. K., Die Lehre, vom Adsorptionvermogens der Bodens, Kolloidchem.
Beihefte, 33,317-448 (1931). Translated by H. Kuron.
29 Le Ch{\telier, H., De l'action de la chaleur sut les argiles, Bull. soc. franc. mineral,
10, 204-211 (1887).

30 Lowenstein, K, Ueber Hydrate deren Dampfspannung sich kontinuerlich mit der
Zusammensetzung andert, Z. anorg. Chem., 63, 69-139 (1909).

31 Hadding, A., Eine rontgenographische Methode kristalline und kryptokristalline
8ubstanzen zu identifizieren, Z. Krist., 58, 108-112 (1923).

32 Rinne, F., Rontgenographische Untersuchungen an einigen feinzerteilten Miner-
alien Kunsprodukten und dichten Gesteinen, Z. Krist., 60, 55-69 (924).

quite independently, published the first X-ray-diffraction analyses of
clay materials. Both these investigators found crystalline material in
the finest fractions of a series of clays and also found that all the samples
studied seemed to be composed of particles of the same small group of
minerals. They did not find a large heterogeneous array of minerals of
a wide variety of types in the fine fractions of the samples studied.
In the early years of the present century, careful researches by some
soil scientists were leading many of them to the idea that soils generally
were composed essentially.of definite compounds and that there were a
limited number of such compounds in soils. Previous mention has been
made of the work of Byers 25 and his colleagues of the U.S. Department of
Agriculture which led them to postulate a few definite compounds, such
as halloysitic acid, montmorillonitic acid, or their salts, as the essential
constituoo.ts of all soils. Another example of this trend on the part of
students of soils is the chemical work of Bradfield 33 on the fine fractions
of the Putnam soil from Missouri, which showed that the clay fractions
below a certaih size resembled each other very strongly but that none of
them behaved like mixed gels of hydrous oxides.
About 1924, Ross and some colleagues 34 .35 of the U.S. Geological Survey
began a study of the mineral composition of clays that led to a series of
monumental papers on the subject. Working particularly with ben-
tonites at first, but within a few years with a variety of clays used in
industry and a variety of soils, it was shown, on the basis of extremely
careful and painstaking optical work with the petrographic microscope
supplemented by excellent chemical data, that the components of clay
materials wt:_re largely essentially crystalline and that there was a limited
number of such crystalline components, to which the name clay minerals
. was applied. A classification of the clay 36 minerals was suggested.
Ross's37 work corrected the erroneous notion, still held in some quarters,
that microscopic studies are of no value in clay researches. Ross38 and
his colleagues later added X-ray analysis to their investigations; in
general it substantiated their earlier findings.
33 Bradfield, R., The Nature of the Acidity of Colloidal Clay of Acid Soils, J. Am.
Chem. Soc., 45, 2669-2678 (1923).

34 Ross, C. S., and Eo V. Shannon, The Chemical Composition and Optical Proper-
ties of Beidellite, J. Wash. Acad. Sci., 15, 467-468 (1925).

35 Ross, C. S., and E. V. Shannon, Minerals of Bentonite and Related Clays, and
Their Physical Properties, J. Am. Ceram. Soc., 9, 77-96 (1926).

36 Ross, C. S., The Mineralogy of Clays, First Intern. Congr. Soil Sci. 4, 555-556
(1928).
37 Ross, C. S., Altered Paleozoic Volcanic Materials and Their Recognition, Bull.
Am. Assoc. Petroleum Geol., 12, 143-164 (1928).
38 Ross, C. S., and P. F. Kerr, The Kaolin Minerl1ls, U.S. Geol. Survey Profess. Paper
105E, pp. 151-175 (1931).

18 Clay Mineralogy
About 1926, Marshall 39 40 began a study of the optical characteristics
of clay-water suspensions when they were placed in an electrical field.
He devised a quantitative method for measuring the birefringence result-
ing from the aggregate orientation that developed in the electrical field.
Marshall's work showed the crystalline nature of the finest fraction of the
soils which he studied. Marshall also showed that the birefringence
exhibited a measurable variation with a variation in the nature of the
exchange cation, indicating, at least in the soils studied, that the sites
of the exchange cations were internal and related to the anisotropy of the
crystal in some definite way. This latter finding did not hold for kaolinite
clays, whose ionic exchanges could then be ascribed to external surfaces
only.
Hendricks and Fry41 in 1930 and Kelley, Dore, and Brown42 in 1931,
presented separate papers from independent work showing, chiefly
on the basis of X-ray-diffraction analyses, that soil materials, even
in their finest size fractions, are composed of crystalline particles and
that the number of different crystalline minerals likely to be found is
limited.
By the early 1930's what has come to be known as the clay-mineral
concept became firmly established in the minds of a great many people
actively studying clay materials. Ross and Kerr 43 in 1931, Endell,
Hofmann, and Wilm 44 in 1933, and Correns,45 in 1936 published par-
ticularly concise statements of this concept. At the present time it has
come to be accepted by almost all students of clays. There are a few'
investigators, notably Mattson,21 who have clung to the notion that the
essence of substantially all clays is an amorphous, extremely variable
material which cannot be revealed by X-ray-diffraction analyses.
According to the clay-mineral concept, clays generally are essentially'
composed of extremely small crystalline particles of one or more members
of a small group of minerals which have come to be known as the clay
minerals. The clay minerals are essentially hydrous aluminum silicates,
with magnesium or iron proxying wholly or in part for the aluminum in
39 Marshall, C. E., The Orientation of Anisotropic Particles in an Electric Field,
Trans. Faraday Soc., 26, 173-189 (1930).
40 Marshall, C. E., Clays as Minerals and Colloids, Trans. Ceram. Soc. (Engl.), 30,
81-96 (1931).
41 Hendricks, S. B., and W. H. Fry, The Results of X-ray and Microscopical Exami-
nations of Soil Colloids, Soil Sci., 29, 457-478 (1930).
42 Kelley, W. P., W. H. Dore, and S. M. Brown, The Nature of the Base-Exchange
Material of Bentonites, Soils, and Zeolites as Revealed by Chemical and X-ray

Analyses, Soil. Sci. 31, 25-45 (1931).
43 Ross, C. S., and P. F. Kerr, The Clay Minerals and Their Identity, J. Sediment.
Petrol., 1, 55-65 (1931).

44 Endell, K., U. Hofmann, and D. Wilm, Ueber die Natur der keramischen Tone,
Ber. deut. keram. Ges., 14, 407-438 (1933).
46 Correns, C. W., Petrographie der Tone, Naturwissenschaften, 24, 117-124 (1936) ..
some minerals and with alkalies or alkaline earths present as essential
constituents in some of them. Some clays are composed of a single clay
mineral, but in many there is a mixture of them. In addition to the clay
minerals, some clay materials contain varying amounts of so-called
nonclay minerals, of which quartz, calcite, feldspar, and pyrite are
important examples Also many clay materials contain organic matter
and water-soluble salts (see page 6).
According to the clay-mineral concept, the crystalline clay minerals
are the essential constituents of nearly all clays and, therefore, the
components which largely determine their properties. As noted before,
the nonclay minerals and some other factors (see page 3) will also
influence properties if they are present in appreciable amounts. In the
e'1rly years of the acceptance of the clay-mineral concept, it was thought
that amorphous material was substantially completely absent in almost
all clay material~ Some amorphous material had been found, but it was
thought to be limited to a few unique clays, e.g., in association with the
halloysite in the so-called indi"anaite from Indiana. 46 Recent work 47
indicates that extremelY,Poorly crystalline material that appears in some
C:1ses to be actually amorphous to X-ray diffraction is not so rare as it
was believed to be earlier. It does not, however, appear to be a very
common component, and the great majority of clay materials appear to be
entirely crystalline. Certainly amorphous material is not present in all
or even in most clay materials, and it cannot be considered a universal
constituent responsible for clay properties. The presence of poorly
crystalline material is revealed by X-ray-diffraction data, but the
presence of definitely amorphous material is usually hard to establish.
Its presence is usually suggested when the analytical data do not indicate
the crystalline constituents to be present in sufficient quantities to add
up to 100 per cent.
Since about 1930 and the general acceptance of the clay-mineral con-
cept, there has been intense interest in the study of clay materials, and a
. very voluminous literature has developed. Workers have approached
the study of clay mineralogy from many different fields-mineralogy,
geology, chemistry, physics, agronomy, etc. Also a tremendous amount
of work has been done on applied clay mineralogy by ceramists, engi-
neers, etc., in a host of university, commercial, and other laboratories.
The clay-mineral literature appears in an extremely wiqe variety of
publications-in chemical, physical, mineralogical, ceramic, and other
journals-as is to be expected because of the wide range of backgrounds
and approaches of the persons working in the field.
46 Ross, C. S., and P. F. Kerr, Halloysite and Allophane, U.S. Geol. Survey, Profe8s.
Paper 185G, pp. 135-148 (1934).

47 Grim, R. E., Some Factors Influencing the Properties of Soil Materials, Second
Intern. Congr. Soil Meeh., 3,8-12 (1948).

20 Clay Mineralogy
In the following statements an attempt will be made to indicate some
of the more important contributions which have appeared in the develop-
ment of clay mineralogy. It is impossible to mention all or even most
of those who have contributed significantly, and an attempt can be made
merely to list some of those who pioneered in the study of the clay
minerals.
Reference has already been made to the monumental work of Ross
and his colleagues, which did much to establish the clay-mineral concept.
The work of Ross and Kerr on kaolinite 38 and halloysite 46 and later of
Ross and Hendricks 48 on montmorillonite provided fundamental data on
the properties of these clay minerals essential to their determination in
clay materials. Hendricks, working alone and with a series of colleagues
from the U.S. Department of Agriculture, has produced a series of out-
standing papers on the structure of the clay minerals and on many
of their physical attributes. Among the many contributions of Hendricks
of particular importance, in addition to his structural studies,49-51
are his works on cation exchange,52 on the reaction of organic ions and the
clay minerals,53 on the hydration characteristics of certain of the clay
minerals,54 and on the nature of the water adsorbed on the surface of the
clay-mineral particle. 55
Since about 1930, Kelley and his colleagues 56 ,57 at the University of
California have contributed immensely to our knowledge of the distribu-
tion of the clay minerals in various soil types and the soil-forming condi-
48 Ross, C. S., and S. B. Hendricks, Minerals of the Montmorillonite Group, U.S.

Geol. Survey Profess. Paper 205B, pp. 23-79 (1945).
49 Hendricks, S. B., On the Structure of the Dickite, Halloysite and Hydrated
Halloysite, Am. Mineral., 23,275-300 (1938).

00 Hendricks, S. B., Polymorphism of the Micas with Optical Measurements by
M. E. Jefferson, Am. Mineral., 24, 729-771 (1939).
01 Hendricks, S. B., Lattice Structure of Clay Minerals and Some Properties of
Clays, J. Geol., 50, 276-290 (1942).
02 Hendricks, S. B., Base-Exchange in the Crystalline Silicates, Ind. Eng. Chem.,
37,625-630 (1945).
03 Hendricks, S. B., Base Exchange of the Clay Mineral Montmorillonite for
Organic Cations and Its Dependence upon Adsorption Due to van der Waals Forces,
J. Phys. Chem., 45, 65-81 (1941).
04 Hendricks, S. B., R. A. Nelson, and L. T. Alexander, Hydration Mechanism of
the Clay Mineral Montmorillonite Saturated with Various Cations, J. Am. Chem.
Soc., 62, 1457-1464 (1936).
05 Hendricks, S. B., and M. E. Jefferson, Structures of Kaolin and Talc-Pyrophyl-
lite Hydrates and Their Bearing on Water Sorption of the Clays, Am. Mineral. 23.
863-875 (1938).
06 Kelley, W. P., W. H. Dore, and A. O. Woodford, The Colloidal Constituents of
California Soils, Soil Sci., 48, 201-255 (1939).

67 Kelley, W. P., W. H. Dore, and J. B. Page, The Colloidal Constituents of Amer-
ican Alkali Soils, Soil Sci., 51, 101-124 (1941).

tions under which various clay minerals form and are stable. Jenny58
of this school has made outstanding contributions to concepts of cation
exchange, and recently Barshad,59 also of this school, has published
valuable data on the vermiculite and chlorite clay minerals.
Beginning about 1931, Grim, Bradley, and their colleagues 60- 62 at
the Illinois State Geological Survey and the University of Illinois have
studied the illite clay minerals and the composition of many clays and
shales. They have been particularly interested in the relation of the
clay-mineral composition to the plastic, burning, strength, and other
properties of clay materials which determine their utility in ceramics, soil
mechanics, oil-well drilling, and other applied fields. They have also
worked with the development of the differential thermal procedure 63 for
the analysis of clay materials and have studied the composition of recent
marine sediments. 64 Bradley,65 working independently, determined the
structure of attapulgite, which provided for the first time an insight into
the structure of some of the fibrous clay minerals.
The classical investigation of the structure of the layer silicates by
Pauling 66 provided the basic id~s permitting the elaboration of the struc-
ture of the layer clay minerltls. Following Pauling's original ideas,
Gruner worked out a structure for kaolinite 67 and vermiculite. 68 The
structural concepts of the latter mineral largely led the way to an under-
I standing of interlayered mixtures of cla~ minerals. Hendricks and Teller69
,
58 Jenny, H., Studies of the Mechanism of Ionic Exchange in Colloidal Aluminum
Silicates, J. Phys. Chem., 36, 2217-2258 (1932).
59 Barshad, 1., The Effect of Interlayer Cations on the Expansion of the Mica Type
of Cry~tal Lattice, Am. Mineral., 35, 275-238 (1950).

I 60 Grim, R. E., R. H. Bray, and W. F. Bradley, The Mica in Argillaceous Sediments"
Am. Mineral., 22, 813-829 (1937).
61 Grim, R. E., Relation of the Composition to the Properties of Clays, J. Am.
Ceram. Soc., 22, 141-151 (1939).

6' Grim, R. E., and F. L. Cuthbert, The Bonding Action of Clays, Illinois State
Geol. Survey Repts. Invest. 102, 110 (1946).
63 Grim, R. E., and R. A. Rowland, Differential Thermal Analysis of Clay Minerals
and Other Hydrous Materials, Am. Mineral., 27, 746-761 (1942).

64 Grim, R. E., W. F. Bradley, and R. S. Dietz, Clay Mineral Composition of Some.
Sediments from the Pacific Ocean off the California Coast and the Gulf of California,
Bull. Ceol. Soc. Am., 60, 1785-1805 (1949).
66 Bradley, W. F., The Structural Scheme of Attapulgite, Am. Mineral., 26,405-410
(1940).
66 Pauling, L., The Structure of Micas and Related Minerals, Proc. Nall. Acad. Sci.
U.S., 16, 123-129 (1930).

67 Gruner, J. W., The Crystal Structure of Kaolinite, Z. Krist., 83, 75-88 (1932).
68 Gruner, J. W., The Structure of Vermiculites and Their Collapse by Dehydration,
Am. Mineral., 19,557-574 (1934).

69 Hendricks, S. B., and E. Teller, X-ray Interference in Partially Ordered Layer
Lattices, J. Chern. Phys., 10, 147-167 (942).

;<1"""'I,.""'l~ r
'"""._,.~,".
22 Clay Mineralogy
later worked out the theory of X-ray diffraction for such inter-
layered mixtures, which has permitted their detailed study and evaluation.
Contributions by Gieseking 70 greatly enhanced our knowledge of the
adsorption of organic ions and their influence on the structure of mont-
. morillonite. Bradley 71 pointed out that certain nonionic polar organic
molecules are also adsorbed and indicated the nature of the bond between
such molecules and the clay-mineral surfaces. Later Jordan 72 in some
extremely interesting and important work showed that multiple adsorp-
tion of organic molecules by montmorillonite was possible and described
organophilic montmorillonite-organic complexes which formed gels in
organic liquids.
At Pennsylvania State College, a group led by H enry73 and Bates 74
have made notable contributions to clay mineralogy. Bates in recent
years has contributed particularly to the knowledge of the halloysite
minerals.
In Germany beginning about 1931 several schools of investigators
began to study the clay minerals. At the Technische Hochschule in
Berlin, Hofmann, Endell, and Wilm 75 began a study of montmorillonite
which resulted in a suggested structure for this mineral which is now
widely accepted in its broader outlines. This same team of workers 76 77
investigated many German clays and were among the first to show the
relationship between the properties of the individual clay minerals and
the larger ceramic and other properties of the clays themselves.
Correns 78 and his associates at the University of Rostock published
valuable data on the X-ray, optical, and chemical specifications of the
various clay minerals. MehmeP9 of this group was the first to indicate
70 Gieseking, J. E., Mechanism of Cation Exchange in the Montmorillonite-Beidel-
lite-Nontronite Type of Clay Mineral, Soil Sci., 47, 1-14 (1939).
71 Bradley, W. F., Molecular Associations between Montmorillonite and Some Poly-
functional Organic Liquids, J. Am. Chem. Soc., 67, 975-981 (1945).
72 Jordan, J. W., Organophilie Bentonites, J. Phys. &: Colloid Chem., 53, 294-306
(1949).
13 Henry, E. C., and A. C. Siefert, Plasticity and Drying Properties of Certain Clays
as Influenced by Electrolyte Content, J. Am. Ger. Am. Soc., 24, 281-285 (1941).
74 Bates, T. F., F. A. Hildebrand, and A. Swineford, Morphology and Structure of
Endellite and Halloysite, Am. Mineral., 35, 463-484 (1950).
75 Hofmann, U., K. Endell, and D. Wilm, Kristallstruktur und Quellung von Mont-
morillonit, Z. Krist., 86, 340-348 (1933).

75 Hofmann, U., K. Endell, and D. Wilm, Rontgenographische und kolloidchemische
Untersuchungen uber Ton, Angew. Chern., 47, 539-547 (1934).
77 Endell, K., U. Hofmann, and E. Maegdefrau, Ueber die Natur des Tonanteils in
Rohstoffen der deutschen Zementindustrie, Zement, 24, 625-632 (1935).
78 Correns, C. W., and M. Mehmel, Ueber den optisehen und rontgenographischen
Nachweis von Kaolinit, Halloysit, und Montmorillonit, Z. Krist., 94,337-348 (1936).

7. Mehmel, M., Ueber die Struktur von HaUoysit und MetahllJloysit, Z. Krist., 90,
35--43 (1935).
clearly the two forms of halIoysite and to show their relations to each
other. They also showed that the indices of refraction of some of the
clay minerals may vary with the nature of the index liquid used to
measure them. Correns,80 studying the marine-bottom samples col-
lected by the Meteor, was among the first to investigate recent marine
sediments by modern analytical techniques.
At Hanover in Germany, N olI sl began about 1933 the study of the
laboratory synthesis of the clay minerals and produced the best data
yet available on this subject.
In the Netherlands at the Agricultural College in Wageningen, Edel-
man and his colleagues S2 began an active study of clay mineralogy in the
early 1930's. They developed a structural concept 83 of montmorillonite
and halloysite, somewhat different from that suggested by Hofmann
et al.,15 which still claims considerable attention. They 84 also determined
the clay-mineral compostion of many types of soil.
In 1927, Orcel 85 in Paris first applied the differential thermal procedure
in its modern form to the study of clay minerals, and its general use today
is due in no small part to Orcel's efforts. In association with Orcel 86 at
first and later independently, Henin and Caillere 87 have actively pursued
a variety of clay-mineral investigations. They have contributed par-
ticularly to our knowledge of the sepiolite-palygorskite clay minerals. 88
Other French investigators, notably Longchambon 89 and De Lapparent,90
80 Correns, C. W., et al., Die Sedimente des aquatorialen atlantischen Ozeans,

Deutsche atlantische Exped. Meteor 1925-1927, Wiss. Ergeb., 3, Teil 3 (1937).
81 Noll, W., Mineralbildung im System Ab03-Si02-H20, Neue8 Jahrb. Mineral.
Geol. Beilage Bd. A, 70, 65-115 (1935).
82 Edelman, C. H-. Relation between the Crystal Structure of Minerals and Their
Base-Exchange Capacity, Trans. Intern. Congr. Soil Sci. 3rd Congr., Oxford, 3, 97-99
(1936).
83 Edelman, C. H., and J. C. L. Favejee, On the Crystal Structure of Montmoril-
lonite and Halloysite, Z. Krist., 102, 417-431 (1940).
84 Edelman, C. H., J. C. 1,. Favejee, and F. A. Van Baren, General Discussion of
the Mineralogical Composition of Clays and Quantitative X-ray Analyses of Dutch
Clays, Overdruk uit Mededeel. Landbouwhoogeschool, 43, 1-39 (1939).
850rceI, J., Researches on the Chemical Composition of the Chlorites, Bull. soc.
fraru;. mineral., 60, 75-450 (1927).
860rcel, J., and S. Calliere, L' Analyse thermique differentielle des argiles a mont-
morillonite (bentonite), Compt. rend., 197, 774-777 (1933).
87 Caillere, S., and S. Henin, Application de ['analyse thermique differentielle de
l'etude des argiles des sols, Ann. agron., 50 pp. (1947).
"8 Caillere, S., Observation sur les composition chemique des palygorskites, Compt.
rend., 199, 1795-1798 (1934).
89 Longchambon, H., Sur certain caracteristiques de la sepiolite d' Ampandandrava
et Ie formule des sepiolites, Bull. soc. franr., mineral., 60, 1-45 (1937).
90 De Lapparent, J., Les Argiles des terres a fouIon, Congr. intern. mines met. et geol.
appl., 1,381-387 (1935).
24 Clay Mineralogy
also have contributed important information regarding this group of clay
minerals.
More recently a second group of French students of clays, led by
.M:ering 91 92 at the National Chemical Laboratory in Paris, has produced
lJxtremely important contributions to the subject. Mering 93 contributed
to the X-ray analysis of the interlayer mixtures of clay minerals, extend-
ing the earlier work of Hendricks and Teller. 69 He has also enhanced our
knowledge of the hydration of montmorillonite and of the catalytic
activity of the same mineral. His colleagues 94 95 have recently con-
tributed to the adsorption phenomena of the clay minerals and their
texture as revealed by the electron microscope.
Among the early contributors to clay mineralogy in Great Britain were
Nagelschmidt and Brammall. Nagelschmidt 96 97 published clay-mineral
analyses of a variety of soils and was among the first to suggest that some
of the clay minerals had an elongate fibrous form. Bramma1l 98 99 con-
tributed particularly to the illite clay minerals and to the possible vari-
ations in the composition of some of the other clay minerals.
More recently a very great interest in clay mineralogy has developed
in Great Britain. MacEwan loo of the Rothampsted Agricultural Experi-
ment Station has studied in detail the reaction vf the clay minerals and
organic compounds. He has presented the analyses of many soil mate-
rials 101 and has contributed to X-ray techniques 102 and the structure of
91 Mering, J., Reactions des montmorillonit, Bull. soc. chim. France, pp. 218-223
(1949).
92 Mering, J., The Hydration of Montmorillonite, Trans. Faraday Soc., 42B,205-
219 (1946).
93 Mering, J., L'Interference des rayons X dans les systemes a stratification disor-
donnee, Acta Crystallog., 3,371-377 (1949).

94 Glaesser, n., On the Mechanism of Formation of Montmorillonite-Acetone Com-
plexes, Compt. rend., 222, 935-938 (1948).

95 Mering, J., A. Mathieu-Sicaud, and I. Perrin-Bonnet, Observation au microscope
electronique de montmorillonite saturce par different cations, Trans. 4th Intern.
Congr. Soil Sci., 3,29-32 (1950).
96 Nagelschmidt, G., Rod-shaped Clay Particles, Nature, 142, 114-115 (1938).
97 Nagplschmidt, G., Identification of Minerals in Soil Colloids, J. Agr. Sci., 29,
477-501 (1939).
98 Brammall, A., J. G. Leech, and F. A. Bannister, The Paragenesis of Cookeite and
Hydromuscovite, Mineralog. Mag., 24,507-520 (19:37).

99 Brammall, A., and .J. G. Leech, Base-Exchange and Its Problems, Sci. J. Roy.
Call. Sci., 7, 69-78 (1937).

100 MacEwan, D. M. C., Complexes of Clays with Organic Compounds, Trans.
Faraday Soc., 44, 349-367 (1948).

101 MacEwan, D. M. C., Les Mincraux argileux de quelques sols ecossais, Verre
silicates ind., 12, 3-7 (1947).

102 MacEwan, D. M. C., Some Notes on the Recording and Interpretation of X-ray
Diagrams of Soil Clays, J. Soil Sci., 1,90-103 (1949).

the clay minerals. Brindleylo3,lo4 at the University of Leeds has studied
the kaolinite minerals in detail and has shown the possible variations
in the degree of crystallinity of these minerals. At the Macauley Agri-
cultural Research Institute in Aberdeen, Scotland, WalkerI 05 and
Mackenzie 106 have studied the composition of Scottish soils and con-
tributed to our knowledge of the biotite-like and vermiculite-like clay
minerals.
Considerable investigation of clay minerals has been made in the
U.S.S.R. by Belyankin, 107 Sedletsky,108 and others. 109 Unfortunately
much of the v'ork of the Russian investigators is not available outside of
that country, so that neither the magnitude nor the worth of it can be
adequately evaluated.
In other countries, notably in Sweden with the work of Forslind,l1O in
Australia with the work of Hoskingll 1 and others,112 in India with the
work of Mukherjee,113 Chatterjee,114 and others, in Italy with the
work of Gallitelli,115 and in Spain with the work of Hoyos,116 notable
contributions to the knowledge of clay mineralogy have been and are
being made.
103 Brindley, G. W., and K. Robinson, The Structure of Kaolinite, Mineralog.
Mag., 27,242-253 (1946).

104 Brindley, G. W., and K. Robinson, Randomness in the Structures of Kaolinitic
Clay Minerals, Trans. Faraday Soc., 42B, 198-205 (1946).

105 Walker, G. F., Trioctahedral Minerals in the Soil Clays of Northeast Scotland,
Mineralog. Mag., 29,72-84 (l,1l50).

106 Mackenzie, R. C., G. F. Walker, and R. Hart, Illite in Decomposed Granite at
Ballater, Aberdeenshire, Mineralog. lYfag., 28, 704-714 (1949).

107 Belyankin, D. S., On Characteristics of the Mineral" Monothermite," Compt.
rend. acado sci. URSS, 18, 673-676 (1938).

108 Sedletsky, 1., Mineralogy of Dispersed Colloids, Acad. sci. URSS, 114 pp. (1945).
109 Pinsker, Z. G., Electronographic Determination of the Elementary Cell and

Space Group of Kaolinite, Dokjady Akad. Nauk. SSSR, 73, 107-110 (1950).
. 110 Forslind, E., The Clay Water System, Swed. Concrete Cement Research Inst. Bull.
11, (1948).
III Hosking, J., The Clay Mineralogy of Some Australian Soils Developed on
Granitic and Basaltic Parent Material, J. Council Sci. Ind. Research, 13, 206-216
(1941).
112 Shearer, J., X-ray Powder Analysis and Its Application to Soil Colloids, Australian
J. Sci., 5, 43-47 (1942).

113 Mukherjee, J. N., and R. P. Mitra, Some Aspects of the Electrochemistry of
Clays, J. Colloid Sci., 1, 141-159 (1946).

114 Chatterjee, B., Tho Role of Aluminum in the Interaction of Clays and Clay
Minerals with Neutral Salts and Bases, J. Indian Chem. Soc., 12,81-99 (1949).
115 Gallitelli, P., Uno Sguardo ad alcuni nuori Aspetti del Problema" Argilla," Soc.
ital. progresso sci. 42, 1-10 (1951).
116 Hoyos de Castro, A., and M. Delgado, Origin of the Bed of Kaolin of Carataunas,
Anales edafol. fisiol. vegetal, 8, 703-785 (1949).

26 Clay Mineralogy
Old Concepts
Ashley, H. E., The Colloidal Matter of Clay and Its Measurement U.S. Geo[. Survey
Bull. 3S8, pp. 1-62 (1909).
Byers, H. G., and M. S. Anderson, The Composition of Soil Colloids in Relation to
Soil Classification, J. Phys. Chem., 36, 348-366 (1932).
Calsow, G., Kaolin und Tone, Chem. Erde, 2, 415-441 (1926).
Davis, N. B., The Plasticity of Clay and Its Helation to Mode of Origin, Trans. Am.
Inst. Mining Met. Engrs., 51, 451-480 (1915).
Dufrenoy, A., "Traite de mineralogie," 2d ed., vol. III, Paris (1856).
Harrassowitz, H., Fossile Verwitterungsdecken, "Handbuch der Bodenlehre," vol. 4,
pp. 225-302, Springer, Berlin (1930).
Hissink, D. J., Base-Exchange in Soils, Trans. Faraday Soc., 20, 551-566 (1924).
Koettgen, P., Die Zusammensetzung der Silikatkomplexe einiger dialytischer Pelite,
.Tahrb. preuss. geol. Landesanstalt (Berlin), 42, 626-656 (1921).
Linck, G., Ueber den Chemismus der toniger Sedimente, Geol. Rundschau 4, 289-311
(1913) .
Merrill, G. P., "Rock-Weathering and Soils," Macmillan, New York (1904).
Robinson, G. 'V., "Soils: their Origin, Constitution and Classification," 2d ed.,
Murby, London (1936).
Roburgh, R. H. J., and H. Kolkmeyer, Ueber die Struktur des Adsorptionkomplexes
der Tone, Z. Krist., 94, 74-79 (1936).
Savelly, T. F., Kaolins and Refractory Clays in Italy, Corriere ceram., 11,299-30::\ (1930).
Schloessing, T., Surla constitution des argiles, Compt. rend., 79,376-380,473-477 (1874).
Van Bcmmelen, J. M., Beitriige zur Kcnntnis der Verwitterungsprodukte des Silikate
in Ton-Vulkanischen und Laterit-Boden, Z. anorg. Chern., 42, 265-314 (1904).
Clay-mineral Concept
Alexander, L. T., S. B. Hendricks, and R. A. Nelson, Minerals Present in Soil Colloids,
II, Estimation in Some Representative Soils, Soil Sci., 48, 273-279 (1939).
Favejee, J. C. L., Quantitative rontgenographische Bodenuntersuchung, Z. Krist.,
101,259-270 (1939).
Grim, R. E., Petrography of the Fuller's Earth Deposit, Olmstead, Illinois with a Brief
Study of Some Non-Illinois Earths, Econ. Geol., 28,344-363 (1933).
Hadding, A., X-ray Investigations of Clays and Some Other Substances, Trans. Ceram.
Soc. (Engl.), 24, 27-32 (1924-25).
Hardon, H. J., and J. C. L. Faveice, Mineralogical Investigations of Clays and Clay
Minerals, III, Quantitative X-ray Analysis of the Clay Fraction of the Principal
Soil Types of Java, Mededeel. Landbouwhoogeschool Wageningen, 43, 55-59 (1939).
Hofmann, U., and K. Giese, Ueber den Kationenaustausch an Tonmineralien, Kol-
loid-Z., 87, 21-36 (1939).
Kelley, W. P., The Evidence as to the Crystallinity of Soil Colloids, Trans. Intern.
Congr. Soil Sci. 3rd Congr., Oxford, 3, 88-91 (1936).
Le ChAtelier, H., Ueber die Konstitution der Tone, Z. physik. Chern., 1,396-402 (1887).
Marshall, C. E., Mineralogical Methods for the Study of Silts and Clays, Z. Krist.,
90,8-34 (1935).
Nagelschmidt, G., On the Lattice Shrinkage and Structure of Montmorillonite, Z.
Krist., 93,481-487 (1936).
Nagelschmidt, G., A. D. Desai, and A. Muri, The Minerals in the Clay Fraction of the
Black Cotton Soils and Red Earths from Hyderabad, .T. Agr. Sci., 30, 639-653
(1940).
CHAPTER 3
Classification and Nomenclature of the Clay Minerals
CLASSIFICATION OF THE CLAY MINERALS
The following classification of the clay minerals is suggested on the

basis of currently available data on the structure and composition of the
various clay minerals. Because such data are far from complete, the
classification can only be tentative, and future investigations may show
that major revisions are necessary. Some classification, however, is
required as a basis for a discussion of the clay minerals.
TABLE 1. CLASSIFICATION OF THE CLAY MINERALS
I. Amorphous
Allophane group
II. Crystalline
A. Two-layer type (sheet structures composed of units of one layer of silica tetra-
hedrons and one layer of alumina octahedrons)
1. Equidimensional
Kaolinite group
Kaolinite, nacrite, etc.
2. Elongate
Halloysite group
B. Three-layer types (sheet structures composed of two layers of silica tetra-
hedrons and one central dioctahedral or trioctahedrallayer)
1. Expanding lattice
a. Equidimensional
Montmorillonite group
Montmorillonite, sauconite, etcl
Vermiculite
b. Elongate
Montmorillonite group
Nontronite, saponite, hectorite
2. Nonexpanding lattice
Illite group
C. Regular mixed-layer types (ordered stacking of alternate layers of different
types)
Chlorite group
D. Chain-structure types (hornhlende-like chains of silica tetrahedrons linked
together by octahedral groups of oxygcns and hydroxyls containing Al and
:\1g atoms)
Attapulgite
Sepiolite
Palygorskite
27
28 Clay Mineralogy
The suggested classification is based on structural attributes of the clay
minerals. The major subdivision into amorphous and crystalline groups
seems to be necessary, even though the amorphous clay minerals are
relatively rare and of little importance.
The separation of the three-layer minerals into expanding and non-
expanding types may ultimately prove unwise, since there may be a
continuous gradation between some nonexpanding and expanding clay
minerals, and some interlayered complexes are difficult to place in either
category. However, the expanding attribute is usually a readily deter-
minable diagnostic criterion, and it imparts generally more or less unique
physical properties to clay materials composed of such minerals.
Also the separation of the expanding minerals into equidimensional
and elongate divisions may ultimately prove to be undesirable, because
there may be a complete gradation between them. At the present time
no separate name has been applied to the elongate varieties; mont-
morillonite has been used as a group to include both varieties. The
division is made because the shape is a readily re~ognizable characteristic
that probably reflects important structural attributes.
No attempt has been made to create new separate categories on the
basis of the population of the octahedral layer, i.e., whether the mineral
is dioctahedral or trioctahedral. Future work may show this to be
desirable and feasible. Examples of both kinds of three-layer types are
known, but their relation to each other and their prevalence are not
known. It is not established, for example, whether there can be com-
plete gradations between them or only limited variations from the exact
dioctahedral and trioctahedral populations. Further, it is frequently
impossible at the present time to determine whether a given clay mineral
which may occur intimately mixed with another mineral is dioctahedral
or trioctahedral. A classification based on this attribute is, therefore,
now unworkable.
At the present time the chlorite minerals are the only known examples
of clay minerals with a regular ordered interstratification of different
types of layers. Future researches may well show that there are other
examples.
No separate category is made for random interlayer mixtures, since
their character varies with the mixing. They must, therefore, be des-
ignated as mixtures of the layers involved.
The clay minerals with the hornblende-like structures are particularly
poorly known. Almost certainly future investigations of these minerals
will make some sort of subdivision of them possible. All that can be
done at this time is to list examples that appear to belong together in this
category.
Other reasons for suggesting the classification given in Table 1 will
Classification and Nomenclature of the Clay Minerals 29
appear in the discussion to follow and in the consideration of the struc-
tures of the various clay-mineral groups. Details of the classification
will also be presented at that time.
I
NOMENCLATURE OF THE CLAY MINERALS I
trhe name allophane was first applied by Stromeyer and Hausmann 1 in

1816 to material lining cavities in marl. The term is derived from Greek
words meaning "to appear" and "other,'! in allusion to its frequent
change on standing from a glassy material to one with an earthy appear-
ance because of the loss of water. Since the work of Stromeyer and
Hausmann l a considerable number of materials have been described as
allophane or classed with it. In general all such material was thought
to be amorphous, and hence the name allophane came to be associated
with amorphous constituents of clay. The study of much of this mate-
1:\.\.\,l, ?\'\'1:t\.~\ll\'\'1:ly by Ro'i','i', \'\'nd K~n,2 'i',h.ow~d that mm~ Q, it wa'i', \V~fv\lalty
crystalline but that much of it was really amorphous to X-ray diffraction.
They suggest that the term be used for all amorphous clay-mineral mate-
rials regardless of their composition. These investigators point out that
the known examples of amorphous material in clays and soils show con-
siderable variation in composition.
As has been indicated (page 13), the word allophaneton has been used
for the portion of a clay soluble in hydrochloric acid. At the time the
term was suggested such material was thought to be amorphous. It is
now known that some of the crystalline cl8-Y minerals, e.g., montmoril-
lonite, are fairly soluble in this acid and that others are somewhat soluble
if present in extremely small particles. The term allophaneton has about
disappeared from usage.
Ross and Kerr 2 discuss in detail the nomenclature of the kaolin min-
erals, and their usage has been generally accepted. According to them,
. . . by kaolin is understood the rock mass which~s composed essentially of a
clay material that is low in iron and usually white or nearly white in color. The
kaolin-forming clays are hydrous aluminum silicates of approximately the com-
position 2H 2 0 AhOa 2Si0 2 and it is believed that other bases if present represent
impurities or adsorbed materials. Kaolinite is the mineral that characterizes
most kaolins.
The name kaolin is a corruption of the Chinese "kauling" meaning "high
ridge," the name of a hill near Jauchau Fu, China, where the material
was obtained centuries ago. Occurrences in many parts of the world are
1 Strom eyer, F., and J. F. L. Hansmann, Gottingische Gelehrte Anziegen, 2, 125
(1816).
2 Ross, C. S., and P. F. Kerr, The Kaolin Mineral$, U.S. Geol. Survey Profess. Paper
165E, pp. 151-175 (1931) 1 r, ..

<r\~~t .... r"Ie of i'-,Qr1('l;',:'i:':,~~
V"'""-'"""6 ,_-'
30 Clay Mineralogy
well known today. Johnson and Blakes in 1867 appear to have first
clearly intended the name kaolinite for the" mineral of kaolin." Ross
and Kerr 2 have shown that the kaolin minerals cannot be assigned to a
single species; i.e., clays of this character are not composed of a single
mineral species, and minerals of that composition also are not all the
same species. They concluded that three distinct species are represented,
namely, kaolinite, nacrite, and dickite. N acrite was proposed by
Brongniart 4 in 1807. Later Des Cloizeaux,5 and much later Dick,6
described a mineral called nacrite from mines in Saxony with sufficient
analytical data to differentiate it from the" Dick mineral" but not from
the" mineral of kaolin." Mellor 7 accepted nacrite as a distinct mineral,
and Ross and Kerr 2 finally established its identity.
Dick 6 in 1908 described a mineral from the island of Anglesey in Wales,
without giving it any specific name, which was referred to with other
clay minerals as a "mineral of kaolin." Ross and Kerr 2 showed that the
"Dick mineral" was a distinct species and first applied the name dickite.
Anauxite was proposed as a mineral name by Breithaupt 8 in 1838 for a
mineral from Bilin, Czechoslovakia. SmirnofP in 1907 gave a detailed
description of it and showed photomicrographs of characteristic" worm-
like" structures resembling those of kaolinite. The mineral from Bilin
had a higher silica-to-alumina ratio than that usually found in the
"mineral of kaolin," and Ross and Kerr 2 proposed that anauxite be
defined as a mineral with essentially the same attributes as kaolinite
but with a higher silica-alumina molecular ratio, frequently approach-
ing 3.
Anauxite is believed to be an interlayer mixture of a double silica layer
and a two-layer type of sheet (see pages 51,52). The mixing of the layers
is not regular, and therefore it should not be placed in category C in
Table l. Further, the dimensions of both layers are substantially the
same, so that it yields X-ray and optical data similar to kaolinite. It
seems desirable to include it in the same category as kaolinite.
Very recently it has been shown, particularly by Brindley and Robin-
3 Johnson, S. W., and J. M. Blake, On Kaolinite and Pholerite, Am. J. Sci., ser. 2,
43,351-361 (1867).
4 Brongniart, A., "Trait6 616mentaire de mincralogie," vol. 1, p. 506 Paris (1807).
Des Cloizeaux, A., "Supplement to Manual of Mineralogy," vol. 1, pp. 548-549

Paris (1862).
6 Dick, A. B., Supplementary Note on the Mineral Kaolinite, Mineralog. Mag., 15,
127 (1908).
1 Mellor, J. W., Do Fireclays Contain Halloysite or Clayite? Trans. Ceram. Soc.
(Engl.), 16,83 (1916).
8 Breithaupt, A., Bestimmung neuer Mineralien J. prakt. Chem., 15, 325 (1838).
9 Smirnoff, W. P., Ueber ein Verwitterungsprodukt des Augits, Z. Krist., 43, 338-346
(1907).
son, 10 that there is cOIlsiderable variation in the perfection of st<wking and
possibly also in the precise positions of the aluminums in the octahedral
sheet of members of the kaolinite group. In the usual mineral the stack-
ing is regular, whereas in some specimens random variations in certain
directions are to be found. Brindleyll at first suggested the name
mellorite for the less well crystallized material, thinking that there was a
specific degree of such disorder, More recent work by Brindley 12 himself
suggests that there is a considerable range of disorder in the poorly crys-
tallized kaolinites and that no new specific mineral name is warranted
at this time.
The name halloysite was given by Berthier 13 in 1826 for material found
in pockets in Carboniferous limestone near Liege, Belgium, in a district
of old zinc and iron mines. It was named in honor of Omalius d'Halloy,
who had observed the mineral several years previously. In the years
prior to the development of X-ray-diffraction techniques many materials
were described as halloysite. Dana 14 lists under halloysite 16 mineral
names that he considers to be synonomous with it. Ross and Kerr 15
studied much of this material by modern methods and in addition
obtained samples from the mineralogical collections of the University of
Liege which are probably as nearly representative of the type material
as can be obtained at the present time. They show that halloysite is
crystalline and state that it is closely related to but distinct from kaolinite.
It is now generally accepted that halloysite is distinct from kaolinite and
warrants a separate specific name.
Ross and Kerr 15 in 1934 pointed out that "there are two types of
halloysite---one that is usually white or light-colored, porous, friable, or
almost cottony in texture; and another that is dense, nonporous and
porcelainlike." In the same year Hofmann, Endell, and Wilm 16 pointed
out that there were two forms of the mineral and also that one form was
more hydrous than the other. They found that the more hydrous form
had a larger c-axis spacing than kaolinite and, when dried at 105C,
experienced a structural change accompanying the dehydration to a
10 Brindley, G. W., and K. Ro~inson, Randomness in the Structures of the Kaolinitic

11 Brindley, G. W., Structural Relationships in the Kaolin Group of Minerals, pre-
sented before International Geological Congress, London, 1948.

12 Brindley, G. W., The Kaolin Minerals, "X-ray Identification and Structure of
the Clay Minerals," Chap. III, pp. 32-75, Mineralogical Society of Great Britain
Monograph (1951).
13 Berthier, P., Analyse de l'halloysite, Ann. chim. et phys., 32, 332-335 (1826).
14 Dana, J. D., "System of Mineralogy," 6th ed., Wiley, New York (1914).
1& Ross, C. S., and P. F. Kerr, Halloysite and Allophane, U.S. Geol. Survey Profess.
Paper 185G, pp. 135-148 (1934).

16 Hofmann, U., K. Endell, and D. Wilm, Rbntgenographische und kolloidchemische
Unteruchungen tiber Ton, Angew. Chem., 47,539-547, (1934).

32 Clay Mineralogy
material similar to kaolinite. MehmeP7 in 1935 and Correns and
MehmeP8 clearly distinguished two forms of halloysite: one with the same
chemical composition as kaolinite, and the other, more hydrated form
with a composition differing from that of kaolinite by having an added
2H 2 0. They showed that the transition from the higher hydrated form
to the lower hydrated form was not reversible, that it could take place at
temperatures at least as low as 60C, and that the lower hydrated form
resembled but was not identical with kaolinite. Mehmel17 suggested the
name halloysite for the highly hydrated form and metahalloysite for the
lower hydration form.
Alexander et al. 19 in 1943 suggested that the name halloysite be applied
to the lower hydration form and the new name endellite be given to the
higher hydration form. They made this suggestion because of their
belief that the original material described by Berthier 13 was the lower
hydration form. MacEwan 20 has disputed this point, claiming that the
original material was the higher hydration form. MacEwan suggests
that halloysite be used as a general term for all naturally occurring
specimens of the mineral regardless of their state of hydration. He
points out that natural materials may be in an intermediate state of
hydration, and he would use the adjectives hydrated, nonhydrated, and
intermediate, when the state of hydration is known and when it is impor-
tant to describe it.
Unfortunately there is no agreement among clay mineralogists at the
present time regarding the proper nomenclature for the halloysite
mineral. The original locality is no longer accessible, and it seems impos-
sible to establish to the satisfaction of everyone, the state of hydration
of the original material described by Berthier. 13 This is particularly
true because of the fact tliat the transition and dehydration may take
place at room temperature over long periods of time.
At the International Congress of Soil Science held in Amsterdam in
1950 a special meeting was held of those present interested in clay
mineralogy to discuss the nomenclature of the clay minerals, particularly
the problem of the halloysites. It was the general feeling of the meeting
that great simplification and clarification would result if the word
halloysite were employed for all forms of the mineral and, when neces-
sary, additional self-explanatory qualifications be used, such as fully
17 Mehmel, M., Ueber die Struktur von Halloysit and Metahalloysit, Z. Krist., 90,
35-43 (1935).
18 Correns, C. W., and M. Mehmel, Ueber die optischen und rontgenographischen
Nachweis von Kaolinit, Halloysit, und Montmorillonit, Z. Krist., 94, 337-348 (1936).
19 Alexander, L. T., G. Faust, S. B. Hendricks, and H. Insley, Relationship of the
Clay Minerals Halloysite and Endellite, Am. Mineral., 28, 1-18 (1943).
20 MacEwan, D. M. C., Halloysite Nomenclature, Mineralog. Mag., 28,36-44 (1947).
hydrated, partially hydrated, dehydrated, halloysite-7 -A, etc.21 It
seems that this usage is the only one with any reasonable chance of
wide acceptance, and it will be followed in the present volume.
Damour and Salvetat 22 proposed the name montmorillonite in 1847
for a mineral from Montmorillon, France, which is a hydrous aluminum
silicate with a silica-to-R 20 3 ratio equal to about 4 and with a small
content of alkalies and alkali earths. Le Chatelier 23 later studied the
material and presented the formula 4Si0 2 AI 2 0 3 H 2 0 + aq. for mont-
morillonite; this was accepted by Dana. 14 Dana listed a considerable
number of names of minerals thought to be similar to montmorillonite
wholly or in part.
Ross and his colleagues in a series of classical studies published from
about 1926 24 to 1945 25 established the identity of montmorillonite as a
valid clay-mineral group. They also indicate the variations in composi-
tion that are to be found in members of this group. For example, they
showed the possible variation in the ratio of silica to R 20 3 and the possible
complete replacement of aluminum by iron and magnesium. They also
emphasized the very frequent presence of magnesium in relatively small
amounts in many specimens apparently as an essential ingredient.
Hofmann, Endell, and Wilm 26 in 1933 published a structure for mont-
morillonite showing the expanding-lattice characteristics of the mineral,
and this attribute is now generally considered to be an essential char-
acteristic of the group. Gruner 27 in 1935 and Marshall 28 in 1935 pointed
out possible replacements within the montmorillonite structure and
emphasized their importance.
Cronstedt 29 in 1788 described a material called" smectis" which seems
21 Brindley, G. W., et al., The Nomenclature of the Clay Minerals, Am. Mineral.
36,370-371 (1951).
22 Damour, A. A., and D. Salvetat, Et analyses sur un hydro silicate d'alumine
trouv6 a Montmorillon, Ann. chim. et phys., ser. 3, 21, 376-383 (1847).
23 Le Chatelier, H., De l'action de la chaleur sur les argiles, Bull. soc.franr;. mineral.,
10, 204-211 (1887).

2' Ross, C. S., and E. V. Shannon, Minerals of Bentonite and Related Clays and
Their Physical Properties, J. Am. Ceram. Soc., 9, 77-96 (1926).
Ceol. Survey Profess. Paper 205B, pp. 23-79 (1945).

morillonit., Z. Krist., 86, 340-347 (1933).

27 Gruner, J. W., Structural Relations of Nontronites and Montmorillonites, Am.
Mineral., 20, 475-483 (1935). ...
28 Marshall, C. E., Layer Lattices and Base-Exchange Clays, Z. Krist., 91, 433-449
(1935).
29 Cronstedt, A., "Mineralogie, " Stockholm (1758). English translation by Magel-
lan, John Hyacinth, London (1788).
34 Clay Mineralogy
to be the same as montmorillonite. Kerr 30 in 1932 showed that certain
clay materials which have been described as smectite are actually mont-
morillonite. The name smectite, therefore, is earlier than montmoril-
lonite. However, the name montmorillonite has been commonly used,
while smectite has fallen into disuse or has been used for a fuller's earth
rather than for a definite mineral. As Kerr 30 points out, "in view of the
large amount of modern literature on montmorillonite it seems in the best
interests of science to continue the use of montmorillonite and drop that
of smectite."
The earliest usage of the name saponite is difficult to establish. The
name, derived from "sapo," meaning soap, was used by Svanberg in
1840,31 and in 1842:l2 he published chemical analyses showing the material
to be essentially a hydrous magnesium silicate. As in the case of many
of the other clay minerals, prior to the development of modern analytical
techniques, the mineral could not be well characterized, and the early
literature includes a considerable variety of materials under this name.
Ross and Kerr 33 in 1931 identified saponite as a member of the mont-
morillonite group with a high content of MgO. Ross and Hendricks'~5 in
1945 defined saponite as a member of the montmorillonite group in which
the replacement of AI3+ by Mg++ is essentially complete and with some
replacement of Si 4 + by AP+.
The name montmorillonite is used currently both as a group name for
all clay minerals with an expanding lattice except vermiculite and also as
a specific mineral name. Specifically it indicates a high-alumina end
member of the montmorillonite group with some slight replacement of
AP+ by Mg++ and substantially no replacement of Si H by AP+. Recently
MacEwan 34 has suggested the term montmorillonoid for the group name
to avoid confusion with montmorillonite as 'a, specific mineral name.
Berthier 35 proposed the name nontronite for a material associated with
manganese ore in the Arrondissement of N ontron near the village of Saint
Pardoux in France. A chemical analysis given by Berthier shows the
mineral to be a hydrous ferric iron silicate. Collins 36 in ]877 appears to
30 Kerr, P. F., Montmorillonite or Smectite as Constituents of Fuller's Earth and
Bentonite, Am. 1l1ineral., 17, 192-198 (1932).

31 Svanberg, A., Saponit, Akad. Handi. Stockholm, p. 153 (1840).
32 Svanberg, A., Saponit, Ann. Physik Chem. (Poggendor.tJ), 67, 165-170 (1842).
33 Ross, C. S., and P. F. Kerr, The Clay Minerals and Their Identity, J. Sediment.
Petrol., 1, 55-65 (1931).

341V[acEwan, D. M. C., The Montmorillonite Minerals, "X-ray Identification and
Structure of the Clay Minerals," Chap. IV, pp. 86-187, Mineralogical Society of
Great Britain Monograph (1951).
36 Berthicr, P., Nontronite nouveau mineral, Ann. chim. et phys., 36, 22-27 (1827).
36 Collins, J. R., Remarks on Gramenite from Smallcombe, and the Chloropal
Group of Minerals, 1l1ineralog. Mag., 1, 67-82 (1877).

have been the first to recognize the asssociation between nontronite and
montmorillonite. The similarity of nontronite and montmorillonite was
established by Larsen and Steiger,37 Ross and Kerr,33 and Gruner.27
Nontronite is now the name generally applied to the iron-rich end mem-
ber of the montmorillonite group.
The name vermiculite from the Latin "vermiculair, to breed worms,"
was first used by Webb 38 in 1824 for some material from Milbury, Massa-
chusetts. Vermiculites have been classed with the micas and have
frequently been listed as alteration products, chiefly of biotite and
phlogopite. They have also been considered as closely related to the
micas and were stated to show a considerable range in chemical composi-
tion. Until the work on the structure of vermiculites by Gruner 39 and
later by Hendricks and Jefferson,40 it was not known whether or not
vermiculite was a distinct mineral. Only very recently41 has the occur-
rence of vermiculite in small particles as a clay-mineral constituent of
clay materials been recognized.
Vermiculites have an expanding lattice, differing from montmorillonite
in that the expansion can take place only to a limited degree. The
possible range of composition of vermiculites is not known, but the
natural materials always seem to contain considerable magnesium
and iron (some of it ferrous), and Mg++ seems to be the characteristic
exchangeable cation. Like the montmorillonites, the vermiculites have
high cation-exchange capacity. The composition of vermiculites can be
the same as that of some montmorillonites, in which case the only differ-
ence would be in the larger particle size of the vermiculites.
The term illite was proposed by Grim, Bray, and Bradley42 in 1937 as
a general term, not as a specific clay-mineral name, for the mica-like clay
minerals. The name was derived from the abbreviation for the state of
Illinois. Prior to 1937 the widespread occurrence of a mica-like clay min-
eral had been recognized by many investigators, and many names had been
suggested, e.g., potash-bearing clay mineral,33 sericite-like mineral, 43
37 Larsen, E. S., and G. Steiger, Dehydration and Optical Studies of Alunogen,
Nontronite and Griffithite, Am. J. Sci., ser. 5, 16, 1-19, (1928).

38 Webb, T. H., Miscellaneous Localities of Minerals, Am. J. Sci., 7,55-61 (1824).
39 Gruner, J. W., Vermiculite and Hydrobiotite Structures, Am. Mineral., 19,557-
575 (1934).
40 Hendricks, S. B., and M. E. Jefferson, Crystal Structure of Vermiculites and
Mixed Vermiculite-chlorites, Am. Mineral., 23,851-862 (1938).

41 MacEwan, D. M. C., Chlorites and Vermiculites in Soil Clays, Verre silicates ind.,
13,41-46 (1948).
42 Grim, R. E., R. H. Bray, and W. F. Bradley, The Mica in Argillaceous Sedi-
ments, Am. Mineral., 22, 813-829 (1937).

43 Grim, R. E., Petrology of Pennsylvanian Shales and Noncalcareous Underclays
Associated with Illinois Coals, Bull. Am. Ceram. Soc., 14, 113-119, 129-134, 170-176
(1935).
36 Clay Mineralogy
glimmerton. 44 Grim et al. 42 pointed out objections to these earlier
names, and the term illite has now been widely accepted for a mica-
type clay mineral with a 10-A c-axis spacing which shows substan-
tially no expanding-lattice characteristics.
Grim, Bray, and Bradley42 gave the general formula for illites as
(OH)4Ky(Si8_yAly)(AI4Fe4Mg4Mg6) 0 20 In muscovite y is equal to 2,
whereas in illite y is less than 2 and frequently equal to 1 to 1.5. Accord-
ing to the formula, illites would include both trioctahedral and diocta-
hedral types, and no attempt was made to differentiate between biotite
and muscovite types of crystallization. At the present time the name
illite is generally used, and will be used herein, for clay-mineral micas of
both dioctahedral and trioctahedral types and of muscovite and biotite
crystallizations. It may well be that future clay-mineral studies will
show it desirable to subdivide the group, or even to use some very
different larger categories.
It has been suggested, chiefly by ROSS25 and his colleagues, that the
name illite should be replaced by bravaisite on the basis that bravaisite is
an earlier name and is a clay-mineral mica. Actually type bravaisite is
a mixture 45 of montmorillonite and a clay-mineral mica and not a specific
mineral and therefore has no standing as a mineral species. Sarospatite
has also been suggested by Hofmann et al. 46 as a substitute for illite, but
the same objection 47 can be raised, namely, that the type material from
Sarospatak, Hungary, is a mixture of clay minerals. It is believed
that experience has shown the desirability of using a new name for
this clay-mineral group rather than attempting a redefinition of an
old name, particularly if the old name originally described a mixture of
minerals.
Werner 48 appears to have first used the name chlorite about 1800. It
has been used for a group of green hydrous silicates in which ferrous iron
is prominent and which are closely related to the micas. A large variety
of materials have been described as chlorites, and there has been much
confusion regarding the identity and validity of species belonging to the
group.
44 Endell, K., U. Hofmann, and E. Maegdefrau, Ueber die Natur des Tonanteils in

45 Grim, R. E., and R. A. Rowland, Differential Thermal Analyses of Clay Min-
erals and Other Hydrous Minerals, Am. Mineral., 27,746-761 (1942).

46 Hofmann, U., J. Endell, and E. Maegdefrau, Specific Identity of the Micaceous
Clay Mineral Sarospatite, Ber. deut. keram. Ges., 24, 339-344 (1943).
47 Grim, R. E., V{. F. Bradley, and G. Brown, The Mica Clay Minerals, "X-ray
Identification and Structure of the Clay Minerals," Chap. V, pp. 138-172, Mineralog-
ical Soeiety of Great Britain Monograph C -n
4. Werner, A. G. See J. D. Dana, "System of J\;iineralogy," 6th ed., p. 643 (1914).
The work of Pauling and McMurchy 50 showed the general structural
19
attributes of the chlorites, and recent works by Barshad 51 and Brindley52

have provided detailed information on the relation of the chlorites to
other mieas and on the specific variations in structure between members
of the group. Structurally the chlorites are regular interstratifications of
single biotite mica layers and brucite layers. The chlorites providei an
early example of interlayering of units into more complex structures, a
phenomenon which is now known to be fairly common in clay materials.
Only recently41 have chlorites been recognized as important con-
stituents of many clay materials, and it is likely that this clay mineral is
much more abundant in clay materials than is now recognized. It is
frequently very difficult to detect small amounts of chlorite when it is
mixed with other clay minerals. The identification of chlorite is par-
ticularly difficult when kaolinite is also a constituent of the clay material
being studied.
Sepiolite and meerschaum have long been considered as synonomous
by mineralogists. The name meerschaum appears to have been first
applied by Werner 53 in 1789 and is German for "sea froth," alluding to
the lightness and color of the material. The term sepiolite was first
applied in 1847 by Glocker 54 and is derived from the Greek for" cuttle-
fish," the bone of which is light and porous. The material is reported by
Dana 14 to vary from compact with a smooth feel to earthy or fibrous.
He gives the formula 2H 2 0 2MgO 3Si0 2 for the mineral. Later Schaller55
stated the formula to be 4H 20 2MgO 3Si0 2
The name palygorskite was first applied by Fersman 56 to a family of
fibrous hydrous siliceous minerals forming an isomorphous series between
two end members: an aluminum end member called paramontmorillonite
because of its resemblance to montmorillonite in all ways except the
49 Pauling, L., Structure of Chlorites, Proc. Natl. Acad. Sci. U.S., 16, 578-582 (1930).
60 McMurchy, R. C., Structure of Chlorites, Z. Krist., 88, 420-432 (1934).
61 Barshad, I., Vermiculite and Its Relation to Biotite as Revealed by Base-
Exchange Reactions, X-ray Analyses, Differential Thermal Curves, and Water Con-
tent, Am. Mineral., 33,655-678 (1948).
52 Brindley, G. \V., and K. Robinson, The Chlorite Minerals, "X-ray Identification
and Structure of the Clay Minerals," Chap. VI, pp. 173-198, Mineralogical Society
of Great Britain Monograph (1951).
53 Werner, A. G., "Letztes mineral System." Notes by Hofmann, Bergm. J., 1,
377 (1789).
64 Glocker, E. F., "Generum et Specierum Mineralium secundum Ordines Naturales
digestorum Synopsis" Halle (1817).

66 Schaller, W. T., The Chemical Composition of Sepiolite (Meerschaum), Am.
Mineral., 21, 202-210 (1936).

"Fersman, A., Studies in Magnesian ~;]; . '!lS, Mem. Acad. Sci. (St. Petersburg),
ser. 8, 32, 323-365 (1913).
38 Clay Mineralogy
fibrous character, and a magnesium end member called sepiolite. Long-
chambon,57 Migeon,58 and De Lapparent 59 have studied these materials
in considerable detail, but much additional work is required before the
minerals can be well understood. The term palygorskite has also been
used in a more specific sense for specimens thought to be like sepiolite
except for some replacement of magnesium by aluminum. Long-
chambon 57 first suggested that the palygorskite-sepiolite minerals have
an amphibole-like structure composed of double chains of silica tetra-
hedrons, whereas De Lapparent 59 believed them to have micaceous
similarities.
The name attapulgite was first applied by De Lapparent 60 in 1935 to a
clay mineral he encountered in fuller's earth from Attapulgus, Georgia,
Quincy, Florida, and Mormoiron, France. Bradley 61 has determined the
structure of the mineral, showing silica chains similar to those in amphi-
bole to be essential components of its structure. It would seem that
attapulgite is to be classed with sepiolite and palygorskite and that the
existence of clay minerals of this type is now definitely established.
Bradley 61 gives the ideal formula of attapulgite as (OH2)4(OH)2Mg5Sis-
02o4H 20, in which there is considerable replacement of magnesium by
aluminum.
As in the case of chlorite, the sepiolite-attapulgite-palygorskite minerals
may well be much more abundant and widespread than is now considered
to be the case. These minerals are easily missed in clay-mineral analysis.
They are very soluble in acids, and they frequently occur in calcareous
material. They would be destroyed if solution of the carbonates by
acids preceded attempts at clay-mineral identification.
QUESTIONABLE AND DISCREDITED CLAY MINERALS
The literature contains a vast number of mineral names applied to

materials which were mainly hydrous silicates of aluminum and iron and
had clay-like properties. These materials were originally believed to
be specific minerals. Many of these names have now been discredited,
because the type material was shown to be a mixture of specific minerals
57 Longchambon, H., Sur certaines caracMristiques de la sepiolite d' Ampandan-
drava et la formule des sepiolites, Bull. soc. fran,. mineral., 60, 232-276 (1937).
58 Migeon, G., Contribution Ii l'etude de la definition des sepiolites, Bull. soc. franr;.
mineral., 59, 6-133 (1936).

59 De Lapparent, J., The Relation of the Sepiolite-Attapulgite Series of Phyllitic
Silicates of the Mica Type, Campt. rend., 203, 482-484 (1936).

60 De Lapparent, J., Formula and Structure of Attapulgite, Campt. rend., 202, 1728-
1731 (1936).
6' Bradley, W. F., The Structural Scheme of Attapulgite,Am. Mineral., 25, 405-410
(1940).
or identical with another species whose name had priority. Most of the
discredited names were suggested prior to the development of X-ray-
diffraction and differential thermal techniques when there were no ade-
quate analytical methods for differentiating the very fine grained min-
erals. These early names were suggested on the basis of optical and
chemical data, and recent work has shown that these data are likely to be
inadequate to differentiate and characterize the clay minerals (see
Chap. 11).
Recently Kerr and Hamilton 62 published a "Glossary of Clay Mineral
Names" in which are listed most of the names that have been given to
minerals which can be classed as clay minerals, together with their
original description. Kerr and Hamilton also record the work which has
established or discredited these names. No attempt will be made here
to discuss all the suggested clay-mineral species found in the literature.
Consideration will be restricted to those which have received some atten-
tion and whose discredited or questionable status has not yet been
implanted in the thinking of mineralogists, geologists, agronomists, and
others.
Leverrierite was first described by Termier 63 in 1890 from material in
black carbonaceous shales near St.-Etienne, France. It was named after
the mining engineer LeVerrier. Dana l4 in 1892 recognized the identity
of leverrierite with kaolinite, as did Cayeux 64 somewhat later (1916).
Ross and Kerr 2 studied the original type material in 1931 and concluded
it was kaolinite. De Lapparent 65 in 1934 restudied more of the type
material and found a mixture of alternating plates of kaolinite and
muscovite. There seems no doubt, therefore, that the name should be
discarded. However, the name still persists in some very recently
published textbooks on mineralogy-in one instance listed as a member
of the montmorillonite group.
Beidellite was first used by Larsen and Wherry 66 in 1925 for a clay
mineral occurring in a gouge clay in a mine at Beidell, Colorado. These
authors 67 first described this same material as leverrierite in 1917 but in
their later work suggested that it was a distinct species. The original
62 Kerr, P. F., and P. K. Hamilton, "Glossary of Clay Mineral Names," Rept. No.1,
American Petroleum Institute Project 49, Columbia University, New York (1948).
63 Termier, P., Etude sur la leverrierite, Ann. mines, 17,372-398 (1890).
64 Cayeux, L., Introduction a la etude petrographique des roches sedimentaires,
Mem. service carte geol. France, pp. 230-232 (1916).

65 De Lapparent, J., Constitution et origine de la leverrierite, Compt. rend., 198,
669-671 (1934).
"Larsen, E. S., and E. T. Wherry, Beidellite-A New Mineral Name, J. Wash.
Acad. Sci., 15, 465-466 (1925).
67 Larsen, E. S., and E. T. Wherry, Leverrierite from Colorado, J. Wash. Acad. Sci.,
7, 208-217 (1917).
40 Clay Mineralogy
material was thought to be a member of the montmorillonite group which
differed from the type montmorillonite in having a lower silica-to-
alumina molecular ratio (about 3 instead of 4) and somewhat higher
indices of refraction.
Larsen and Wherry did their work before the development of X-ray
and differential thermal techniques, and their identification was based on
optical, chemical, and dehydration data. Grim and Rowland,45 on the
basis of X-ray-diffraction and thermal data, concluded that the type
material was a mixture of clay minerals and showed that the other
samples listed as beidellite in the U.S. National Museum collections were
also mixtures. For example, a sample from N amiquipa, Mexico, listed
as beidellite, which had appare'1tly the appropriate optical properties,
proved to be composed of a mixture of halloysite, limonite, and a small
amount of illite.
In early clay-mineral studies in the 1930's a large amount of material
was identified as beidellite on the basis of analytical data now known to
be inadequate. Many of these identifications were based on X-ray-
diffraction patterns which did not include low-angle reflections, which are
now known to be most critical, and/or on optical measurements, which
are now also known to be inadequate. Also such identifications were
made before the development of the glycol technique for the detection
of small amounts of montmorillonite in mixtures with other clay minerals.
Recent studies show that a very large amount of this early so-called
beidellite is actually a mixture, in many cases an interlayered mixture of
illite and montmorillonite. Many clay-mineral investigators feel, there-
fore, that the term beidellite should be dropped from clay-mineral
terminology, and the author subscribes to their view. Recently, Ross
and Hendricks 25 have published a definition of beidellite as a member
of the montmorillonite group in which there is substantially no magnesium
or iron present and in which replacement of Si H by AP+ accounts for the
cation-exchange capacity. This is substantially in accord with the
original definition of Larsen and Wherry. 66 However, because of the
past usage of the term, it is believed desirable to drop it entirely in order
to avoid confusion.
1l1onothermite was suggested by Belyankin 68 in 1938 for a material from
Chassov-Yar, U.S.S.R., that showed a single high-temperature thermal
reaction at about 550C and had a composition 2RO Ab03 3Si0 2 1.5H 20
0.5Ag. The name has not been used outside of the U.S.S.R. and only
recently has come to be used in that country by others besides Belyankin.
Sedletsky69 has recently stated that it occurs in the weathered products
68 Belyankin, D. S., On the Characteristics of the Mineral Monothermite, Compt.

rend. acado sci. URSS, 18, 673-676 (1938).
69 Sedletsky, 1. D., Mineralogy of White Clays of the Rostov Region, Doklady
Akad. Nauk. SSSR, 69, 69-72 (1949).
of some slates from the Don Basin. It is impossible to determine posi-
tively from the published analytical data whether or not the mineral is a
valid species and, if so, its proper classification. However, the data
strongly suggest that the mineral is a mixture in which illite and kaolinite
are important components.
Gedroizite was first suggested as a mineral name by Sedletsky 70 in 1939
and later 71 (1941) described in more detail for a material believed to be a
specific mineral and characteristic of many alkali soils. The name is
after Gedroiz, the famous Russian soil scientist. Sedletsky gives the
formula 6(KNa)20 5Al 20 3 14Si0 2 12H 20 for the mineral and states
that it belongs to the vermiculite group. He considers it to be a vermicu-
lite in which the magnesium is replaced by potassium and sodium. As
in the case of monothermite, material called gedroizite has not been
studied outside of the U.S.S.R., and the validity or possible classifica-
tion of the mineral cannot be determined.
Pkolerite was originally used by Gui1lemin~2 in 1825 for a mineral from
Rive-de-Gier in France. The original descriptions are somewhat vague,
but the same material was described later in more detail by Des Cloizeaux73
in 1862, and his descriptions apply to kaolinite. Ross and Kerr 2 in 1931
showed definitely that the mineral is identical with kaolinite, or one of the
other kaolin minerals, and point out that the name should be abandoned.
M orencite was first described from Morenci, Arizona, by Lindgren and
Hillebrand 74 in 1904 as a brownish-yellow hydrous silicate of ferric
iron with magnesium, calcium, and aluminum. Gruner 27 has shown the
mineral to be structurally the same as nontronite.
Celadonite was proposed in 1847 by Glocker n for a soft gray-green
mineral that was a hydrous silicate of iron, magnesium, and potassium.
The name, meaning "sea-green" in French, refers to the color of the
mineral. Similar material appears to have been described earlier as
terra verti by DeLish76 in 1783 and as grunerde by Hofmann 77 in 1788.
Hendricks and Ross7s in 1941 showed that celadonite and glauconite have
70 Sedletsky, 1. D., Gedroizite in the Alkali-Soils, Compt. rend. acado sci. URSS,
23,565-568 (1939).
71 Sedletsky, 1. D., Characteristics of the Mineral Gedroizite, Compt. rend. acado
sci. URSS, 27, 308-411 (1941).
72 Guillemin, A., Sur la pholerite, Ann. mines, 11, 489 (1825).
73 Des Cloizeaux, A., "Manuel de mineralogie," vol. 1, p. 190 Paris (1862).
74 Lindgren, W., and W. F. Hillebrand, Minerals from the Clifton, Morenci Dis-
trict, Arizona, Am. J. Sci., ser. 4, 18, 448-460 (1904).
n Glocker, E. F., "Generum et Specierum Mineralium Secundum ordines Naturales
Digestorium Synopsis," p. 193, Halle (1847).
76 DeLish, R., "Cristallographie ou description des formes propres a to us les corps
du regne mineral," vol. 2, p. 502, Paris, (1783).

77 Hofmann, C., Berfmannisches Journal, p. 519 (1788).
a Hendricks, S. B., and C. S. Ross, Chemical C()mposition and Genesis of Glau-

conite and Celadonite, Am. Mineral., 26, 683-708 (1941).
42 Clay Mineralogy
a very similar structure. The original celadonite came from amygda-
loidal fillings, and Kerr and Hamilton 62 have suggested that it may
be desirable to retain the name celadonite because of its different origin
from that of glauconite.
Faratsihite was first described by Lacroix 79 in 1914 from Faratsiho,
Madagascar. The original material was pale yellow in color and was
described as a hydrous silicate of aluminum and iron. On the basis of
X-ray-diffraction data Gruner 27 showed that some specimens of the
mineral were identical with nontronite, and Hendricks so has shown that
other specimens are mixtures of kaolinite and nontronite.
Sedletsky and Yusupova 81 in 1940 described a mineral from Jurassic
beds of Tashkent, U.S.S.R., as ablykite. The material was said to
resemble halloysite in its dehydration characteristics but to differ from
it in its X-ray-diffraction properties. Analytical data are insufficient to
determine the validity or possible classification of the material.
Chloropal, meaning" green opal," was first described by Bernhardi and
Brandes 82 in 1822 for material from Unghwar, Hungary. Gruner 27 and
Ross and Hendricks 25 have shown that chloropal and the montmoril-
lonites rich in iron are the same. The name chloropal has priority over
nontronite, but nontronite has come to be the accepted name for the
mineral. Further, the name chloropal is unsatisfactory, since it errone-
ously implies that the mineral is an opal.
79 Lacroix, A. A., Sur l'opale et sur une nouvelle espice minerale (faratsihite) de
Faratsiho, Madagascar, Bull. soc. fran,. mineral., 37, 231-236 (1914).

80 Hendricks, S. B., Random Structures of Layer Minerals as Illustrated by Cron-
stedite (2FeOFe z03Si0 2 2H 2 0). Possible Iron Content of Kaolin, Am. l"lineral.,
24, 329-539 (1939).
81 Sedletsky, r. D., and S. Yusupova, Argillaceous Minerals Closely Approaching
Halloysite, Compt. rend. acado sci. URSS, 26, 944-947 (1940).

82 Bernhardi, J. J., and R. Brandes, Mineralogische-chemische Untersuchungen
zwei neuer ungarisches Mineralien des Muschligen und des erdigen Chloropals, J.
Chern. Physik, 6, 29 (1882).
, . \~
. 11'--
.f,"
CHAPTER 4
Structure of the Clay Minerals
ill GENERAL STATEMENT

II
The atomic structures of the common clay minerals have been deter-
mined in considerable detail by numerous investigators based on the
generalizations of Pauling l for the structure of the micas and related layer
minerals.
Two structural units are involved in the atomic lattices of most of the
clay minerals. One unit consists of two sheets of closely packed oxygens
or hydroxyls in which aluminum, iron, or magnesium atoms are embedded
in octahedral coordination, so that they are equidistant from six oxygens
(0) (b)
o and ( ) =Hydroxyfs Aluminums, magnesiums, etc.
FIG. 1. Diagrammatic sketch showing (a) single octahedral unit and (b) the sheet
structure of the octahedral units.
or hydroxyls (Fig. 1). When aluminum is present, only two-thirds of

the possible positions are filled to balance the structure, which is the
gibbsite structure and has the formula Al 2 (OH)6. When magnesium is
prcsent, all the positions are filled to balance the structure, which is the
brucite structure and has the formula Mg 3 (OH)6. The normal O-to-O
distance is 2.60 A, and a common OH-to-OH distance is about 3 A, but
in this structural unit the OH-to-OH distance is 294 A, and the space
available for the ion in octahedral coordination is about 0.61 A. The
thickness of the unit is 5.05 A in clay-mineral structures.
1 Pauling, L., The Structure of Micas and Related Minerals, Proc. Natl. Acad. Sci.
U.S., 16, 123-129 (1930).
" LIBRARY
'C'.:'l~~ge of AgrkultUSr8
..!r::;lm~ P1'~esh
;'.':"ic': :ural Univc:'sity
., .. -:,""1:
~-,
44 Clay Mineralogy
The second unit is built of silica tetrahedrons. In each tetrahedron a
silicon atom is equidistant from four oxygens, or hydroxyls if needed to
balance the structure, arranged in the form of a tetrahedron with a
silicon atom at the center. The silica tetrahedral groups are arranged to
form a hexagonal network, ,vhich is repeated indefinitely to form a sheet
of composition Si 40 6 (OH)4 (Fig. 2). The tetrahedrons are arranged so
that the tips of all of them point in the same direction, and the bases of
(0) (b)
o and ( ) = Oxygens o and = Si/icons
FIG. 2. Diagrammatic sketch showing (a) single silica tetrahedron and (b) sheet
structure of silica tetrahedrons arranged in a hexagonal network.
all tetrahedrons are in the same plane. The structure can be considered
as made of a perforated plane of oxygens which is the plane of the base
of the tetrahedral groups; a plane of silicon atoms with each silicon in the
cavity at the junction of three oxygen atoms and therefore forming a
hexagonal network; and a plane of hydroxyl atoms with each hydroxyl
directly above the silicon at the tip of the tetrahedrons. The open hexag-
onal network can be considered as composed of three strings of oxygen
atoms intersecting at angles of 120. The 0-0 distance in the silica
tetrahedral sheet is 2.55 A, and the space available for the ion in tetra-
hedral coordination is about 0.55 A. The thickness of the unit is 4.93 A
in clay-mineral structures. Each of these units presents a center-to-
center height of about 2.1 A.
Some of the clay minerals are fibrous and are composed of different
structural units from those noted above. These minerals resemble the
amphiboles in their structural characteristics, and the basic structural
unit is composed of silica tetrahedrons arranged in a double chain of
composition Si 40 u , as shown in Fig. 3. The structure is similar to that
of the sheet of silica tetrahedrons in the layer minerals except that it is
continuous in only one direction. In the other direction it is restricted
to a width of about 11.5 A.
The chains are bound together by atoms of aluminum and/or mag-
nesium pJacpd so that each such atom is surrounded by six "active"
Structure of the Clay Minerals 45
oxygen atoms. The active oxygens are those with only one link to
silicon and hence are those at the edges of the chains and at the tips of the
tetrahedrons.
(b)
FIG. 3. Diagrammatic sketch of double chains of silica tetrahedrons, as in the
amphibole structural type of clay minerals; (a) in perspective, (b) projected on the
plane of the base of the tetrahedrons.
ALLOPHANE MINERALS
By definition the allophane clay minerals are ..amorphous 2 to X-ray

diffraction. It seems probable that allophanes are random arrangements
of silicon in tetrahedral coordination and metallic ions in octahedral
coordination without any symmetry. Such a random arrangement of
silica tetrahedrons, alumina octahedrons, and occasional other units
such as phosphate tetrahedrons appears more likely than a completely
unorganized assembly.
Very little careful study has been made of allophanes, and their vari-
ation in composition is unknown. Their study is particularly difficult
2 Ross, C. S., and P. F. Kerr, Halloysite and Allophane, u.s. Geol. Sllrvey,Pro!ess.
Paper 185G, pp. 135-148 (1934).
46 Clay Mineralogy
because of the problem of separating them from intermixed crystalline
material.
KAOLINITE MINERALS
The structure of kaolinite was first suggested in general outlines by

Pauling. 3 It was worked out in some detail by Gruner 4 and later
revised by Brindley and his colleagues. 5,6 The structure is composed of a
o Oxyqt'ns
@ Hydroxyls
Aluminums
. 0 Sit/cons
FIG. 4. Diag;rammatic sketch of the structure of the kaolinite layer, after Gruner.'
single silica tetrahedral sheet and a single alumina octahedral sheet

combined in a unit so that the tips of the silica tetrahedrons and one of
the layers of the octahedral sheet form a common layer (Fig. 4). All the
tips of the silica tetrahedrons point in the same direction and toward the
center of the unit made of the silica and octahedral sheets. The dimen-
sions of the sheets of tetrahedral units and of the octahedral units are
closely similar in their a and b dimensions, and consequently composite
octahedral-tetrahedral layers are readily formed. Only minor changes
in interatomic distances are necessary, and there would seem to be little
3 Pauling, L., The Structure of the Chlorites, Proc. Natl. Acad. Sci. U.S., 16,5i8-
582 (1930).
4 Gruner, J. W., The Crystal Structure of Kaolinite, Z. Kri8i., 83, 75-88 (1932).
5 BrimUey, G. W., and K. Robinson, The Structure of Kaolinite, JI.[ineralog. Jl[ag.,
27,242-253 (1946).
6 Brindley, C. \V., The Kaolin Minerals, "X-Ray Identification and Structure of
the Clay Minerals," Chap. II, pp. :32-75, Mineralogical Society of Great Britain Mono-
graph (1951).
tendency for the unit to bend in the stacking arrangement of kaolinite,
although large crystals would be unlikely.
In the layer common to the octahedral and tetrahedral groups, two-
thirds of the atoms are shared by the silicon and aluminum, and then
they become 0 instead of OR. Only two-thirds of the possible positions
for aluminum in the octahedral sheet are filled, and there are three
possible plans of regular population of the octahedral layer with alum-
inums. The aluminum atoms are considered to be so placed that two
aluminums are separated by an OR above and below, thus making a
hexagonal distribution in a single plane in the center of the octahedral
sheet. The OR groups are placed so that each OR is directly below the
perforation of the hexagonal net of oxygens in the tetrahedral sheet.
The charge distribution in the layers is as follows:
60-- 12-
4Si H 16+
40-- + 2(OR)- 10 - (Layer common to tetrahedral and octahedral
sheets)
4AlH 12+
6(OR)- 6-
The charges within the structural unit are balanced. The structural
formula is (OR)8Si 4AI 4 0 1o , and the theoretical composition expressed
in oxides is Si0 2 46.54 per cent; Al 2 0 2 39.50 per cent; R 2 0 13.96 per cent.
The analyses 7 of many samples of kaolinite minerals have shown that
there is very little substitution within the lattice. In a few instances,
the evidence suggests a very small amount of substitution of iron and/or
titanium for aluminum in the relatively poorly crystalline variety (see
page 49).
The minerals of the kaolinite group consist of sheet units of the type
just described continuous in the a and b directions and stacked one above
the other in the c direction. The variation between members of this
group consists in the way in which the unit layers are stacked above each
other and possibly in the position of the aluminum atoms in the possible
positions open to them in the octahedral layer.
In the case of kaolinite itself Brindley 6 has investigated the stacking
of the unit layers, and the following statements are taken largely from
his work. The mineral is triclinic; a = 5.16 A; b = 8.94 A; c = 7.38 A;
a = 91.8; {3 = 104.5; 'Y = 90; and the space group is C!-Cc. Succes-
sive unit layers are so arranged that oxygen atoms and OR groups of
adjacent layers approach one another in pairs. This disposition can be
obtained in a variety of ways, by placing one layer directly above another
7 !loss, C. S., and P. F. Kerr, The Kaolin Minerals, U.S. Geol. Survey Profess. Paper
165E, pp. 151-176 (H)31).

.;,. ,
48 Clay Mineralogy
so that the c axis is perpendicular to the ab plane or by the displacement
of one layer relative to its neighbor along either a or b so that the displace-
ment is nao/6 or mb o/6, where nand m are integral and ao and bo are
unit-cell dimensions. The displace-
ment for kaolinite in accordance with
a and f3 angles is m = 0 and n = 2.
The relative position of basal 0
atoms and uppermost OH groups
of two adjacent kaolinite layers is
shown in Fig. 5. The stacking of
the units along the a and b axes is
shown in Fig. 6.
Because of the superposition of 0
and OH planes in adjacent units, the
units are held together fairly tightly
by hydrogen bonding between the
layers. The plane between the unit
layer is a cleavage plane. However,
in kaolinite the cleavage is not so pro-
nounced as it is in other clay minerals
where 0 planes are adjacent at unit
boundaries so that there is no hydro-
o Oxygen GHydroxyls
gen bonding.
Numerous investigators 8- 1o of clays
Silicon have reported the finding of a kaolinite
FIG. 5. Projection of 0 atoms and mineral of lower crystallinity than that
OH groups of adjacent kaolinite of the well-crystallized material just
layers on (001) to show the stacking in
noted. Brindley and his colleagues 6 ,8
kaolinite, after Brindley and Robin-
son.S have investigated in detail some exam-
ples of rather poorly crystallized kao-
linite and have shown that their examples from English fire clays contain
fewer reflections than normal kaolinite. According to them the reflections
can be indexed as monoclinic or psuedo-monoclinic, with a unit cell of the
same dimensions as kaolinite but with a = 90 0 Also, they state that the
structure is highly disordered along the b axis with the unit layers randomly
displaced by multiples of bo/3. The arrangement along the a axis is like
that of kaolinite. They suggest that there is some randomness in the
8 Brindley, G. W., and K. Robinson, Randomness in the Structures of Kaolinitic

Grimshaw, R. W., E. Heaton, and A. L. Roberts, Constitution of Refractory
Clays, Trans. Brit. Ceram. Soc. 44, 69-92 (1945).
,0 Grim, R. E., Differential Thermal Curves of Prepared Mixtures of Clay Minerals,
Am. Mineral., 32, 493-501 (1947).
distribution of aluminum atoms among octahedral positions. The first-
order spacings are slightly higher (7.15 to 7.20 A) for the poorly crystalline
than for the well-crystallized mineral, suggesting some occasional inter-
layer water between the kaolin units. Dehydration data tend to confirm
the presence of such water (see page 212). X-ray-diffraction character-
Kaolinite
FIG. 6. The stacking of unit layers of kaolinite along the a and b axes, after Brindley. 6
istics of the poorly crystallized kaolinite as contrasted to well-crystallized

kaolinite are given in Fig. 7. Brindleyll suggested mellorite as a name
for such poorly crystallized kaolinite, and Roberts 12 has suggested levi site
for similar material. As there are probably all gradations from well-
crystallized kaolinite to that of complete randomness in the b direction
and in the population of aluminum positions, it appears doubtful that a
specific name should be applied.
Available data suggest that there may be some slight substitution of
titanium or iron for aluminum in the poorly crystallized kaolinite. Such
substitutions appear to be restricted to the poorly crystallized kaolinite.
It is not necessarily established that any of the titanium is in the struc-
ture, as anatase is an almost universal minor component of kaolins.
11 Brindley, G. W., Structural Relationships in the Kaolin Group of Minerals, pre-

sented before International Geological Congress, London, 1948.
12 Roberts, A. L., and R. W. Grimshaw, The Principal Clay Mineral in Certain
Refractory and Bond Clays, presented before International Geological Congress, Lon-
don, 1948.
50 Clay Mineralogy
13 14
Dickite and nacrite have structures somewhat similar to that of
kaolinite and are usually listed as clay minerals, although they are rarely
found in clay materials. They are made up of unit layers of an alumina
octahedral sheet and a silica tetrahedral sheet like that in kaolinite and
differ only in the stacking of the layers. In dickite, 13,15,16 the unit cell is
I~
Kaolinite (0)
Kaolinite (b)
Firecloy (c)
Hoiloysite (eI)
I I I I I I I
7 5 4 3 2.5 2.2 2 1.5 d,in A
FIG. 7. Diagrammatic representation of X-ray photographs of (a and b) well-
crystallized kaolinite, (c) poorly crystallized kaolinite, and (d) halloysite 2H 2 0, after
Brindley and Robinson. s
made up of two unit layers. The exact stacking of successive layers is

not well established. The parameters of the mineral are best accounted
for by shifts along the a axis of n = 1 in nao/6 and along the b axis of
m = 3 in mb o/6, with the shift being positive and negative in successive
layers. A shift of m = 1 is also a possibility but seems less likely.
However, shifts of m = 1 or 3 are not completely in accord with the
diffraction data. According to Gruner,13 the mineral is monoclinic
a = 5.15 A, b = 8.96 A, c = 14.45 A, (3 = 9650', and the space group is
C!-Cc.
In nacrite, the unit cell is made up of six unit layers, each having the
13 Gruner, J. W., Crystal Structure of Dickite, Z. Krist., 83, 394-404 (1932).

14 Gruner, J. W., The Crystal Structure of Nacrite and a Comparison of Certain
Optical Properties of the Kaolin Group with Its Structure, Z. Krist., 85, 345-354
(1933).
15 Ksanda, C. J., and T. F. W. Barth, Note on the Structure of Dickite and Othe~
Clay Minerals, Am. Mineral., 20, 631-637 (1935).
16 Hendricks, S. B., On the Structure of the Clay Minerals: Dickite, Halloysite,
and Hydrated Halloysite, Am. Mineral., 23, 295-301 (1938).
~
arrangement described for kaolinite. According to Hendricks the unit '7
dimensions are a = 5.15 A, b = 8.96 A, c = 43 A, and (3 = 9020', so

that the structure approaches rhombohedral symmetry. In nacrite, suc-
cessive layers are stacked one above the other so that the c axis is about
perpendicular to the ab plane; i.e., nand m are equal to zero. The space
group is C;-Cc, according to Hendricks. 17
Occasionally kaolinite types of clay minerals are found which appear to
be monomineral species on the basis of X-ray diffraction but have a con-
siderably higher silica-to-alumina molecular ratio, which often approaches
l
.5J
(I).
CIl
f.IJ
CJ
()
a:
f
Fw.~. Diagrammatic sketch of the double silica layer in the anauxite structure,
after Hendricks.'s
3. Such silica-rich kaolinites are called anauxite. The structure of

anauxite has been the subject of considerable controversy, but the recent
suggestion of Hendricks '8 seems most plausible. According to Hendricks,
anauxite consists of kaolinite unit layers between which units composed
of double silica tetrahedral sheets are interlayered randomly. In the
double silica units, the tetrahedrons of each sheet point toward the
center, and a common 0 forms the tip of the tetrahedrons in both sheets
(Fig. 8). The charge distribution is as follows:
60-- 12-
4Si4+ 16+
40-- 8 - (Sheet common to both tetrahedral sheets)
4Si H 16+
60-- 12-
17 Hendricks, S. E., Crystal Structure of N acrite and the Polymorphism of the
Kaolin Minerals, Z. Krist., 100, 509-518 (1938).
18 Hendricks, S. B., Lattice Structure of Clay Mineral::. ~~~o~~ ~operties of
Clays, J. Geol., 60,276-290 (1942).
52 Clay Mineralogy
The dimensions and characteristics of the double silica sheet are about
those of kaolinite, and hence an interlamination with kaolinite sheets is
possible without changing materially the diffraction characteristics.
Also the optical properties of the two units would not be very different.
The silica-to-alumina molecular ratio of anauxites is known to vary, and
this may be accounted for by variations in the relative abundance of the
interlayering of the double silica sheets.
HALLOYSITE MINERALS
As indicated earlier, there are two forms of halloysite (see page 31),
one with the composition (OHhSi 4Al 40 1o and the other with the com-
position (OH) 8Si4A1401o4H20. The latter form dehydrates to the
former irreversibly at relatively low temperatures. Various structures
for the halloysite minerals have been suggested by Mehmel,19 Edelman
and Favejee,20 Stout,21 and Hendricks. 16
Hendricks has shown that the earlier suggested structures are not in
accord with the observed intensities of the basal reflections or with the
very easy dehydration of the mineral. The basal spacing of the dehy-
drated form is about 7.2 A or about the thickness of the kaolinite layer,
and the basal spacing of the hydrated form is about 10.1 A. The differ-
ence, 2.9 A, is about the thickness of a single molecular sheet of water
molecules. Hendricks suggested therefore that the highly hydrated form
consists of kaolinite layers separated from each other by single molecular
layers of water. Later Hendricks and Jefferson 22 suggested that the
water molecule in this layer had a definite configuration (Fig. 9). Dif-
fraction data for halloysite are not suited for detailed structural study,
but the intensities of the basal reflections are in accord with Hendricks's
suggested structure. The transition to the dehydrated form is due to the
loss of the interlayer of water molecules.
In the halloysite minerals, the successive kaolinite layers are displaced
randomly in both the a and b directions. According to Brindley6 the
probable displacements are simple fractions of the cell dimensions, such
as mao/6 and nb o/6. He states that" the experimental results require
19 Mehmel, M., Ueber die Struktur von Halloysit und Metahalloysit, Z. Krist.,
90, 35-43 (1935).
20 Edelman, C. II., and J. C. 1,. Favejee, On the Crystal Structure of Montmoril-
lonite and Halloysite, Z. Krist., 102, 417-431 (1940).
21 Stout, P. P., Alterations in the Crystal Structure of the Clay Minerals as a Result
of Phosphate Fixation, Soil Sci. Soc. Am. Proc., 4, 177-182 (1939).

22 Hendricks, S. B., and M. E. Jefferson, Structures of Kaolin and Talc-Pyrophyllite
Hydrates and Their Bearing on Water Sorption of the Clays, Am. Mineral., 23,863-
875 (1938).
only that the displacements along the a and b axes shall be random with
respect to each other."
Brindley and his colleagues 23- 25 have studied the transition of the
hydrated to the dehydrated form of the mineral in great detail. Accord-
ing to them only partial dehydration takes place at low temperatures
FIG. 9. Diagrammatic sketch of a portion of the halloysite 4H 20 structure showing

a single layer of water molecules with the configuration suggested by Hendricks and
Jefferson. 22
(60 to 750), and temperatures of the order of 4000 are necessary for
complete removal of the interlayer water and the development of the
7.2-A spacing. At temperatures of 60 to 750 or at a lower temperature
for longer time, a partially dehydrated form develops which tends to
23 Brindley, G. W., and K. Robinson, X-ray Studies of Halloysite and Metahalloy-
site, I, The Structure of Metahalloysite, Mineralog. Mag., 28, 393-407 (1948).
24 Brindley, G. W., and J. Goodyear, X-ray Studies of Halloysiteand Metahalloy-
site, II, The Transition of Halloysite to Metahalloysite in Relation to Relative Humid-

ity, Mineralog. Mag., 28,407-422 (1948).
26 Brindley, G. W., K. Robinson, and J. Goodyear, X-ray Studies of Halloysite and
Metahalloysite, III, Effect of Temperature and Pressure on the Transition from

Halloysite to Metahalloysite, Mineralog. Mag., 28, 423-428 (1948).
54 Clay Mineralogy
persist and has, therefore, considerable stabjIity. The partially hydrated
form has a basal spacing of 7.36 to 7.9 A, corresponding to a hydration of
about 0.5 to 1.5H 20. The partially hydrated form consists of a statis-
tical distribution of hydrated layers and nonhydrated layers; the maxi-
mum value would correspond to a little more than 1 water layer for every
4 kaolinite layers. Brindley 6 states that examples with hydrations rang-
[ . .1.
ing between 1.5H 20 and about 3H 20 have not been observed.
4
~~;
4'51~A ~ O:(~!H: 0 ~ ~i'51 A
2.8 ,0
2.99 A
(8.62 A) t --0 8.62 A -
O.ISA
.l --8.93 A--
0
0
~/P$#~$/#&_60 ---------4 H20-------5IA
60Wdd/'//dd/d#M
(b)
~$g~~
":)\ .. o'/.
!><. t>, ...,. 6'(6~.1 <?(Oly.l
{I".':;' \ \
\- a-axis
(c)
FIG. 10. Diagrammatic representation of the structure of kaolinite and halloysite
4H 20 after Bates et al." (a) Arrangement of layers in kaolinite; (b) arrangement of
Layers in halloysite according to Hendricks; (c) proposed arrangement of halloysite
layers.
Bates et al. 26 have shown the tubular nature (see Chap. 6) of the
halloysite minerals from electron micrographs and have suggested that
the 4H 20 form consists of tubes made up of overlapping, curved sheets
of the kaolinite type with the c axis for any point on the tube nearly
perpendicular to a plane tangent to the tube at that point. The axis of
the tube may be parallel to either the a or b axis, or possibly to any inter-
mediate crystallographic direction in the plane of the sheets. In the
dehydration to the 2H 20 form the tubes frequently collapse, split, or
unroll.
26 Bates, T. F., F. A. Hildebrand and A. Swineford, Morphology and Structure of
Endellite and Halloysi+e, Am. M-ineral., 35,463-484 (1950) .

., .._
The difference in effective size of the 0 associated with the silicon and
of the OH associated with the aluminum has been stated previously
(pages 43,44). Bates 26 has shown that the a dimension in the 0 plane is
5.14 A and in the OH plane is 5.06 A, and that the b dimension in the 0
plane is 8.93 A and in the OH plane is 8.62 A. According to him this
difference in the dimensions of the upper and lower plane of the silica-
alumina layer will cause a curvature of the layers with a radius in accord-
ance with the dimensions of the observed tubes (see Fig. 10). The
curvature can develop in hydrated halloysite because of the irregular
stacking of the layers and the interlayer water molecules, which cause a
weak bond between successive layers. It does not develop in kaolinite
because the regular stacking and close spacing of layers causes a relatively
strong bond between successive layers. The electron micrographs of
halloysite satisfactorily document Bates's morphology, but no detailed
reconciliation of diffraction data with the curved structure has yet been
accomplished.
MONTMORILLONITE MINERALS
The montmorillonite minerals occur only in extremely small particles

so that single-crystal X-ray-diffraction data cannot be obtained. Struc-
tural concepts, therefore, must be deduced from powder data and infer-
ences from better known structures, with the result that there is consider-
able uncertainty regarding details of their structure. Currently the
generally accepted structure for the montmorillonite minerals follows the
original suggestion made in 1933 by Hofmann, Endell, and Wilm 27 modi-
fied by later suggestions of Maegdefrau and Hofmann,28 Marshall,29 and
Hendricks. 18 According to this concept, montmorillonite is composed of
units made up of two silica tetrahedral sheets with a central alumina
octahedral sheet. All the tips of the tetrahedrons point in the same
direction and toward the center of the unit. The tetrahedral and octa-
hedral sheets are combined so that the tips of the tetrahedrons. of each
silica sheet and one of the hydroxyl layers of the octahedral sheet form a
common layer. The atoms common to both the tetrahedral and octa-
hedrallayer become 0 instead of OR. The layers are continuous in the
a and b directions and are stacked one above the other in the c direction.
In the stacking of the silica-alumina-silica units, 0 layers of each unit
are adjacent to 0 layers of the neighboring units with the consequence
morillonit, Z. Krist., 86, 340-348 (1933).

28 Maegdcfrau, E., and U. Hofmann, Die Kristallstruktur des Montmorillonits, Z.
Krist., 98, 299-323 (1937).

29 Marshall, C. E., Layer Lattice~ and Base-Exchange Clays, Z. Krist., 91, 433-449
(1935).
\I'AU CS: 13:ilil{ARY
Ace' --,
Date: 6"'>~
56 Clay Mineralogy
Excharlqeable Cations
nH 2 0
o Oxyqens 8 Hydroxyls Aluminum, Iron, maqneslum

o and. Silicon, occasionally aluminum
FIG. 11. Diagrammatic sketch of the structure of montmorillonite according to
Hofmann, Endcll, and 'Vilm, 27 Mar~hall, 29 and Hendricks. ls
that there is a very weak bond and an excellent cleavage between them.
The outstanding feature of the montmorillonite structure is that water
and other polar molecules, such as certain organic molecules (see Chaps.
8 and 10), can enter between the unit layers, causing the lattice to
expand in the c direction. The c-axis dimension of montmorillonite is,
therefore, not fixed but varies from about 9.6 A, when no polar molecules
.-::re between the unit layers, to substantially complete separation of the
individual layers in some cases. Figure 11 shows a diagrammatic sketch
of ' this structure of montmorillonite.
Exchangeable cations occur between the silicate layers, and the c-axis
spacing of completely dehydrated montmorillonite depends somewhat on
the size of the interlayer cation, being larger the larger the cation. In the
case of adsorption of polar organic molecules between the silicate layers,
the c-axis dimension also varies with the size and geometry of the organic
molecule. The thickness of the water layers between the silicate units
depends on the nature of the exchangeable cation at a given water-vapor
pressure (see Chap. 7). Under ordinary conditions a montmorillonite
with N a+ as the exchange ion frequently has one molecular water layer
and a c-axis spacing of about 12.5 A; with Ca++ there are frequently two
molecular water layers and a c-axis spacing of about 15.5 A. The expan-
sion properties are reversible. However, when the structure is com-
pletely collapsed by removal of all of the interlayer polar molecules,
reexpansion may proceed with difficulty.
Experiments by Mering 30 and others 31 with montmorillonite in the
presence of large quantities of water suggest that with certain adsorbed
cations, e.g., Na+, the unit layers completely separate but that with other
cations, e.g., Ca++ and H+, the separation is not complete.
It appears from the work of Bradley, Grim, and Clark,32 that the thick-
ness of the water layers between successive silicate layers is an integral
number of molecules. That is, the water layer is one, two, three, or four
molecular water layers thick. A natural montmorillonite may have a
regular ordering of a single thickness of water layers, or it may be a
random mixture of different "hyc.rates." Roth 33 has shown that impor-
tant physical characteristics of clays composed of montmorillonite are
related to the regularity or randomness of the interlayer water layers.
Mering 30 has demonstrated that the diffraction effects of Ca montmoril-
lonite prepared at low relative humidities (50 % ) can be explained by
a mixing of hydrates with thicknesses of 14 and 15 A. Similar clays pre-
pared at 90 per cent relative humidity show only a 15-A reflection, 15 A
being the thickness of the double-layer hydrate of Bradley et al.
Hofmann, Endell, and Wilm 27 described the montmorillonite structure
as generally similar to that of pyrophyllite except for the expanding J
30 Mering, J., On the Hydration of Montmorillonite, Trans. Faraday Soc., 42B.
205-219 (1946).
31 Bradley, W. F., and R. E. Grim, Colloidal Properties of Layer Silicates, J. Phys.
& Colloid Chem., 52, 1404-1413 (1948).

32 Bradley, W. F., R. E. Grim, and G. L. Clark, A Study of the Behavior of Mont-
morillonite on Wetting, Z. Krist., 97, 216-222 (1937).

33 Roth, R. S., The Structure of Montmorillonite in Relation to the Properties of
Certain Bentonites, Ph.D. thesis, University of Illinois (1951).
58 Clay Mineralogy
lattice characteristic. They postulated a stacking of layers to give an
orthorhombic unit.
Later Maegdefrau and Hofmann 28 showed that as a general thing the
unit layers are stacked without any fixed periodical arrangement in the
a and b directions. Occasional montmorillonites do seem to have some
regular periodicity in the horizontal crystallographic directions.
The theoretical charge distribution, without considering lattice substi-
tutions within the layer, is as follows:
60-- 12-
4Si H 16+
40--2(OH)- 10 - (Layer common to tetrahedral and octahedral
sheets)
4AIH 12+
40--2(OH)- 10 - (Layer common to tetrahedral and octahedral
sheets)
4Si H 16+
60-- 12-
Interlayer layer of H 20 or other polar molecules
The theoretical formula without considering lattice substitutions is
(OH)4Si8AI402o'nHiO(interlayer), and the theoretical composition without
the interlayer material is Si0 2 66.7 per cent, A1 2 0 3 28.3 per cent, H 2 0
5 per cent.
As first emphasized by Marshall 29 and Hendricks, I~ montmorillonite
always differs from the above theoretical formula because of substitution
within the lattice of aluminum and possibly phosphorus for silicon in
tetrahedral coordination and/or magnesium, iron, zinc, nickel, lithium,
etc., for aluminum in the octahedral sheet. In the tetrahedral sheet the
substitution of A13+ for Si H appears to be limited to less than about 15
per cent. In the formula noted above, only two-thirds of the possible
positions in the octahedral sheet are filled. The substitution of Mg++ for
AP+ can be one for one, or three Mg++ for two AP+, with all possible
octahedral positions being filled in the latter case. Substitutions within
the octahedral sheet may vary from few to complete. Total replacement
of 2AIH by 3Mg++ yields the mineral saponite; replacement of aluminum
by iron yields nontronite; by chromium, volkonskoite; by zinc, sauconite.
The Mg++ ion of diameter 0.65 Aand the Fe3+ ion of diameter 0.67 A
are somewhat large to fit into the lattice, and as a consequence montmoril-
lonite minerals with large substitutions of these ions are subject to a
directional strain which results in elongate lath- or needle-shaped units
(see Chap. 6).
Layer minerals of this general type in which all the possible octahedral
positions are filled are called octaphyllites, or trioctahedral, and those in
which only two-thirds of the possible positions are filled are called
heptaphyllites, or dioctahedral. Numerous analyses of montmorillonite
have shown that the substitutions within the octahedral sheet are such
that the mineral is either almost exactly trioctahedral or dioctahedral
and not intermediate. Ross and Hendricks 34 have computed the struc-
tural fit for a large number of chemical analyses of montmorillonite.
Unfortunately there are no accompanying X-ray data, and some of the
samples may be mixtures of clay minerals. However, the number of
analyses is so great that their conclusion seems well established, namely,
that the number of ions in six coordination, i.e., in the octahedral posi-
tions, lie in two distinct ranges, 4.00 to 4.44 and 5.76 to 6.00. Further,
it appears that, if the mineral is dioctahedral, there may be considerable
variation in the exact position of the aluminums or other atoms in the
possible octahedral positions.
A further way in which montmorillonite always differs from the theoret-
ical formula is that the lattice is always unbalanced by the substitutions
noted above, i.e., Mg++ for AP+, AP+ for Si H , etc. The unbalanc-
ing may result from substitution of ions of different valence in the tetra-
hedral or octahedral sheet or both. Unbalancing in one sheet may be
compensated for in part, but only in part, by substitution in the other
sheets of the unit layer. Thus, the substitutions of AP+ for Si H may in
part be compensated by filling slightly more than two-thirds of the octa-
hedral positions. Compensation also may be by substitution of OH for
o in the octahedral layer. It is significant that the substitutions in the
montmorillonite lattice, with the internal compensating substitutions,
always result in about the same net charge on the lattice. Many analyses
have shown this to be about 0.66 - per unit cell. This net-charge
deficiency is balanced by exchangeable cations adsorbed between the
unit layers and around their edges (see Chap. 7, Ion Exchange). This
charge deficiency corresponds to about two-thirds unit per unit cell. It
would require the substitution of one Mg++ for every sixth AI3+, for
example, or about one out of every six Si H by AP+.
Montmorillonite has been synthesized from pure hydrous mixtures of
magnesia and silica, in which case the charge deficiency cannot be due to
lattice substitutions. It must be the result of vacancies in the lattice,
and such vacancies probably also occur in natural minerals.
Ross and Hendricks 34 have computed the structural formula from the
chemical composition of many montmorillonites, and these have shown
something of the range of substitutions within the lattice and the partial
compensating substitutions within the structure. The nomenclature cf
the members of the montmorillonite group depends on the substitution
34 Ross, C. S., and S. 13. Hendricks, Minerals of the Montmorillonite Group, U.S.
Ceol. Survey Profess. Paper 205B, pp. 23-77 (1945).

60 Clay Mineralogy
within the lattice, and Ross and Hendricks 34 have assigned the names in
Table 2 to montmorillonites with the compositions indicated. The
names conform closely to general usage with the exception of beidellite,
which many investigators consider to be discredited. Arrows are placed
TABLE 2. FORMULAS FOR SOME MEMBERS OF THE MONTMORILLONITE GROUP
SUGGESTED BY Ross AND HENDRICKS3.
Dioctahedral (heptaphyllitic) montmorillonites
Montmorillonite (OH),Si 8 (AI 3.3,Mg. 66 )020
1
Na .66
Beidellite (OH).(Si6.3,All.66)Al,.3,020
1
or Na .66
Beidellite (OH),(Si 6A!.)Al,.,,020
1
Na .66
Nontronite (OH),(Si 7 3,AI. 66 )Fe!+020
1
or Na .66
Nontronite (aluminian) (OH) ,(Si 6. 34 AJ,.66)Fc!1,020
1
Na .66
Trioctahedral (octaphyllitic) montmorillonites
Hectorite (OH).Si s (Mgs. 34 Li. 66 )020
1
Na .66
Saponite (OH),(Si7.3,Al.66)Mg6020
1
or Na .66
Saponite (aluminian) (OH) ,(Si 6. 66 All. 34) (Mg,. 34' AI. 66) 0 20
1
Na .66
under the group having the charge deficiency, which requires the addition
of a cation external to the silicate layer to balance the structure. In each
case the balancing external cation has been indicated as Na+ for con-
venience. The water or other polar molecules between the silicate layers
are omitted from the formulas.
MacEwan 3 has considered the relation of the a and b dimensions of
the montmorillonite unit to variations in its chemical composition. He
has pointed out the conclusion that these axial lengths would increase in
the order montmorillonite ---7 nontronite ---7 saponite. He has offered the
following formula for computing bo, from which ao, of course, can also be
computed:
bo = 8.91 + 0.061' + 0.0348 + 0.048t A
35 MacEwdn, D. M. C., The Montmorillonite Minerals, "X-ray Identification and
Great Britain Monograph, (1951).
}Vhere r = number of Al ions in tetrahedral coordination
s = number of Mg ions in octahedral coordination
t = number of Fe ions in octahedral coordination (per unit cell, in
each case)
The formula is based on observed values for muscovite, talc, and
nontronite with the assumption that the effects of the above substitu-
tions are proportional and additive. MacEwan 65 states that the formula
gives reasonably good agreement with the observed values. It appears
from MacEwan's data that changes in bo are much greater for
variations in the population of octahedral positions than for tetrahedral
positions.
The ratio of silicon to aluminum in the montmorillonite lattice appar-
ently can vary from about 1: 1 to 1 :3. In the former case there is a
maximum filling of octahedral positions by AP+ (about 4.44 per unit)
with added replacement of Si H by AP+. The replacement of Si H by
AlH causes the unbalanced charge and also compensates for the excess
charge in the octahedral layer. For the highest ratio all tetrahedral
positions are populated by Si H , and the mineral is dioctahedral with the
maximum replacement of AIH by a divalent ion to cause the charge
deficiency.
Iron can apparently proxy for aluminum in all positions in the octa-
hedrallayer and not at all in the tetrahedral layer. Iron-rich varieties,
i.e., nontronite, on the basis of currently available analyses, show little
replacement of Fe H by Mg++, and the charge deficiency seems to result
mainly from substitutions of AIH for Si 4+. In the trioctahedral mont-
morillonites the charge deficiency results mainly from substitutions of AIH
for Si H It would seem from data of Ross and Hendricks 34 that up to
about one atom of aluminum or iron per unit cell may be present in the
octahedral layer of trioctahedral montmorillonites. The excess charge
in the octahedral layer is balanced by deficiencies in the tetrahedral layer
due to larger replacements of AIH for Si H .
The opinion has frequently been expressed that the structure of
Hofmann et al.,27 Marshall,29 and Hendricks 18 does not adequately
account for all the properties of montmorillonite, notably its ion-exchange
capacity. Edelman and Favejee 20 have suggested an alternative struc-
ture that, it is claimed, explains these properties more adequately. This
structure differs from that of Hofmann et al. 27 in that every other silica
tetrahedron in both silica sheets is inverted, so that half of them point in
the opposite direction. Those that point away from the silicate sheet
would have the tip 0 replaced by OH. In this structure the silicon atoms
are not all in a single plane in the silica tetrahedral sheets, and there must
be some substitutions of OH for 0 in the octahedral layers to balance the
structure (Figs. 12 and 13). The charge distribution is as follows:
62 Clay Mineralogy
2(OH)- 2-
2Si H 8+
60-- 12-
2Si H 8+
20--4(OH)- 8-
4AI3+ 12+
20--4(OH)- 8-
2Si H 8+
60-- 12-
2Si H 8+
2(OH)- 2-
Interlayer H 20 or other polar groups.
The structural formula is (OH) 12SisAI4016H20(interlayer). In this

structure, no substitution is necessary within the lattice to account for the
exchange capacity. The lattice may be completely balanced. The
exchange reaction is believed due primarily to replacement of the H of the
OH of the projecting tetrahedrons. As originally postulated, there are
far more available OH groups than required by the exchange capacity.
Edelman and Favejee 20 postulated that only some of them were available
for exchange. Evidence from X-ray-diffraction data based on Fourier
syntheses,36 chemical data indicating lattice substitution, and careful
dehydration studies, all are strongly against the original Edelman and
Favejee 20 structure. Edelman 37 has recently suggested a modified struc-
~t-UrelnWh-l"ch only about 20 per cent of the tetrahedrons are inverted,
which would satisfy the discrepancy with the exchange capacity. It is
questionable whether or not X-ray data would be adequate to provide
evidence for or against the modified structure. However, the modified
structure still does not seem to fit with chemical data indicating lattice
replacements and with the geometry of the adsorption of organic mole-
cules. Studies of the methylation of certain organics during their
adsorption by montmorillonite have led some workers (Berger,3S Due1l 39 )
to favor this structure, since more OH groups seem to be called for on the
surface of the montmorillonite silicate unit than are provided for by the
structure of Hofmann et al. 27 Further studies are necessary before the
possible structural implications of such organic-montmorillonite reactions
become clear.
36 Brown, G., A Fourier Investigation of Montmorillonite, Mineralog. Soc. Gr.
Britain Clay Mineral Bull. 4, pp. 109-111 (1950).

37 Edelman, C. H., Relations entre les propri6t6s et la structure de quelques mineraux
argileux, Verre silicates ind., 12,3-6 (1947).

38 Berger, G., De Struktuur van Montmorilloniet, Chem. Weekblad,38,42-43 (1941).
39 Duell, H., G. Huber, and R. Iberg, Organische Derivate von Tonmineralien.

Helv. Chim. Acta, 33, 1229-1232 (1950).
Structure of the Clay Minerals
n H2 0
Exchongeob~ Co~ons
,-..
, ,
t,', r"\
,, 2(OH)
2 Si
60
2 Si
<!
o
<i
+1
(J) . /
/
x
<!
u
2 Si
60
2 Si
, , ,
,_/,
/
{ (
'j
(
2(OH)
- - - - - - - b Axis ------~.
Montmorillonite (OH)12 AI4 Sis 0 16 n H20

FIG. 12. Schcmatic presentation of the structure of montmorillonite suggested by
Edelman and Favejee. 2o
McConnell 40 has recently suggested a reVISIOn in the structure of

Hofmann ct al. 27 whereby some of the silica tetrahedrons are replaced
by (OH)4 tetrahedrons. This is essentially a hole in the silicon positions
40 McConnell, D., The Crystal Chemistry of Montmorillonite, Am. Mineral., 35
166-172 (1950).
Clay Mineralogy
Exchangeable Cations
nH2 0
FIG. 13. Diagrammatic sketch of the structure of montmorillonite 3uggested by

Edelman and Favejee.2
with adjacent balancing by OR for O. This would provide the needed

quantity of surface OR required by certain organic-adsorption character-
istics and, according to McConnell, is in accord with the dehydration of
the mineral. Further investigations are necessary before the feasibility
of McConnell's suggestion can be evaluated.
ILLITE MINERALS
The structure of the micas, the variations in their composition, and

their polymorphic variations have been worked out in considerable detail
by Pauling,l Mauguin,4l,42 Jackson and West,43,44 Winchell,45 and
Hendricks and Jefferson. 46 The basic structural unit is a layer composed
of two silica tetrahedral sheets with a central octahedral sheet. The tips
of the tetrahedrons in each silica sheet point toward the center of the unit
and are combined with the octahedral sheet in a single layer with suitable
replacement of OH by O. The unit is the same as that for montmoril-
lonite except that some of the silicons are always replaced byaluminums
and the resultant charge deficiency is balanced by potassium ions. In
many of the well-crystallized micas one-fourth of the silicons are replaced
by aluminums. The unit layers extend indefinitely in the a and b direc-
tions and are stacked in the c direction. The potassium ions occur
between unit layers where they just fit into perforations in the surface
o layers. Adjacent layers arE; atacked in such a way that the potassium
ion is equidistant from 12 oxygens, 6 of each layer (Fig. 14).
Muscovite is dioctahedral; i.e., only two-thirds of the possible octa-
hedral positions are filled, and the octahedral sheet is populated only by
aluminums. The structural formula is (OH)4K2(Si6'Ab)AI402o, and the
theoretical composition is K 20 11.8 per cent, Si0 2 45.2 per cent, Ah03
38.5 per cent, H 20 4.5 per cent. The mineral is monoclinic, and the unit
cell is composed of two silica-alumina-silica layers; the dimensions are
a = 5.2 A, b = 9.0 A, c = 20.0 A, f3 = 9530' with space group C~h-C2/c.
The biotite micas are trioctaliedral with the octahedral positions
populated mostly by Mg++, Fe++, and/or Fe H . Examples are biotite,
(OH)4K 2(Si eAh) (MgFe)e020, with the relative abundance of the iron and
magnesium varying widely, and phlogopite, (OH)4K 2(Si 6AI 2)Mge()2o.
Hendricks 46 has shown the existence of at least six polymorphic variations
of biotite due to variations in the number of silica-alumina-silica units
per unit cell and in the manner of stacking of the unit cells. Unit cells
41 Mauguin, C. H., Etude des micas au moyens des rayons X, Compt. rend" 185,
288-291 (1927).
42 Mauguin, C. H., Etude des micas au moyens des rayons X, Bull. soc. franc.
mineral., 51, 285-332 (1928).
43 Jackson, W. W., and J. West, The Crystal Structure of Muscovite, Z. Krist., 76,
211-227 (1930).
44 Jackson, W. W., and J. West, The Crystal Structure of Muscovite, Z. Krist., 85,
160-164 (1933).
45 Winchell, A. N., Studies in the Mica Group, Am. J. Sci., ser. 5, 9,309-327,415-
430 (1925).
46 Hendricks, S. B., and M. Jefferson, Polymorphism of the Micas with Optical
Measurements, Am. Mineral., 24, 729-771 (1939).
66 Clay Mineralogy
composed of 1, 2, 3, 6, and 24 silica-alumina-silica units are known, with
stackings yielding monoclinic, rhombohedral, or triclinic forms.
o Oxygens, e Hydroxyls, Aluminum, 0 Potas$ium

o and. Si/icons (one fourth replaced by aluminums)
FIG. 14. Diagrammatic sketch of the structure of muscovite.
It is important that, in the large well-crystallized micas with no imper-

fections in the regularity of stacking, there seem to be no isomorphous
gradations from dioctahedral to trioctahedral types.
The charge distribution within the layers of the well-crystallized micas
is as follows:
K+ 1+
60-- 12-
3Si4+1AP+ 15+
40--2(OH)- 10-
4A13+ (Dioctahedral) or
6R (Trioctahedral, R = Mg++, Fe++, Fe H , Li+, TiH) 12+
40--2(OH)- 10-
3Si4+1AP+ 15+
60-- 12-
K+ 1+
The illite clay minerals differ from the well-crystallized micas in
47
several possible ways, all of which may be exhibited by a given sample;

these are as follows: There is less substitution of AP+ for Si4+; in the well-
crystallized micas one-fourth of the Si H are replaced, whereas in the
illites frequently only about one-sixth are replaced. As a consequence of
this smaller substitution, the silica-to-alumina molecular ratio of the
illites is higher than that of the well-crystallized micas, and the net-
unbalanced-charge deficiency is reduced from 2 per unit cell to about 1.3
per unit cell. The potassium ions between the unit layers may be
partially replaced by other cations, possibly Ca++, Mg++, H+. There is
some randomness in the stacking of the layers in the c direction, and the
size of the illite particles occurring naturally is very small, of the order of
1 to 2 microns or less.
Because of the foregoing variations, illite clay minerals show differ-
ences in their characteristics from those of the well-crystallized micas. 48
The lO-A diffraction line of the micas is shown by illite usually as a
characteristic diffraction effect, but it is often modified into a band which
tends to tail off slightly toward the low-angle region as a consequence of
the small particle size, variation in the interlayer cation, and occasional
slight interlayer hydration. The occasional layers may simply be the
nontypical features associated with any particle surface. Also because
of these defects, hkl reflections are weak, or absent in the case of those
normally weak in the well-crystallized micas. .Most of the illite clay
minerals so far described are dioctahedral, but some are known that are
trioctahedral with unit cells composed of one, two, or three silica-alumina-
silica layers. 48,49 Careful X-ray-diffraction data (see page 93) will
permit the identification of the polymorphic form of illite if the sample is
reasonably pure and if the crystallinity is not too poor. In some cases
it is not possible with currently available analytical techniques to identify
the polymorphic forms of the illites.
47 Grim, R. E., R. M. Bray, and W. F. Bradley, The Mica in Argillaceous Sedi-
ments, Am. Mineral., 22,813-829 (1937).

48 Grim, R. E., W. F. Bradley, and G. Brown, The Mica Clay Minerals, "X-ray
ical Society of Great Britain Monograph (1951).
49 Walker, G. F., Trioctahedral Minerals in Soil-Clays of Northeast Scotland,
Mineralog. Mag., 29,72-84 (1950).

68 Clay Mineralogy
Dioctahedral illites are known in which the aluminum is replaced by
some iron and magnesium. Whether or not there is any isomorphous
relationship between the dioctahedral and trioctahedral illites is not
definitely known, but none has been established thus far.
Glauconite 50 is a dioctahedral illite with considerable replacement of
A}Z+ by Fe H , Fe++, ana Mg++, frequently with even less than two-thirds
of the possible positions filled. There is consequently a charge deficiency
in the octahedral sheet as well as in the tetrahedral sheets, and the inter-
layer cations seern to balance both these charges. Often Ca++ and Na+
as well as K+ are the interlayer cations. Hendricks and ROSS50 give
the following formula as characteristic of many glauconites: (OH) 4-
(K,Ca7'i,N a) 1.68 (Si7.3 oAl o. 70) (Alo.94Fei~4Fet.t8Mgo.80)02o. A further char-
acteristic of glauconite is that the unit cell is composed of a single silicate
layer rather than the double layer of most other dioctahedral micas.
Grim, Bradley, and Brown48 have presented the following formula
relating the bo dirnension to the chemical composition of the micas. The
formula is based on the measured bo dimensions of pyrophyllite, musco-
vite, and talc with the assumptions that the effect of lattice substitutions
is additive and linear and that the increase per ion is proportional to the
ionic diameter. Brown states that better correlations between observed
and calculated results are obtained if all the iron is assumed to have the
radius of the ferrous ion and if the lithium ion is treated as if it had the
same effect as the aluminum ion. The Brown formula is
bo = 8.91 + 0.12Al + 0.067Mg + 0.084Fe + 0.022Ti A
where .4.1 = number of Al3+ in tetrahedral coordination per haJJ ceJJ
Mg = number of Mg++ in octahedral coordination per half cell
Fe = number of iron ions in octahedral coordination per half cell
Ti = number of Ti4+ ions in octahedral coordination per half cell
The structure of illite differs from that of montmorillonite in several
important ways ai'3 follows: The charge deficiency due to substitutions pE;)r
unit cell layer is about 1.30 to 1.50 for illite and 0.65 for montmorillonite.
The seat of this charge deficiency in illite is largely in the silica sheet and
therefore close to the surface of the unit layer, whereas in montmorillonite
it is frequently, perhaps chiefly, in the octahedral sheet at the center of
the unit layer. Also in the case of illite the balancing cation between the
unit layers is chiefly or entirely potassium. Because of these differences
the illite structural unit layers are relatively fixed in position, so thn,t
polar ions cannot enter between them and cause expansion. Also the
interlayer balancing cations are not exchangeable (see Chap. 7).
It is conceivable that all gradations can exist between illite and well-
50 Hendricks, S. B., and C. S. Ross, The Chemical Composition and Genesis ()f
Glauconite and Celaclonite, Am. Mineral., 26, 683-708 (1941).
crystallized muscovites and biotites on the one hand and montmorillonite
on the other hand. Data are not yet available to settle this point posi-
tively, but it is noteworthy that the illites so far studied in detail have
shown similar and distinctive attributes rather than complete gradational
variations with muscovite or biotite. With regard to gradations with
montmorillonite the data are less clear. Many clays are known which
seem to be gradational between illite and montmorillonite but which are
actually composed of interlayered mixtures of true illite and mont-
morillonite (see mixed-layered minerals, pages 79, 80). It is possible, even
likely, that clay minerals are to be found with a stacking of layers differing
slightly from illite in having less unbalanced charge and with substantial
replacement of K+ by another cation like Na+. It wou~d be expected,
that in such specimens some water could penetrate between the unit
layers with some accompanying expansion. In such specimens there
might well be some variation in the unbalancing between different layers,
and the nature of the baJa.ncing cations might vary from layer to layer as
well as between the same layer. In such material, expansion between
all layers would not be the same and might even vary within a single
interface between adjacent units. The classification of such material
would be difficult, and the separation of illite and montmorillonite must
frequently be purely arbitrary with, in general, definitely expanding
material being called montmorillonite and nonexpanding being called
illite.
CHLORITE MINERALS
The structure of chlorite was first suggested by Pauling. 3 Later

McMurchy 51 examined the mineral in great detail, verifying Pauling's
suggested structure and providing additional information on its sym-
metry and dimensions. Recently Brindley and Robinson 52 ,53 have further
studied the mineral and considerably amplified knowledge of the general
chlorite structure and the variations of different members.
All the true chlorites have the same general structural framework.
The structure consists of alternate mica-like and brucite-like layers.
The layers are continuous in the a and b dimensions and are stacked in the
c direction with basal cleavage between the layers. The mica-like layers
are trioctahedral with the general composition (OH)4(SiAl)s(MgFe)G02o.
The brucite-like layer has the general composition (MgAl)u(OH)12. The
61 McMurchy, R. C., Structure of Chlorites, Z. Krist., 88, 420-432 (1934).

62 Robinson, K., and G. W. Brindley, Structure of Chlorites, Proc. Leeds Phil. Lit.
Soc. Sci. Sect., 6, 102-108 (1948) .
3 Brindley, G. W., and K. Robinson, The Chlorite Minerals, "X-ray Identifica-
tion and Structure of the Clay Minerals," Chap. VI, pp. 172-198, Mineralogical
Society of Great Britain Monograph (1951).
70 Clay Mineralogy
mica layer is unbalanced by substitution of AP+ for Si H , and this defi-
ciency of charge is balanced by an excess c:harge in the brucite sheet as a
cwnsequence of substitution of AlH for Mg++. The general chlorite
FIG. 15. Diagrammatic sketch of the structure of chlorite, after McMurchy.!j1
structure is illustrated in Figs. 15 and 16. The charge distribution in the

layers is as follows, where X = 1 or 2:
6(OH)- 6-
(6 - 2X)Mg++'2XAP+ [2(6 - 2X) + 3(2X)] +
6(OH)- 6-
60-- 12-
(4 - X)SiH'XAP+ [4(4 - X) + 3X] +
40-- 2(OH)-- 10-
6R(Trioctahodral, R = Mg++, Mn++, Fe++,
Fe H , Cr H , TiH) 12 +
40-- 2(OH)-- 10-
(4 - X)SiHXAP+ [4(4 - X) + 3X] +
60-- 12-
Te:::~{--
---0
Talc
Layer """"'"' {
,:~~.:~~--~--~----~----~~~~~~~~~~~~~~
FIG. 16. Projection of the chlorite structure on the ac plane, after McMurchy&1 from
Brindley and Robinson. 53
Two brucite-like and two mica-like layers form the unit cell. Mc-
Murchy 51 gives the following specifications for the chlorite he studied:
monoclinic, a = 5.3 A, b = 9.3 A, c = 28.52 A, {3 = 978', and space
group C~h-C2/c. Subsequently Robinson and Brindley 52 have stated
"the displacements are exactly ao/3 and since bo/ao = yi3, the unit cell
may be regarded as rhombohedral with the trigonal axis perpendicular
to the cleavage plane, or alternatively as an orthohexagonal cell with a
volume three times that of the monoclinic prism and with dimensions in
kX units ao = 5.32, bo = 9.21, Co = 3 X 14.2."
72 Clay Mineralogy
Various members of the chlorite group differ from each other in the
kind and amount of substitutions within the brucite layer and the tetra-
hedral and octahedral positions of the mica layer. They also differ in
the detailed orientation of successive octahedral and tetrahedral layers,
in the relation of the mica to brucite layers, and in the stacking of succes-
sive chlorite units.
Substitutions within the tetrahedral sheets vary from about Si3AI to
Si 2Al 2 and substitutions within the octahedral layers from Mg5AI to
Mg4Ab, with Fe++ and Mn++ partially replacing Mg++ and Fe H or Cr H
partially replacing AP+. The substitutions of AP+ for Si H in the tetra-
hedral sheets expands them sufficiently to accommodate the somewhat
larger octahedral layers between them.
By detailed study of single crystals, Brindley53 and his colleagues have
shown that chlorites exhibit polymorphic forms similar to the biotite
micas. They have found structures consisting of several different stack-
ing arrangements of the chlorite layers. Units with three, six, and nine
chlorite layers in the orthohexagonal cell have been described. For
details of these polymorphic forms and other details of the chlorite struc-
ture the original work of Brindley et al. 53 should be consulted.
Fine-grained chlorites are known in some clay materials. The range
of polymorphic forms of the chlorite minerals and the variations in struc-
ture between clay-mineral chlorite and well-crystallized chlorite occurring
in large units have not been established. It is not definitely known, for
example, whether somewhat more hydrous forms of chlorite, like illite in
relation to muscovite and biotite, are to be found in clays and soils.
Available data seem to suggest that clay-mineral chlorites differ from well-
crystallized material in l1 somewhat random stacking of layers and per-
haps in some hydration.
Some evidence is available that some chlorites in sediments may be
dioctahedral.
VERMICULITE
The structure of vermiculite was first worked out by Gruner. suo

Later Hendricks and Jefferson 56 .57 confirmed Gruner's general conclusions

575 (1934).
55 Gruner, J. W., Water Layers in Vermiculite, Am. Mineral., 24, 428-433
(1939).
56 Hendricks, S. B., and M. E. Jefferson, Crystal Structure of Vermiculites and
Mixed Vermiculite-chlorites, Am. Mineral., 23, 851-863 (1938).
57 Hendricks, S. B., and M. E. Jefferson, Structures of Kaolin and Talc-pyrophyllite
Hydrates and Their Bearing on Water Sorption of the Clays, Am. Mineral., 23,
863-875 (1938).
but changed certain structural details. Recently Barshad5s- 60 and
Walker 61 .62 independently have added much to our knowledge of the
structure of the mineral. Gruner 54 also showed that many materials
classed as vermiculite are mixed-layer mica-vermiculite structures.
Gruner 54 showed that the structure consists of sheets of trioctahedral
mica or talc separated by layers of water molecules occupying a definite
space (4.98 A) which is about the thickness of two water molecules. In
its natural state, therefore, the mineral consists of an alternation of mica
and double water layers. Gruner 54 assigned vermiculite to the space
group C~,,-C2/c.
Based on single-crystal data, Hendricks and Jefferson 56 assigned the
following crystallographic constants to the mineral: c = 28.91 A,
b = 9.20 A, a = 5.34 A, {3 = 9315/. They showed that vermiculites
have the same types of shifts along the a axis as muscovite, talc, and
pyrophyllite and have a partially random displacement of structural
layers parallel to the b axis. They give the space group as Cc-C; and
point out the psuedo nature of the space group because of the random
displacement parallel to the b axis.
The structure is unbalanced chiefly by substitutions of AIH for Si H .
These substitutions may be partially balanced by other substitutions
within the mica lattice, but there is always a residual net-charge deficiency
of 1 to 1.4 per unit cell. The charge deficiency is satisfied by cations
which occur chiefly between the mica layers and are largely exchangeable.
In the natural mineral, which has the same cation-exchange capacity as
montmorillonite, or somewhat higher, the balancing cation is Mg++,
sometimes with a small amount of Ca++ also present.
A general formula for natural vermiculite is (OHMMgCa)x(Sis_xAl x )
(MgFe)602oyH 20 with X = 1 to 1.4 and y = about 8. The Mg++ and
Ca++ are the balancing and largely exchangeable cations. The charge
distribution in the layers is as follows:
yH 20 double water layers
X(Mg++'Ca++) X +
60-- 12-
(4 - X)SiHXAP+ (16 - X)+
68 Barshad, 1., Vermiculite and Its Relation to Biotite, Am. Mineral., 33, 655-678
(1948).
69 Barshad, 1., The Nature of Lattice Expansion and Its Relation to Hydration in
Montmorillonite and Vermiculite, Am. Mineral., 34, 675-684 (1949).

Crystal Lattice, Am. Mineral., 35, 225-238 (1950).

61 Walker, G. F., Water Layers in Vermiculite, Nature, 163, 726 (1949).
62 Walker, G. F., Vermiculite and Some Related Mixed-Layer Minerals, "X-ray
Ide~tificati~n and Structure. of. the Cl!Jl Minerals," Chap. VII, pp. 199-223, Miner-
aloglCal SOCiety of Great Bntam ~oriograph (1951).
/"'-
74 Clay Mineralogy
40--2(OH)- 10-
6(MgFe)++ 12+
40--2(OH)- 10-
(4 - X)SiHXAP+ (16 - X)+
60-- 12-
X(Mg++Ca++) X+
yH 20 double water layers
On heating vermiculite to temperatures as high as 500C, the water is
driven out from between the mica layers, but the mineral quickly rehy-
drates on exposure to moisture at room temperature. The mineral,
therefore, has an expanding lattice, but the expansion is restricted to
about 4.98 A, or two water layers. If the mineral is heated to 700C,
there is no expansion again. In such material the 14-A line and higher
orders of it disappear, and a new line at 9.3 A with other new mica lines
appear on the diffraction pattern.
There has been considerable discussion regarding the exact structure
of the water layers in vermiculite, and as yet there is no general agree-
ment among investigators. Gruner 55 suggested that the water might
occur as charged hydronium (H 30)+ groups. Hendricks and Jefferson 57
postulated an extended hexagonal network of water molecules (Fig. 17)
(see Chap. 8 for a further discussion). It has recently been pointed out
by Barshad 59 ,60 and Walker 61 62 independently that the nature of the
exchangeable cations must influence the state of the interlayer water
since they occur between the mica layers. These investigators have
shown this to be the case (Table 3).
Walker 62 views the state of the water in natural vermiculite as follows:
TABLE 3. THICKNESS OF WATER LAYERS IN VERMICULITE SATURATED WITH VARIOU~
CATIONS
(After WalkerS')
Cation d(002), A Thickness of water layer, A
Mg++ 14.39 5.11

Ca++ 15.0 5.75
Sr++
Ba++
15.0
12.3 -" 5.75
3.04
Li+ 12.2 2.04
Na+ 14.8 5.55
K+ 10.6 diffuse About 1.34
NHt 10. 8 diffuse About 1.54
The water exists in two forms, "bound" and "unbound." The bound
water occurs around the Mg++ as a hydration envelope of six water
molecules per Mg++, probably ta-king the form of an octahedral grouping.
The available space, 4.98 A, is consistent with this grouping if the octa-
hedral groups lie on their sides so that they present their minimum width
to adjacent sheets. The unbound water fills in the space between the
octahedral water units. Walker computes the ratio of unbound water to
FIG. 17. Diagrammatic sketch of the vermiculite structure showing layers of water,
after Hendricks and Jefferson. 67
total interlayer water to be about 8: 14. Walker shows that, if vermicu-

lite is heated to about lloDe in a closed tube, so that rehydration is not
possible, partial but regular collapse of the structures takes place to
11.8 A, representing a water-layer thickness of 2.54 A, which is equivalent
to one molecular water layer. Walker reasons that in such partial col-
lapse the unbound water is driven off and also the octahedral net around
the Mg++ is broken, with these water molecules pressed flat between the
mica layers. Such partially collapsed vermiculite has a strong tendency
76 Clay Mineralogy
to rehydrate, which may be due to the strong tendency of Mg++ to
rebuild its octahedral envelope.
In the case of vermiculite prepared with exchangeable Ba++ and Li+,
there seems to be a single water layer between the mica layers. For
vermiculites with K+ and NHt, the space is smaller than that required
for a single molecular water layer, and Barshad 59 considers the mineral
to be dehydrated. WaLker,6:1 however, considers that some water is
present with the 1(+ and NHi and that the reduced space is due to a
partial embedding in the silicon-oxygen sheets. He points out that,
(0) (b) (c)
Silicate LaYE?r
TI T Silicate Layer
I I T Silicate Layer
I I I
I I I
I
14.2A OO~ ~ ~ I
11.8A~ 0 0
I
9.26A
: ~~~ I
I
I
I
I I
I I
I I l. Silicate Layer
I .I. Silicate Layer
i Silicate Layer
Magnesium ion @j)"Unbound" Water o ''Bound'' Water

FIG. 18. Configuration of the interlayer water in vermiculite (diagrammatic),
from \Yalker. 62 (a) Pully hydrated. (b) About half hydrated after removal of the
unbound water. Of the six water molecules in the original hydration shell of mag-
nesium (see a), only four are now in actual contact with the ion. (c) Fully dehy-
drated with magnesi1Hns in holes in the silicate layers.
when K+-saturated vermiculite is heated, the basal spacing is further

reduced to 10.1 A. Walker's concept is illustrated in Fig. 18. (For a
further discussion ()f the hydration of vermiculite, see Chap. 8.)
Since vermiculite, has an expanding structure and an unbalanced lattice
with high cation-exchange capacity, it is important to consider its relation
to montmorillonite, which also has these characteristics. Vermiculite
differs from montrnorillonite in that the expansion with water is limited
to about 4.98 A. Vermiculites adsorb certain organic molecules (see
Chap. 10) between the mica layers but differ from montmorillonite in
that the adsorbed layer is thinner and less variable. These character-
istics may be the result of the relatively larger particle size of the vermicu-
lite layers, and possibly also of the fact that there is considerably less
randomness in the stacking of the vermiculite layers. In fact, in ver-
miculite the randomness may be in only one horizontal direction. All
rows of tetrahedral apexes at alternate intervals in the top and bottom of
the sheets. The ribbed sheets are arranged so that the apexes of succes-
sive sheets point together, and the sheets are held together by aluminum
and/or magnesium in octahedral coordination between the apex oxygens
of successive sheets (Fig. 19). The octahedral layer is similar to that in
the layer clay minerals, but it is continuous in only one direction. The
octahedral layer is completed by central OH groups and by hydroxyls at
the open sides. The mineral has good cleavage parallel to (110) caused
by the weak link through 0 of the silica chains in the ribbed layer.
-Chains of water molecules running parallel to c fill the interstices between
the amphibole chains. The cavities will accommodate four molecules
of water per unit cell, and this water would account for the dehydration
loss at low temperatures. The dimensions of the cell suggested by
Bradley72 are ao sin ex = 12.9 A, bo = 18 A, and Co = 5.2 A. There are
two molecules in the unit cell, and the space group is probably C~h-C2/m.
The structure is balanced, and the composition of the ideal cell is (OH 2)4-
(OH)2Mg5Si802o.4H20, in which trivalent cations are considered equiva-
lent to 1.5Mg++. Some substitution of Aj3+ for SiH is considered
probable. Bradley72 points out that substitution of Aj3+ for either Mg++
or Si H or both should weaken the structure, so that it appears doubtful
that extensive replacement takes place, and an aluminum end member
would not be expected.
There are general similarities but also distinct differences in the diffrac-
tion data for attapulgite and some sepiolites (see the tables of Chap. 5).
It appears likely, therefore, that the general structural framework of all
these clay minerals is the same but that there are distinct differences in
details. Considerably more work is necessary before these details
become known. Structural studies of the sepiolites are particularly
needed so that structural variations between them and attapulgite can
become known.
MIXED-LAYER MINERALS
Many clay materials are composed of more than one clay mineral, and
the clay minerals may be mixed in several ways. The mixture may be
of discrete clay-mineral particles in which there is no preferred geometric
orientation of one particle with respect to its neighboring clay-mineral
particles.
Another type of mixing is the interstratification of the layer clay
minerals in which the individual layers are of the order of a single or a
few aluminosilicate sheets. These so-called mixed-layer structures are a
consequence of the fact that the layers of the different layer clay minerals
are very similar, all being composed of silica tetrahedral-hexagonal layers
and closely packed octahedral layers of oxygens and hydroxyl groups.
80 Clay Mineralogy
Mixed-layer structures as stable as those composed of a single kind of
layer are therefore possible.
Mixed-layer structures are of two different types. The interstratifica-
tion may be regular; i.e., the stacking along the c axis is a regular repeti-
tion of the different layers. In such cases, the resulting structure has
distinctive characteristics; the unit cell is equivalent to the sum of the
component layers, and regular (001) reflections are obtained. An
example of a regular mixed-layer mineral is chlorite composed of a regular
alternation of mica and brucite layers (see pages 69-72).
Another kind of mixed-layer structure is due to a random irregular
interstratification of layers in which there is no uniform repetition of
layers. The importance of such random mixing was pointed out by
Gruner 54 and later elaborated by Hendricks and Teller 73 and Bradley.74
Mixed-layer structures appear to be very common in clay materials.
Their study is particularly difficult, and in many cases investigators have
failed to recognize their presence. Mixed layers of illite and mont-
morillonite and of chlorite and vermiculite are particularly common.
Clay-mineral occurrences which are probably mixtures of illite and
kaolinite have also been described. 75
As pointed out previously (pages 38--42), the literature contains the
names of many discredited or questionable clay-mineral species. Proba-
bly a great many of these "species" are mixed-layer combinations.
Careful X-ray-diffraction techniques (page 102) are required to detect
the occurrence of mixed-layer minerals. Sometimes differential thermal
procedure will aid in their detection. Very frequently mixed layering
cannot be detected solely on the basis of optical measurements, and the
widespread occurrence of mixed-layer minerals is an important reason
why clay-mineral identifications based solely on optical measurements
must be made with great caution.
Since random mixed-layer minerals have an inherent variability, they
cannot be given specific names. They can only be designated as mixtures
of the layers involved.
Kaolinite Minerals
Alvfeldt, 0., X-ray Study on Kaolinite as an Alteration Product of Spodumene from
Varutrask, Geol. Fiiren. i. Stockholm Fiirh., 63, 49-51 (1941).
Brindley, G. W., and K. Robinson, X-ray Study of Some Kaolinitic Fireclays, Trans.
Ceram. Soc. (Engl.), 46, 49-62 (1947).
fa Hendricks, S. 13., and E. Teller, X-ray Interference in Partially Ordered Layer
Lattices, J. Chem. Phys., 10, 147-167 (1942).
74 Bradley, W. F., Diagnostic Criteria for Clay Minerals, Am. Mineral., 30, 704-
713 (1945).
75 De Lapparent, J., Constitution et origine de la leverrierite, Compt. rend., 198,
669-671 (1934), "
Gruner, J. W., Density and Structural Relations of Kaolinites and Anauxites, Am.
Mineral., 22,855-860 (1937).
Gruner, J. W., Progress in Silicate Structures, Am. Mineral., 33, 679-691 (1948).
Hendricks, S. B., Concerning the Crystal Structure of Kaolinite, Ah032SiOz2H20
and the Composition of Anauxite, Z. Krist., 95, 247-252 (1936).
H alloysite Minerals
Alexander, L. T., G. Faust, S. B. Hendricks, anc'_ H. Insley, Relationship of the Clay
Minerals Halloysite and Endellite, Am. Mineral., 28, 1-18 (1943).
Brammall, A., The Layer-lattice in Relation to Min~ral Chemistry, Science Progress,
31, no. 124 (April, 1937).
Correns, C. W., and M. Mehmel, Ueber die opti~chen und riintgenographischen
Nachweis von Kaolinit, Halloysit und Montmorillonit, Z. Krist., 94, 337-348
(1936).
Hofmann, U., Neues aus der Chemie des Tons, Die Chemie, 55, 283-289 (1942).
Nagelschmidt, G., Riintgenographische Untersuchungen an Tonen, Z. Krist., 87, 120~
146 (1934).
Ross, C. S., Minerals and Mineral Relationships of the Clay Minerals, J. Am. Ceram.
Soc., 28, 173-183 (1945).
Montmorillonite
Cornet, 1., Expansion of the Montmorillonite Lattic~ on Hydration, J. Chem. Phys.,
18, 623-626 (1950).
Forslind, E., The Clay-Water System, I, Crystal Structure and Water Adsorption of
th~ Clay Minerals, Swed. Cement Concrete Resea~ch Ins!. Bull. 11 (1948).
Ginsburg, 1. 1., Nontronites from the Southern Ural~, Compi. rend. acado sci. URSS,
pp. 41-61 (1946).
Gruner, J. W., The Structural Relationships of NOI:ttronites and Montmorillonites,
Am. Mineral., 20,475-483 (1935).
Hendricks, S. B., and C. S. Ross, Lattice LimitatioI(s of Montmorillonite, Z. Krist.,
100,251-264 (1938).
Hofmann, U., and W. Bilke, Ueber die innerkristlilline Quellung und das Basen-
austauschvermiigens des Montmorillonits, Kolloid-Z., 77,239-251 (1936).
Hofmann, U., and A. Hausdorf, Kristallstruktur uI(d innerkristalline Quellung von
Montmorillonit, Z. Krist, 104, 265-293 (1942).
Jackson, M. L., and N. N. Hellman, X-ray Diffracticm Procedure for Positive Differ-
entiation of Montmorillonite from Hydrous Mtca, Soil Sci. Soc. Am. Proc., 6,
133-145 (1941).
Kelley, W. P., Calculating Formulas for Fine-Grained. Minerals on the Basis of Chem-
ical Analyses, Am. Mineral., 30, 1-26 (1945).
MacEwan, D. M. C., A Trioctahedral Montmorillonit\) Derived from Biotite, Abstracts,
18th Intern. Geol. Congr., p. 128 (1948).
Marshall, C. E., The Constitution of the Clay Minerll,ls, Science Progress, 30, 422-433
(1936).
Marshall, C. E., The Colloidal Properties of Clays as Related to Their Crystal Struc-
ture, J. Phys. Chem., 41, 935-942 (1937).
Krist. A, 93, 481-487 (1936).
Nagelschmidt, G., On the Atomic Arrangement and. Variability of the Members of
the Montmorillonite Group, Mineralog. Mag., 26, 140-155 (1938).
Noll, W., Zur Kenntniss des Nontronits, Chem. Erde, 5,373-384 (1930).
82 Clay Mineralogy
Ross, C. S., and S. B. Hendricks, Clay Minerals of the Montmorillonite Group, Soil
Sci. Soc. Am. Proc., 6, 58-62 (1941).
Sedletsky, 1. D., Kolloidno-Despersnaja Mineralogia, Izvesl. Akad. Nauk SSSR (1945).
Stresse, H., and U. Hofmann, Synthese von Magnesium-Silikatgelen mit zweidimen-
sional regelmassiger Struktur, Z. anorg. u. allgem. Chern., 247, 65-95 (1941).
Winkler, H. G. F., Kristallstruktur von Montmorillonit, Z. Krist., 105, 291-303
(1943).
Illite
Andreatta, C., ANew Type of Illite-Hydromica in a Hydrothermal Deposit, M ineralog.
Soc. Gr. Bri(ain Clay Mineral Bull. 1, pp. 96-99 (1949).
Bannister, F. A., Brammallite (Sodium Illite): a New Mineral from Llandebie, South
Wales, Mineralog. Mag., 26, 304-307 (1943).
Grim, R. Eo, Petrology of the Pennsylvanian Shales and Noncaleareous Underclays
Associated with Illinois Coals, Bull. Am. Cerarn. Soc., 14, 18-25 (1934).
Gruner, J. W., The Structural Relationship of Glauconite and Mica, Am. Mineral.,
20, 699-714 (1935).
Hellman, N. N., D. G. Aldrich, and M. L. Jackson, Further Note on an X-ray Diffrac-
tion Procedure fOJ: the Positive Differentiation of Montmorillonite and Hydrous
Mica, Soil Sci. Soc. Am. Proc., 7, 197-199 (1943).
Hofmann, U., J. Endell, and E. Maegdefrau, Specific Identity of the Clay Mineral,
Sarospatite, Ber. deut. keram. Ges., 24, 339-344 (1942).
Hofmann, U., K. Endell, and D. Wilm, Riintgenographisehe l:ntersuchungen an
Tonen, Angew. Chern., 47, 539-547 (1934).
Maegdefrau, E., and U. Hofmann, Glimmerartige Mineralien als Tonsubstanzen, Z.
Krist., 90, 31-59 (1937).
Nagelschmidt, G., X-ray Investigation of Clays, III, Z. Krist., 97, 514-521 (1937).
Nagelschmidt, G., Roentgen-ray Diffraction Experiments on Illite and Bravaisite,
Mineralog. Mag., 27, 59-61 (1944).
Chlorite
Brindley, G. W., B. M. Oughton, and K. Robinson, Polymorphism of the Chlorites,
I, Ordered Structures, Acta Crystallog., 3,408-416 (1950).
Mauguin, C. H., X-ray Studies of Chlorites, Compt. rend., 186, 1852-1855 (1928).
Mauguin, C. H., Unit Cell of Chlorites, Bull. soc. fran~. mineral., 53,279-300 (1930).
Oreel, J., Chemical Composition of the Chlorites, Bull. soc. fran!;. mineral., 50, 70-454
(1927).
Serdyuchenko, D. P., Chemical Constitution of the Chlorites, Doklady Akad. N auk
SSSR, 60, 1561-1564 (1948).
Stephen, I., and D. M. C. MacEwan, Swelling Chlorite, Mineralog. Soc. Gr. Britain
Clay Mineral Btdl. 4 (1949).
Vermiculite
Barshad, L, Vermiculite in Soil Clays, Soil Sci., 61, 423-442 (1946).
Caillere, S., and S. Henin, Transformation of ,Minerals of the Montmorillonite Family
into 10 A Micas, Mineralog. Mag., 28, 606-611 (1949).
Caillere, S., and S. Henin, Experimental Formation of Chlorites from Montmoril-
, nite, Mineralog. Mag., 28, 612-620 (1949).
K "ntzer, V. P., Vermiculite Structure, lJtern. soc. russe mineral., ser. 2, 63,464-480
(1934).
MacEwan, D. M. C., Chlorites and Vermiculites in Soil Clays, Verre 8ilicate8 ind., 13,
41-46 (1948).
Wager, L. R., Hydrobiotite, Proc. Yorkshire Geol. Soc., 25,366-372 (1944).
Walker, G. F., The Decomposition of Biotite in Soils, Mineralog., Mag., 28, 693-703
(1949).
Sepiolite-Attapulgite-Palygorskite
Caillere, S., and S. Renin, New Observations on the Mineral from Allevard-a Type
of Very Aluminous Palygorskite, Compt. rend., 222, 328-329 (1946).
Caillere, S., and S. Henin, Application de l'analyse thermique diff6rentielle i11'etude
des argiles des sols, Ann. agron., 17, Part 1, pp. 23-72 (1947).
Caldwell, O. G., and C. E. Marshall, A Study of Some Chemical and Physical Proper-
ties of the Clay Minerals Nontronite, Attapulgite and Saponite, Call. Agr. Univ.
Missouri Research Bull. 354 (1942).
De Lapparent, J., Formula and Structure of Attapulgite, Compt. rend., 202, 1728-1731
(1935).
De Lapparent, J., The Constituents of Fuller's Earths, Ann. office nail. combustibles
liquides, pp. 863-943 (1936).
De Lapparent, J., A propos de l'attapulgite, Z. Krist., 97, 237-248 (1937).
De Lapparent, J., Formules structurales et classification des argiles, Z. Krist., 98,
233-258 (1938). '
Deribere, M., Le Mineral papyrace de la table (Savoie), Bull. soc. franr;. mineral., 66,
222-237 (1943). .",-
Kaufman, A. J., Fibrous Sepiolites from Yavapai County, Arizona, Am. Mineral., 28,
512-520 (1943). ~
Longchambon, II., and G. Migeon, Sur la definition des sepiolites, Campt. rend., 200,.
1946-1949 (1935). _
Longchambon, H., and G. l\1igeon, Sur les sepiolites, Campi. rend., 204, 431-434
(1936). /'
Longchambon, H., Sur les caracteristiques des palygorskites, Campt. rend., 204,55-58
(1937). /"
Marshall, C. E., and O. G. Caldwell, The Colloid Chemistry of the Clay Mineral
Attapulgite, J. Phys. Chern., 51, 311-320 (1947).
Urbain, P., Classification of Hydrated Aluminum Silicates, Compt. rend. soc. geol.
France, pp. 147-149 (1936).
CHAPTER 5
X-ray-diffraction Data
GENERAL STATEMENT
In this chapter X-ray-diffraction data are given for the clay minerals,
together with a discussion of the application of these data in their
identification. The diffraction characteristics of many of the clay
minerals have considerable similarity so that identifications based on
diffraction data are frequently complicated.
X-ray-diffraction data are given in this chapter and the tables to follow
in kX units, i.e., they are based on comparisons with the values for calcite.
Recent determinations of Avogadro's number, on which the calculated
value of the absolute calcite spacing depends, show that values expressed
in kX units are converted to angstroms by multiplying by 1.00202. 1- 3
The difference between kX and angstrom units is significant only when
measurements are accurate to 0.5 per cent.
The diffraction data are given in kX units since the most complete data
available are in this unit in the recent volume" X-ray Identification and
Structure of the Clay Minerals" edited by G. W. Brindley and published
as a monograph by the Mineralogical Society of Great Britain. The
diffraction data here are taken from this volume, for which permission was
kindly granted.
Clay minerals exist for the most part only in very fine particles, and
some form of the powder method usually must be used to obtain the
diffraction data. Because of certain inherent characteristics of the clay
minerals, special cameras and special techniques are frequently required
for clay-mineral work. Cameras and techniques entirely adequate for
other materials have been found to be unsatisfactory for clay-mineral
investigations. Brindley 4 has recently considered X-ray methods as
applied to clay-mineral researches, and the following discussion is largely
1 Bearden, J. A., Evaluation of Nand e by X-rays, J. Applied Phys., 12, 395-403

(1941).
2 Siegbahn, J., X-ray Wavelengths, Natnre, 151,502 (1943).
3 Bragg, W. L., Conversion of kX to A units, J. Sci. Instruments, 24, 27 (1947).

4 Brindley, G. W., Experimental Methods, "X-ray Identification and Structure of
the Clay Minerals," Chap. I, pp. 1-31, Mineralogical Society of Great Brit.ain M('""-
graph (1951).
84
X-ray-diffraction Data 85
taken from his work. For a consideration of X-ray methods in general,
the standard works of Bragg,5,6 Buerger,7 James,s Wilson,9 et al. IO should
be consulted.
In the clay minerals there are few, if any, reflections obtained from
spacings less than about 1 kX, and therefore with copper Ka radiation, it
is unnecessary to extend either the film or the camera to values of 2()
greater than about 90 0 or 100 0 The specimen, therefore, can be brought
nearer the focus of the X-ray tube and the exposure time shortened by
omitting that part of the usual cylindrical camera corresponding to
90 < 2() < 180 0
The first-order basal reflections of the clay minerals are frequently
their most important reflections. Such reflections may correspond to
spacings as large as 20 to 30 A, and it is to obtain clear recordings of such
low-angle reflections that special cameras and special techniques have
been devised. Further, it may be necessary to study the exact shape or
profile of such a reflection, and for this purpose cameras capable
of recording spacings of the order of 50 A are requiredy,12 Cameras
with diameters of 9 to 20 em are commonly employed in clay-mineral
work.
Most of the clay minerals occur in flake-shaped units, and consequently
when a powder is packed, an aggregate orientation of the flakes develops
readily. Thus, care must be taken in the preparation and mounting of
clay samples in the camera, or enhanced basal reflections will be obtained
from an aggregate arrangement. For some types of work where basal
reflections are particularly desirable, as in the study of mixed-layered
structures, advantage may be taken of this tendency to develop oriented
aggregates. Bradley, Grim, and Clark l3 have described a method of
forming well-oriented aggregates and using them in X-ray-diffraction
studies.
In the montmorillonite and vermiculite clay minerals the basal spacings
vary with the humidity of the atmosphere to which they are subjected.
'Bragg, W. L., "The Crystalline State," G. Bell, London (1933).
6 Bragg, W. L., "Atomic Structure of Minerals," Oxford, New York (1937).
1 Buerger, M. L., "X-ray Crystallography," Wiley, New York (1942).
8 James, R. W., Optical Principles of the Diffraction of X-rays, "The Crystalline

State," vol. 2, G. Bell, London (1948).
9 Wilson, A. J. C., "X-ray Optics," Methuen, London (1949).
10 Bragg, W. H., M. Laue, and C. Hermann, Eds., "International Tables for the
Determination of Crystal Structure," rev. ed., 1. Bell, London (1944).
11 MacEwan, D. M. C., The Identification and Estimation of the Montmorillonite
Group of Minerals, J. Soc. Chem. Ind. (London), 65, 298-306 (1946).
12 MacEwan, D. M. C., Some Notes on the Recording and Interpretation of X-Ray
Diagrams of Soil Clays, J. Soil Sci., 1,90-103 (1949).
13 Bradley, W. F., R. E. Grim, and G. F. Clark, X-Ray Study of Montmorillonite,
Z. Krist., 97, 216-222 (1937).

86 Clay Mineralogy
It may, therefore, be necessary to control the humidity of the atmosphere
within the camera. Recently techniques using clay-mineral-organic
complexes 14 ,15 have been devised for these minerals (see pages 93 and 99)
which are superior for identification purposes.
Imperfections of crystals affect their diffraction characteristics, and
they are particularly important for the clay minerals. The imperfections
may be of several kinds, as follows:
Because of the highly symmetrical arrangement of the atoms in the
various layers and the relatively weak binding force between them, the
layers can be displaced with respect to one another. The geometrical
relation between adjacent sheets of atoms in contact may remain the
same, although the relation between more distant atoms is changed.
The result of such imperfections is the absence of certain types of reflec-
tions. Two-dimensional diffraction effects become the major feature
when the layers are displaced randomly in two directions. If, for
example, displacements occur along both the a and b axes, the only hkl
reflections which are possible are the basal (OOl) reflections. Two-
dimensional diffraction effects then arise from the regularity within
the layers and require only two indices, namely, hk. They give
broad diffraction bands rather than sharp lines and are generally easily
recognized.
Another type of imperfection is found in the mixed-layered structures
when layers of different types are interstratified. Because of the fact
that many claY-illlineral layers are very similar, being composed of silica
tetrahedral sheets and closely packed layers of oxygen and hydroxyls,
they can interstratify to build structures as stable or almost as stable as
structures composed of a single layer. When the alternation of different
layers is regular, the result is equivalent to a single unit of larger unit-
cell dimensions, e.g., chlorite. When the alternation of different layers
is irregular, new diffraction effects arise, in particular a nonintegral series
of reflections from the basal planes (see pages 102-104).
The shape and size of the crystal particles exposed to X ray influence
the diffraction effects. Thus, if the crystal particles have a distinct
shape, e.g., plate-like or rod-like, certain reflections may be broader than
others. Theoretically a study of the diffuseness of the reflections and
their profile should provide information on the shape and size of the
crystals. However, because the lattice imperfections of the clay minerals
cause somewhat similar diffraction effects, only qualitative results could
be hoped for at best, and much more useful results on size and shape
are likely to be revealed by the electron microscope.
14MacEwan, D. M. C., Identification of Montmorillonite, Nature, 154,577 (1944).
15Bradley, W. F., Molecular Associations between Montmorillonite and Some Poly-
functional Organic Liquids, J. Am. Chern. Soc., 67, 975-981 (1945).
KAOLINITE AND HALLOYSITE MINERALS
Diffraction data for kaolinite and halloysite (low-hydration type) are

given in Table 4 from Brindley and Robinson 16- 18 taken from Brindley's
text. 19 Diffraction data for nacrite and dickite are given in Table 5 from
Gruner 20 ,21 and Brindley,19 also taken from Brindley's text.19 In general
the determination of the kaolinite group by X-ray diffraction is simple,
but the identification of the particular members of the group may be more
difficult. The prominent basal reflections at about 7.14 kX (00l) and
3.57 kX (002) are usually adequate for identification.
The chloritic clay minerals may be confused with the kaolinite minerals,
especially if the experimental technique is not adequate to reveal the
14-kX chlorite reflection. However, usually the third order at 4.7 kX
is seen to indicate the presence of chlorite. Chlorites rich in iron fre-
quently give weak first- and third-order reflections, and differentiation
from kaolinite is particularly difficult. In such cases only a little chlorite
may be suggested by the diffraction data, and further tests are necessary
to estimate the abundance of the mineral accurately. Such a further test
may be treatment with warm HCl to take advantage of the greater solu-
bility of chlorite in this acid. Also kaolinite on heating to 600C tends
to lose its crystalline eharacter, whereas chlorite at this temperature is
only partially dehydrated, causing increased intensity of the 14-kX
reflection. Frequently other analytical methods, such as differential
thermal analyses, are necessary to supplement the X-ray data to estab-
lish the presence of chlorite.
Nacrite, dickite, and kaolinite can be differentiated from each other on
the basis of X-ray-diffraction data if the technique used provides ade-
quate resolution (Table 5).
Diffraction characteristics of well-crystallized versus poorly crystallized
kaolinite are given in Fig. 7 of Chap. 4. The transition from well-
crystallized to poorly crystallized kaolinite is shown by a broadening and
16 Brindley, G. VV., and K. Robinson, The Structure of Kaolinite, Mineralog. Mag.,

27,242-253 (1946).
17 Brindley, G. W., and K. Robinson, Randomness in the Structure of the Kaolinitic
18 Brindley, G. W., and K. Robinson, X-ray Studies of Halloysite and Metahalloy-
site, I, The Structure of Metahalloysite, Mineralog. Mag., 28, 393-407 (1948).

19 Brindley, G. W., The Kaolin Minerals, "X-ray Identification and Structure of
the Clay Minerals," Chap. II, pp. 32-75, Mineralogical Society of Great Britain
Monograph (1951).
20 Gruner, J. W., Crystal Structure of Dickite, Z. Krist., 83,394-404 (1932).
21 Gruner, J. W., The Crystal Structure 0{ Nacrite and a Comparison of Certain
Optical Properties of the Kaolin Group with Its Structure, Z. Krist., 85, 345-354
(1933).
88 Clay Mineralogy
TABLE 4. LATTICE SPACINGS, INDICES, AND INTENSITIES IN POWDER DIAGRAMS OF
WELL-CRYSTALLIZED KAOLINITE, POORLY CRYSTALLIZED KAOLINITE, AND THE
LOW-HYDRATION VARIETY OF HALLOYSITE
(From Brindley and Robinson 16-18 taken from Brindley's text l9 )
Well-crystallized Poorly crystallized Halloysite, low-

kaolinite kaolinite hydration variety
(001)
I d hkl hkl d I or hk d I
10+ 7.15 001-- 001 7.15 1 0 - 001 7.2-7.5 8

4 4.45. 020 --02, 11 4.45 5 8 --..... 02,11
110 __ 110* 4.422 10+
6 4.35 4.36 2-
6 4.17 111 ----- 11 1* 4.145 2
3 4.12 lIT ___.."....
4 3.837 021
2 3.734 021
10+ 3.566 002 ----002 3.57 10 - 002 3.578 8
4 3.365 111
2 3.138 112
2 3.091 112
2 2.748 022
rOT 201}
7~
2.55.
8 2.553 130/130
130
4 2.521 {l~T~ _ 13,20 2.559 7
1I2/131}
131 (
9 2.486 112 2.50 7 (2.48) 2
{z00_200 _______
7 2.374 003 -=-003 2.375 7 - 003 2.403 2
12Q2_202} (2.33) 2
2.325 8b
iO
9
2.331
2.284
113/
131-131
{ 113
131
~
1 2.243 132,040
-- - - - --------- 04,22 2.218 1
3 2.182 132, 220 ------~
2 2.127 023,041
1- 2.057 222
7 1.985 {2132/132
Q3 --- 203} 1.977 41
4 1.935 {132
221
2 1.892 133
1 1.865 042
4 1.835 133,202,223
1 1.805 114,223
5 1.778 004 ~004 1.7B5 3 --004 1.800
..
X-fay-diffraction Data 89
TABLE 4. LATTICE SPACINGS, INDICES, AND INTENSITIES IN POWDER DIAGRAMS
OF WELL-CRYSTALLIZED KAOLINITE, POORLY CRYSTALLIZED KAOLINI'l'E, AND THE
LOW-HYDRATION VARIETY OF HALLOYSITE. (Continued)
Well-crystallized Poorly-crystallized Halloysite, low-

kaolinite kaolinite hydration variety
(001)
I d hkl hid d I or hk d I
1.704 222
2 150,241,311 - - - - - - - - - ~
1.682 { 3IT,222
rIO, 151 -----~
8
1.659 (!!f =l~~~~~~ ___ 2__ j31, 15,24
1.678 5
6 1.616 2~2, 310 -------

151, 242
4 1.581 {i~i',:: ________ _
134-
5 1. 539 203 _ 203 1. 541 1
{ 241, 224
114,313
060 - - 06~}
9 _ 1.486 8 --06,33 1.481 8
1.486 331 _ 331
{
331--
223,115
2 1.464 061
!3~~332}
_332_~
(
1.457 2
_______ 061
061
4 1.449 { 153,152,330
115,223,330 - - - - - - - -
4 1. 426 005 005 1. 428 2
Values obtained with filtered Cu Ka radiations; camera radius 100 mill.

* Weak lines not always observed; may be due to an impurity.
b = broad lines.
weakening of the reflections with the complete elimination of the weaker

ones. There is a tendency for adjacent reflections to fuse into one. The
basal reflection also increases from 7.14 to as much as 7.20 kX. The
group of lines from (020) (d = 4.46) to (002) (d = 3.57) particularly
reflect the change to lower crystallinity. In this region the clearly
resolved doublet (111) and (111) of well-crystallized material gives way
to a single band, affording good evidence for the decrease in crystallinity.
It is frequently difficult to determine precisely the crystallinity of very
90 Clay Mineralogy
TABLE 5. LATTICE SPACINGS AND ESTIMATED INTENSITIES IN POWDER DIAGRAMS OF
NACRITE AND DlCKITE
(From Gruner 20 ,21 and Brindley and Robinson,16 taken from Brindley's textl 9 )
Nacrite* Dickite
Brindley and Brindley and

Gruner'l Gruner 20t
Robinson 16t Robinson 16 t
d I d I d I d T
-~-- --_.--
7.08 10 7.12 10 7.15 10 7.16 10
4.40 8 4.43 4 4.432 6 4.443 7
4.13 3 4.361 6 4.366 7
3.93 2b 4.255 4 4.265 5
3.58 9 4.15 3 4.126 6 4.121 I 7
3.44 1b 3.95 5 3.962 2 3.\154 3
3.04 Jb 3.79 1-2 3.783 [> 3.790 8
2.537 1b 3 ..578 10 3.576 10+ 3.575 10-
2.418 10 3.4:~:J ),2-1 3.430 4 3.425 5
I I
2.3\13 1-2 3.257 1 3.247 I
2
2.319 I? I
3.096 2 3.097 I 2
2.263 >-2- 1b 2.933 3 2.931 2
2.069 Ivb 2,786 2 2.790 2
1.902 2-3vb 2.634 1-2 I 2.647 1-2 2.651 72
1.795 1 2.566 3-4 I 2.560 4
1. 744 ;J,i-1b 2.514 4 2.518 1
1.675 2? 2,498 6 2.501 6
1.616 2b 2.384 5 2,381 3 2.383 3
1.583 1 2.328 5 2.315 7 2.318 8
1.486 8 2.180 72 d 2.206 2 2.207 2
1.455 4b 2.099 }-2 2.081 1 2.09\1 72
1.434 72-1 1.976 4
1.358 1-2vb 1.895 72
1.314 }-2 1.859 1
1.284 1 1.792 3
1.263 3 1.718 72
1.230 2vb 1.649 4
1.208 }-2-1 1.555 3 \
1.195 1 1.490
1.455
5
2 I~
1.431 2
1.391 1
1.374 1 I
I
I
1.317 4
i
----_.-
* Nacrite; unfiltered Cu K radiations, camera radius 57.3 mm.
t Dickite; unfiltered Fe K radiations, camera radius 57.3 mm.
t Dickite; filtered Cll Ka radiations, camera radius 100 mm .
. b = broad lines; vb = very broad lines.
,
fine kaolinite, since the small size would also cause a broadening and
weakening of reflections.
In the case of halloysite (2H 20) the broadening of reflections, the
development of bands replacing adjacent lines, and the elimination of
weak reflections is carried to the stage (see Fig. 7, Chap. 4) where differ-
entiation is usually not difficult.
Halloysite (4H20) gives a basal reflection at about 10.1 kX, a second-
order reflection at about 5 kX, which is usually unobservable, and a
third-order reflection almost coinciding with the second order for halloy-
site (2H 20). Except for these few basal reflections, the patterns of the
two forms of halloysite consist of identical bands. Brindley 19 and Brown
and MacEwan 22 have shown in detail the diffraction effect resulting from
mixed layering of 4H 2 0 and 2H 2 0 forms when the interlayer water is
gradually eliminated. The 1O.1-kX spacing of halloysite (4H 20) occurs
at about the same place as the (001) basal spacing of the illites, but other
reflections usually serve to differentiate these minerals easily. Also the
reflection can be shifted in the halloysite mineral by drying at low tem-
perature, and the halloysite will adsorb certain organic molecules (see
Chap. 10), so that auxiliary procedures are available to identify the
mineral.
MONTMORILLONITE MINERALS
The diffractions shown by powder diagrams of the montmorillonite

minerals can be placed in two categories. One class consists of basal
reflections, which vary with the state of hydration of the mineral, i.e.,
with the thickness and regularity of the water layers between the silicate
sheets. The c-axis spacing, the diffuseness of the reflections, and the
number of orders shown vary from sample to sample depending on the
thickness of the water layers and their regularity, which factors in turn
are dependent on the exchangeable cation present and the conditions,
e.g., water-vapor pressure, under which the sample has been prepared.
Because of the variability of the basal spacings no table can be given for
them.
The second class of diffractions consists of general diffractions which
are characteristic of the structure of the montmorillonite layers them-
selves and are not dependent on the interlayer hydration. These are
hk bands, which in general are the same for all montmorillonites. The
differences shown by various montmorillonites are in details of spacings
22 Brown, G., and D. M. C. MacEwan, X-ray Diffraction by Structures with
Random Interstratification, "X-ray Identification and Structure of the Clay Min-

erals," Chap. XI, pp. 266-284, Mineralogical Society of Great Britain Monograph
(1951).
92 Clay Mineralogy
and relative intensities. Table 6, after MacEwan,23 gives data for the
general diffractions of some montmorillonite minerals.
TABLE 6. DATA FOR GENERAL DIFFRACTIONS OF THE MONTMORILLONITE MINERALS
(After MacEwan 23 )
Montmorillonite Trioctahedral
(Wyoming Hectorite soil mont-
Indices d(calc) bentonite) morillonite
d I S d I S d I
_- - - ~-- -- -- --
11,02 4.49 4.60 vs vbr 4.44 ms vbr 4.49 ms
4.42 mw
(3.03) (v)
13,20 2.59 2.55 s vbr 2.60} 8 br 2.64 mwbr
2.50
22,04 2.25 2.217 v10 br 2.294 v
31,15,24 1.703 1.689 m br 1.714 vw br 1. 736} v
1.679
33,06 1.500 1.489 V8 vsh 1.511 s vsh 1.533 m
(1.480) (v)
26,40 1.298 1.286 m br 1.302 ms sh 1.323 v
35, 17, 42 1.249 1.241 m sh 1.257 v10 sh
08,44 1.125 1.118 v10 sh
37,28,51 1.034 1.027 vw br 1.040 v10 sh
19,53,46 0.982 0.9745 10 br 0.986 mw br
55 0.900
39, 60 0.8665 0.8625 10 sh
-- --
0.873
~--
__
mw sh
-- ._--
ao 5.20 5.16 5.23 5.31
bo 9.00 8.93 9.07 9.20
S = shape of band; V8 = very strong; s = strong; m = medium; 10 = weak;

vw '= very weak; vvw = very, very weak; v = just visible; vbr = very broad; br
= broad; 8h = sharp; vsh = very sharp.
The ao and b o parameters are deduced from measurements and adjusted to make
b o = ao V3.
The second column gives the calculated edges of the bands with the ao and bo axes
given at the bottom of the column. ~
"'here two or more spacings are given against a particular index, or grou of indices,
it is to be understood that they probably represent subsidiary maxima or apparent
edges of bands (usually rather uncertain, except for the innermost on). Spacings
enclosed in parentheses are uncertain and may be due to impurities in some cases.
By using oriented aggregates, as MacEwan 23 has emphasized, patterns

which contain either the (OOl) or the hk lines can be obtained by suitably
mounting the specimen in front of the X-ray beam.
23 MacEwan, D. M. C., The Montmorillonite Minerals, "X-ray Identification and
As will be shown elsewhere (see Chap. 10), montmorillonite possesses
the property of adsorbing certain organic molecules between the indi-
vidual silicate layers with a consequent shift in the c dimension depending
on the size and configuration of the organic molecule. Such resulting
montmorillonite-organic complexes, as Bradleylo showed, have a high
degree of regularity, yielding sharp (OOl) lines of many orders. For
example, treatment of a sample of montmorillonite with ethylene glycol
provides a sharp (001) reflection at 17 kX. Treatment with organic
compounds permits the detection of small amounts of the mineral which
would otherwise be missed in complex mixtures.
ILLITE MINERALS
Diffraction data for muscovite and biotite 24 and three illites 25- 27 are
given in Table 7, from Grim, Bradley, and Brown.28 The illite minerals
can usually be identified on the basis of X-ray diffraction by their (OOl)
spacings with the first order at about 10 kX. Some difficulty may be
encountered with completely collapsed montmorillonite, but the slightly
lower (001) reflection for it and the characteristics of the remainder of the
pattern are generally adequate to distinguish it from illite.
Differentiation of the polymorphic forms of mica by X-ray diffraction
is not always an easy matter. Grim, Bradley, and Brown 28 have shown
that the reflections in the region between 4.4 and 2.6 kX can usually be
used to distinguish micas of one-, two-, and three-layer unit cells (see
Table 8). This region includes only (02l) and (lll) reflections except for
one strong basal order at 3.33 kX.
Grim, Bradley, and Brown,28 Walker,27 and others29 have pointed out
that the position of the (060) reflection and the intensity of the second-
order basal reflection can usually be used to distinguish between diocta-
hedral and trioctahedral micas. For the dioctahedral forms, (060) is
close to 1.50 kX and (002) is strong. For the trioctahedral forms, (060)
24 Nagelschmidt, G., X-Ray Identification of Clays, III, Z. Krist., 97, 514-521
(1937).
25 Grim, R. E., R. H. Bray, and 'V. F. Bradley, The Mica in Argillaceous Sediments,
Am. Mineral., 22, 813-829 (1937).
26 Mackenzie, R. C., G. F. Walker, and R. Hart, Illite in Decomposed Granite at
Ballatcr, Aberdeenshire, Mineralog. Mag., 28, 704-714 (1949).
27 Walker, G. F., Trioctahedral Minerals in Soil-Clays of Northeast Scotland,
Mineralog. Mag., 29, 72-84 (1950).

28 Grim, R. E., W. G. Bradley, and G. Brown, The Mica Clay Minerals, "X-ray
ical Society of Great Britain Monograph (1951).
29 MacEwan, D. M. C., Les Mineraux argileux de quelques sols ecossais, Verre
silicates ind., 12, 3-7 (1947).

94 Clay Mineralogy
TABLE 7. LATTICE SPACINGS AND ESTIMATED INTENSITIES IN POWDER DIAGRAMS OF
MUSCOVITE, BIOTITE, AND V ARlO US ILLITES
(From Nagelschmidt,'4 Grim et ai.,'5 Mackenzie et ai.' 6 and Walker,27 taken from
Grim et ai. 2B )
Muscovite * Biotite * Illitet TIlitet Illite
d I d I d I d I d I
--- --- --- ---
9.98 8 10.1 V8 9.98 8 9.9 8 10.0 10
5.0 8 4.97 w 4.9 m 4.94 2
4.49 s 4.58 1.0 4.47 s 4.45 vs 4.47 9
4.28 1.0
4.11 v1.o 4.10 1.0
3.91 1.0 3.87 m
3.73 1.0 3.7 v1.o 3.64 m1.O 3.68 2b
3.5 m 3.4 v1.o
3.33 vs 3.36 vs 3.31 m 3.35 V8 3.32 9
3.2 m 3.15 v1.o 3.2 v1.o 3.16 0.5
3.00 m 2.98 1.0 3.09 m1.Od
2.88 m 2.91 v1.o 2.84 v1.o 2.85 md 2.86 1
2.0 m
2.57 vs 2.65 s 2.56 s 2.56 V8 2.60 6
2.51 1.0 2.50 1
2.475 1.Ob 2.45 s 2.44 1.0 2.45 m1.O 2.41 4
2.385 m 2.38 m 2.39 m
2.28} 2.282 v1.o 2.24 m 2.235 m1.O
1.0
2.19 2.183 s 2.18 1.0 2.158 2
2.134 s 2.11 1.0 2.14 m
1.995 vs 2.002 s 1.98 m 1.988 m 1.982 1
1.911 v1.o 1. 94 1.0
1. 73 v1.O 1.752 v1.O 1.689 } 3
1.651 sb 1.672 s 1.65 1.0 1.647 md 1.639
1.546 v1.O 1 ..551} 1.64 m
s
1.523 vw 1.527 1.53 6
1..50 s 1.50 s 1.497 8
1.45 v1.O 1.474 v1.O 1.446 0.5
1.421 vw 1.428 v1.o
1.:~77} m 1.361 m
1.356
1 ..'334 1.0 1.34 v1.o 1.342 m1.Od
1.31 1.0 1.320 0.5
1.296 m 1.29 m 1.294 m
1.266 1.0
1.247 m 1.24 1.0 1.243 m1.O
Valnes for biotite and Walker's illite with Fe Ka. radiations; other samples with
Cu Ka. radiations.
* Muscovite and biotite from Nagclschmidt."4
t Illite, Gilead, Illinois, from Grim, Bray, and Bradley."5
tIllite, Ballater, Scotland, from Mackenzie, 'Walker, and Hart.26
Illite, Carden Wood, Scotland (tri,!ctahedral), from WalkerY
TABLE 8. X-RAY-DIFFRACTION DATA FOR DISTINGUISHING POLYMORPHIC FORMS OF
MICA
(From Grim, Bradley, and Brown 28 )
Two-layer Single-layer Three-layer hcx-

Muscovite
phlogopite lepidolite agonal zwinnwaldite
Indices d I d I Indices d I Indices I_Ii ---

I
III 4.29 w
022 4.11 w 4.17 021 4.14 013
112 3.95 vw
113 3.87 m 3.95 111 014 3.91 w
023 3.72 m 3.79 w
113 3.55 vw 3.63 vw 112 3.60 8 015 3.65 m
114 :3.475 m 3.53 w
024 3.32 * 3.39 * 022 3.34 t 016 3.39 rn
114 3.20 m-8 3.27 w
115 3.10 vw 3.15 112 3.10 8 017 3.15 8
025 2.98 8 3.03 w-rn
115 2.86 m 2.91 vw 113 2.93 rn 018 2.90 8
116 2.78 m 2.83 w
026 2.68 2.73 vw 023 2.71 vw 019 2.70 m
* Coincident with (006); t coincident with (003).

lies between 1.525 and 1.535 kX and (002) is weak or absent. There are
some exceptions to this rule, e.g., glauconite, and it must be used with
caution. The (060) line is useful in identifying other clay minerals; for
kaolinite it is at 1.485 kX and for chlorites at 1.53 to 1.56 kX. This line
cannot be used to separate illites and montmorillonites, and care must be
taken to avoid confusion with quartz, which can give a reflection at
1.53 kX, if present in abundance of more than about 10 per cent.
Grim, Bradley, and Brown 28 have attempted a correlation of the effect
of major substitutions in the mica lattice and the intensities of the (OOl)
reflections. They have made no computations for substitutions of Li+
and TiH. Substitutions of AP+ for SiH in the tetrahedral sheet make
little or no difference in basal intensities, whereas substitutions in the
octahedral sheet cause large intensity variations. The results of Brown's
analyses are given in Figs. 20 and 21. He points out that a comparison
of calculated and observed values should make possible a guess at the
composition of the octahedral layer, and hence a guess at whether the
mica is dioctahedral or trioctahedral (if Li+ and TiH are absent or very
scant).
Suppose the mica is dioctahedral, then if the second order basal reflection is of
the same order of strength as the first and third order, the mica is a highly alumi-
nous member. If the second order basal reflection is weak, or missing, then the
LIBRARY
-, of Agriculture
. ~ "hra Pradesh
L.J...A-J. 11 'I. ~ .......
96 Clay Mineralogy
300~-----------------------------------------,
200
001
003
100
~~004 005
o 002
2AQ.3+ ItAQ.3,+ 1- Fe3 + IA23i'Fe3 + t A2 3,+ It Fe3 + 2 Fe3+
FIG. 20. Graph showing the change in intensities of the (001) reflections with varying
composition of the octahedral layer for a dioctahedral mica, after Grim, Bradley, and
Brown.2s
500.----------------------------------------,
001
400
300
200
o~==~~d
3Mg 2tMg,tFe 2Mg,IFe I~Mg,ltFe IMg,2Fe kMg,2tFe 3Fe
FIG.21. Graph showing the change in intensities of the (001) reflections with varying
composition of the octahedral layer for a trioctahedral mica, after Grim, Bradley and
Brown.' s
mica is mainly ferriferous in octahedral positions. If the mineral is known to be
of the trioctahedral variety the third and fifth orders should be compared. If the
third and fifth orders are of similar intensity, then the mica is of the magnesium
end of the series, but if the third order is stronger than the fifth order, then the
mica is highly ferriferous.
The intensities shown in Figs. 20 and 21 were calculated from the
formula for a single crystal
1001 = F~o/;e
where FooL = the structure factor for a particular plane (OOl)

Z = Lorentz and polarization factor for a single crystal =
1 + cos 2 2(}
sin 20
o = Bragg reflection angle
The radiation is assumed to be Cu Ka with A = 1.54 kX.
The intensities for the corresponding power diagrams can be calculated
by multiplying each intensity value by the factor
1
Eaingle crysta.l - sin ()
for each plane.
CHLORITE MINERALS
X-ray-diffraction data for several chlorite minerals are given in Table 9

after McMurchy 30 and Brindley and Ali taken from Brindley and
Robinson. 31 The variation from one chlorite to another is to be found
in slight modifications of intensities or lattice spacings due to small
variations of unit-cell dimensions and population of various cation posi-
tions. The identification of the various forms of chlorite is very difficult
and may be impossible unless chemical and optical data are available to
supplement the X-ray analysis. The small size of the clay-mineral
chlorite particles and their less regular crystallinity cause some diffuse-
ness of reflections and the absence of some ordinarily weak reflections.
As a consequence it is possible now to identify a clay mineral only as
belonging to the chlorite group of clay minerals.
Care must be taken in the identification of the chlorite minerals by
X-ray diffraction, particularly to avoid confusion with montmorillonite,
vermiculite, and kaolinite. Most chlorites give a clear sequence of four
or five basal reflections from the 14-kX basal spacing. Chlorites rich in
iron, however, give weak first and third orders and strong second and
30McMurchy, R. C., The Structure of the Chlorites, Z. Krist., 88, 420-432 (1934).
31Brindley, G. W., and K. Robinson, The Chlorite Minerals, "X-ray Identification
and Structure of the Clay Minerals," Chap. VI, pp. 173-198, Mineralogical Society
of Great Britain Monograph (1951).
98 Clay Mineralogy
fourth orders. Such chlorites may be readily confused with kaolinite
unless supplementary techniques are used, as described on page 87.
TABLE 9. DIn'RAcTION DATA FROM POWDER DIAGRAMS OF SOME CFlLORITE

MINERALS
(After McMurchy30 and Brindley and Ali taken from the text by Brindley and
Ro biIlson 31)
Penninitet
Sheridanitc * Sheridanite t Penninitet
after 600C
hkl
I
d I d I d I d I
-- ~--_
001 13.7 7 13.6 6 14.3 6 14.2 10

002 7_04 8 7.01 7 7.17 10 7.07 4
003 4.68 9 4.69 9 4.78 10 4.70 6
020 4.60 2
004 3.509 10 3.518 10 3.585 10 3.527 6
005 2.828 7 2.816 6 2.867 6 2.823 9
131, 202 2.578 7 2.565 1 2.585 3 2.599 2
132,201 2.542 5 2.521 2 2.543 4 2.562 2
132,203 2.430 6 2.421 2 2.444 4 2.453 3
133,202 2.387 3
006 2.370 3 2.348 2
133,204 2.247 4 2.240 2 2.264 2 2.269 3
007 2.021 7 2.012 4 2.049 2
135, 204 1.998 6 1.989 4 2.014 4 2.009 2
135,206 1.881 4 1.868 3 1.892 1 1.882 2
136,205 1.825 4 1.811 3 1.834 3
136,207 1.706 3 1.705 7-2 1.726 72 1.714 3
137,206 1.659 2 1.650 H 1.672 72
137,208 1.562 10 1.556 4 1.579 4 1.564 3
060, 331, 331 1.534 10 1.527 3 1.536 2 1.546 3
062, 331 1.502 2 1.502 1
0010 1.417 3 1.409 3 1.434 1 -
139,208 1.390 10 1.386 5 1.406 3 1.389 2
400 1.319 4 1.312 1 1.320 72
I /
* Sheridanite from Me Murch y 30., unfiltered Cu K radiations' camera radius 57.3
mm.
t Sheridanite, penninite, from Brindley and Ali; filtered Cu Ka radiations; camera
radius 100 mm.
The characteristic 14-!(X spacing may be confused with that of mont-

morillonite or vermiculite even though it is generally sharper in the
chlorites. For chlorites the 14-kX spacing is not changed on moderate
heatings to about 200C, as it is in vermiculite, nor is it changed by
treatment with a polar organic molecule (like glycol), as it is in the
montmorillonites.
VERMICULITE MINERALS
Diffraction data for two vermiculites are given in Table 10 after

Gruner. 32 Diffraction data for another sample not heated and heated to
successively higher temperatures are given in Table 11 after Walker. 33
The strong reflection at 14 kX is characteristic of the mineral, but care
must be used to differentiate vermiculite from chlorite, which also has a
reflection at 14 kX, and from montmorillonite, which may show a reflec-
tion at this position.
Vermiculite can usually be distinguished from chlorite on the basis of
the intensity of the (OOl) reflections: In vermiculite the 14-kX line is the
strongest reflection, and the 7-kX line and subsequent basal orders are
relatively weaker; whereas in chlorite several basal orders may occur of
approximately equal intensity. Vermiculite can be further distinguished
from chlorite by the fact that the basal spacings are more readily shifted
by heating. Also, treatment with certain organic molecules causes a
shift in the basal reflections for vermiculite and not for chlorite.
Possible confusion of kaolinite with minerals giving a 14-kX reflection
have already been considered (see page 87). Care must be taken not to
confuse mixtures of kaolinite and vermiculite with chlorite, and this
usually requires that diffraction data be obtained for samples which have
been heated as well as for the natural sample.
Vermiculites and montmorillonite may both show a 14-kX reflection.
A distinction can usually be made because this reflection changes readily
for slight variations of relative humidity in montmorillonite, whereas the
vermiculites are stable over a wide range of moisture contents. Further,
it appears that vermiculite forms only a single molecular layer of organic
molecules between the mica sheets, whereas montmorillonites may form
multimolecular layers with certain organic molecules.
SEPIOLITE-ATTAPULGITE-PALYGORSKITE MINERALS
Diffraction data for sepiolite, after Longchambon,34 and attapulgite,

after Bradley,35 are given in Tables 12 and 13, respectively. The general
575 (1939).
33 'Walker, G. F., Vermiculites and Some Related Mixed-Layer Minerals, "X-Ray
Identification and Structure of the Clay Minerals," Chap. VII, pp. 199-223, Mineral-
ogical Society of Great Britain Monograph (1951).
34 Longchambon, H., Sur certaines caracteristiques de la sepiolite d' Ampandandrava
et la formule des sepiolites, Bull. 80C. jrant;. mineral., 60, 232-276 (1937).
35 Bradley, W. F., The Structural Scheme of Attapulgite, Am. Mineral., 25, 405-413
(1940).
100 Clay Mineralogy
TABLE 10. DIFFRACTION DATA FOR VERMICULITE FROM POWDER DIAGRAMS
(After Gruner 32 )
Baltimore, Maryland 'Webster, North Carolina

Indices
d I d I
002 13.4 8 13.6 10

004 7.03 1 7.12 5
006 4.75 1
4A5 Ii 4.45 2b
008{3 3.91 1 3.94 Ib
008 3.537 4 3.562 5
0010{3 3.126 1 3.162 0.5i
0010 2.838 3 2.847 Ii
130, 202, 132, 200, 136{3, 204{3 Ii 1
2.6 ) 2.6 )
i vb
132, 204, 134, 202 2.521 2b 2.48 3
136,204 2.366 4 2.386 1b
2.192 1
206, 138 2.129 0.5 2.133 0.5i
2010, 138 2.009 0.5
0014 2.035 1
1310,208 2.002 0.5
131:2,2010 1.833 1
0016 1. 774 1
1.732 0.5 1.739 0.5i
1314, 2012 1.664 2 1.676 2i
1316{3,2018{3 1.589 0.5
0018, 1314, 2016 1.570 0.5
060, 332, 330, 334, 1316, 2014 1.526 4 1.526 6b
332, 336 1.496 1 1.494 2b
1.462 0.5
1316, 2018 1.441 2 1.441 0.5i
0020 1.429 1
338, 3312 1.352 1
1318,2020 1.321 3 1
1.312 1 ivb
0022 1.289 1 1.300)
1.270 1
I
Photographs taken with unfiltered Fe K radiations. Because of the m~thof of
mounting, the basal reflections are somewhat enhanced relatively in intensity.
i = indistinct edge of line; b = broad line; vb = very broad line; {3 = a 'ne pro-
duced by K{3 radiation of iron.
TABLE 1l. DIFFRACTION DATA FOR VERMICULITE FROM 'VEST CHESTER,
PENNSYLVANIA, HEATED AND UNHEATED
(After Walker SS )
Heated to 200C Heated to 500C Heated to 700C

Unheated I for 2 hr for 24 hr for 24 hr
d I d I d I d I
~---
14.2 10 14.0 10 14.1 4d

9.2 6d 9.23 8
7.1
4.75 ~
, .!y'
5
4.58 5 4.56 4 4.56 6 4.50 2
4.38} 3
3.75
3.51 6 3.50 4
3.42 5 3.44 6
3.00 10 3.02 10
2.81 4 2.86 2 2.86 4
2.74 1 2.75 2
2.63 5 2.61 4 2.63 8 2 61 8
2.53 3 2.59 2
2.49 0.5 2.47 2 2.46 10 2 46 10
2.373 4b 2.38 2
2.291 0.5 2.29 2
2.187 0.5 2.18 2 2.18 4
2.084 0.5 2.090 2 2.08 2 2 08 4
2.006 1
1.963 0.5 1.96 0.5
1.895 1 1.89 2
1.736) 1 1.73 1 1.73 1 1. 74 1
1.67 0.5
1.649 0.5 1.65 1 1. 64 Id
1.530 6 1.528 6 1.528 8 1.523 6
1.501 0.5
1.438 } 0.5 1.439 2
1.406 0.5
1.360 2d 1.366 2
1. 325} 2 1.321 4 1.319 1
1.311 2
1.285 1
1.270 0.5 1.269 0.5 1.273 0.5
Photographs taken with Fe KOl radiations.

After heating, samples cooled without sealing to prevent rehydration.
Relative intensity estimated visually; d = diffuse; b = broad.
102 Clay Mineralogy
difference in the diffraction patterns of these minerals as compared to
those of other clay minerals makes them readily distinguishable.
Sepiolite shows a characteristic diffraction at 12.15 kX and attapulgite
at 10.48 kX. These spacings do not change with variations in relative
humidity or moderate heating up to about 300C. Also they are not
TABLE 12. LATTICE SPACINGS AND ESTIMATED INTENSITIES OF SEPIOLITE

(After Longchambon 34)
d I
12.15 .15 10
7.6 .10 3
5.05 .105 2
4.60 .10 10
3.82 .02 5
3.60 .02 1
3.40 .02 2
3.22 .02 5
2.61 .02 8
2.45 .02 7
2.28 6
2.09 5
1.97 1
1.88 2
1.71 2
1.60 4
1.56 4
1.52 2
1.38 2
1.36 2
1.31 3
changed by treatment with organic polar molecules, and therefore they

should not be confused with the expanding, layer clay minerals or with
halloysite. They do exhibit a change in diffraction characteristics on
heating to over 300C (see Chap. 9), which can be used as a further means
of identifying them.
MIXED-LAYER STRUCTURES
In the case of mixtures of discrete clay-mineral particles, a diffraction

pattern will contain reflections typical of each component. The char-
acteristics will be no different from that resulting from a mixturlil of clay
minerals and nonclay minerals, for example, when quartz is present in
the clay material.
In the case of a mixture which is a regular interstratific~f layer
clay minerals, the diffraction effect is equivalent to a larger unit cell
which is a multiple of the individual layers. The regularity of the struc-
tures is maintained on a larger scale, and an integral series of (OOl)
reflections is obtained from the larger unit cell. The structure of chlorite
is an example of this type of mixing.
If the mixture is a random interstratification of layer clay minerals, and
only a very few layers of a seqond type are present, the reflections will
differ very little from those of the dominant layer. If, however, the
TABLE 13. X-RAY-DIFFRACTION DATA FOR ATTAPULGITE
(After Bradle y 3.)
Data calculated from proposed structure Experimental data
hkl d I d I
110 10.48 330 10.50 10

200 6.45 17 6.44 6
130 5.44 12 5.42 5
220 5.24 1
040 4 50 66 4.49 8
310 4.18 13 4.18 3
240 3.69 18 3.6\l 5
330
150
3.49
3 47
I;} 3 .~o 3
400 3.23 120 3.23 10
420 3.04 3 3.03 1
350 2.76 2
440 2.62 43 2.61 8
510 2.56 8 2.55 3
530 2.:38 5 2.38 3
080 2.25 1
600 2.15 15 2.15 5
550 2.10 1
480 1.845 1
3\l0 1.815 3 1.82 1
660 1. 75 1
800 1.615 3 1.62 1
680 1.555 17 1.56 3
0120 1.50 40 1.50 5
Data obtained with eu Ka radiations.
second type of layer is present in considerable abundance (> 10% ),

new diffraction effects will arise; in particular, a nonintegral series of
reflections is obtained from the basal planes. Also the layers can con-
tribute to the scattering only as individuals. Prism zone reflections and
the scattering distribution from the bases are to be expected, but special
sequences such as the (lll) and (02l) interference for muscovite, which
depend on a specific mutual orientation of neighboring layers, cannot be
given by a mixed structure. There is, therefore, likely to be no diffrac-
tion distinction between muscovite and biotite in mixed-layer structures.
104 Clay Mineralogy
In mixed-layer structures the basal reflections are composites of
adjacent reflections of the same orders of the different layers and at an
intermediate position between them, or a composite of overlapping
reflections of different orders of the different layers. The position and
intensity of the composite reflections will vary with the relative abun-
dance of the different individual layers. For example, in a random
mixed-layer structure of lO-kX and 15-kX layers, the first observed
reflection will be a composite of (001) of both layers and will have a
spacing intermediate between 10 and 15 kX, the exact position depending
on the relative abundance of the individual layers. A strong reflection
at 5 kX will appear; this will be a composite of (002) of the lO-kX layers
and (003) of the 15-kX layers. Only if the relative amount of the 15-kX
layer is large will a slight reflection be found at about 7.5 kX.~ ~_~ /
In the case of random mixture of a 17.7-kX layer and a 10-kX layer,
where both are present in considerable abundance, the first reflection will
be a composite of (001) of both layers, and the second reflection will be a
composite of (00l) of the lO-kX layer and (002) of the 17.7-kX layer.
The position and relative intensity of the composite peaks, or, more
properly, bands, will vary with the relative abundance of the two layers.
It should be pointed out that the position and intensity of the composite
peaks do not vary with the relative abundance of the constituent layers
in a necessarily straight-line relationship.
The above illustrations will serve to show how a nonintegral series of
reflections arises from a random mixing of layers. The mathematical
theory of X-ray diffraction for such statistical structures has been dis-
cussed by Hendricks and Teller 36 and later by Mering. 37 Brown and
MacEwan 38 have very recently presented a further discussion of this
matter and an extremely useful series of graphs, showing the expected
basal reflections of mixtures of various proportions of a variety of layers.
The reader is referred to these works for further details of mixed-layer
structures.
Brindley, G. W., X-ray Diffraction by Randomly Displaced Layer Lattice Minerals,

"X-ray Identification and Structure of the Clay Minerals," Chap. XII, pp. 285-
305, Mineralogical Society of Great Britain Monograph (1951).
Brindley, G. W., and D. M. C. MacEwan, The Interpretation of Composite X-ray
36 Hendricks, S. B., and E. Teller, X-ray Interference in Partially Ordered Layer
Lattices, J. Chem. Phys., 10, 147-167 (1942).

37 Mering, J., L'Interference des rayons X dans les systemes a stratification dis-
ordonnee, Acta Crystallog., 2, 371-377 (1949).

38 Brown, G., and D. M. C. MacEwan, X-ray Diffraction by Structures with Ran-
dom Interstratifications, "X-ray Identification and Structures of the Clay Minerals,"

Chap. XI, pp. 266-284, Mineralogical Society of Great Britain Monograph (1951). \
,', .
Powder Diagrams, "X-ray Identification and Structure of the Clay Minerals,"
Chap. XIV, pp. 314-320, Mineralogical Society of Great Britain Monograph
(1951).
Brindley, G. W., and J. Mering, Diffraction by Random Layers, Nature, 161, 774
(1948).
Correns, C. W., and M. Mehmel, Ueber den optischen und rontgenographischen
Nachweis von Kaolinit, Halloysit, und Montmorillonit, Z. Krist., 94, 337-348
(1936).
De Lapparent, J., Formules structurales et classification des argiles, Z. Krist., 98,
233-248 (1938).
Favejee, J. C. L., Zur Methodik der rontgenographischen Bodenforschung, Z. Krist.,
100, 425-436 (1939); 101, 259-270 (1939).
Favejee, J. C. L., Quantitative rontgenographische Bodenforschung, Z. Krist., 101,
259-269 (1939).
Favejee, J. C. L., Quantitative X-ray Analysis of Some Dutch Soils, Overdruk uit
Mededul. Landbouwhoogeschool, 43,43-51 (1939).
Guinier, A. J., X-ray Focusing Cameras, Ann. phys., 12, 161-237 (1939).
Guinier, A. J., Imperfections of Crystal Lattices, Proc. Phys. Soc. (London), 57, 310-
324 (1945).
Hendricks, S. B., Continuous Scattering of X-rays from Layer Silicates, Phys. Rev.,
57,448-454 (1940).
Kerr, P. F., Attapulgus Clay, Am. Mineral., 22, 534-550 (1937).
Kerr, P. F., P. K. Hamilton, and R. J. Pill, X-ray Diffraction Measurements, "Analy-
tical Data on Reference Clay Minerals," pp. 1-37, Prelim. Rept. 7, American
Petroleum Institute Project 49, Columbia University, New York (1950).
MacEwan, D. M. C., The Identification and Estimation of the Montmorillonite
Group of Minerals with Special Reference to Soil Clays, J. Soc. Chem. Ind. (Lon-
don), 65, 298-304 (1946).
MacEwan, D. M. C., and J. L. Amoros, Roentgenographic Investigation of Clays,
Anales edafol. fisiol. vegetal., 9, 363-379 (1950).
Mehmel, M., Datensammlung zum Mineralbestimmung mit Rontgenstrahlen,
Fortschr. Mineral. Krist. Petrog., 23,91-118 (1939).
Nagelschmidt, G., Identification of Minerals in Soil Colloids, J. Agr. Sci., 29,477-501
(1939).
Nagelschmidt, G., Aggregate Technique, J. Sci. Instruments, 18, 100-101 (1941).
Nagelschmidt, G., "Mineralogy of Soil Colloids," Imp. Bur. Soil Sci. Tech Com. 42,
Harpenden, England (1944).
Nagelschmidt, G., and D. Hicks, The Mica of Certain Coal Measure Shales in South
Wales, Mineralog. Mag., 26, 297-303 (1943).
Shearer, J., An X-ray Camera for Clay Studies, J. Sci. Instruments, 21,198-200 (1944).
Van der Marel, H. W., Identification of Minerals in Soil Clays by X-ray Diffraction,
Soil Sci., 70, 109-136 (1950).
Von Englehardt, W., Ueber silikatische Tonminerale, Fortschr. Mineral. Krist.
Petrog., 21, 276-340 (1937).
Winkler, H., Zur Kristallstruktur von Montmorillonit, Z. Krist., 107,291-303 (1943).
Zvyagin, B. B., and Z. G. Pinsker, Electronographic Determinations of the Structure
of Montmorillonite, Doklady Akad. Nauk SSSR, 68, 65-67 (1949).
Many of the references given with Chap. 4 on Structure of the Clay Minerals also
contain X-ray-diffraction data.
\ .
CHAPTER 6
Shape and Size-Electron Micrographs
GENERAL STATEMENT
The use of the electron microscope has permitted the precise determina-
tion of the shape of the particles of various ,clay minerals and has shed
light on the range of particle size of the components of clay and on the
degree to which the particle size can be reduced when the clay is worked
mechanically with water. Numerous investi~~ators have publis]jJ.ed
electron micrographs of the clay minerals and discussed electron-micro-
scopic techniques as applied to clay-mineral researches. Particularly
worthy of mention are the contributions of Ardenne,1 Eitel,2 Middel,3
Humbert,4 Shaw,5 Marshall,6 Alexander,7 Moore,8 Bates,9 Kerr,10 and
their various collaborators. For general details of technique, the works
of these authors should be consulted, as well as the texts by Zworykin
et al. 11 and Cosslett. 12
1 Ardenne, M. von, K. Endell, and U. Hofmann, Investigation of the finest Frac-
tion of Bentonite and Clay Soil with the Universal Electron Microscope, Ber. deut.
keram. Ges., 21, 209-227 (1940).
2 Eitel, W., and O. E. Radczewski, On Recognition of Montmorillonite Clay
Minerals in Supermicroscope Pictures, Naturwissenschaften, 28,397-398 (1940).

3 Middel, V., R. Reichmann, and G. S. Kausche, Supermicroscopic Investigation of
the Structure of Bentonites, Wiss. Ver6ifentl. Siemens-Werken, pp. 334-341 (1\l40).
4 Humbert, R. P., and B. Shaw, Studies of Clay Particles with the Electron Micro-
scope, I, Shape of Clay Particles, Soil Sci., 52, 481-487 (1\l41).

6 Shaw, B. T., The Nature of Colloidal Clay as Revealed by the Electron Micro-
scope, J. Phys. Chem., 46,1032-1043 (1\l42).

6 Marshall, C. E., R. P. Humbert, B. T. Shaw, and O. G. Caldwell, Studies of Clay
Particles with the Electron Microscope, II, Beidellite, Nontronite, Magnesium Ben-
tonite, and Attapulgite, Soil Sci., 54, 149-158 (1942).
7 Alexander, L. T., G. T. Faust, S. B. Hendricks, H. Insley, and H. F. McMurdie,
Relationship of the Clay Minerals Halloysite and Endellite, Am. Mineral., 28, 1-18
(1943).
8 Moore, C. A., Some Geological Applications of the Electron Microscope, Proc.
Oklahoma Acad. Sci., 27, 86-90 (1947).

Endellite and Halloysite, Am. Mineral.; 35, 463-484 (1950).

10 Kerr, P. F., P. K. Hamilton, D. W. Davis, T. G. Rochow, F. G. Rowe, and M. L.
Fuller, "Electron Micrographs of Reference Clay Minerals," Prelim. Rept. 6, Amer-

ican Petroleum Institute Project 49, Columbia University, New York (1950).
11 Zworykin, V. K., C. A. Morton, E. G. Remberg, J. Hillier, and A. W. Vorce,
"Electron Optics and the Electron Microscope," \Viley, New York (1945).
12 Cosslett, V. E., "Introduction to Electron Optics," Oxford, New York (1946).
106
Shape and Size-Electron Micrographs 107
Some of the electron micrographs presented herein are reproduced
with permission from the reporVo of Project 49 of the American Petro-
leum Institute. Other electron micrographs published in this vol-
ume were kindly obtained by Prof. T. A. Bates of Pennsylvania
State College expressly for this volume. It is desired to acknowledge
the great kindness and courtesy of Professor Kerr and the American
Petroleum Institute in permitting the reproduction of their micrographs
and of Professor Bates for supplying and permitting the publication of
his work.
In electron microscopy, the image formation is due to the scattering
of electrons as the electron beam passes through the sample. In the
early work, the clay-mineral partieles were mounted directly in the
electron beam without any treatment. In later work the shadow method
of preparing samples has been used frequently in order to show the thick-
ness of the units better and to increase the contrast of the particles against
the background. In the shadow method, mounted specimens are placed
in a metal evaporator, where some metal, frequently uranium or beryl-
lium, is evaporated onto the specimen mounts obliquely. The angle of
deposition of the metal is varied according to the thickness of particles
in the sample. Angles of about 10 are used for thin particles, while
angles up to 30 are used for thicker particles. Direct magnifications of
from 3,000 to 15,000 are obtained, and these are increased from three to
five times by photographi'p enlargement.
Both positive and negative prints are used in this volume. The posi-
tive prints cast white shadows, and the negative prints cast black
shadows.
In electron microscopy the specimens become heated by energy
absorption. Since heating to low temperature (60 to 100C) causes
dehydration reactions of some of the clay minerals, e.g., halloysite,
vermiculite, and montmorillonite, care must be taken to register the
original character of these minerals.
HasV3 has applied the replica technique to the study of clay. In his
method, a replica peel of cellulose nitrate is obtained from a freshly frac-
tured surface of the clay. The peel is obtained by coating the surface of
the clay with cellulose nitrate dissolved in amyl acetate and allowing the
acetate to evaporate. The peel is coated with vaporized beryllium or
aluminum 25 to 50 A thick, after which the cellulose nitrate film is dis-
solved away in amyl acetate. The metal replica is then mounted in the
electron beam. Applying the technique to bentonites, HasV3 has found
evidence of a "thread-like" stacking arrangement of the montmorillonite
particles. Further work is necessary before the apparent results yielded
by the method can be evaluated.
13 Hast, N., Structure of Clays, Nature, 159,354-357 (1947).
108 Clay Mineralogy
DATA FOR THE CLAY MINERALS
Allophane. As would be expected, this amorphous material is found in

particles without any definite and regular shape. The electron micro-
graph of the sample from Indiana (Fig. 22) shows a rounded nodular
appearance which is perhaps characteristic of that developed by amor-
phous material on drying. Occasional holes are to be seen in the rounded
masses; these may be attributed to shrinkage on drying. The sample
from Woolwich, England (Fig. 23) shows particles somewhat rounded,
but in general with an irregular shape.
Kaolinite. Electron micrographs of well-crystallized kaolinite (Fig.
24) show well-formed six-sided flakes, frequently with a prominent
elongation (Fig. 26) in one direction. The elongation is parallel to either
(010) or (110). Certain of the edges of the particles are beveled instead
of being at right angles to the flake surface. Occasionally the particles
appear to be twinned. In poorly crystallized kaolinite, the particles
show less distinct six-sided flakes. The edges of the flakes are som~what
ragged and irregular, and the hexagonal outline is only crudely shown
(Fig. 25).
Anauxite from the type locality (Fig. 27) is in hexagonal flakes, which
are frequently somewhat ragged and irregular.
Electron micrographs of various kaolinite sa'!pples have shown par-
ticles with maximum dimensions of flake surfaces from 0.3 to about
4 microns and thicknesses from 0.05 to about 2 microns. This does not
mean that larger particles of kaolinite are not present-in some clays, since
such larger particles may have been split or otherwise reduced in size in
the preparation of the sample for the electron microscopy. It does mean,
however, that kaolinite particles are easily reduced to this size but are
not reduced to smaller sizes except with difficulty, e.g., after much
mechanical working.
In general, poorly crystallized kaolinite occurs in smaller particles than
the well-crystallized mineral. It does not always follow that kaolinite
with a very small particle size has a low degree of crystallinity. For
example, the kaolinite component of some so-called flint clays is extremely
fine-grained and extremely well crystallized.
Dickite. Dickite occurs in well-formed, six-sided, flake-shaped par-
ticles frequently showing a definite elongation in one direction (Fig. 28).
Samples which have been examined show flake surface dimensions ranging
from about 2.5 to 8 microns and thickness dimensions of 0.07 to 0.25
micron. Dickite particles are often large enough to be studied with the
light microscope. ..
Nacrite. Electron micrographs of a few samples ~acrite show some-
what irregular, rounded, flake-shaped units. In some \articles a crude
.",'~_ ,
", .
FIG. 22. Electron micrograph of allophane, Lawrence County, Indiana, from Kerr.
109
FIG. 34. Electron micrograph of nontronite, Sandy Ridge, North Carolina, from
Bates.
FlG. 35. Electron micrograph of saponite, Death Valley, California, from Ee.IT.
118
FIG. 36. Electron micrograph of hectorite, Hector, California, from Bates.
Some of the magnesium-rich members of the montmorillonite group

appe'a r to be composed of equidimensional flake-shaped units about
like those of the aluminum-rich mOlltmorillonites. The saponite from
Death Valley, California, illustrates such material (Fig. 35). How-
ever, hectorite, the fluorine-bearing magnesium-rich montmorillonite,
is found in thin laths, which often lose their identity when gathered
into aggregates (Fig. 36). The laths attain a length of about 1
micron and a width of about 0.1 micron. The thinnest laths appear
to be only 12 to 18 A thick. Some of the particles are terminated by
edges intersecting the long edges at distinct angles which,' however, ar3
not of uniform size.
Sauconite (Fig. 37) appears to occur in broad laths, which are approxi-
mately 50 A thick. The particles in Fig. 37 show a variety of habits, and
some of them may be other montmorillonite clay minerals.
Illite. The electron micrograph of Fithian illite (Fig. 38) shows small,
poorly defined flakes, commonly grouped together in irregular aggregates.
Some of the flakes have a distinct hexagonal outline. The thinnest
flakes are approximately 30 A thick.' Many of the flakes have a diameter
of 0.1 to 0.3 micron. Some other illites (see, for example, Fig. 39) show
no evidence of flakes with hexagonal ou lines. Such flakes have irregular
but well-defined outlines and are characterized by a uniform thickness.
The electron micrographs of illite resemble those of some montmoril-
FIG. 37. Electron micrograph of sauconite, FriedensYiIle Ridge, Pennsylvania, from
Bates.
38. Electron micrograph of illite, Fithian, Illinois, from Bates.
120
(brammallit ), outh " 'al s, from Bates.
121
122 Clay 1\lineralogy
FIG. 41. Electron micrograpb of attapulgite, Attapulgus, Georgia, from Bates.
lon.ites. The particles of illite are, however, larger and thicker and have
better-defined edges.
Figure 40 shows lath-shaped particles of illite obtained in studies by
Dr. C . E. Weaver 16 of Pennsylvania State College. With the exception
of morphology, this material has aU the attributes of other illites. It
occurs in the -I-micron fraction of a sandstone. The hexagonal fiakes
associated with the laths are probably kaolinite. The significance of an
elongate form of illite has not been determined.
Vermiculite and Chlorite. No information is available regarding the
electron micrography of the vermiculite and chlorite clay minerals. It
appears probable from structural considerations that electron micrographs
of these minerals would be similar to those of the illites, except that
vermiculite might occur in tmnner Bakes.
Attapulgite-Sepiolite-Palygorskite . Attapulgite is shown by electron
micrographs (Fig. 41) to occur in single laths and bundles of laths.
There appears to be no evidence of a tubular form like that found for
halloysite. The laths attain a maximum length of about 4 to 5 microns,
a maximum thickness of 50 to 100 A, and a width only two or three times
the thickness. There appears to be a tendency for two or three laths to
form a parallel bundle or for a great many laths to form thick bundles.
16 Weaver, C. E., manuscript in preparation.
FlO. 42. Electron micrograph of alpha sepiolite, Uttle Cottonwood, Utah, from Kerr.
123
'. .I
,
.&
\ ~
,
FIG. 43 . Electron micrograph of beta sepiolite, Balclissaro, Italy, from Kerr.
FIG. 44. Electron micrograph of mixed-layer metabentonite, Oak Hill, Pennsylvania,

frQll Bates.
124
Shape and Size- Electron i\'firrographs 125
Some sepiolites are elongate and fibrous like attapulgite (Fig. 42), and
this type has been called alpha-sepiolite or parasepiolite by Fersmann. 17
Other sepiolites, called beta-sepiolite by Fersmann, are composed of
extremely small flake-shaped units of irregular outline (Fig. 43) . In
some cases the elongate variety appears to have a tubular character, but
this attribute has not yet been definitely established. The sepiolite
fibers are about the same size as those of attapulgite. The dimensions
of the flaky variety are not well known, but the maximun dimension is
less than about 1 micron. The cause of the difference between the two
types is not known definitely, but it may be that the flake shape results
from some replacement of magnesium by aluminum in the structure.
Mixed-layer Minerals. Adequate data for any generalizations on the
electron micrography of mixed-layer minerals are not available. Figure
44 was obtained from a metabentonite composed of a random inter-
stratification of about 80 per cent illite (nonexpanding, dioctahedral) and
20 per cent expandable clay mineral. The micrograph shows thin flakes
with irregular but well-defined outlines. The flakes are quite thin, and
there is rather little tendency for them to aggregate.
Eitel, W., The Electron Microscope and Its Use in Ceramic Problems, Ber. deut.
keram. Ges., 24,37-53 (1943).
Endell, J ., The Study of the Fine Structure of Clays by Means of the Supermicro-
scope, Keram. Rundschau, 49, 23-26 (1941).
Jackson, M. L ., W. J. Mackie, and R. P. Pennington, Electron Microscope Applica-
t,ions in Soils Research, Soil Sci. Soc. Am. Proc., 22, 57-63 (1947 ).
Kelley, O. J ., and B. T. Shaw, Studies of Clay Particles with the E lectron Micro-
scope, III, Hydrodynamic Considerations in Relation to Shape of Particles, Soil
Sci. Soc. Am. Proc., 7, 58-62 (1942).
Noll, W., The Electron Microscope in the Study of Hydrothermal Silicate Reactions,
Kolloid-Z., 107, 181-190 (1944) .
Prebus, A. F., The Electf(m Microscope, Ohio State Univ. Studies Eng. Expt. Sta. News,
14,7-32 (1942).
Waterman, A. T. , The Electron Microscope, Am. J. Sci., 239, 386-388 (1941).
17 Fersmann, A. E., Parasepiolite, Bull. Acad. Sci. St. Petersburg, ser. 6, 2, 263
(1908).
110 Clay Mineralogy
FIG . 23. Electwn micrograph of allophanc, Woolwich, England, from Kerr .
hexagonal outline can be seen (Fig. 29). In these samples, the flakes are
generally less than about 1 micron in diameter, and the thickness is about
0.025 to 0 .15 micron.
Halloysite . Numerous investigators have shown that the morphology
of halJoysite is quite different from that .of kaolinite, the former being
elongate instead of flake-shaped (Figs. 30 and 31). Pr~marily because of
the work of Bates,9 halloysite is now known to appear in electron micro-
graphs as elongate tubular particles. Convincing evidence.of the tubular
development is furnished by the doughnut forms, believed to be the ends
of tubular particles, that are seen in some electron micrographs. Some-
times it is difficult to determine if the halloysite particles are actually
tubes or flake-shaped units which have curled and rolled up at their
edges. Bates 14 states his belief that the rolling-up process is possible in
very thin flakes where the number of silicate layers- is so few that the
strain produced by the misfit of silicon-oxygen and aluminum-hydroxyl
sheets in each layer is not overcome by the cumulative effect of interlayer
bonds between a large number of successive layers.
Bates has determined the dimensions of the halloysite' tubes from the
electron micrographs of many samples. The outside diameters were
found to range from 0.04 to 0.19 micron with a median value of 0.07
micron. Inside diameters were found to have a median value of 0 .04
14 Bates, T. F., Personal Communication.
FIG. 24. Electron micrograph of kaolinite, l\1acon, Georgia, from Kerr.
III
Ele tron micrograph of kaolinite, Mesa Alta, New Mexico, from Kerr .
Electron micrograph of kaolinite, Zettlitz, Czechoslovakia, from Bate.
112
t--_ _I..:.
M__.__ .;..f
FIG. 27. Electron micrograph of anauxite, Bilin, Czechoslovakia, from Kerr.

Fw. 29. Electron micrograph of nacrite, Brand, Saxony, from Kerr.
114
FIG. 30. Ele troll micrograph of haHoysite, British Guiana, from Bates .
FIG. 31. Electron micrograph of halloysite, Wendover, Utah, from Bates.

115
116 Clay ~ioeralog;r
micron with a range from 0.02 to 0.1 micron. The average wall thickness
was found to be 0.02 micron. Figure 31 shows a high degree of uni-
formity of size among the halloysite particles; this feature is not char-
acteristic of many halloysites from other localities.
Some of the electron micrographs of halloysite (2H 20) show split ends
and partially unrolled tubes. This phenomenon appears to develop as a
consequence of the change from the .J:H 20 to the 2H 20 form of the
mineral.
Montmorillonite. Numerous electron micrographs of montmorillonite
(Figs. 32 and 33) show irregular fluffy masses of extremely small par-
ticles. Frequently, details concerning the individual particles are not
revealed, but in some cases the larger masses appear to be stackings of
flake-shaped units without regular outlines. Some of the individua1
particles appear to be about 0.002 micron thick, from which it can be con-
cluded that at least some montmorillonites break down relatively easily
to flakes approaching unit-cell thickness. Accurate estimations of the
areal dimensions of the flakes are difficult to (lbtain because of their
irregularity but are probably of the order of 10 to 100 times the thickness.
Mathieu-Sicaud, Mering, and Perrin-BonneV have shown that
variations in the exchangeable cation carried by montmorillonites are
reflected in their appearance in electron micrographs. According to
them, Na montmorillonite appears as a more or less continuous, nebulous,
film-like assemblage, frequently showing cracks intersecting at 120.
H montmorillonite appears as a mosaic of aggregate masses with distinct
hexagonal outlines, which are often 300 A in diameter and 50 to 80 A
thick. Ca montmorillonite appears as irregular aggregates, which
incl'ease in size as the preparatory suspension is aged. At concentl"a-
tions of Ca++ at least equal to the cation-exchange capacity, the aggre-
gates grow in thickness as well as laterally. According to Mathieu-
Sicaud et al.15 these differences are due to variations in the attraction
between the montmorillonite particles and in the relative strength of the
attractive forces at the edges of the particles and on their basal planes.
Electron micrographs of nontronite show that the iron-rich member of
the montmorillonite group tends to occur in elongate lath-shaped units
(Fig. 34), although flakes, needles, and rods of varying size are also
evident. In som-e samples the particles appear to show striations pamllel
to the maximum dimension. The length of the laths may reach several
mkrons and is frequently about five times the width dimension. The
thickness of the laths has a minimum dimension of the order of one or a
few unit cells.
15 Mathien-Sicaud, A., J. Mering, and I. Perrin-Bonnet, Etude au microscope
electronique de Ill. montmorillonite et de l'hectorite saturees par differents cations,
B1.Jl. $()C. iran!;. mi1~eral., 74, 439-456 (1951).
FIG. 32. Electl'On micrograph of montmorillonite, Yavapai County, Arizona, from
Bates.
FIG. 33. Electron micrograph of montmorillonite, upton, \Yyoming, from Kerr.
117
108 Clay Mineralogy

from Woolwich, England (Fig . 23) shows particles somewhat rounded,
24) show well-formed sLx-sided flakes, frequently with a prominent
show less distinct six-sided flakes . The edges of the flakes are somewhat
(Fig. 25).
Electron micrographs of various kaolinite samples have shown par-
4 microns and thicknesses from 0.05 to about 2 microns. This does n ot
mean that larger particles of kaolinite are not present in some clays, since
mechanical working.
t he well-crystallized mineral. It does not always follow that kaolinite
Dickite. Dickite occurs in welt-formed, six-sided, flake-shaped par-
micron. Dickite particles are often large enough to be studied with the
light microscope.
Nacrite. Electron micrographs of a few samples of nacrite show some-
what irregular, rounded, flake-shaped units. In some particles a crude
116 Clay Mineralogy
Some of the electron micrographs of halloY8ite (2H 20) show split ends
consequence of the change from the 4H 2 0 to the 2H 2 0 form of the
mineral.
flake-shaped units without regular outlines. Some of the individual
to flakes approaching unit-cell thickne8s. Accurate estimations of the
areal dimensions of the flakes are difficult to obtain because of their
irregularity but are probably of the order of 10 to 100 times the thickness.
Mathieu-Sicaud, Mering, and Perrin-BonneV 5 have shown that
variations in the exchangeable cation carried by montmorillonites are
them, N a montmorillonite appears as a more or less continuous, nebulous,
film-like assemblage, frequently showing cracks intersecting at 120.
increase in size as the preparatory suspension is aged. At concentra-
Sicaud et al. 15 these differences are due to variations in the attraction
attractive forces at the edges of the particles aljld on their basal planes.
(Fig. 34), although flakes, needles, and rods of varying size are also
evident. In some samples the particles appear to show striations parallel
microns and is frequently about five times the width dimension. The
few unit cells.
16 Mathieu-Sicaud, A., J. Mering, and 1. Perrin-Bonnet, Etude au microscope
electronique de la montmorillonite et de I'hectorite saturees par diilerents cations,
Bull. soc. jrant;. mineral., 74, 439-458 (1951).
FIG. 22. Electron micrograph of allophane, Lawrence County, Indiana, from Kerr.
109
FIG. 34. Electron micrograph of nontronite, Sandy Ridge, North Carolina, from
Bates.
FlO. 35. Electron micrograph of saponite, Death Valley, California, from Ken.
118
FIG. 36. Electron micrograph of hectorite, Hector, California, from Bates.
Some of the magnesium-rich members of the montmorillonite group

appear to be composed of equidimensional flake-shaped units about
like those of the aluminum-rich mOl1tmorillonites. The saponite from
Death Valley, California, illustrates such material (Fig .. 35). How-
ever, hectorite, the fluorine-bearing magnesium-rich montmorillonite,
is found in thin laths, which often lose their identity when gathered
into aggregates (Fig. 36). The laths attain a length of about 1
micron and a width of about 0.1 micron. The thinnest laths appear
to be only 12 to 18 A thick. Some of the particles are terminated by
edges intersect ing the long edges at distinct angles which,'however, ar3
not of uniform size.
Sauconite (Fig. 37) appears to occur in broad laths, which are approxi-
mately 50 A t hick. The particles in Fig. 37 show a variety of habits, and
some of them may be other montmorillonite clay minerals.
Illite. The electron micrograph of Fithian illite (Fig. 38) shows sma1l,
poorly defined flakes, commonly grouped together in irregular aggregates.
'Some of the flakes have a distinct hexagonal outline. The thinnest
flakes are approximately 30 A t hick.' Many of the flakes have a diameter
of 0.1 to 0.3 micron. Some other illites (see, for example, Fig. 39) show
no evidence of flakes with hexagonal outlines. Such flakes have irregular
but well-defined outlines and are characterized by a uniform thickness.
The electron micrographs of illite resemble those of some montmoril-
Bates. Electron micrograph of sallconite, FriedensYiIle Ridge, Penn ylv:1Uia, from
28. Electron micrograph of illite, Fithian, liIinois, from Bates.
120
(brammallitc) outh IYaleR, from Bates.
121
122 Clay Mineralogy
FIG. 41. Electron micrograph of I1ttapulgite, Attapulgus, Georgia, from Bates.
Ionites. The particles of illite are, however, larger and thicker and have
better-defined edges.
Figure 40 shows lath-shaped particles of illite obtained in studies by
Dr. C. E. Weaver 16 of P ennsylvania State College. With the exception
of morphology, this material has all the attributes of other illites. It
occurs in th e -I-micron fraction of a sandstone. The hexagonal flakes
associated with the laths ani probably kaolinite. The significance of an
elongate form of illite has not been determined.
Vermiculite and Chlorite. No information is available regarding the
electron micrography of the vermiculite and chlorite clay minerals. It
appears probable from structural considerations that electron micrographs
of these minerals would be similar to those of the mites, except that
vermiculite might occur in thinner flakes.
Attapulgite-Sepiolite-Palygorskite. Attapulgite is shown by electron
micrographs (Fig. 41) to occur in single laths and bundles of laths.
There appears to be no evidence of a tubular form like that found for
halloysite. The laths attain a maximum length of about 4 to 5 microns,
a maximum thickness of 50 to 100 A, and a width only two or three times
the thickness. There appears to be a tendency for two or three laths to
form a parallel bundle or for a great many laths to form thick bundles.
16 Weaver, C. E., manuscript in. preparation.
FlO . 42. Electron micrograph of alpha sepiolite, Little Cottonwood, Utah, from Kerr.
123
FIG. 43. E lectro n micrograph of beta sepiolite, Baldissaro, Italy, from Kerr.
FIG. 44. Electron micrograph of mixed-layer metabentonite, Oak Hill, Pennsylvania,

from Bates.
124
Shape and Size- Electron Micrographs 125
Other sepioli"tes, called beta-sepiolite by Fersmann, are composed of
extremely small flake-shaped units of irregular outline (Fig. 43). In
with irregular but well-defined outlines. The flakes are quite thin, ane!
keram. Ges ., 24, 37-53 (1943).
Endell, J ., The Study of the Fine Structure of Clays by Means of the Supermicro-
scope, Keram. Rundschau, 49, 23-26 (1941) .
J ackson, M. L., W. J. Mackie, and R. P. Pennington, Electron Microscope Applica-
t.ions in Soils Research, Soil Sci. Soc. Am. Proc., 22,57-63 (1947 ).
Kelley, O. J., and B. T. Shaw, Studies of Clay Particles with the Electron Micro-
Sci. Soc. Am. Proc., 7, 58-62 (1942).
Noll, W., The Electron Microscope in the Study of Hydrothermal Silicate Reactions,
Kolloid-Z., 107, 181-190 (1944).
Prebus, A. F., The Electrcm Microscope, Ohio State Univ. Studies Eng . Expt. Sta. News,
14,7-32 (1942 ).
Waterman, A. T., The Electron Microscope, Am. J. Sci., 239, 386-388 (1941).
17 Fersmann, A. E ., Parasepiolite, Bldl. Acad. Sci. St. Petersburg, ser. 6, 2, 263
(1908).
110 Clay Mineralogy
FIG . 23. Electron micrograph of allophanc, \Yool"ich, England, from Kerr .
h exagonal outline can be seen (Fig. 29). In these samples, the flakes are
generally less than about 1 micron in diameter, and the thickness is about
0.025 to 0 .15 micron.
Halloysite . Numerous investigators have shown that the morphology
of haUoysite is quite different from that .of kaolinite, the former being
elongate instead of flake-shaped (Figs. 30 and 31). Pr~marily because of
the work of Bates,9 halloysite is now known to appear in electron micro-
graphs as elongate tubular particles. Convincing evidence.of the tubular
development is furnished by the doughnut forms, believed to be the ends
of tubular particles, that are seen in some electron micrographs. Some-
times it is difficult to determine if the halloysite particles are actually
tubes or flake-shaped units which have curled and rolled up at their
edges. Bates 14 states his belief that the rolling-up process is possible in
very thin flakes where the number of silicate layerS" is so few that the
strain produced by the misfit of silicon-oxygen and aluminum-hydroxyl
sheets in each layer is not overcome by the cumulative effect of interlayer
bonds between a large number of successive layers.
Bates has determined the dimensions of the halloysite tubes from the
electron micrographs of many samples. The outside diameters were
found to range from 0.04 to 0.19 micron with a median value of 0.07
micron. Inside diameters were found to have a median value of 0.04
14 Bates, T. F., Personal Communication.
FIG. 24. Electron micrograph of kaolinite, l\1&con, Georgia, from Kerr.
III
FIG. 25. Electron micrograph of kaolinite, Mesa Alta, New Mexico, from Kerr.
Electron micrograph of kaolinite, Zettlitz, Czechoslovakia, from Bat e .
112
t -_ _I...:;.fJo.
_ _ __ 4
F IG. 27. Electron micrograph of anauxite, BiEn, Czechoslovakia, from Kerr.
FIG. 28. Electron micrograph of dickite, San Juanito, Mexico, from Kerr.
Fw. 29. Electron micrograph of nacrite, Brand, Saxony, from Kerr.
114
FIG. 30. Electron micrograph of halloysite, Briti, h Guiana, from Bate .
FrG. 3l. Electron micrograph of halloysite, Wendover, Utah, from Bates.

115
116 Clay Mineralogy
Some of the electron micrographs of halloysite (2H 20) show split ends
consequence of the change from the 4:H 20 to the 2H 20 form of the
mineral.
flake-shaped units without regular outlines. Some of the individua1
to flakes approaching unit-cell thickness. Accurate estimations of the
areal dimensions of the flakes are difficult to Gbtai.n because of their
irregulari ty but are probably of the order of 10 to 100 times the thickness.
Mathieu-Sicaud, Mering, and Perrin-Bonnet 16 have shown that
variations in the exchangeable cation carried by montmorilionites are
them, Na montmorillonite appears as a more or less continuous, nebulous,
film-like assemblage, frequently showing cracks intersecting at 120,
incl'ease in size as the preparatory suspension is aged. At conceutl'a-
Sicaud et al. 15 these differences are due to variations in the attra.ction
attractive forces at the edges of the particles and on their basal planes.
(Fig. 34), although flakes, need1es, and rods of varying size are also
evident. In som-e samples the particles appear to show striations pamllel
microns and is frequently about five times the width dimension. The
few unit celis.
15 Mathie\l-Sicaud, A., J . Mering, and I. Perrin-Bonnet, Etude au microscope
electronique de la montmorillonite ct de l'hectorite saturees pa.r differents cations,
B1ta. $()C. fran!;. mineral., 74, 439-456 (1951).
FIG. 32. Electron micrograph of montmorillonite, Yavapai County, Arizona, from
Bates.
FIG. 33. Electron micrograph of montmorillonite, upton, Wyoming, from Kerr.
117
108 Clay Mineralogy

from Woolwich, England (Fig. 23) shows particles somewhat rounded,
24) show well-formed six-sided flakes, frequently with a prominent
show less distinct six-sided flakes. The edges of the flakes are somewhat
(Fig. 25).
Electron micrographs of various kaolinite samples have shown par-
4 microns and thicknesses from 0.05 to about 2 microns. This does not
mean that larger particles of kaolinite are not present in some clays, since
mechanical working.
the well-crystallized mineral. It does not always follow that kaolinite
Dickite. Dickite occurs in well-formed, six-sided, flake-shaped par-
mi cron. Dickite particles are often large enough to be studied with the
light microscope.
Nacrite. Electron micrographs of a few samples of nacrite show some-
what irregular, rounded, flake-shaped units. In some particles a crude
Other sepiolites, called beta-sepiolite by Fersmann, are composed of
extremely small flake-shaped units of irregular outline (Fig. 43). In
with irregular but well-defined outlines. The flakes are quite thin, and
keram. Ges., 24, 37-53 (1943).
Endell, J., The Study of the Fine Structure of Clays by Means of the Supermicro-
scope, Keram. llundschau, 49, 23-26 (1941).
Jackson, M. L., 'V. J. Mackie, and R. P. Pennington, Electron Microscope Applica-
tions in Soils Research, Soil Sci. Soc. Am. Proc., 22,57-63 (1947).
Kelley, O. J., and B. T. Shaw, Studies of Clay Particles with the Electron Micro-
Sci. Soc. Am. Proc., 7, 58-62 (1942).
Noll, W., The Elcctron Microscope in the Study of Hydrothermal Silicate Reactions,
Ko/loid-Z., 107, 181-190 (1944).
Prebus, A. F., The Electron Microscope, Ohio State Univ. Studies Eng. Expt. Sta. News,
14,7-32 (1942).
Waterman, A. T., The Electron Microscope, Am. J. Sci., 239, 386-388 (1941).
17 Fersmann, A. E., Parasepiolite, Bull. Acad. Sci. St. Petersburg, ser. 6, 2, 263
(1908).
CHAPTER 7
Ion Exchange
The clay minerals have the property of'!orbing certain anions and
cations and retaining these in an exchangeable state; i.e., these ions are
exchangeable for other anions or cations by treatment with such ions in a
water solution [the exchange reaction also takes place sometimes in a
nonaqueous environment (see page 141)]. The exchange reaction is
stoichiometric. The exchangeable ions are held around the outside of
the silica-alumina clay-mineral structural unit, and the exchange reaction
generally does not affect the structure of the silica-alumina packet. A
simple and well-known example of the ion-exchange reaction is the soften-
ing of water by the use of zeolites, permutites, or carbon exchangers.
The property of exchange capacity is measured in terms of milli-
equivalents per gram or more frequently per 100 g. One equivalent of
Na+ expressed as Na 2 0 would be a combining weight of 31 and 1 meq per
100 g would be equal to 0.031 per cent Na 20. Exchange capacity is
determined at neutrality, i.e., pH 7.
Vastly more information is available regarding cation exchange than
anion exchange, and although they will be considered separately, an
elaborate discussion of all the aspects of anion exchange is not possible at
this time. In clay materials the commonest exchangeable cations are
Ca++, Mg++, H+, K+, NHt, Na+, frequently in about that order of
general relative abundance. The common anions in clay materials
are sn;-, CI-, PO~-, NOs. The general relative abundance of the
anions is not yet known.
IMPORTANCE OF ION EXCHANGE
The property of ion exchange and the exchange reaction are of very
great fundamental and practical importance in all the fields in which clay
materials are studied and used. The significance of exchange reactions
in many fields has not always been appreciated even by those actively
working in these fields. Therefore, some significant applications of ion
exchange will be given so that its importance will be apparent.
In the field of soils, plant foods are frequently held in the soils as
exchangeable ions, and consequently their persistence in the soil and their
126
Ion Exchange 127
availability for plant growth depends on exchange reactions. For
example, the retention and availability of potash added in fertilizers
depends on cation exchange between the potash salt and the clay mineral
in the soil. Further, the tilth of the soil is frequently determined by the
character of the exchangeable ion, and it may be controlled by an
exchange reaction. Thus the presence of appreciable Na+ in a soil
makes it unsuitable for agriculture. The replacement of the N a+ by
another ion, usually Ca++, will generally make the soil suitable for
agriculture.
In the field of geology many examples could be given, but two will
suffice. Weathering processes involve the liberation of alkalies and
alkaline earths, which mayor may not be retained in the secondary
material, depending on exchange reactions. The nature of the weather-
ing product depends to a very great extent on whether or not the alkalies
and alkaline earths are retained and on which of them are preferentially
retained. Weathering is not simply the breakdown of the primary
minerals followed by leaching.
Variations in the amount and kind of certain ions in the environment of
accumulation of sedimentary rocks must be reflected in the exchangeable-
ion composition of argillaceous sediments. Therefore, exchangeable-ion
data should be of significance in determining the environment of accumu-
lation of ancient sediments.
In oceanography the concentration of sodium in sea water is to a con-
siderable extent a consequence of the cation-exchange properties of clay
materials which have accumulated in the sea. The relative exchange-
ability of the common cations brought to the sea and the property of some
clay minerals to fix K+ would lead to a concentration of sodium (see
pages 144-152).
Ion exchange is of very great importance in all the applied arts where
clay materials are used, or where the properties of clays are important,
because the physical properties of clay materials are frequently dependent
to a large extent on the exchangeable ions carried by a clay. Again many
examples of the great importance of this fact could be given, but a few
will suffice. In general the plastic properties of a clay or soil will be very
different depending on whether N a+ or Ca++ is the excharrgeable cation.
The ceramist, therefore, can change the plastic characteristics of many
clays to meet his needs by carrying out a base-exchange reaction. Thus,
it is common practice in the brick industry to add soda ash to the plastic
clay to improve its properties. The construction engineer, also, can
sometimes vary to suit his needs the property of a soil material on which,
through which, or with which he proposes to work.
Sometimes the construction engineer inadvertently causes an ion-
exchange reaction, by a shift of water table, emplacement of a mass of
128 Clay Mineralogy
concrete, etc., with an unexpected change in the properties of the soil.
If the changes in the plastic, compaction, and shrinkage properties result-
ing from such exchange reactions are not foreseen, the consequences may
be disastrous.
Bentonite clay is widely used for many purposes. For certain uses it
must form thixotropic suspensions in water. Only bentonites composed
of the clay mineral mOl)tmorillonite carrying N a+ as the exchangeable
ion mix with water to give suspensions of pronounced thixotropic
character.
CATION EXCHANGE
History. According to Kelley,! who reviewed the history of cation

exchange in detail, the discovery that soils have the power of exchanging
cations with solutions containing other cations was the outgrowth )f
observations dating back into the remote past. For example, it has been
known for centuries that liquid manures become decolorized and deodor-
ized when filtered through soils. Thompson 2 is generally credited with
being the first person who systematically studied cation exchange. The
term base exchange was used to describe the reaction for many years,
even long after it was established that the hydrogen ion may take part
in the exchange reaction. In experiments begun in 1845 and published
5 years later, Thompson showed that, when soils were mixed with
ammonia and then leached with water, the greater part of the ammonia
was held back.
Following Thompson, W ay 3.4 began a detailed study of the phenome-
non and began to publish his results in 1850. Way showed that cation
exchange in soils was restricted to the clay fraction and that it was con-
nected with the silicate compounds in the soil. Also in 1850 Forschamer li
showed that calcium and magnesium are released from soil by leaching
with sea water. Following the pioneering labors of Thompson, Way, and
Forschamer, a large number of investigators, particularly in the field of
soil chemistry, have studied all aspects of the exchange reaction.
Cation-exchange Capacity. The range of the cation-exchange capacity
of the clay minerals is given in Table 14.
It follows from a consideration of the factors influencing cation-
1 Kelley, W. P., "Cation Exchange in Soils," Reinhold, New York (1948).
Thompson, H. S., On the Absorbent Power of Soils, J. Roy. Agr. Soc. Engl., 11,
68-74 (1850).
3 Way, J. F., On the Power of Soils to Absorb Manure, J. Roy. Agr. Soc. Engl., 11,
313-379 (1850).
4 Way, J. F., On the Power of Soils to Absorb Manure, J. Roy. Agr. Soc. Engl., 13,
123-143 (1852).
5 Forschamer, G., reference in Wiklander, Studies on Ionic Exchange, Ann. Roy.
Agr. Coll. Sweden, 14, pp. 1-171 (1946).

Ion Exchange 129
exchange capacity (see pages 132-136) that there is no single capacity
value that is characteristic of a given group of clay minerals. A range of
capacities must be shown for each group. Since the cation-exchange
capacity of a given mineral type may vary with so many factors, capacity
values are rigorously comparable only if they have been obtained by the
same standard procedure on material of comparable textural and struc-
tural attributes.
TABLE 14. CATION-EXCHANGE CAPACITY OF CLAY MINERALS, IN lVIILLIEQUIVALENTS
PER 100 G.
Kaolinite. . .. . ....................... . 3-15
Halloysite 2H 20 ........................ . 5-10
Halloysite 4H 2 0 ........................ . 40-50
Montmorillonite ........................ . 80-150
Illite. . . . ......................... . 10-40
Vermiculite ............................. 100-150
Chlorite. . . ............... , 10-40
Sepiolite-attapulgite-palygorskite. . . . . . . . .. 20-30
The cation-exchange capacities in Table 14 are taken at pH 7. Titra-

tion curves showing the relation between pH and milliequivalents of
added NaOH are given in Figs. 45 to 48. Marshall,6,7 Mukherjee,8
Mitra,9 and others have considered in detail the significance of such
titration curves. At relatively low and high pH values, attack of acids
and alkalies on the clay-mineral structure is the factor largely controlling
the shape of the curves. It can be seen from Figs. 45 to 50 that the
character of the clay mineral, its concentration, and the base used 10, 11 also
affect the nature of the curves. Other attributes to be discussed later
(see pages 132 and 137), such as variations in degree of crystallinity and
particle size, also affect the shape of the curves.
Marshall 6 and others8 have shown that a single cation may be sorbed
by a clay mineral with a wide range of bonding energies, and that this is
fundamentally related to the position on the silica-alumina packet at
6 Marshall, C. E., "The Colloid Chemistry of the Silicate Minerals," Academic
Pre>;s, New York (1949).

7 Marshall, C. E., and C. A. Krinbill, The Clays as Colloidal Electrolytes, J. Phys.
Chem., 46, 1077-1090 (1942).

8 Mukherjee, J. N., and R. P. Mitra, Some Aspects of the Electrochemisty of the
Clays, J. Colloid Sci., 1, 141-159 (1946).
9 Mitra, R. P., Electrochemical Aspects of Ion Exchanges in Clays, Bentonites, and
Clay Minerals, Indian Soc. Soil Sci. Bull. 4, pp. 41-61 (1942).
10 Mukherjee, J. N., R. P. Mitra, and D. K. Mitra, Electrochemical Properties of
Clay Minerals and the Differentiation of Hydrogen Clays and Bentonites by Electro-
chemical Methods, I, J. Phys. Chem., 47, 543-549 (1943).
11 Mitra, R. P., S. N. Bagchi, and S. P. Ray, Electrochemical Properties of Clay
Minerals and the Differentiation of Hydrogen Clays and Bentonites by Electrochem-

ical Methods, II, J. Phys. Chem., 47, 549-553 (1943).
130 Clay Mineralogy
rr II
~
~
/ /' ..,/
/; v
10 10
~
...0
9 p 9
8 P 8
III
7 n-- 4 7
_) )d
# ~
I-- ,/
V
..-0::::: ~
pH ~. pH
5
6
~~
1(/ Illite < 2)-l
r-- 6
5 r Kaolinite < 2/;L

--
4 .10% - -r - 4 .10% I--

f/ o 5%
o 1%
02%
3 3
2 2
o 5 10 15 20 25 30 35 o 123 456
Milliequivalents No (OH) Milliequivalents No (OH)
per 100 Grams per 100 Grams
FIG. 45. Titration curves for hydrogen- FIG. 46. Titration curves for hy-
illite, after Marshall and Krinbill. 7 drogen-kaolinite, after Marshall and
Krinbill. 7
10.5 12
9.5 II
8.5 10
9
7.5
6.5
pH
5.5
Bentonite
Bentonite <.2j-t b. 0.25%
3.5 .2.0% 00.98%
01.0% 02.90%
00.3% 8.80%
2.5
1.5!:::-~:--~--=~--,.;!-~~-;;!
o 25 50 75 100 125 150 20~~2~0~4~0~6~0~8~0~10~0~12~0~14~0~16~0~180
Milliequivalents No (OH) Milliequivolents No (OH)
FIG. 47. Titration curves for hy- FIG. 48. Titration curves for hydrogen-
drogen-montmorillonite (Wyoming montmorillonite (Indian bentonite), after
bentonite), after Marshall and Krin- Mukherjee and Mitra. 8
bill. 7
Ion Exchange 131
which the cation is sorbed, e.g., whether it is held between the sheets of
the layered minerals or around their edges. According to these workers,
this is an important cause of the variation in titration curves. The
possible positions for sorbed cations vary with the lattice structure of the
minerals, and consequently the variation of bonding energy for a given
cation would not be expected to be the same for all types of clay minerals
(see page 134).
It can be seen from Figs. 45 to 48 that the cation-exchange capacity of
montmoriilonite may vary with the concentration of the clay, particu-
larly at relatively high concentrations. Thus the cation-exchange
10
10.5 9
9.5 8
8.5 pH
pH 7
7.5
D.No (OH) 6.No (OH)
o Bo(OH)z o Bo (OH)z
o Co (OH)z o Co (OH)z
5.5
4.5 4L_____ ~ ____ ~ ______ ~
0 10 20 30 40 o 50 100 150
Milliequivalents of Base Milliequivalents of Bose
FIG. 49. Titration curves for hydrogen- FIG. 50. Titration curves for hy-
kaolinite, after l\:fukherjee, Mitra, and drogen-bentonite (montmorillonite),
Mitra. 10 after Mitra, Bagchi, and Ray."
capac,ity based on N aOH for an Indian bentonite (montmorillonite) calcu-

lated at the inflection point of its potentiometric-titration curve is shown
to vary from 81 to 103 meq per 100 g as the clay concentration increased
from 0.25 to 8.80 per cent. The variation with clay concentration is con-
siderably less for kaolinites and illites than for montmorillonites.
The exchange capacity may vary also with the nature of the cation.
Figures 49 and 50, showing titration curves for kaolinite and mont-
morillonite with cations of different valence, illustrate the relatively gre'at
possible differences for monovalent and divalent cations. In both cases
the exchange capacity is considerably larger when determined with Ca++
than with Na+. As will be shown (pages 137-144), other factors such
as particle size, lattice distortion, clogging of exchange positions, etc.,
may also affect the cation-exchange capacity.
Other Minerals with Cation-exchange Capacity. The clay minerals
are not the only components of clay materials that have cation-exchange
capacity. All inorganic minerals of extreme fineness have a small cation-
132 Clay Mineralogy
exchange capacity as a result of broken bonds around their edges. This
capacity increases as the particle size decreases, but even in the small size
in which nonclay minerals occur in clays, the exchange capacity is
generally insignificant. Zeolite minerals, which are occasionally found
in some clays, have cation-exchange capacities of the order of 100 to
300 meq per 100 g.
Some organic materials have cation-exchange capacity, and values
ranging from 150 to 500 meq per 100 g are reported 12 ,13 for the organic
fraction of some soils. In general, organic material with high exchange
capacity is restricted to Recent sediments and to soils. The organic
material in ancient sediments which has undergone even a small amount
of metamorphism is not likely to have significant cation-exchange capacity.
Causes of Cation Exchange. There are three causes of the cation-
exchange capacity of the clay minerals:
1. Broken bonds around the edges of the silica-alumina units would give
rise to unsatisfied charges, which would be balanced by adsorbed cations.
The broken bonds would tend to be on noncleavage surfaces and hence
on the vertical planes, parallel to the c axis, of the layer clay minerals and
on horizontal planes, perpendicular to the c axis, of the sepiolite-palygor-
skite-attapulgite minerals.
The number of broken bonds and hence the exchange capacity due to
this cause would increase as the particle size decreased. Also lattice
distortions would tend to increase broken bonds, and the exchange
capacity would be expected to increase as the degree of crystallinity
decreased.
In the kaolinite and halloysite minerals broken bonds are the major
cause of exchange capacity. In the illite, chlorite, and sepiolite-palygor-
skite-attapulgite minerals, broken bonds are an important cause of
exchange capacity, and when these minerals are well crystallized and
have relatively low capacity, it may be the major cause. Substitutions
within the lattice, particularly in poorly crystalline examples of illite,
chlorite, and sepiolite-palygorskite-attapulgite, may partially explain
their exchange capacity.
In montmorillonites and vermiculites, broken bonds are responsible for
a relatively small portion (20 % ) of cation-exchange capacity, the
remainder probably resulting from substitutions within the lattice.
Johnson 14 has recently argued that the total exchange capacity of
12 Malquori, A., Behavior of Humus in Clay-Bearing Soils, II, Base Exchange

Capacity of Organic Substances, Ann. chim. applicata, 23, 111-126 (1944).
13 Francis, M., Sur la matiere organique dans les argiles, Verre silicates ind., 14,
155-158 (1949).
14 Johnson, A. L., Surface Area and Its Effect on Exchange Capacity of Montmoril-
lonite, J. Am. Cerarn. Soc., 32, 210-213 (1949).

Ion Exchange 133
montmorillonites re~lts from broken bonds, but his evidence is not
convincing.
2. Substitutions within the lattice structure of trivalent aluminum for
quadrivalent silicon in the tetrahedral sheet and of ions of lower valence,
particularly magnesium, for trivalent aluminum in the octahedral sheet
result in unbalanced charges in the structural units of some clay minerals.
Sometimes such substitutions are balanced by other lattice changes, e.g.,
OH for 0, or by filling more than two-thirds of the possible octahedral
positions, but frequently they are balanced by adsorbed cations.
Exchangeable cations resulting from lattice substitutions are to be
found mostly on cleavage surfaces, e.g., the basal cleavage surfaces of the
layer clay minerals. Since the charges resulting from substitutions in
the octahedral sheet would act through a greater distance than the
charges resulting from substitutions in the tetrahedral sheet, it would be
expected that cations held because of the latter substitutions would be
bonded by a stronger force than those held by forces resulting from sub-
stitutions in the octahedral sheet. In some cases, cations held by forces
due to substitutions of aluminum for silicon seem to be substantially
nonexchangeable, e.g., the potassium in the micas. In the clay minerals,
replacements in the octahedral layer are probably the major substitutions
causing cation-exchange capacity.
In montmorillonite and vermiculite, substitutions within the lattice
cause about 80 per cent of the total cation-exchange capacity. In a
montmorillonite with a cation-exchange capacity of about 100 meq per
100 g (1 equivalent per 1,000 g) and a molecular weight of about 720, sub-
stitutions of less than one-sixth of the aluminum by magnesium or one-
twelfth of the silicon by aluminum would account for the capacity if there
were no internal balancing of charges.
3. The hydrogen of exposed hydroxyls may be replaced by a cation which
would be exchangeable. Some hydroxyl groups would be exposed
around the broken edges of all the clay minerals, and cation exchange due
to broken bonds would, in part at least, be replacement of the hydrogens
of exposed hydroxyls. This cause of exchange capacity would be impor-
tant for kaolinite and halloysite because of the presence of the sheet of
hydroxyls on one side of the basal cleavage plane.
Edelman and Favejee 15 have suggested structures for montmorillonite
and halloysite, and McConnell 16 has suggested an alternative structure
for montmorillonite which would have hydroxyls in basal plane cleavage
surfaces. On the basis of these structural concepts a considerable amount
16 Edelman, C. H., and J. C. L. Favejee, On the Crystal Structure of Montmoril-

lonite and Halloysite, Z. Krist., 102,417-431 (1940).
16 McConnell, D., The Crystal Chemistry of Montmorillonite, Am. Mineral., 35,
166-172 (1950).
134 Clay Mineralogy
of the cation-exchange capacity of these minerals would be due to this
cause. However, neither of these sugge'sted structures have been gen-
erally accepted.
Position of Exchangeable Cations. In the clay minerals in which the
cation exchange results from broken bonds, the exchangeable cations are
held around the edges of the flakes and elongate units. In the clay
minerals where the exchange is due to lattice substitutions, the cations
are mostly on the basal plane surfaces. Thus in kaolinite and halloysite
the cations are at the edges, whereas in montmorillonite and vermiculite
about 80 per cent are on basal plane surfaces, with the remainder on the
edges. In the case of the illite, chlorite, and sepiolite-palygorskite-
attapulgite minerals, most of the cations are at edges with a relative few
on cleavage surfaces.
In a montmorillonite with a cation-exchange capacity of 100 meq per
100 g and a unit-cell weight of 720, and assuming that 80 per cent of the
exchange positions are on basal plane surfaces, there would be 1 equiva-
lent for 1%: unit cells. With a basal surface area of 92.26 A 2 per unit cell,
there would be a total surface area per exchange position of about 160 A2,
or an area. per exchange position on each basal plane surface of about
80 A2.
In masses of clay with relatively small amounts of adsorbed water,
i.e., with no more water than is required to develop plasticity, it is likely
that the adsorbed cations around the edges of the flakes are held directly
in contact or at least very close to the clay-mineral surface. Brown 17 has
present X-ray data based on Fourier syntheses indicating that, for mont-
morillonite under such conditions of clay-water concentration, the
adsorbed cations between the basal plane surfaces are held midway
between the clay-mineral surfaces.
In clay-water systems in which the amount of water is at least greater
than that required for the plastic state, the exchangeable cations may be
at greater distances from the clay-mineral surfaces and separated from
them by water molecules. In any given system of this kind, the position
of all the exchangeable cations with respect to the clay-mineral surface
is not the same, and even the relative positions of all of the same sort of
cation probably will not be the same; some cations of a given type will be
closer to the clay-mineral surface than others. Very careful investigation
of the ionization of exchangeable cations by Marshall 6 and his colleagues 18
has thrown much light on this problem. Only a portion of the adsorbed
17 Brown, G., A Fourier Investigation of Montmorillonite, .Mineralog. Soc. Gr.

18 McLean, E. 0., and C. E. Marshall, Reciprocal Effects of Calcium and Potassium
as Shown by Their Cationic Activities in Montmorillonite, Soil Sci. Soc. Am. Proc.,
lS, 179-182 (1948).
Ion Exchange 135
cations are likely to be ionized, and the percentage ionized depends on
the particular clay mineral, the amount of water, i.e., the concentration
of the clay-water system, the nature of the cations, the relative concen-
tration of the cations, and the nature of the adsorbed anions.
Marshalls has shown that, for alkali cations and for constant concen-
tration of cations, the order of ionization is kaolinite> montmorillonite
> illitc. An investigation 'of Na+, K+, NHt showed that the order of
these three cations is not consistently the same, although NHt generally
shows a lower fraction active. Attapulgite appears to give a higher
degree of ionization than kaolinite. With decreasing concentration of
a given clay the fraction active falls, and some data suggest that there is
a reversal at very low concentrations. Recent data from Marshall 19 in
Table 15, show the fraction of monovalent cations active in relation to
clay mineral, clay concentration, degree of saturation, and monovalent
cation employed.
The divalent ion Oa++ shows a lesser fraction ionized than Na+, and
less than K+ by a considerable factor. This factor is about ~~ for
kaolinite and H 0 to ~~ 0 for clays of the montmorillonite groups. Chat-
terjee and Marsha1l 20 have recently arrived at the following conclusions
regarding the ionization of Oa++, Ba++, and Mg++ for kaolinite, illite, and
montmorillonite: For these divalent ions, illite> montmorillonite, with
kaolinite varying in its relative position according to the nature of the
cation and the extent of neutralization. Recent data from Marshall 19
are given in Table 16, showing fraction active of Mg++, Oa++, and Ba++
in relation to clay mineral, degree of saturation, clay concentration, and
divalent cation employed. These data show, for example, that at 50 per
cent concentration the fraction of Oa++ active is slightly less for kaolinite
than for montmorillonite and considerably less than for illite. At 100
per cent concentration, the fraction of Ca++ active varies as follows:
illite < montmorillonite < kaolinite. At 50 per cent concentration, the
fraction of Mg++ active varies in the order kaolinite < montmorillonite
< illite, whereas at 100 per cent concentration the order is montmoril-
lonite < illite < kaolinite.
Data are not at hand for the ionization characteristics of the other clay
minerals. The complete structural implications of ionization studies
have not yet been worked out, but it seems clear that a single cation can
be held by a given clay mineral with different bonding energies. The
bonding energy is dependent in part on the position of the adsorbed
cation on the clay-mineral unit. Thus, cations between the basal plane
\.
19 Marshall, C. E., The Electrochemistry of the Clay Minerals in Relation to
Pedology, Trans. 4th Intern. Congr. Soil Sci., 1,71-82 (1950).
20 Chatterjee, B., and C. E. Marshall, Studies in the Ionization of Mg++, Ca++,
Ba++ Clays, J. Phys. & Colloid Chern., 64, 671-681, (1950).

136 Clay Mineralogy
surfaces are likely to be held more strongly than those around the edges
of units. This would account in part for the variations in ionization
shown by the clay minerals. It does not, however, fit all of Marshall's-
data, and it must be concluded that the significance of many of these
data is not yet clear.
TABLE 15. IONIZATION OF SOME EXCHANGEABLE UNIVALENT CATIONS FOR SEVERAL
CLAY MINERALS
(After Marshall")
Fraction active
Conccn- Assumed
Clay tration, meq/100 Cation Saturation
% g
50% 75% 100%
--------- ---
Montmorillonite ............. 2.8 100 Na 0.377 0.258 0.381
(Wyoming bentonite) ...... 3.3 100 K 0.295 0.271 0.297
3.0 100 NH4 0.264 0.249 0.245
Illite ....................... 10.0 28 Na 0.073 0.076 0.123
10.0 28 K 0.144 0.127 0.155
10.0 28 NH4 0.144 0.130 0.134
Kaolinite ................... 10.0 2.75 Na 0.263 0.246 0.312
10.0 2.75 K 0.192 0.215 0.297
10.0 2.75 NH4 0.234 0.239 0.263
TABLE 16. CATIONIC FRACTIONS ACTIVE IN RELATION TO CLAY l\1INERAL, DEGREE

OF SATURATION, AND DIVALENT CATION EMPLOYED
(After Marshall 19 )
Fraction active
Concen- Assumed
Clay tration, meq/100 Cation Saturation
% g
50% 75% 100%
Montmorillonite .......... 1.04 100 Mg 0.0122 0.0086 0.0085

(Wyoming bentonite) ... 1.07 100 Ca 0.0175 0.0172 0.053
1.07 100 Ba 0.0036 0.0063 o 0235
Illite .................... 4.9 28 Mg 0.047 0.030 0.023
4.9 28 Ca 0.048 0.032 0.034
. "
4.9 28 Ba 0.0036 0.0027 0.0087
Kaolinite ................ 9.0 2.75 Mg 0.0100 0.0059 0.042
9.0 2.75 Ca 0.0144 0.0170 0.086
9.0 2.75 Ba 0.{)l93 0.038 0.105
Rate of the Exchange Reaction. The rate of cation exchange varies

with the clay mineral, the concentration of the cations, and the nature
and concentration of the anions. In general the reaction for kaolinite
Ion Exchange 137
is most rapid, being almost instantaneous. It is slower for montmoril-
lonites and for attapulgite and requires an even longer time, perhaps I
hours, to reach completion for illites. Apparently exchange on the edge

of the particles, as in kaolinite, can take pla.ce quickly, but penetration
between the sheets of montmorillonite or in the channels of attapulgite
requires more time. In the case of illite a small part of the exchange is
between basal flake surfaces firmly held together, and this is likely to
be slow, causing a long time to complete the reaction.
Chloritic clay minerals are likely to have a rate of exchange similar to
the illites. Vermiculite is likely to be similar to montmorillonite, except
that it may be somewhat slower because of the larger areal size of the
flakes of vermiculite, so that more time is required for penetration
between them.
Exchange in some of the zeolite minerals is even slower than that in the
clay minerals, many hours being required for completion. Apparently
much time is required to penetrate the channel-like openings of some of
these minerals.
Variation Due to Particle Size. As shown in Tables 17 and 18, the
cation-exchange capacities of kaolinite and illite increase as the particle
TABLE 17. VARIATIONS IN THE CATION-EXCHANGE CAPACITY OF KAOLINITE WITH
PARTICLE SIZE
(After Harmon and Fraulini 21)
Particle size, microns ............ 10-205-102-41-0.50.5-0.250.25-0.10.1-0.05
Cation-exchange capacity, meq/

100 g ........................ 2.4 2.63.6 3.8 3.9 5.4 9.5
TABLE 18. VARIATIONS IN THE CATION-EXCHANGE CAPACITY OF ILLITE WITH

PARTICLE SIZE
(After Grim and Bray")
Particle size, microns .................. . . . . . . , . 1-0.1 0.1-0.06 -0.06

Cation-exchange capacity, meq/l00 g .... Sample A 18.5 21.6 33
Sample B 13.0 20.0 27.5
Sample C 20.0 30.0 41. 7
size decreases. It has generally been considered that the cation-exchange

capacity of montmorillonite does not change substantially with particle
size, and Hauser and Reed23 have shown that there is no variation in the
21 Harmon, C. G., and F. Fraulini, Properties of Kaolinite as a Function of its

Particle Size, J. Am. Ceram. Soc., 23, 252-258 (1940).
22 Grim, R. E., and R. H. Bray, The Mineral Constitution of Various Ceramic Clays,
J. Am. Ceram. Soc., 19,307-315 (1936) .
3 Hauser, E., and C. E. Reed, Studies in Thixotropy, II, The Thixotropic Behavior
and Structure of Bentonite, J. Phys. Chem., 41, 911-934 (1937).

)
138 Clay Mineralogy
87- to 14-mtL particle range. Caldwell and Marshall~4 (see Table 19)
show no variation in the cation-exchange capacity of nontronite in the
particle range from 2 to 0.05 micron, but they do show a slight increase
in this same range for saponite. The same authors show a slight increase
for attapulgite with decreasing particle size.
TABLE 19. VARIATION IN THE CATION-EXCHANGE CAPACITY OF N ONTRONITE,
ATTAPULGITE, AND SAPONITE WITH PARTICLE SIZE
(After Caldwell and Marshall 24 )
Particle size, microns ...... ........ 2-1 1-0.5 0.5-0.2 0.2-0.05 -0.05
Cation-exchange capacity,
meq/100 g .............. Nontronite 60.8 61.0 64.3 57.0
Attapulgite 18.0 19.0 22.2 ~
Saponite 69.3 76.0 81. 5 86.3 81.4
In minerals such as kaolinite and illite, in which the exchange capacity

is due primarily to broken bonds, an increase is to be expected with
decreasing particle size. In the case of expanding-lattice minerals,
where most of the exchange is on accessible basal plane surfaces, it seems
that particle size should make little difference. However, it would be
expected that, in some montmorillonites, perhaps because of the location
of lattice substitutions, nature of exchangeable ion, areal size of the
flakes, etc., the accessibility of the basal plane surfaces would increase
with decreasing particle size, and therefore some types of montmorillonite
would be expected to show moderate variation of cation-exchange
capacity with particle size. Johnson 14 has recently presented data show-
ing a variation of cation capacity with partiele size for several mont-
morillonites and concludes that the capacity is entirely derived from
broken bonds. This conclusion, however, does not necessarily follow
from Johnson's data.
Effect of Grinding. Kelley and Jenny25 showed that grinding of the
clay minerals, as well as many non clay minerals, caused an increase in
cation-exchange capacity, as is shown in Table 20. The experiments of
these investigators were carried out in a rubber-lined ball mill, using
polished agate balls. The grinding causes a variation in particle size, an
increase in surface, and an increase in the number of broken bonds.
X-ray examination of the ground material showed a progressive broaden-
ing and diffuseness of the diffraction bands, with their final disappearance
24 Caldwell, O. G., and C. E. Marshall, A Study of Some Chemical and Physical
Properties of the Clay Minerals Nontronite, Attapulgite, and Saponite, Missouri Agr.
Expt. Sta. Research Bull. 354 (1942).
26 Kelley, W. P., and H. Jenny, The Relation of Crystal Structure to Base-Exchange
and Its Bearing on Base-Exchange in Soils, Soil Sci., 41, 367-382 (1936).
--
~ f
Ion Exchange 139
TABLE 20. CATION-EXCHANGE CAPACITY, IN MILLIEQUIVALENTS PER 100 G, IN
RELATION TO GRINDING
(After Kelley and Jenny2S)
Minerals 100 mesh Ground 48 hr Ground 72 hr Ground 7 days
Muscovite ............... 10.5 ... 76.0

Biotite ..... ............. 3.0 62.0 72.5
Kaolinite ................ 8.0 57.5 70.4 100.5
Montmorillonite .......... 126 238.0
after very long grinding indicating a gradual breaking down of the

structure.
Recently Laws and Page 26 have found that there is little further increase
in the cation-exchange capacity of kaolinite after 96 hr of grinding, and
they present evidence to indicate that the kaolinite structure has been
destroyed at this point and that a new permutite-like structure has
developed. Perkins 27 has shown that only 48 hr of grinding is enough to
seriously disrupt the structure of both kaolinite and muscovite.
Relation to Temperature. According to Kelley! the temperature effect
of cation exchange is generally small. Wiegner 2cl has reported a small
negative temperature coefficient, and various other workers 29 have found
that the exchange reaction is accelerated somewhat by raising the tem-
perature. Chapman and Kelley 30 point out that the disadvantages of
heating may outweigh the advantage because of the increased solubility
of certain constituents at the higher temperatures.
The change in cation-exchange capacity of montmorillonite saturated
with Ca++, Na+, and Li+ on heating to various temperatures is given in
Table 21. The exchange capacity is reduced on heating, but the reduc-
tion is not uniform and varies with the (:ation present. Thus the Ca
montmorillonite shows a gradual loss of cation-exchange capacity on
heating to 300C (93 to 41 meq per 100 g) and an abrupt drop from 41 to
12 meq between 300 and 390C, which is the temperature interval
during which inner crystalline swelling is lost. N a montmorillonite,
unlike the calcium variety, shows only a slight drop in exchange capacity
up to 300C (95 to 90 meq), and a moderate drop from 300 to 390C.
26 Laws, 'vV. D., and J. B. Page, Changes Produced in Kaolinite by Dry Grinding,
Soil Sci., 62,319-336 (1946).
21 Perkins, A. T., Kaolin and Treated Kaolins and Their Reactions, Soil Sci., 65,
185-192 (1948).
28 Wiegner, G., Zum Basenaustausch in der Ackererde, J. Landw., 60, 11-150 (1912).
29 Kelley, 'vV. P., and S. M. Brown, Replaceable Bases in Soils, Cal1j. Agr. Expt.
Sla. Tech. Paper 15 (1924).

3. Chapman, H. D., and 'vV. P. Kelley, The Determination of the Replaceable Bases
and the Base-Exchange Capacity of Soils, Soil Sci., 30,391-406 (1930).
140 Clay Mineralogy
)
TABLE 21. EFFECT OF HEATING ON THE CATION-EXCHANGE CAPACITY OF
l\10NTMORILLONI'l'E AND ILLITE
(After Hofmann and Klemen 32)
d(OOI), A Water
Heating Exchange-
Drying content *
temper- able
Mineral time, over
ature, Dry vVetted cations,
C days 35%
clay clay meq/lOOg
H 2SO 4
Ca montmorillonite ....... 105 2 10.2 20 33 93

300 2 9.8 20 26 41
390 14 9.6 9.6 16 12
490 14 9.6 9.6 5.7 6.1
700 2 9.6 9.6 4.7 2.6
N a montmorillonite ....... 105 2 9.8 30 24 95
300 2 9.8 30 22 90
390 14 9.6 21 16 68
490 14 9.6 9.6 10 39
700 2 9.6 9.6 1.5 3.4
Li montmorillonite ........ 20 .. 10.2 30 14.7 98
105 2 10.5 30 12.5 56
125 2 10 10 10 31
200 2 10 10 10 20
Illite, Sarospatak, Hun- 105 2 .. .' ... . . ... 17
gary. 300 2 .... ... . . ... 14 ;
500 2 ... . .... . ... 11

700 2 .. .
' 9
* Computed on dry basis on heating to 900 D C.
Between 390 and 490C the reduction is only to 39 meq, even though the
property of inner crystalline swelling is lost in this temperature interval.
At 700C, following the loss of lattice OH water from the Na montmoril-
lonite, the exchange capacity drops to 3.4 meq. For Li montmorillonite,
the exchange capacity is reduced to 56 meq on heating to only 105C,
which is below the point of loss of the swelling characteristic. Swelling
of this montmorillonite is lost at 125C, with a drop in exchange capacity
to 31 meq per 100 g.
The data of Hofmann and Endell 31 and Hofmann and Klemen 32 show
that the exchange capacity of montmorillonite is reduced to a considerable
extent by heating before the swelling property is lost. The amount of
reduction is large for montmorillonite saturated with Li+ and Ca++ and
31 Hofmann, U., and J. Endell, Die Abhangigkeit des Kationenaustausches und der
Quellung bie Montmorillonit von der Vorerhitzung, Ver. deut. Chemiker Beihefte, 35,
10 (1939).
32 Hofmann, U., and R. Klemen, Verlust der Austauschfiihigkeit von Lithiumionen
an Bentonit durch Erhitzung, Z. anorg. Chem., 262,95-99 (1950).

Ion Exchange 141
small when it is saturated with Na+. Complete loss of swelling follows
the reduction in exchange capacity. Hofmann and Endell 31 interpret
their data to mean that, when the clay is heated, the exchangeable cations
tend to move inside the montmorillonite lattice. Since Li+ is a small
ion, it can fit easily into the structure, possibly in vacancies in the octa-
hedral sheet, and consequently only a low temperature is required for the
shift into the structure. The move into the lattice is followed by a loss
of swelling. Because the Na+ ion is a lal'ge one, it would fit with great
difficulty into the structure, and a high temperature would be required
for the move. The exchange capacity and swelling of Na montmoril-
lonite would persist, therefore, to a relatively high temperature.
In the case of nonexpanding clay minel'als the available data indicate
a gradual reduction in cation-exchange capacity with increasing tem-
perature of heating.
Environment of the Exchange Reaction. Cation exchange usually
takes pJace in an aqueous envjmnment., .llJJd the ions geneTJl-]]Y hJl-ve con-
siderable solubility. However, it has bef;)n shown that clays may take
ions from water suspensions of very insoluble substances and resistant
minerals by means of ionic sorption reactions, and it is probable that the
reaction can take place in suspensions of high concentration, i.e., in the
presence of relatively little water. Indeed this is a means of the natural
disintegration of minerals that is important and is not generally appre-
ciated. Thus, Bradfield 33 found that sodium-saturated clays were able
to take enough barium from barium sulfate to fill about one-fifth of the
exchange positions of the clay. Graham 34 has shown that H clay will
extract calcium from anorthite by cation exchange. This exchange has
been explained 35 as follows: The resistant minerals in water suspensions
are in equilibrium with traces of ions whi(~h dissolve from their surfaces.
The clays destroy this equilibrium by sorbing the ions, and if the equilib-
rium is to be maintained, ions must move from the resistant mineral into
the solution.
It was suggested, by Kelly 36 and by Jenny and Overstreet,37 and shown
later, by Jenny and his coworkers/ 8 Using tracer elements, that ion
33

Bradfield, R., The Concentration of Cation:g in Clay Soils, J. Phys. Chem., 36,
340-347 (1932).
34 Graham, E. R., Calcium Transfer from MinElral to Plant through Colloidal Clay,
Soil Sci., 51,65-71 (1941).
35 Gieseking, J., The Clay Minerals in Soils, A4vances in Agron., 1, 159-204 (1949).
36 Kelley, W. P., A General Discussion of the Chemical and Physical Properties of

Alkali Soils, Proc. 1st Intern. Congr. Soil Sci., 4, 183-489 (1927).
37 Jenny, H., and R. Overstreet, Cation Interchange between Plant Roots and Soil
Colloids, Soil Sci., 47, 257-272 (1939).

38 Jenny, H., R. Overstreet, and A. D. Ayers, Contact Depletion of Barley Roots
as Revealed by Radioactive Indicators, Soil Sci., 48, 9-24 (1939).
142 Clay Mineralogy
exchange can take place directly between plant roots and clays without
the intermediate solution of the ions. The cation moves directly from
the clay to the plant in return for another ion, which moves directly from
the plant to the clay. This probably requires that ions be able to migrate
on clay-mineral surfaces from exchange spot to exchange spot. Jenny 39
postulated that exchangeable ions are in a continuous state of thermal
agitation, and when neighboring zones of agitation overlap, there should
be an opportunity for a given cation to jump from one spot to another,
provided that there is another ion of like charge simultaneously jumping
in the opposite direction. It may well be that direct exchange can take
place between clay minerals and inorganic materials as well as between
clay minerals and plants.
Buswell and Dudenbostel 40 have shown that H montmorillonite will
react with dry NH3 gas to form NH! montmorillonite. CometH has
shown the same thing and has gone further to indicate that some of the
NH~ is taken up hetween the basal plane surfaces of the montmorillonite.
The present use of dry gaseous ammonia as a fertilizer makes use of this
sorption of NH3 by the clay minerals.
Magistad and Burgess 42 have shown that the cation-exchange reaction
can take place in alcohol.
Hydrogen Clays. It has been shown by Paver and Marshall,43 Chat-
terjee and Paul,44 Mukherjee and others 5 that hydrogen montmoril-
Ionites and hydrogen kaolinites are in reality hydrogen-aluminum sys-
tems. It is substantially impossible to prepare a clay in which all the
exchange positions are occupied by H+, since A13+ moves from the lattice
to exchange positions before saturation with H+ becomes complete.
These conclusions are probably applicable to the other clay minerals as
well and apply to the greatest degree to the expanding-lattice minerals
and sepiolite-attapulgite-palygorskite. Kaolinite is affected to the least
39 Jenny, H., Simple Kinetic Theory of Ionic Exchange: Ions of Equal Valency, J.
Phys. Chem., 40, 501-517 (1936).

40 Buswell, A. M., and B. F. Dudenbostel, Spectroscopic Studies of Base-Exchange
Materials, J. Am. Chem. Soc., 63, 2554-2.559 (1941).

41 Cornet, 1., Sorption of NH3 on Montmorillonite Clay, J. Chem. Phys., 11,217-226
(1943).
42 Magistad, O. C., and P. S. Burgess, The Use of Alcoholic Salt Solutions for the
Determination of Replaceable Bases in Calcareous Soils, Ariz. Agr. Expt. Sta. Tech.
Bull. 20 (1928). .
43 Paver, H., and C. Eo Marshall, The Role of Alumina in the Reactions of Clays,
J. Soc. Chem. Ind. (London), 53, 750-760 (1934).

44 Chatterjee, B., and M. Paul, Interaction between Hydrogen Clays and Neutral
Salts, Indian J. Agr. Sci., 12, 113-120 (1942).
45 Mukherjee, J. N., B. Chatterjee, and P. C. Goswami, Limiting Exchange of
Aluminum Ions from Hydrogen Clays on the Addition of Neutral Salts, J. Indian
Chem. Soc., 19, 40-407 (1942).
Ion Exchange 143
degree, with the illite and chlorite minerals showing an intermediate
effect.
Paver and Marsha1l 43 have shown that the electrolysis of clays removes
not only the mobile cations, although this predominates in the earlier
stages, but also basic constituents, such as iron and magnesium, which
are frequently found at the anode, while silicic acid moves in considerable
quantity with the bulk of the bases to the cathode. Electrodialyzed
clays, on treatment with neutral salts, liberate aluminum, and the amount
increases with the concentration of the salt to a maximum which is
approximately equivalent to the exchange capacity of the clay.
Mukherjee and his coUeagues 46 ,47 have studied in detail the changes
which occur when the clay minerals undergo repeated desaturation and
treatment with barium chloride. With montmorillonite, repeated cycles
of desaturation and neutral salt treatment caused a steady decrease in the
aluminum and a smaller decrease in the hydrogen released by exchange.
After four cycles the ratio of hydrogen to aluminum released becomes
very large. Successive treatments caused a large reduction in the cation-
exchange capacity of the montmorillonite, indicating considerable
decomposition, which was eonfirmed by the appearance of soluble silica
in the earlier desaturations. With kaolinite, aluminum and hydrogen
were similarly brought into solution by the neutral salt treatment, but
there was no reduction in cation-exchange capacity, and no soluble silica
was found. Probably the attack on kaolinite is restricted to the alumi-
num exposed on external surfaces. It is likely that the sepiolite-
palygorskite and vermiculite minerals would act like the montmorillonite
and that illites and chlorites would be intermediate, but more like the
kaolinite.
Unpublished work by Michelson 48 suggests that the movement of
aluminum from positions in the lattice to exchange positions is facilitated
by drying. So long as the sample is not dried, the amount of movement
is relatively small. Michelson's work was done on montmorillonite, but
it probably applies also to the other clay minerals.
The fact that hydrogen clays are in reality hyd.rogen-aluminum
systems is of great importance in clay and soil investigations. The
failure to reeognize this faet has eaused many erroneous conclusions to
be reported and much confusion in interpreting results. For example,
numerous studies have been made of the physical properties of supposedly
46 Mukherjee, J. N., and B. Chatterjee, Liberation of H+, AI+++ and Fe+++ Ions
from Hydrogen Clays by Neutral Salts, Nature, 155,268-269 (1945).

47 Mukherjee, J. N., B. Chatterjee, and A. Ray, Liberation of H+, Al+++ and Fe+++
Ions from Pure Clay Minerals on Repeated Soil Treatment andDesaturation, J.

Colloid Sci., 3, 437-446 (1946).
48 Michelson, G., personal communication.
144 Clay Mineralogy
monoionic clays prepared by treating hydrogen clays with various
cations. Actually in many cases monoionic clays were not attained, and
the reported results are to a considerable extent a consequence of the
presence of AIH and not, as supposed, of the cation with which the clay
was treated. Further it makes extremely difficult the study of the
properties of calcareous clays, since the carbonate cannot readily be
removed without damaging the clay-mineral component and altering its
properties.
Clogging of Cation-exchange Positions. It has just been pointed out
that the development of aluminum on exchange positions reduces the
exchange capacity of montmorillonite clays. This is in part due to
damaging of the montmorillonite lattice, but also in part to clogging
of the exchange positions by aluminum. Dion 49 has pointed out that
Fe 20s alone or in the hydrated form may serve to reduce the cation-
exchange capacity of clay minerals by a clogging action.
A somewhat similar effect may be produced by organic ions. Hen-
dricks 50 has pointed out that large, flat organic ions adsorbed on the basal
surface of montmorillonite may be of sufficient size to blanket more than
one exchange position and thereby seem to reduce the exchange capacity
of the montmorillonite. Organic molecules with an area greater than
about 80 A2 and flat-lying could spread over more than one exchange
position.
Other effects may serve to reduce the exchange reaction. Sulfur com-
pounds may form on the adsorbing surface 5 ! of the clay minerals, which
would serve to reduce exchange. This matter is not well understood,
but studies of clay-mineral catalysts reveal that they can be poisoned by
sulfur, probably because of sulfur compounds which form on the clay-
mineral surface and reduce the sorption activity.
Replaceability of Exchangeable Cations. Very early studies of cation
exchange showed that under a given set of conditions, various cations
were not equally replaceable and did not have the same replacing power.
Way4 concluded that the replacing power of the common ions was
Na < K < Ca < Mg < NH 4 , which means, for example, that in general
Ca++ will more easily replace Na+ than Na+ will replace Ca++.
As the cation-exchange reaction was studied, it became obvious that
there was no single universal replaceability series. The series varied
depending on the experimental conditions, on the cations involved, and
49 Dion, H. G., Iron Oxide Removal from Clays and Its Influence on Base-Exchange
Properties and X-ray Diffraction Patterns of the Clays, Soil Sci., 58, 411-424 (1944) .
0 Hendricks, S. B., Base Exchange of the Clay Mineral Montmorillonite for Organic
Cations and Its Dependence on Adsorption Due to van der Waals Forces, J. Phys.
Chem., 45, 65-81 (1944).
Dl Davidson, R. C., Cracking Sulphur Stocks with Natural Catalysts, Petroleum
Refiner, September, 1947.

Ion Exchange 145
52
on the kind of clay material. Gedroiz in 1922, for example, gave the
following order of replacing power based on the replacement of Ca++ by
0.01 N solution of chlorides: Li < N a < K < Mg < Rb < NH4 < Co
< AI. In the replacement of Ba++ by 0.1 N solution, using the same soil
material, Gedroiz found Li < N a < NH4 < K < Mg < Rb < Ca <
Co < AI.
The matter of cation replace ability is of very great importance,
and a very large amount cf research on the problem has been reported.
Replaceability is not yet completely understood, but it is known that
it is controlled by a considerable number of factors, and much has been
learned about the nature and the influence of each of the factors. In the
following discussion of the factors involved. no attempt is made to place
them in an order of importance.
Effect of Concentration. Kelley and Cummins 53 found that the replace-
ment of Ca++ and Mg++ by Na+ in the Yolo soil of California increased
as the concentration of Na+ in the solution increased. Gedroiz 52 pointed
out that the replacement of Ca++ and Mg++ by NHt in a Chernozem soil
increased as the concentration of NHt increased. This is to be expected,
since cation exchange is a stoichiometric reaction and the laws of mass
action would hold. In general, therefore, increased concentration of the
replacing cation causes greater exchange by that cation.
Gedroiz 52 in his experiments found that the increased replacement of
Ca++ and Mg++ by NHt was not in direct proportion to the concentration
and that the ratio of Ca++ to Mg++ was also not in direct proportion.
That is to say, the concentration of replacing cation is important, but it is
not the sole factor. Schachtschabe1 54 demonstrated conclusively that
the effects of concentration depend on the kind of cation that is being
replaced and also on the valence of the cation, as well as on other
factors.
The complexity of this factor of concentration is shown by the following
data brought out by Kelley:l "With cation pairs of about similar replac-
ing power and of the same valence such as K+ vs. NHt or Ca++ vs. Ba++,
dilution has relatively little effect on exchange, while with cations of
different replacing power and different valence, for example N a+ vs. Ca++
or NHt vs. Ca++, dilution produces marked effect on exchange." Kelley
points out that a fully satisfactory explanation of these facts cannot now
be given.
52 Gedroiz, K., On the Absorptive Power of Soils. Translated by S. Waksman
and distributed by U.S. Department of Agriculture (1922).
53 Kelley, W. P., and A. B. Cummins, Chemical Effects of Salts on Soils, Soil Sci.,
li, 139-159 (1921).
I 54 Schachtschabel, P., Untersuchungen tiber die Sorption der Tonmineralien und
organischen Bodenkolloide und die Bestimmung des Anteils dieser Kolloide an der
Sorption in Boden, Kolloid-Beihefte, 61, 199-276 (1940).
146 Clay Mineralogy
Population of Exchange Positions. Jenny and Ayers 55 and later
Wiklander,66 in considerable detail, have shown that the ease of release
of an ion depends not only on the nature of the ion itself but also upon the
nature of the complementary ions filling the remainder of the exchange
positions, and on the degree to which the replaced ion saturates the ex-
change spots. Thus, Wiklander shows that, as the amount of exchange-
able calcium on the clay mineral becomes less, the calcium becomes more
and more difficult to release. Sodium, on the other hand, tends to
become easier to release as the degree of saturation with sodium ions
becomes less. Magnesium and potassium are not affected by the degree
of saturation to the same extent as calcium and sodium.
Nature of Anion in Replacing Solution. Numerous investigators have
found that the replaceability of a given ion varies with the nature of the
anion which may be present. Thus Neznayko 57 found considerable
variation in the replaceability of N a+ from montmorillonite by Ca++
depending on whether calcium hydroxide or calcium sulfate was used. J
Marshall 6 reports that considerable variation in the cation-exchange
capacity is obtained by leaching with different neutral salts of a given
ion. The whole matter of the effect of anions is complicated by the
possibility of the formation of "basic" salts with the clay and a soluble
anion, e.g., clay-(ZnOH)+,5 8 and, at relatively high concentration of
anions, by the probability of reactions with the clay mineral, altering its
structure and forming new compounds.
Nature of the Ion. Other things being equal, the higher the valence of
the ion, the greater its replacing power and the more difficult it is to
displace when already present on the clay. Hydrogen is an exception
since for the most part it behaves like a divalent or trivalent ion.
Kelleyl points out that the 'Teplacing power' increases qualitatively
with atomic number in ions of the same valence.
Also, in ions of the same valence, replacing power tends to increase as
the size of the ion increases; i.e.,.the smaller ions are less tightly held than
the larger ions. An exception to the effect of ion size occurs in those ions
which have almost exactly the correct size to fit into the hole in the basal
oxygen sheet of the layer clay minerals. Thus potassium, as pointed
out by Page and Baver,s9 has an ionic diameter of 2.66 A, which is about
66 Jenny, H., and A. D. Ayers, The Influence of the Degree of Saturation of Soil
Colloids on the Nutrient Intake by Roots, Soil Sci., 48, 443-459 (1939).
66 Wiklander, L., Studies of Ionic Exchange with Special Reference to the Condi-
tions in Soils, Ann. Roy. Agr. Coli. Sweden, 14 (1946).

67 Neznayko, M., personal communication.
6. Jenny, H., and M. M. Engabaly, Cation and Anion Interchange with Zinc
Montmorillonite Clays, J. Phys. Chem., 47, 399-410 (1943).
5. Page, J. B., and L. D. Baver, Ionic Size in Relation to Fixation of Cations by Col-
loidal Clay, Soil Sci. Soc. Am. Proc., 4, 150-155 (1939).
Ion Exchange 147
the same as the diameter of the cavity in the oxygen layer, so that the
potassium ion can ju:ot fit into one of these cavitie:o. As a consequence
the potassium ion is relatively very difficult to replace.
Wiegner and J enny 60 suggested that the size of the hydrated ion,
rather than the size of the nonhydrated ion, controIls replaceability.
According to Wiegner,61 for ions of equal valence, those which are least
hydrated have the greatest energy of replacement and are the most
difficult to displace when already present upon the clay. Thus, lithium,
although a very small ion, is considered to be highly hydrated and, there-
fore, to have a very large hydrated size. The low replacing power of Li+
and its ready replaceability are said to be a consequence of this large
hydrated size.
The magnitude of the computed hydration of the various ions varies
with the base chosen for comparison. Hydration values suggested by
several investigators and obtained in different ways are given in Table 22
to show the order of magnitude of the hydration generally given for the
common IOns.
The matter of the hydration of the adsorbed ions in clay-water systems
is a matter of much controversy at the present time. Wiegner and
Jenny 60 and later Alten and Kurmies 62 have presented strong evidence
that all the common cations are hydrated and that hydration is important
in exchange reactions. The evidence is based on investigations carried
out in alcohol rather than water. It is stated that alcohol, in which the
ions would not be hydrated, reverses the usual lyotropic replacement
series, so that ions are held in the order of their true ionic size and not
in the order of their hydrated size.
Bernal and Fowler 63 some years ago presented data that threw doubt
on the hydration of some of the cations. Bar and Tenderlo0 64 have
enumerated difficulties encountered by explanations of day properties
based on cation hydration, and Baver 65 has summarized experimental
data on the swelling and heat of wetting of soil colloids that do not
fit well with cation hydration. Recently, Hendricks and his col-
60 Wiegner, G., and H. Jenny, Ueber Basenaustausch an Permutiten, Kolloid-Z.,
43,268-272 (1927).
61 Wiegner, G., Ionenumtausch und Struktur, Trans. Intern. Congr. Soil Sci. 3rd
Congr., Oxford, 3, 5-28 (1935).

62 Alten, F., and B. Kurmies, "Handbuch der Bodenlehre," vol. 8, Springer, Berlin
(1931).
63 Bernal, J. D., and R. H. Fowler, A Theory of Water and Ionic Solution with
Particular Reference to Hydrogen and Hydroxyl Ions, J. Chem. Phys., 1, 515-548

(1933).
64 Bar, A. L., and H. J. Tenderloo, Ueber die Doppelschicht der Tonkolloide, Kol-
loid-Beihefte, 44,97-124 (1936).
66 Baver, L. D., "Soil Physics," Wiley, New York (1940).
148 Clay Mineralogy
TABLE 22. ION SIZES AND IONIC HYDRATION
Ionic radii, A
Hydration, moles H 2 O
Ion Not hydrated Hydrated
A B C D E ]I' G H I
------ --- --- --- ~---
--- --- --- ---
Li 0.68 0.78 10.03 7.3 12.6 10 15 11-13 13-14
Na 0.98 0.98 7.90 5.6 8.4 5 8 9-11 8-9
K 1.33 1.33 5.32 3.8 4.0 1 4 5-6 5
NH4 .... 1.43 5.37 4.4 .. . . .. 2-3
Rb ... . 1.49 5.09 3.6 '" . 0.5
Cs .... 1.65 5.05 3.6 .... 0.2
Mg 0.89 0.78 ..... 10.8 13.3 33 21 20-23
Ca 1.17 1.06 ..... 9.6 10.0 22 22 19-22
Sr 1.34 1.27 ..... 9.6 8.2 21
Ba 1.49 1.43 ..... 8 8 4.1 17 14 18-20
Al 0.79 0.57 .. . .. . ' " . . .. 57
La 1.30 1.22 ..... 30.5
A. Zachariassen, W. H., Z. Krist., 80, 137 (1931).

B. Goldschmidt, V., Norske. Videnskaps- Akad. Oslo, Skrifter. Mat.-Naturv. Klasse,
7,8 (1926, 1927).
C. Jenny, H., J. Phys. Chem., 36, 2217-2258 (1935).
D and F. Pallman, H., Bodenk und Forsch., 6, 21 (1938).
E. Remy, H., Z. Phys. Chem. A., 39, 467 (1915).
G. Brintziner, H., and C. Ratanarat, Z. anorg. u. allgem. Chem., 222, 119 (1935).
H. Bourion, F., E. Ronyer, and O. Hun, Compt. rend. 204, l420 (1937).
l. Baborovski, J., J. Velisch, and A. Wagner, J. Chem. Phys., 25,452-482 (1928).
leagues 66 have presented compelling evidence, based on careful dehydration

studies, that Na+, H+, and K+ and the trivalent ions are not hydrated
when adsorbed by the clay minerals. According to them, Ca++ and
Mg++ are hydrated to 6H 2 0, and Li+ is hydrated to 3H 2 0. This is
directly contrary to conclusions of those favoring the hydration concept,
who consider Li+ and Na+ to be among the most highly hydrated ions.
According to Bar and Tenderloo,64 replaceability is related to the
polarization of the ion, with increasing polarization being accompanied
by increasing difficulty of exchange. Also, highly polar ions are thought
to be held closer to the adsorbing surface. Polarization increases as the
valence increases and as the size of the ion decreases.
Effect of Heating. The work of Hofmann and J. Ende1l 31 has indicated
that heating to moderate temperatures not only reduces the cation-
exchange capacity but changes the relative replaceability of the cations
the Clay Mineral Montmorillonite Saturated with Various Ions, J. Am. Chem. Soc.,
62, 1457-1464 (1940).
Ion Exchange 149
(see page 140). Thus, for montmorillonite, heating to 125C tends to
fix Li+ in an unreplaceable form and not to affect the replaceability of
Na+, whereas at ordinary temperatures Li+ is more easily replaced than
Na+. The extensive and excellent work of Marshall and his colleagues 67
in developing and using clay membranes as electrodes to measure ion
activity also has shown the great variation of ion replace ability with
increasing temperatures. Andrews and Maldonad0 68 have shown that
the relative amounts of the cations which are replaceable change when
the clay fraction of the Crowley silt loam soil is heated up to 100C.
They found that the relative amounts of replaceable K+, Ca++, and H+
decreased on prolonged heating, whereas Na+ and Mg++ increased.
It seems likely that the change in replaceability of cations when heated
would be relatively greater for the expanding-lattice minerals than for
those minerals in which the capacity is due largely to broken bonds. In
the former case, most of the adsorbed cations are held between the basal
layers, and the change in replaceability is probably largely in these ions.
At elevated temperatures, when there is little or no water present between
basal layers in addition to the sorbed cations, the size of the ion and its
geometrical fit into the structure of the oxygen layers are probably major
factors in determining replaceability.
Nature of the Clay Mineral. Early investigations of many soils sug-
gested that the replace ability of various cations varied with the nature
of the adsorption complex, all other factors being equa1.' The work of
Jarusov,69 Bar and Tenderloo,64 and Gieseking and J enny70 particularly
showed this to be true, and the careful work on pure kaolinite, muscovite,
and montmorillonite by Schachtschabe1 54 clearly demonstrated that there,
was not a single replace ability series characteristic of all clay materials,
but separate replace ability series for the various clay minerals.
The replaceability of NHt from kaolinite, montmorillonite, and mus-
covite by various cations at various concentrations is shown in Figs. 51
to 53, taken from Schachtschabe1. These data show, for example, that,
for NH4 montmorillonite, H+ and K+ are about equally exchangeable,'
that all the univalent cations except Rb+ and Cs+ are more exchangeable
than the divalent ions, and that all the divalent ions have about the same
replaceability. For NH4 kaolinite the exchangeability of the cations is
about the same as for NH4 montmorillonite. NHt, however, is more
67 Marshall, C. E., et al., The Electrical Properties of Mineral Membranes, series
of papers beginning in 1941 in J. Am. Chem. Soc. and J. Phys. Chem.
68 Andrews, J. S., and J, F. Maldonado, Effect of Temperature on the Base-Ex-
change Capacity of Clays, J. Agr. Univ. Puerto Rico, 24, 133-142 (1940).
69 Jarusov, S. S., Mobility of Exchangeable Cation$ in the Soil, Soil Sci., 43, 285-303
(1937).
70 Gieseking, J. E., and H. Jenny, Behavior of Polyvalent Cations in Base Exchange,
Soil Sci., 42, 273-280 (193(i). . - - -- . - ~.
'150 Clay Mineralogy
lightly held by kaolinite. For NH4 muscovite, H+ and K+ are more
lightly held than the divalent ions, and NHt is even more lightly held
v than in kaolinite. When a mixture of montmorillonite and muscovite is
treated with mixtures of calcium
and ammonium acetate, the mica
adsorbs relatively much more of the
NHt, and the montmorillonite ad-
sorbs relatively more of the Ca++.
Also Ca++ in competition with
equivalent concentrations of K+
would be taken up preferentially
by montmorillonite, while the oppo-
site is the case for muscovite.
Schachtschabel 54 used this latter be-
havior as the basis for an analytical
method for the determination of the
relative abundance of montmorillo-
nite and the mica-like clay minerals
when mixed together in soils.
10 Hendricks and Alexander 71 have
also studied the relative acceptance
0~_LI--L2--~3--~4 100 of Ce 3+ and H+ by montmorillonite
Me. of Cations in 100 ee. and illite and have found that the
of Water per Me. of illites preferentially take up H+,
NH4 Kaolinite
while the montmorillonites take up
FIG.51. NH, kaolinite; ionic exchange
with alkali and alkaline-earth chlorides,
'much more Ce H . They use this
after Schachtschabcl. 5, difference as the basis for a method
of estimating the amount of mont-
morillonite present in mixtures of these two minerals .
./ Barshad 72 has recently investigated the cation-exchange characteristics
of vermiculite and has shown that the exchange process is reversible
between Na+, Ca++, Mg++, and K+ but is not completely reversible
between K+, NHt, Rb+, and Cs+. The latter ions tend to become fixed
and relatively nonreplaceable. Walker and Milne 73 have also investi-
gated the replaceability of the exchangeable cations of vermiculite with
much the same results. They show that K+ and NHt replace Mg++ with
difficulty, whereas Mg++ is readily replaced by Li+ and Na+. However,
71 Hendricks, S. B., and L. T. Alexander, Semi qualitative Estimation of Montmoril- ..
lonite in Clays, Soil Sci. Soc. Am. Proc., 6,95-99 (1940).
72 Barshad, 1., Vermiculite and Its Relation to Biotite as Revealed by Base Exchange
Reactions, Differential Thermal Curves, and Water Content, Am. M1:neral., 33, 65S-
678 (1948).
,. Walker, G. F., and A. Milne, Hydration or Vermiculite Saturated with Varivus
Cations, Trans. 4th Intern. Congr. Soil Sci., 2, 62-67 (1950).
Ion Exchange 151
100 K
H
g;, 90 No
c
0
Li
.:
U
X
W
0
;
.._
0
+-
C
<lJ
U
~ 40
.~
30
<lJ
(J>
C
0
.:
20
U
X
W
10
0
0 2 3 4 5 6 7 8 9 10 100
Me. of Cations in 100 cc. of Water
per Me. of NH4 Montmorillonite
FIG. 52. NH4 montmorillonite; ionic exchange with alkali and alkaline-earth
chlorides, after Schachtschabel. 64
~ 90 H
c
0
.:
u
x
80
W Rb K
0 K
; Cs No
80
.....0 Sr
..... Li
C
<lJ
Co
U Mq
....
<l> 40
a..

30
<l>
(J>
C
0 20
.:
u
x
, W
10
0
0 2 3 4 5 6 7 8 9 10
Me. of Cations in 100 cc. of Water
per Me. of NH4 Mica
FIG. 53. NH, muscovite; ionic exchange with alkali and alkaline-earth chlorides,
after Schachtsrihabel. 64
152 Clay Mineralogy
a vermiculite prepared with Li+ or N a+ readily loses these ions by replace-
ment with K+ or NHt. Any difference in replaceability between
vermiculite and montmorillonite is of interest, since the only major
difference between these two minerals seems to be in the size of the layers
in the a and b crystallographic directions.
Allaway74 has investigated the replace ability of Ca++ in various clay
minerals by determining its availability to plants. He finds that for'
Ca++ the availability series is peat> kaolinite> illite> montmoril-
lonite and that the availability increases as the saturation of the Ca++
Increases.
Marshall's6 work on the ionization of the cations sorbed by the clay
minerals has also shown a difference in relative cation replaceability for
the various clay minerals (see pages 134-136). In general the higher
the degree of ionization, the greater the ease of replaceability.
Fixation of Cations. It is well known by soil investigators that certain
cations may be sorbed by the clay minerals in a nonexchangeable or
difficultly exchangeable state. Potassium is the commonest ion that is
"fixed" to a considerable degree and the ion that has been most studied
because of its importance in soil fertilization. Wood and DeTurk/"
Bray and DeTurk,76 Stanford,77 and others have shown that illite will
fix potassium, and most of the K+ fixed in soils is probably due to the
action of this clay mineral. Volk 78 showed that potash fixation is often
accompanied by the formation of mica in the soil. The potassium
fixation in illite occurs by the emplacement of K+ between the basal
surfaces of the mineral in the positions normally occupied by K+ in this
mineral. Much illite, particularly in soils which are undergoing leaching,
is somewhat degraded, i.e., it is potash-deficient in that leaching has
removed some K+ from positions between the layers. On the addition
of potassium to such material, K+ goes back into these normal positions,
and the illite is rebuilt. Under natural soil conditions, fixation appears
to be completed in a matter of months following its addition as fertilizer.
j Wiklander 79 has shown that the amount of fixation in illite-type material
74 Allaway, W. H., Availability of Replaceable Calcium from Different Types of
Colloids as Affected by Degree of Calcium Saturation, Soil Sci., 59, 207-217 (1945).
75 'Vood, L. K., and E. E. DeTurk, The Chemical Effect of Soluble Potassium
Salts in Some Illinois Soils, Soil Sci. Soc. Am. Proc., 5, 152-161 (1941).
76 Bray, R. H., and E. E. DeTurk, The Release of Potassium from Non Replaceable
Forms in Illinois Soils, Soil Sci. Soc. Am. Proc., 8, 101-106 (1939).
77 Stanford, G., Fixation of Potassium in Soils under Moist Conditions and on Dl'Y-
~g in Relation to the Type of Clay Mineral, Soil Sci. Soc. Am. Proc., 12, 167-171
{::.\)47) .
8 Yolk, G. W., Nature of Potash Fixation in Soils, Soil Sci., 45, 263-276 (1938).
79 Wiklander, L., Fixation of Potassium by Clays Saturated with Different Cations,
Soil. Sci., 69,261-271 (1950).
Ion Exchange 153
varies with the cations already adsorbed. Thus fixation is relatively high
for materials saturated with Ca++ and Na+ and low for those saturated
with H+ and NHt. Stanford 77 also has shown that the presence of H+,
Fe H , and AF+ tends to block the fixation of K+ by illite. Drying tends
to increase the fixation by illite. -
According to Raney and Hoover 8 () and others, montmorillonite has
some power to fix potassium, but less than illite. Other investigators,
notably Stanford,77 state that K+ is fixed in montmorillonite only if the
material is dried. Barshad 72 has shown that vermiculite can fix some
potassium. In the case of montmorillonite and vermiculite, as with
illite, any fixation of K+ would take place on the basal planes between the
unit layers.
Kaolinite has substantially no power to fix K+ either in a moist condi-
tion or after drying. Joffe and Levine 81 have reported K+ fixation by
very finely ground kaolinite, but the grinding may well have destroyed
the kaolinite structure.
Bray and DeTurk 76 and Wood and DeTurk,75 working with illitic
Prairie soils, have shown that the K+ is held in various degrees of replace-
ability and availability to plants and that some substantially non-
replaceable K+ may be available to plants. These authors have shown
that an equilibrium exists in such soils between the exchangeable and.
nonexchangeable potassium.
NHt is very similar to K+ in its ion-exchange properties, and Chami-
nade,82 Page and Baver,83 and Stanford and Pierre 84 have demonstrated
the fixation of NHt by clays in difficultly exchangeable form. The latter
workers have concluded that the ammonium ion and the potassium ion
are fixed in soils by the same mechanism.
Barshad 72 has recently investigated vermiculite and has found that
this clay mineral can fix at least a part of any adsorbed Rb+ or Cs+.
Although the matter has not been established experimentally, it seems
likely that Mg++ could in some conditions be fixed by "degraded chlorite"
by a mechanism similar to that operating for K+ and "degraded illite."
80 Raney, W. A., and C. D. Hoover, The Release of Artificially Fixed Potassium

from Kaolinitic and Montmorillonitic Soils, Soil Sci. Soc. Am. Proc., 11, 231-237
(1946).
81 Joffe, J. S., and A. K. Levine, Fixation of Potassium in Relation to the Exchange
Capacity of Soils, Soil Sci., 63,241-247 (1947).

82 Chaminade, R., Fixation d'l'ion NH4 par les colloides argileux des sols sous forme
non exchangeable, Compt. rend., 210, 264-266 (1940).

83 Page, J. B., and L. D. Baver, Ionic Size in Relation to Fixation of Cations by
Colloidal Clay, Soil Sci. Soc. Am. Proc., 4, 150-155 (1939).

"Stanford, G., and W. H. Pierre, The Relation of Potassium Fixation to Ammo-
nium Fixation, Soil Sci. Soc. Am. Proc., 11, 155-160 (1946).
154 Clay Mineralogy
Also the synthesis experiments of Caillere, Henin, and Mering 85 indicate
that Mg++ can be fixed by montmorillonite with the development of a
chlorite type of structure under certain conditions of concentration of
the Mg++.
The fixation of magnesium and potassium is of considerable importance
in diagenetic processes in sediments accumulating at the present time
(see pages 351-352).
Theory of Cation Exchange. Numerous attempts have been made to
develop a theory of cation exchange which would permit the quantitative
expression of exchange data by an equation. Kelley,1 Marshall,6 and
Du Rietz 86 have reviewed this matter in detail and have commented on
the deficiencies of the theories that have been proposed.
Equations based on three considerations have been suggested. Wieg-
ner and J~ny,60 and Vageler and Woltersdorf87 derived empirical equa-
tions from, or connected with, adsorption isotherms. The general basis
of these theories is Freundlich's88 adsorption equation or modifications of
it. Thermodynamic arguments involving more or less crude approxima-
tions have led to various mass-action equations by Kerr,89 Vanselow, 90
and Rothmund and Kornfeld. 91 Davis 92 has criticized these latter
formulations, and Kelly1 in turn has objected to some of Davis's remarks.
A third approach to the problem based on kinetic considerations has been
made by Gapon 93 and Jenny. 39 Davis 94 has recently attempted to
improve Jenny's equation.
As Kelley 1 has shown, none of these equations is completely satis-
factory, and indeed they cannot be universally applied because of the
large number of variations dependent on the nature of the clay mineral,
nature of ion, concentration of ion, concentration of clay, etc. The
86 Caillere, S., S. Henin, and J. Mering, Experimental Transformation of Mont-
morillonite to a Phyllite of Stable c-Distance of 14 A, Compt. rend., 224, 842-843 (1947).

86 Du Rietz, C., "Das Ionenhindungsvermogen fester Stoffe," Stockholm(1938).
87 Vageler, P., and J. vVoltersdorf, Beitrage zur Frage des Basenaustausches und
der Aziditaten, Z. Pjlanzenerniihr. Diingung u. Bodenk., 15, 329-342 (1930).

88 Freundlich, H., "Kappillarchemie," 2d ed., Akademische Verlagsgesellschaft,
Leipzig (1922).
8, Kerr, H. W., The Nature of Base-Exchange and Soil Acidity, J. Am. Soc. Agron.,
20,309-335 (1928).
90 Vanselow, A. P., Equilibria of the Base-Exchange Reactions of Bentonites,
Permutite, Soil Colloids, and Zeolites, Soil Sci., 33, 95-113 (1932).
91 Rothmund, V., and G. Kornfeld, Der Basenaustausch im Permutit, Z. anorg. u.
allgem. Chem., 103, 129-162 (1918).

92 Davis, L. E., Theories of Base-Exchange Equilibrium, Soil Sci., 59,379-395 (1945).
93 Gapon, E. N., Theory of Exchange Adsorption in Soils, J. Gen. Chem. U.S.S.R.,
3, 144-152, 153-158 (1933).

94 Davis, L. E., Simple Kinetic Theory of Ionic Exchange for Ions of Unequal
Charge, J. Phys. Chem., 49, 473-479 (1945).
~'. '~
- .... -
Ion Exchan~
application of any such equations would probably need to be restricted to

similar kinds of exchange material, to a given range of concentrations,
- 155
and perhaps to other factors.

Determination of Cation-exchange Capacity and Exchangeable Cat-
ions. It should be obvious from the foregoing discussion of the factors
controlling cation exchange in clay minerals that accurate determinations
of cation-exchange capacity and exchangeable cations are very difficult
to accomplish. Literally dozens of methods have been suggested, and '"
tremendous amount of time has been spent on this problem. Kelley!
and Peech et al. 95 ,96 have considered in detail the various methods and
have pointed out various pitfalls.
The determination of cation-exchange capacity is at best a more or less
arbitrary matter, and no high degree of accuracy can be claimed. The
measurement is generally made by saturating the clay with NHt or Ba++
and determining the amount held at pH 7. In the absence of water-
soluble or slightly soluble salts, and for a clay mineral which itself is not
moderately soluble, the determination is not very difficult. However,
even for such materials, if the clay mineral is degraded illite or degraded
chlorite, it is difficult frequently to separate exchangeable from non-
exchangeable ions. Also in acid clays the presence of AP+, or in fer-
ruginous clays the presence of Fe3+, tends to partially clog some of the
exchange positions and make the determination difficult. The determi-
nation is particularly difficult for relatively soluble clay minerals, such as
the high-magnesium montmorillonites and the palygorskites. Trust-
worthy determinations of cation-exchange capacity can be accomplished
only by a skilled analyst who is aware of the fundamental causes of the
difficulties that beset the problem.
Determination of the exchangeable-cation composition is also a very
difficult problem. It involves the complete replacement of all exchange-
able cations by some cation which is not present in the sample, accurate
analysis of the solution obtained, and determination of and suitable
correction for the cations that pass into solution from any soluble sub-
stances that may be present or from the decomposition of some insoluble
material in the sample. Again the problem is particularly difficult for
clay minerals of relatively high solubilities, for acid clays, and for mate-
rials with a fairly high content of soluble or moderately soluble salts.
The use of the flame photometer for the analysis of the ions in the
replacing solution has in recent years provided satisfactory data in a very
short time, particularly for some of the alkali ions and calcium.
96 Peech, M., L. T. Alexander, and L. A. Dean, Methods of Soil Analysis for Soil-
Fertility Investigations, U.S. Dept. Agr. Dept. Circ. 757 (1947) .
6 Peech, M., Determinations of Exchangeable Cations and Exchange Capacity of
Soils, Soil Sci., 59, 25-38 (1945) ..

156 Clay Mineralogy
ANION EXCHANGE
It has been shown by Mattson,97 Ravikovitch,98 Scarseth,99 Toth,to

and many others 101 ,27 that the constituent minerals of many soil clays
exhibit anion-exchange reactions. The investigations of anion exchange
in clay materials have been to a considerable extent associated with
studies of the adsorption of phosphate by soils.
The investigation of anion-exchange reactions in soils is very difficult,
primarily because of the possibility of the decomposition of the clay
minerals in the course of the reaction. Thus, in studies of the adsorption
of phosphate by kaolinite, there has been considerable argument as to
whether many of the results observed are due to adsorption, to replace-
ment of OR ions in the kaolinite lattice by phosphate ions, or to a reaction
between the phosphate !wd alumina produced by the destruction of the
kaolinite lattice.
There seem to be two, and possibly three, types of anion exchange in the
clay minerals, as follows:
1. Replacement of on ions, as has been suggested by many authors
for the phosphate adsorption by kaolinite. Buswell and Dudenbostel 40
presented strong evidence, based on infrared absorption, that this can
take place, and recently McAuliffe and coworkers,I02 using deuterium-
tagged hydroxyls, have shown conclusively i;hat the OR ions of clay-
mineral surfaces can enter into exchange reactions.
Dickman and BraylOS have presented very clear evidence for the
replacement of hydroxyls by fluorine in kaolinite. They showed that
the liberation of hydroxyls caused a well-marked increase in the alkalinity
of the suspensions. The OI;[ and fluorine ions are of about the same size,
and the exchange would involve no lattice rearrangement. In the case
of exchange due to replacement of OR ions, the extent of the reaction
depends on the accessibility of the OR ions, and in general the only factor
97 Mattson, S., The Laws of Soil Colloidal Behavior, VI, Amphoteric Behavior, Soil
Sci., 32,343-365 (1931).

98 Ravikovitch, S., Anion Exchange, I, Adsorption of Phosphoric Acid Ions by Soils,
Soil Sci., 38, 219-239, 279-290 (1934).

99 Scarseth, G. D., The Mechanism of Phosphate Retention by Natural Alumina-
Silicate Colloids, J. Am. Soc. Agron., 27, 596-616 (1935).

100 Toth, S. J., Anion Adsorption of Soil Colloids in Relation to Changes in Free
Iron Oxides, Soil Sci, 44, 299-314 (1937).

101 Dean, L. A., and E. J. Rubins, Anion Exchange in Soils, I, Soil Sci., 63, 377-406
(1947).
1M McAuliffe, C. D., M. S. Hall, L. A. Dean, and S. B. Hendricks, Exchange Reac-
tions between Phosphates and Soils, Soil Sci. Soc. Am, Proc., 12, 119-123 (1947),
103 Dickman, S. R., and R. H. Bray, Replacement of Adsorbed Phosphate from
Kaolinite by Fluoride, Soil Sci., 62, 263-275 (1941).

Ion Exchange 157
preventing complete substitution is the fact that many OH ions are
within the lattice and, therefore, not accessible.
2. Hendricks 104 has suggested that another factor in anion exchange is
the geometry of the anion in relation to the geometry of the clay-mineral
structure units. Anions such as phosphate, arsenate, borate, etc., which
have about the same size and geometry as the silica tetrahedron, may be
adsorbed by fitting on to the edges of the silica tetrahedral sheets and
growing as extensions of these sheets. Other anions such as sulfate,
chloride, nitrate, etc., because their geometry does not fit that of the
silica tetrahedral sheets, cannot be so adsorbed.
In the case of both types 1 and 2 listed above, anion exchange would
take place around the edges of the clay minerals. In contrast to cation
exchange, as in montmorillonite, there would be substantially no anion
exchange on the basal plane surfaces of any of the clay minerals. For
these types of anion exchange, and in the case of the clay minerals like
kaolinite in which cation exchange is due to broken bonds, the cation- and
anion-exchange capacities should be substantially equal, and Dean and
Rubins 101 have shown this to be the case. In the case of montmoril-
lonite and vermiculite, in which cation exchange is due mostly to lattice
substitutions, anion capacity should be only a small fraction of the cation-
exchange capacity. In the case of illite, chlorite, and the sepiolite-
palygorskite minerals, anion-exchange capacity should be slightly less
than cation-exchange capacity.
Dean and Rubins 101 have shown that anion-exchange capacity of the
clay mineral in the Sassafrass soil is proportional to the surface area.
In addition, anion-exchange capacity would be expected to vary with the
degree of crystallinity of the clay minerals. Thus, there should be a
difference in the anion-exchange capacity between well-crystallized and
poorly crystallized kaolinite. In the latter, because of imperfections in
the stacking of the alumina-silica layers, there would be considerably
more exposed OR ions available for anion exchange.
Schofield 105, 106 has suggested a third manner of anion exchange.
According to Schofield, the clay minerals may have anion-exchange spots,
as well as cation-exchange spots, on basal plane surfaces. Such active
anion-exchange positions would be due to unbalanced charges within the
lattice, e.g., an excess of aluminum in the octahedral positions. It is
difficult to see how this could be, since positive and negative deficiencies
would tend to balance each other, unless they recurred at considerable
104 Hendricks, S. B., personal communication.

105 Schofield, R. K., Clay Mineral Structures and Their Physical Significance,
Trans. Ceram. Soc. (Engl.), 39, 147-158 (1940).
106 Schofield, R. K., Calculation of Surface Area of Clays from Measurements of
Negative Adsorption, Trans. Ceram. Soc. (Engl)., 48, 207-213 (1949).
158 Clay Mineralogy
distances from each other. That is to say, there would be either positive
or negative deficiencies, but not both. Schofield has indicated that in
strongly acid solutions clays can take up both potassium and chloride
ions, the former greatly predominating, and that both the K+ and Cl-
are exchangeable for other cations and anions. Mattson,107 on the other
hand, has inter-preted experiments on the Donnan equilibrium in such
a way as to show an apparent negative adsorption of chloride and sulfate.
Further experimentation is necessary to resolve this matter, and indeed
much remains to be learned about the whole subject of anion exchange
in the clay minerals.
A further factor complicating anion-exchange studies is that any free
or exchangeable iron, aluminum, or alkaline earths present in the clay
may form insoluble salts with the anions, and it is very difficult to separate
the effects due to such reactions from those which may be due to reaction
with the clay minerl1ls.
Phosphate Fixation. Students of soils have known for a long time that
some of the phosphate added to soils in fertilizers is frequently converted
to an insoluble form and fixed in the soil. Such fixation can be due to the
formation of insoluble salts of iron, aluminum, or the alkaline earths.
The formation of phosphates of aluminum and iron probably explains
to a considerable extent the results of soil-phosphate fixation experiments
carried out at low pB, and the formation of phosphates of alkaline earths
probably accounts for many of the results obtained at a high pH.
Black/os Murphy,109 Scarseth,99 and Stout llO have shown that phos-
phate fixation also is carried out by the clay minerals, and this probably
accounts lor the pho\30phate that is fixed at about neutrality. The fixation
power of the various clay minerals is relatively low, and it would seem
from the work of Murphy that kaolinite, or possibly halloysite, has the
highest power of all the clay minerals for phosphate fixation. Data
presented by Steele 1ll indicate that the phosphate retained at pH 7 for a
montmorillonitic soil is about 0.03 millimole per gram. For a kaolinitic
soil it is about 0.07 millimole per gram. The other clay minerals would
probably retain about as much phosphate as the montmorillonites.
In the case of kaolinite, phosphate fixation increases with decreasing
107 Mattson, S., The Laws of Soil Colloidal Behavior, I, Soil Sci., 28, 179-201 (1929).
108 Black, C. A., Phosphate Fixation by Kaolinite and Other Clays as Affected by
pH, Phosphate Concentration, and Time of Contact, Soil Sci. Soc. Am. Proc., 7, 132-
137 (1943).
109 Murphy, H. F., ClfLY Minerals and Phosphate Availability, Proc. Oklahoma Acad.
Sci., 20, 79-81 (1940).

110 Stout, P. P., Alterations in the Crystal Structure of the Clay Minerals as a
Result of Phosphate FiJ{ation, Soil Sci. Soc. Am. Proc., 4, 177-182 (1939).
111 Steele, J. G., The Effect of Other Anions on the Retention of Phosphate by Col-
loidal Clays, Ph.D. thesis, Ohio State University (1934).
t.
Ion Exchange 159
particle size, and Stout llO has shown that fixation greatly increases in
finely ground kaolinite. However, it may well be that the fine grinding
of the kaolinite has so disrupted its structure that the results obtained are
not a simple matter of clay-mineral fixation.
The manner in which phosphate fixation takes place in clay minerals
is probably a matter of sorption of the phosphate ion in the way developed
in discussing anion exchange, with the portion that is fixed being adsorbed
in relatively inaccessible positions in the clay-mineral structure.
Bower, C. A., and E. Truog, Base Exchange Capacity Determinations as Influenced

by Nature of Cation Employed and Formation of Base-Exchange Salts, Soil Sci.
Soc. Am. Proc., 5, 86-90 (1940).
Brammall, A., and J. G. Leech, Base-Exchange and Its Problems, Sci. J. Roy. Call.
Sci., 8, 43-51 (1938).
Chaminade, R., and G. Drouineau, Recherches sur la mecanique chemique des cations
exchangeables, Ann. agron., 6,677-690 (1936).
Ganguly, A. K., Reactivity of Exchange Spots of Silicate Minerals, Science and Culture,
14,341-342 (1949).
Gedroiz, K., Die Lehre von Adsorptionsvermiigen der Boden, Kolloidchem. Beihefte,
33, 317-448 (1931).
Hendricks, S. B., Base Exchange of Crystalline Silicates, Ind. Eng. Chem., 37, 625-630
(1945).
Hissink, D. J., Beitrag zur Kenntnis der Adsorptionvorgange im Boden, Intern. Mitt.
Bodenk., 12,81-172 (1922).
Hissink, D. J., Base Exchange in Soils, Trans. Faraday Soc., 20, 551-569 (1925).
Hofmann, U., and W. Bilke, Ueber die innerkristalline Quellung und das Basenaustau-
schvermiigens des Montmorillonits, Kolloid-Z., 77,238-251 (1936).
Hofmann, U., and A. Hausdorf, Kristallstruktur und innerkristalline Quellung der
Montmorillonits, Z. Krist., 104,2.05-293 (1942).
Kelley, W. P., Modern Concepts of Soil Science, Soil Sci., 62, 469-476 (1946).
Kelley, W. P., and S. M. Brown, Replaceable Bases in Soils, Calif. Agr. Expt. Sta.
Tech. Paper 15 (1924).
Kielland, J., Thermodynamics of Base Exchange Equilibria of Some Different Clays,
J. Soc. Chem. Ind. (London), 54, 232-247 (1935).
Kurtz, T., E. E. DeTurk, and R. H. Bray, Phosphate Adsorption by Illinois Soils,
Soil Sci., 61, 111-124 (1946).
Lemberg, J., Ueber Silicatumverwandlungen, Z. deut. geol. Ges., 28, 519-621 (1876).
Marshall C. E., and R. S. Gupta, Base Exchange Equilibria in Clays, J. Soc. Chem.
Ind. (London), 52, 433-443 (1933).
Mattson, S., The Laws of Soil Colloidal Behavior, V, Ion Adsorption and Exchange,
Soil Sci., 31, 311-331 (1931).
Mattson, S., and V. Gustafson, The Electrochemistry of Soil Formation, I, Gel and
Soil Complex, Ann. Roy. Agr. Call. Sweden, 4, 1-54 (1937).
Mehlich, A., Soil Properties Affecting the Proportionate Amounts of Calcium, Mag-
nesium, and Potassium in Plants and HCI Extracts, Soil Sci., 62,373-405 (1946).
Mitchell, R. L., Base Exchange Equilibria in Soil Profiles, J. Agr. Sci., 27, 557-568
(1937).
160 Clay Mineralogy
Mitra, R. P., Broken Bonds as Seats of Ion Exchange in Crystalline Silicates, J.
Indian Chem. Soc., 23,386-387 (1947).
Murphy, H. F., The Role of Kaolinite in Phosphate Fixation, Hilgardia, 12,341-382
(1939).
Peters, C., Ueber die Absorption von Kali durch Ackererde, Landw. Verso Sta., 2, 131-
151 (1860).
Schallenberger, C. J., and R. H. Simmons, Determination of Exchange Capacity and
Exchangeable Bases in Soils, Soil Sci. 59, 13-24 (1945).
Sieling, D. H., Role of Kaolin in Anion Sorption and Exchange, Soil Sci. Soc. Am.
Proc., 11, 161-170 (1946).
Vageler, P., "Der Kationen und Wasserhaushalt des Mineralbodens," Julius Springer,
Berlin (1932).
Van der Muelen, J. B., Relation between the Phenomenon of Cation Exchange with
Silica-Alumina Complexes and Their Crystal Structure, Rec. trav. chim., 54, 107-
113 (1935).
Weyl, 'V. A., Surface Structure and Surface Properties of Crystals and Glasses, J.
Am. Ceram. Soc., 32, 367-374 (1949).
Wiegner, G., Disperstat und Bascnaustausch (Ionenaustausch), Kolloid Z., 36, 341-
369 (1925).
Wiegner, G., Some Physico-Chemical Properties of Clays, J. Soc. Chem. Ind. (Lon-
don), 50,65-71, 105-112 (1931).
'f
CHAPTER 8
Clay-Water System
This chapter is concerned with water which can be held by clay mate-
rials only at relatively low temperatures and is driven off by heating to
about 100 to 150C. The OH lattice water which is lost at temperatures
above about 300C is considered elsewhere (Chap. 9). The nature of the
low-temperature water and the factors that control its characteristics are
of great importance, since they largely determine the plastic, bonding,
compaction, suspension, and other properties of clay materials, which in
turn frequently control their commercial utilization. Thus, an under-
standing of this low-temperature water in relation to the clay minerals
must precede an understanding of the plastic and many other properties
of clay materials.
The water lost at low temperatures may be classed in three categories,
depending on its relation to the mineral components and to the texture
of the clay materials, as follows: (1) the water in pores, on the surfaces,
and around the edges of the discrete particles of the minerals composing
the clay material; (2) in the case of vermiculite, montmorillonite, and the
hydrated form of halloysite, the interlayer water between the unit-cell
layers of these minerals (this is the water which causes the swelling of
montmorillonite); and (3) in the case of the sepiolite-attapulgite-
palygorskite minerals, the water which occurs within the tubular opening
between the elongate structural units. Type 1 water requires generally
very little energy for its removal, and drying at only a little above ordi-
nary room temperatures is adequate for its substantially complete elimi-
nation. Water of types 2 and 3 requires definite energy for its complete
removal. In the case of the hydrated form of halloysite, drying at room
temperature is adequate to remove most of the interlayer water, but
higher temperatures (see pages 217-218) are required for total removal.
In the case of the vermiculite and montmorillonite minerals, temperatures
at least approaching 100C are necessary for substantially complete
elimination of interlayer water. At this temperature some little time is
required for complete removal of this water, whereas at somewhat higher
temperatures it is driven off rapidly. In the case of halloysite, the
reaction is not reversible and the hydrated mineral ordinarily cannot be
fmmed again. Vermiculites and montmorillonites rehydrate with
161
162 Clay Mineralogy
difficulty if the dehydration is absolutely complete, but easily if only a
trace of water is allowed to remain between the unit layers. The water
in the channels of the sepiolite-attapulgite-palygorskite minerals is lost
at about the same temperature as the interlayer water of the layer clay
minerals. The channel water is regained readily if removal is by drying
only at low temperatures. The energy necessary for dehydration and the
precise temperatures at which it occurs are shown for the various clay-
mineral groups by differential thermal and dehydration analyses in
Chap. 9.
NATURE OF ADSORBED WATER
Langmuir,! Terzaghi,2 Hardy,3 Baver,4 Winterkorn,5 and many others 6

have presented evidence to show that the water held directly on the
surfaces of the clay particles is in a physical state different from that of
liquid water. It has frequently been considered to be denser and more
viscous than ordinary liquid water. The specific characteristics of this
water which delimit it from ordinary water would probably be restricted
ordinarily to relatively short distances from the clay-particle surfaces,
distances generally of the order of three to ten molecular layers of water,
i.e., 8 to 28 A. The possible thickness of the nonordinary water can vary
a good deal even for a given clay mineral, and the transition from non-
ordinary water to ordinary water can be abrupt or gradual, depending
on factors that will be considered presently.
It appears certain that the possible thickness of the nonordinary, or
so-called nonliquid, water is relatively small on irregular surfaces, such as
those around the edges of clay-mineral particles, and relatively large on
the flat surfaces of the clay minerals. The film of non ordinary water is
best developed and appears to reach its greatest thickness on the basal
plane surfaces of the expanding-Iatti.ce minerals of the montmorillonite
group. It follows that the water in pores would be substantially liquid
water, with nonliquid water forming only a thin film on the surface of the
pores and where adjacent clay-mineral particles come together.
1 Langmuir,!., The Constitution and Fundamental Properties of Solids and Liquids,

J. Am. Chem. Soc., 39,1848-1906 (1917).
2 Terzaghi, K., The Physical Properties of Clays, Tech. Eng. News MIT, 9, 10, 11,
36 (1928).
~ Hardy, W. E., Friction, Surface Energy, and Lubrication, "Colloid Chemistry,"
J. Alexander, ed., vol. 1, pp. 301-330, Reinhold, New York (1926).
4 Baver, L. D., and H. W. Winterkorn, Sorption of Liquids by Soil Colloids, II,
Surface Behavior in the Hydration of Clays, Soil Sci., 40, 403-418 (1936).
5 Winterkorn, H. W., The Condition of Water in Porous Systems, Soil Sci., 55, 109-
115 (1943).
6 Grim, R. E., and F. L. Cuthbert, Some Clay-Water Properties of Certain Clay
Minerals, J. Am. Ceram. Soc., 28, 90-95 (1945).

Clay-Water System 163
Terzaghi 2 and many others have explained the initially adsorbed water
on the basis of the dipole character of the water molecule; the latter
possesses positive and negative charges, the centers of which do not
coincide. Since the surface of the clay particle is normally negatively
charged, the positive ends of the water molecules are considered to lie
toward the clay surface with the negative ends extending outward. The
initial layer of water is believed to consist of water molecules all oriented
in the same direction. According to this concept, the first layer of
oriented-dipole water molecules forms another surface of negative
charges on which can be built another layer of completely oriented water
molecules. This process of building up layers could be continued
indefinitely were it not for the fact that the water molecules possess
thermal energy and tend to be in a state of continuous motion. In
accordance with classical concepts of colloidal theory, the motion due to
thermal energy will oppose the regular orientation. Therefore, at any
given instant and at a certain distance from the surface some of the
molecules will be oriented at right angles to the surface. An instant
later these molecules will have moved, but others will have become
oriented in their place. At the actual clay-mineral surface the molecules
will be highly oriented, and the degree of orientation will decrease going
outward, as the relative effect of thermal movement becomes greater.
Macey 7 has pointed out difficulties encountered by this concept, par-
ticularly in view of the facts that the clay-mineral surface is not a uni-
formly charged plane and that the water molecules strictly do not act as
little rods with positive and negative ends.
Earlier investigations, mainly by Bernal and FowlerS and Bernal and
Megaw,9 of the hydrogen bond and of the distribution of charges about a
water molecule led Hendricks and Jefferson 10 to a concept of the nature
of the initially adsorbed water that is based on an orientation of water
molecules with a structure tied to the configuration of the oxygens or
hydroxyls in the basal surface layers of the unit cells of the clay minerals.
In the language of Hendricks and Jefferson, 10 a water layer is composed of
water molecules joined into hexagonal groups of an extended hexagonal
net, as shown in projection in Fig. 54. This arrangement is partly a
result of a tetrahedral distribution of charge about a water molecule; two
7 Macey, H. H., Clay-Water Relationships and the Internal Mechanism of Drying,
Trans. Ceram. Soc., (Engl.) 41,73-121 (1942).
8 Bernal, J. D., and A. H. Fowler, A Theory of Water and Ionic Solution with Par-
ticular Reference to Hydroxyl Ions, J. Chem. Phys., 1, 515-548 (1933).
9 Bernal, J. D., and H. D. Megaw, The Function of Hydrogen in Intermolecular
Forces, Proc. Roy. Soc. (London) A, 161, 384-420 (1935).

10 Hendricks, S. B., and M. E. Jefferson, Structure of Kaolin and Talc-Pyrophyllite
Hydrates and Their Bearing on Water Sorption of Clays, Am. Mineral. 23, 863-875
(1938).
164 Clay Mineralogy
--~~~~G--~-'
Q\
----~~---
'.I r -- r ,
--
\
, ,
I
\ ,
, \
, , ,
\
\ _-~-~L_ \ --~~_j__ \_
- Z}-o------_\W -O-o--~--- \ -(}..o_-
I -~-'-.r I . ----~-, - :-

I
I
"
I
.
Q', ,
I
I
FIG. 54. Configuration of water adsorbed directly adjacent to the basal plane sur-
faces of the clay minerals, after Hendricks and Jeffe!son.'o
---< ~~---
'~--~
Q '
I
o"?"j
\ ~j 'M ~I
,'po--'1; ,~--Q
, :~--<% \
---~ ~---
\\ '
o Oxygens
o Hydrogens in plane of oxy;ens
() Hydrogens below plane of oxygens
FIG. 55. Arrangement of oxygens and hydrogens in water net, after Hendricks and
Jefferson. 10
corners of the tetrahedron being occupied by hydrogen atoms and the

other two corners by an excess of electrons. Each side of the hexagon
(Fig. 54) must correspond to a hydroxyl bond, the hydrogen bond of one
water molecule being directed toward the negative charge of a neighboring
molecule. One-fourth of the hydrogen atoms, or a hydrogen atom of
half of the water molecules, are not involved in bonding within the net
(K, M, and 0 of Fig. 55). The net is tied to the surface of the clay
minerals by the attraction between those hydrogen atoms not involved in
bonding within the net and the surface oxygen layer of the clay-mineral
unit (Figs. 55 and 56). When the surface of the clay mineral contains
hydroxyl groups (e.g., halloysite and kaolinite), part of the hydroxyls are
free for bonding through hydrogen to oxygen atoms in the water layer.
Silica Loyer
of Cloy
Mineral o 0
OoQ
. . . I
I
\ _
I
\ \
I
I
. --~j-~----ln,j ~ ~~
--~~~~~~~--
Oriented Water " "
I
Molecules
--~~----~
--~~~~~----~~-1-~-C~----~--
I I I I I
I ' I I ,
, , I I I I I
, 60 10 I 00 1 f
Silko Loyer o I 0 0 /000
of Cloy
Mineral
o 0
FIG. 56. Configuration of water net proposed by Hendricks and Jefferson,'o showing
the binding through hydrogens to the adjacent clay-mineral surfaces.
The suggested 10 superposition of the water net and basal layers of oxygens
and hydroxyls, together with the type of bonding between them, is
shown in Figs. 57 and 58, respectively.
The net has just the a and b dimensions of the silicate layer minerals,
if the separation of the oxygen atoms of the water molecules is about 3.0 A
in projection. It is assumed that the oxygen atoms are in one plane.
In this configuration, there is relatively loose packing of the water mole-
cule, there being four water molecules per molecular layer per unit cell
of the clay mineral instead of six, as would be the case in close packing.
The stability of the layer of water molecules arises from the geometrical
166 Clay Mineralogy
FIG. 57. Superposition of water net on the basal oxygen layer of vermiculite, sug-
gested by Hendricks and JeffersonY Large dltshed circles (A, B, C, etc.) are oxygen
atoms of clay-mineral surface, 2.73A below plane of water molecules,

FIG. 58. Superposition of water net on the basal hydroxyl layers of two-layer clay
minerals and the type of bonding between hydroxyl groups and water layers, after
Hendricks and JeffersonY At L, N, and P hydrogens of hydroxyl groups are free
for binding to oxygells of water molecules.
relationship to the oxygen atoms or hydroxyl groups of the silicate frame-
work. The presence of the first layer would favor the formation of a
second layer, and the water structure would thus be propagated away
from the clay-mineral surface. At some distance from the surface the
dissociation pressure of successive layers, considered as hydrates, would
finally approach the vapor pressure of water at the temperature of
observation, and the oriented water net would cease to develop.
In the Hendricks and Jefferson 10 configuration every other water
molecule lies about over an oxygen of the surface layer of the three layer
clay minerals (Fig. 57). Since half of the water molecules have hydrogen
available for vertical bonding, the directly superimposed oxygens and
water molecules may be assumed to be tied together through these
hydrogen bonds. As successive molecule layers develop, the exterior
layer can be tied to the layer next below by a hydrogen from every other
water molecule. According to this scheme no hydrogens would be
available for tying together series of water layers growing outward from
two neighboring clay-mineral surfaces, unless the directly adjacent water
layers of two water envelopes were tied together by relatively fewer
bonds. A plane or planes of relatively weakly bonded water molecules
would exist, therefore, at the junction of two water envelopes.
The Hendricks and Jefferson 10 concept of the structure of the adsorbed
water has been criticized by Hofmann and Hausdorf,l1 Mackenzie,12 and
Walker,13 because it neglects the probable influence of adsorbed cations,
which in many cases are to be found directly on or at least very close to
the oxygen surface on which the water configuration develops. Further,
some of these cations are hydrated and, therefore, would probably
significantly influence the arrangement of the water molecules directly
adjacent to them. Mackenzie has particularly emphasized this point
and has indicated that the matter of cation hydration would be particu-
larly important in the part of the adsorbed water net closest to the clay-
mineral surfaces. In the case of montmorillonite, there would be about
one monovalent cation for each two hexagonal configurations of water
in the first molecular water layer. According to Mackenzie, there is a
space problem in fitting such a number of ions into the water net. Mac-
kenzie presents computations of the energy of hydration of the sheet
surface of montmorillonite carrying various ions; these data are in accord
with his suggestion that, at low water contents, water sorption depends
primarily on the exchangeable ion present, the sheet surface being of
11 Hofmann, U., and A. Hausdorf, Kristallstruktur und innerkristalline Quellung
der Montmorillonits, Z. Krist., 104, 265-293 (1942).

12 Mackenzie, R. C., Some Notes on the Hydration of Montmorillonite, Mineralog.
Soc. Gr. Britain Clay Mineral Bull. 4, pp. 115-120 (1950).

18 Walker, G. F., 'Vater Layers in Vermiculite, Nature, 163,726 (1949).
168 Clay Mineralogy
subsidiary importance. Others, notably Williamson,14 have objected
to the Hendricks and Jefferson concept, because of their feeling that the
sorption forces are strong enough to cause a dense packing of at least the
initially adsorbed water molecules.
@ @ @
@ 0 @
@ @ @
0 @ 0
@ @ @
0 ~-i~
Ilo
@ ~ @
(0)
Oxygen layers forming surface of sllicote

OOxygen atoms of ice
FIG. 59. (a) Superposition of the ice lattice on the surface oxygen layer of the clay
minerals, after Macey.7 (b) Structure of ice.
Macey,7 arguing from the similarity between the structure of ice and
of the oxygen atoms exposed at a sheet surface of the layer clay minerals,
has postulated that the initially adsorbed water has the structure of ice.
He considers that it fits on top of the oxygen net of the basal plane of
14 Williamson, W.o., The Physical Relationship between Clay and Water, Trans.
Ceram. 801)., (Engl.) 50, 10-34 (1951).
the three-layer clay minerals, as shown in Fig. 59A. The fit of the water
molecules with the oxygen net as suggested by Macey is different from
that suggested by Hendricks and Jefferson. According to the Macey
concept, the distance between oxygens in the water layer would be 4.52 A,
and the packing would be even looser than that suggested by Hendricks
and Jefferson.lO There would be three molecules of water per unit cell
per molecular layer. Such loose packing of water in a given layer would
be predicated on hydrogen binding to additional water not in the layer
under consideration. As illustrated in Fig. 59B, showing the structure
of ice, other water molecules would be just out of the plane in contact
with the silicate structure, and there is no epitaxial arrangement between
them and the silicate surface. In the Hendricks and Jefferson concept,
the ice structure is stretched so that the offset water molecules come into
the same plane and there is no change in the hydrogen binding. Also
the stretch permits a complete epitaxial arrangement of all the water
molecules and the silicate surface. The same objections raised against
the Hendricks and Jefferson structure can be raised against the Macey
structure. However, the space problem caused by the presence of
adsorbed cations might be less acute. Forslind 15 has reported electron-
diffraction data which seem to indicate a structure at least similar to ice
in the initially adsorbed water of montmorillonite. Very recently
DeWit and Arens 16 have published some density measurements that
seem to be in agreement with the Macey concept.
Barshad 17 has very recently suggested another concept of the nature
of the adsorbed water on the basis of careful dehydration determinations.
According to him, at very low states of hydration for montmorillonite, the
water molecules tend to form tetrahedrons with the oxygens of the top
layer of the linked silica tetrahedrons of the lattice. This type of packing
would give rise to hexagonal rings of water molecules which are similar
to the hexagonal rings of oxygens of the vertices of the linked silica tetra-
hedrons of the individual silicate sheets. In Fig. 60, a to f represent such
water molecules forming tetrahedral units with oxygens of the underlying
silica tetrahedral network. The packing in this configuration would be
loose, as there would be only four molecules of water per unit cell per
molecular layer, and the height added for a single layer of water mole-
cules would be 1.78 A, according to Barshad. Ii appears that this value
16 Forslind, E., The Crystal Structure and Water Absorption of the Clay Minerals,
Trans. 1st Intern. Ceram. Congr., pp. 98-110, (1948).
16 DeWit, C. T., and P. L. Arens, Moisture Content and Density of Som.e Clay
Minl'rals and Some Remarks on the Hydration Patterns of Clay, Trans. 4th Intern.
Congr. Soil Sci., 2, 59-62 (1950).
17 Barshad, I., The Nature of Lattice Expansion and Its Relation to Hydration in
.tiontmorillonite and Vermiculite, Am. Mineral., 34,675-684 (1949).

170 Clay Mineralogy
should be 2.1 A, using Barshad's value of 2.55 A for the water molecules,
which reduces the :1greement of Barshad's computed and analytical data
for c dimensions at certain stages of hydration. Barshad postulates that
at higher states of hydration the water adsorbed by montmorillonite
tends to form hexagonal rings of water
ill.\)\'<~'li\.<'b., tb.<'b~ ""y~ 'b)ill.)\.'b,y t\) tb.l
hexagonal rings of the oxygens of the
montmorillonite basal plane which
forms the bases of the linked silica
tetrahedrons. In Fig. 60, 1 to 6 repre-
sent such water molecules. In this
configuration the packing is more
dense, and there are six molecules
of water per unit cell per layer of
water molecules. The height added
for a single water layer would be
FIG. 60. Arrangement of the water
about 2.55 A, since the water mole-
molecules in the interlayer space of
montmorillonite and vermiculite, after
cules would be directly superimposed
Barshad. '7 on the oxygens. At still higher states
of hydration, it is believed that water
molecules fill everJ. the centers of these hexagonal water rings and
the centers of the hexagonal oxygen rings of the linked silica tetrahedrons
not occupied by exchangeable ions. Such water layers would consist of
closely packed water molecules, and there would be eight water molecules
per unit cell per molecular layer. Barshad 17 states that adsorption does
nl)t "prl)gn~BB beYl)no tb.~ 'bt""t~ \)l t,,'V\) ill.\)\.~~'li\.'b,y \.'b,y~y'b 'lint)\. \\~ll'b~ I>",,~k)w~,
is accomplished. It must be pointed out that Barshad's interpretations
are based on dehydration data, which require extremely pure mono-
mineral material. They would be subject to another interpretation if
the montmorillonite and vermiculite investigated consisted of mixed
layers with hydration layers of different thicknesses. Investigations by
Hendricks et al. 1S aod Mering 19 suggest the possibility of this alternative
explanation.
Thus, Mering has suggested that, in Ca montmorillonite, the initially
adsorbed water is packed octahedrally about the Ca++ and that, with
increasing available water, a double water layer of superimposed water
molecules first develops, followed by successive layers of superimposed
water molecules. In N a montmorillonite, increasing amounts of water
the Clay Mineral Montmorillonite Saturated with Various Ions, J. Am. Chem. Soc.,
62, 1457-1464, (1940).
19 Mering, J., The Hydration of Montmorillonite, Trans. Faraday Soc., 42B, 205-219
(1946).
develop successive layers of superimposed water molecules. It appears
that Barshad's concept is not substantiated geometrically as well as
those of Mering and Hendricks et al.
It is obvious that students of clay mineralogy are not in agreement
regarding the precise nature of the configuration of the initially adsorbed
water. There is, however, general agreement that this adsorbed water
has some sort of definite configuration.
DENSITY OF INITIALLY ADSORBED WATER
As indicated in discussing the configuration of the adsorbed water

molecules on the clay-mineral surfaces, rather large differences in the
denseness of packing are possible. The density of the adsorbed water
would vary, of course, depending on the nature of the packing. Accurate
determinations of the density of adsorbed water would provide evidence
for or against a particular configuration. Unfortunately, the determina-
tion of the density of the adsorbed water is a very difficult problem
experimentally, and although numerous determinations have been pub-
lished, none are completely satisfactory. Indeed the very wide variation
in the results obtained emphasizes their unsatisfactory nature.
Tscapek 20 has arrived at a value of 1.7, and Hauser and Le Beau,21
studying bentonite suspensions, present values greater than 1, suggesting
a dense packing. On the other hand Nitzsche 22 has presented determina-
tions giving values less than 1, and recently De Wit and Arens 16 have
obtained values as low as 0.73 for the density of the first few molecular
layers of water adsorbed by montmorillonites. The configurations which
have the greatest similarity in geometry to that of the oxygen layer of the
surface of the clay minerals, namely, those of Hendricks and Jefferson,l
Macey,7 and the low-hydration types of Barshad,17 require a specific
gravity less than 1.
EVIDENCE FOR THE CRYSTALLINE STATE OF THE INITIALLY

ADSORBED WATER
There is abundant evidence for some sort of definite configuration of

the water molecules initially adsorbed on the surfaces of the clay minerals.
The characteristics of water molecules themselves afford evidence for a
20 Tscapek, VV., The Density of Adsorbed Water in Soil3, Z. Pjlanzenerndhr. Diingung

u. Bodenk., 34, 265-271 (1934).
21 Hauser, E., and D. S. Ie Beau, Studies of Colloidal Clay, I, J. Phys. Chern., 42,
1031-1050 (1938).
22 Nitz~che, \V., On the Structure of the Hydration Hull of Inorganic Soil Colloids,
Kolloid-Z., 93, 110-115 (1940).
172 Clay Mineralogy
grouping into a definite network. Gieseking 23 has summarized this
point as follows:
The oxygen atom in water according to the work of Bernal and Fowler 8 and of
Cross et al. 24 tends to direct its attractive forces toward the four corners of a
regular tetrahedron. In the water molecules, most of the attractive force of the
oxygen atom will be directed towards the two corners of the tetrahedron where the
hydrogen atoms will reside, .but a small residual force will be directed towards
the two other corners of the tetrahedron. The total positive attractive force on
the two hydrogen atoms should exactly equal the total negative attractive force
on ~he divalent oxygen atom. For steric reasons according to Bernal and Fowler,
the two hydrogens cannot exactly neutralize the four spots on the oxygen atoms
so that the spots where the hydrogen atoms reside will carry a slight residual
positive charge and the other two corners of the tetrahedron should carry an
equally small residual negative charge. Water molecules, therefore, tend to
attract or sorb each other. In this way a molecule of water is subjected to less
strain if the residual positive side of the molecules has this positive charge neutral-
ized by the negative side of a neighboring molecule. This process, whereby
residual positive hydrogen rich spots of one molecule neutralize residual nega-
tive spots in neighboring molecules is called hydrogen bonding. According to
_ Bernal and Fowler, hydrogen bonding does not stop with the union of only two
molecules of water in the liquid phase, but from X-ray diffraction patterns of
water Fowler and Bernal 25 have concluded that a few tens or hundreds of water
molecules tend to be hydrogen bonded. . . . The oxygen ions in the surface of
the clay mineral crystals, like the oxygen atoms in the water molecules, will also
be weakly charged. These layers of oxygen atoms will direct most of their forces
backward into the crystal towards the positive silicon or metallic ions in the
interstitial spaces of the clay mineral crystal. Weak residual negative charges
will result on the outside layer of oxygen atoms in the clay mineral crystals,
which will be free to take part as donors in hydrogen bonding. Each oxygen if
free from other sorbed substances could attract the hydrogenated or positive side
of a water molecule.
The oxygen atoms in the clay-mineral surface have a definite pattern, and
consequently there would be a pattern of charges on this surface that
could be carried over into the adsorbed water molecules. Therefore, the
natures of the clay-mineral surface and of the water molecules both favor
the development of a definite configuration in the initially adsorbed water.
Strong evidence that the character of initially adsorbed water is
different from that of liquid water comes from a study of the properties
of clay-water systems. Thus Grim and Cuthbert2 6 have suggested a
23Gieseking, J. E., The Clay Minerals in Soils, Advances in Agron., 1159-202 (1949).
24Cross, P. C., P. Burnham, and P. A. Leighton, The Raman Spectrum and the
Structure of Water, J. Am. Chem. Soc., 59, 1134-1147 (1937).
26 Fowler, A. H., and J. D. Bernal, Notes on the Pseudo Crystalline Structure of
Water, Trans. Faraday Soc., 29, 1049-1056 (1933).
26 Grim, R. E., and F. L. Cuthbert, The Bonding Action of Clays, I, Clays in Green
Molding Sands; Illinois State Geol. Survey Rept. Invest. 102 (1945).
theory of the bonding action of clay and water in molding sands, based
on the nonliquid nature of initially adsorbed water, which accounts for
the strength and other properties of such materials and explains many
of their attribute@ which heretofore have not been understood. Grim~7
and many others 1.l 4 have shown that the plastic properties of clays are
d.\ft\<.'o\l\t ()\: \llij)()'6~ib\~ \,() ""'~~()\l\\\' l()\: 'Ni..\,h.()\\\, j)()'6\'\\\""'\'i..Wi!, '6()lli~ <.'oh."",\\'i!,~ i..R
the physical state of the adsorbed water. Thus, in the case of sand-
clay-water mixtures, the maximum bonding force is developed when the
sand and clay are wet with an extremely restricted amount of wate).".
With a very small amount less or more than the optimum amount of
water, the bond is greatly reduced. This can readily be explained by
considering that the maximum bond is developed when the mixture coI).-
tains the maximum amount of water that can develop a definite COI).-
figuration. Additional water will, at least in part, partake of the nature
of liquid water and greatly weaken the bond between the sand graini3.
Similarly, clays develop optimum plasticity when a definite amount of
water is added to a dry clay. If water is added in small increments to a
dry powdered clay, almost no experience is necessary to determine just
when enough water has been added to the clay to develop plasticity.
Again this phenor1lenon is readily explained by considering that plasticity
develops when jUi3t enough water is added to satisfy the requirements of
all available surf&ces for water with a definite configuration plus a little
more water which would develop little or no definite configuration of
water molecules. For a further discussion of this point, see Grim. 27
Russell 28 has pointed out that the structural characteristics of soils
\:~<\\li\:~ \'h."",t th.e i..nitiaUy ad.,arbed water be iR a RaRli<\\lid state.
The investigators of the phenomenon of heat of wetting in clay mate-
rials (see page 186) have generally attributed it, at least in part, to a
change in the nature of the water adsorbed on the surfaces of the clay-
mineral particles. Many soil investigators have shown that the freezing
point of the initially adsorbed water in soils is depressed. In other words
the soil contains a certain amount of water in a physical state that makes
it difficult to freeze. Bodman and D ay 29 have shown that mont
morillonite deprei3ses the freezing point to a greater degree than kao-
linite, which is ill accordance with the high water-sorption powers of
montmorillonite.
On the basis of an analysis of the phase relations of water, Winterkorn 6
has concluded thttt the water held directly on the clay-mineral surfaces
27 Grim, R. E., Some Fundamental Factors Influencing the Properties of Soil
Materials, Proc. 2d Intern. Congr. Soil Mech., 3,8-12 (1948).

28 Russell, E. W., Soil Structure, Imp. Bur. Soil Sci. Tech. Commun. 37 (1938).
29 Bodman, G. B., and P. R. Day, Freezing Points of a GrouD of California Soils
and Their Extracted Clays, Soil Sci., 66, 225-246 (1943).

174 Clay Mineralogy
must be solid and not liquid and that the change in state of the water with
distance from the clay-mineral surface is an exponential one.
THICKNESS OF ADSORBED NONLIQUID WATER
Because of the kinetic nature of the water molecules, adsorbed water

with a definite configuration of water molecules could be expected to
extend for only a limited distance from the clay-mineral surface. Numer-
ous attempts have been made to estimate the thickness of adsorbed
water with definite nonliquid characteristics, without very good agree-
ment. However, the values probably do indicate the correct order of
magnitude.
Houwink 30 has arrived at a value of 25 A, which is equivalent to about
10 molecular layers of water, and Mattson 31 has given the figure of 40 A,
equal to about 16 water layers. Grim and Cuthbert,26 from their study
of the relation of bonding strength to water content, concluded that, for
montmorillonite with N a + as the, exchangeable cation, the nonliquid
water has a thickness of three molecular layers and, for Ca montmoril-
lonite, the thickness is four molecular layers, or about 10 A. These same
authors also suggest that for N a m,ontmorillonite there is a gradual
transition to liquid water beyond the 7.5-A thickness, whereas in Ca
montmorillonite the transition to liquid water is abrupt. In N a mont-
morillonite some orientation of water molecules may persist to a distance
of a hundred or more angstroms from the montmorillonite surface,
whereas in Ca montmorillonite little or none persists beyond about 15 A.
An explanation of the variation in physical properties due to variations
in the exchangeable cations seems to require that the cations influence
the thickness of the oriented layers of water molecules as well as the per-
fection of the arrangement, and perhaps also the actual nature of the
configuration of the water molecules. It is a well-known fact that, other
things being equal, N a clays require less water to develop plasticity than
do Ca clays. This may be explained by the reduced thickness require-
ment for oriented water when Na+ is the dominant adsorbed cation.
In contrast to the relatively small thickness of nonliquid water sug-
gested by the previous authors, Jaeger32 and later SpieP3 have concluded
30 Houwink, R., "Elasticity, Plasticity and Structure of Matter," Cambridge, New
York (1937).
31 Mattson, S., Laws of Soil Colloidal Behavior, VII, Form and Function of Water,
Soil Sci., 33,301-322 (1932).

32 Jaeger, F. M., Viscosity of Liquids in Conncction with Their Chcmical and
Physical Constitution, Verhandel. Koninkl. Nederland. Akad. Wettenschap. Afdeel.

Natuurk., Sec. II, 16 (1938).
33 Spiel, S., Effect of Adsorbed Electrolytes on the Properties of Monodisperse
Clay-water Systems, J. Am. Ceram. Soc., 23,33-38 (1940).
that the thickness of adsorbed water at optimum plasticity for kaolinite
clays is of the order of several hundred angstroms. Opposed to these
very large values, De Wit and Arens 16 have recently concluded that the
thickness of adsorbed water held by kaolinite is to be measured in a very
few molecular layers.
Reasoning on structural grounds and froln the properties of clay-
water systems, it would seem unlikely that kaolinite would adsorb water
with a definite configuration to a greater thickness than would mont-
morillonite. The low values for the thickness of the nonliquid water for
kaolinite therefore seem most reasonable. It seems likely that the illite,
chlorite, and possibly also the vermiculite minerals would yield values
similar to kaolinite. No information is available on this characteristic
for the palygorskite-attapulgite-sepiolite minerals.
TIME FACTOR
Grim and Cuthbert 34 have shown that, when rammed samples of sand,
water, and certain clays are allowed to stand under certain conditions,
there is an increase in the compression strength without any loss of water.
This "air-set" strength is explained by them on the basis that a certain
amount of time is required for the water to penetrate to some of the
surfaces of the clay-mineral particles and for the water molecules to
assume their orientation. According to them, the compressive strength
is dependent on the development of nonliquid water.
These authors 6 emphasize the general fact that the properties of clay-
water systems may change with time without a change in water content
because of a change in the nature of the water. This is a matter of very
great importance in certain fields of the utilization of clays. For example,
in engineering practice, it makes it difficult to translate laboratory find-
ings on strength, sensitivity, etc., to actual field application because of
the difficulty of duplicating field conditions and of evaluating the time
- factor. Engineers are aware of this factor in their constant use of undis-
turbed samples with carefully retained natural moisture content for their
laboratory tests. Its significance is illustrated by the very different
values obtained for such properties as compressive strength and plasticity
on undisturbed and remolded samples.
INFLUENCE OF CATIONS ANI) ANIONS
The ions adsorbed on the surfaces of the clay minerals may affeCt the
adsorbed water in several ways:
34 Grim, R. K, and F. L. Cuthbert, The Bonding Action of Clays, II, Clays in Dry
Molding Sands. Illinois State Geol. Survey Rept. Invest. 110 (1946).
176 Clay Mineralogy
1. A cation may serve as a bond to hold the clay-mineral particles
together or to limit the distance to which they can be separated.
In general, multivalent cations have a greater tendency than univalent
cations to tie clay-mineral flakes together. The potassium ion is an
exception, probably because its size permits it to fit snugly in the hexag-
onal perforation of the oxygen net of the surface of the three-layer clay
minerals and because its coordination number is favorable. The ammo-
nium ion acts like the potassium ion. In the case of other common
monovalent cations, there would probably be little tendency for the ion
to tie the particles together, and under some conditions it might enhance
the repulsive force between the particles.
2. Certain of the adsorbed cations may become hydrated. The
hydrated ions would be enclosed in an envelope of water molecules prob-
ably havipg some sort of configuration. This configuration and its rela-
tion to the arrangement of water molecules on adsorbing surfaces some
distance from the hydrated ions would influence the over-all nature of the
arrangement of the water molecules and the thickness to which the
orientation CQuld develop. It is of interest and importance that the
hydrated calcium and magnesium ions apparently restrict the thickness
of the adsorbed water with a well-developed configuration, whereas in the
case of the sodium ion, which may not be hydrated,I8 oriented water can
grow to very great thicknesses. However, the lithium ion, which is also
hydrated, permits oriented water of great thickness to develop. Accord-
ing to Hendricks et al.,18 Li+ is surrounded by only three molecules of
hydration water, whereas Ca++ and Mg++ are surrounded by six mole-
cules of hydration water. Because of this difference in the amount of
hydration and perhaps also because of the small size of the lithium ion,
it would be expected to influence the normal net of water molecules to a
smaller degree than would calcium and magnesium ions.
3. The size of the adsorbed ions is of importance, as has been empha-
sized by Sullivan,35 Forslind,36 and others.12 Forslind, for example, has
suggested that small ions, which would fit into the net of water m~lecules
without disrupting it, would be expected to have a tightening effect on
the net which would enhance its development. Ions which would not
fit into the water net, because of their size, would be expected to retard
or even prevent its development.
4. The geometry of the adsorbed ions is probably also of importance,
in relation to the possible fit of the adsorbed ions into the water net
35 Sullivan, J. D., Physical Chemical Control of Properties of Clays, Trans. Electro-

chem. Soc., 75,71-98 (1939).
36 Forslind, E., Some Remarks on the Interaction between the Exchangeahle Ions
and the Adsorbed Water Layers in Montmorillonite, Trans. 4th Intern. Congr. Soil Sci.,
1,110-113 (1950).
'
and resulting disruption or weakening of the water configuration. It
seems that, in the case of some anions, such as the phosphate ion, this
matter of geometry is of particular importance. For example, it is well
known in the drilling-mud art that an extremely small amount of the
phosphate ion may greatly influence some of the thixotropic character-
istics of montmorillonite-water suspensions. The probable explanation
resides in the similarity between the tetrahedral nature of the phosphate
ion and the configuration of the water molecules; this similarity would
permit the phosphate tetrahedrons to almost fit into the water structure.
They would fit without completely disrupting the water structure but
would cause some strain and weakening of the water configuration, which
would account for the alteration of the thixotropic characteristics of the
system.
Hendricks, Nelson, and Alexander 18 have made a very careful study of
the water adsorbed by montmorillonite saturated with various cations.
They have determined the amount of water adsorbed at various relative
humidities, the resulting variation in the c-axis spacing of the mineral,
and the corresponding differential thermal curves. The differential ther-
mal curves show an endothermic peak due to the energy necessary to
drive off the adsorbed waters from the montmorillonite. Some of the
results of the work of Hendricks and his colleagues 18 are shown in
Figs. 61 and 62. The endothermic peak due to loss of adsorbed water
is single, double, or triple for certain of the cations prepared at
certain relative humidities. These workers interpret the composite
nature of the multiple peaks-and their interpretation is in accord with
their other data-as follows: In the case of dual peaks of samples prepared
at comparatively low relative humidities, the larger of the maxima is
attributed to the dehydration of the cation, and the lesser of the maxima
to the loss of water from the surface of the clay mineral away from the
hydrating cation. At higher relative humidities the dual peaks may
become triple, and the third maximum is attributed to the development
of an additional layer of oriented water molecules. Thus Mg mont-
morillonite shows a dual peak after adsorbing water at 5 per cent relative
humidity and a triple peak after being treated at only 10 per cent relative
humidity. Ba and Li montmorillonites show no evidence of the develop-
ment of a second water layer unless the relative humidity is at least
40 per cent.
C~rtain of the univalent ions, e.g., Na+, K+, and C;s+, show a single
peak after treatment at lew relative humidities, and this is interpreted
to mean that these ions are not hydrated. As has been mentioned else-
where, there is other evidence for the nonhydration characteristics of these
cations.
Hendricks, Nelson, and Alexander 18 conclude from their investigation
178 Clay Mineralogy
00
0",0<1" 00
.. ...
00<1"
000
",0<1"
_"'''' ~Nt\J _"''''
III III III
Magnesium
~,~
," '.
.085g .115g .160g .195 9 .22g .27 9 .34 9
.045 .040 .005 D70.o40 .040.000 .040 .D75.080.Q40
13.5 A 13.5 A 14.1 A 14.4 A 15.5 A 15.8A 15.8 A
III III III II I II I II I

Calcium
.070g .105 g .160 9 .20 9 .22 9 .27 9 .36 9

.035 .035 .060 .045 .Q35 .075.05(J [)75 .075 .050
12.8 A 13.2 A 14.9 A 15.0 A 15.0 A 15.SA
III III III "I
J\
Strontium
.050 9 .085 9 .140 9 .200g .22 9 .26 9 .34g

.010 .040 .035 .050 .02 .07 .05
12.1 A 12.3 A 13.2 A IS.I A IS.SA 15.SA IS.S A
0 0 o 0 o 0 0
00 000 000
08~
o
00<1" 00<1" 00<1"
~C\lC\J ~(\j(\j ~C\JC\l ~C\JC\J
III I II III II I
Barium
.050g .0759 .110g .140 9 .160g .23 9 .32 9

.005 .045 .030 .045 .065 .04~
11.9 A 12.0 A 12.IA 12.4A 13.2A 15.7 A 16.6A
5% 10% 25% 40% 50% 70% 90%
FIG. 61. Thermal-analysis curves, amounts of adsorbed water, and values of tl:
apparent cleavage space of magnesium, calcium, strontium, and barium salts (
Mississippi montmorillonite, after Hendricks, Nelson, and Alexander.'
o 0 0
000
<0 0 <;)"
_C\JC\J
III III III III

Lithium
A ~ N\ (\/l
A, "..-IX \ ) /{, \...)'-)iJ,) - ~
0f
.0659 .OS59 .1209 .155 9 .1709 . 24 9 .34 9
.015 .050 .030 .055 .060 .060 .030 .070 .055
12.3 A 12.3 A 12.3 A 13.0 A 13.0 A 15.0 A 15.5 A
III
.0 45 9 075 9 .130 9 .165 9 1909 .25 9 ,36 9

035 .095 .055 .110 080 ,110
Diffuse 12.3 A Diffuse 13.8 A 14.4 A 15S A 16.6 A
I I I III I I I II I I I I I I I [ [ I
.025 9 .040 9 .065 9 105 9 .120 9 .1759 .2S 9

.010 .055 .035 .070 .040 .080 .095 .080
9,SA Diffuse 9.SA Diffuse Diffuse Diffuse Diffuse 14.4A Diffuse 15.5 A
I[[ j III III II I I I

I
Potossium !\ ff\ {J.
./'- -"- A _j_ "--j\\....J/\~
0159 .030 9 .055 9 .OSO 9 .100 9 140 9 20 9
.005 .075 .020 .080 .060 .080
11.4 A 11.9 A 11.9 A 11.9A Diffuse 11.9ADiffuse
;:u;_;_;;~ AAI /(
0209 .040g .065g .OS09 .0959 ,125g .IS 9
.015 .065 .030 .065
Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse
5% 10 'Yo 25 % 40 'Yo 50 % 70 % 90 %
FIG. 62. Thermal-analysis curves, amounts of adsorbed water, and values of the
apparcnt cleavage space of lithium, hydrogen, potassium, and cesium salts of Mis-
sissippi montmorillonite, after Hendricks, Nelson, and Alcxander.18
180 Clay Mineralogy
that, for montmorillonite carrying magnesium and alkaline-earth ions,
the first step of water sorption is the hydration of the cation with six
molecules of water; this is followed by completion of a water layer having
a hexagonal type of structure. An additional water layer of similar
structure is taken up at higher relative humidities. Similar results were
obtained for Li montmorillonite, except that only three molecules of
water were required for the hydration of the lithium ion. In mont-
morillonites carrying sodium, potassium, and cesium, the cation appar-
ently was not hydrated. Mackenzie 12 has recently questioned the inter-
pretation of the single endothermic peak of Na montmorillonite as
evidence for the nonhydration of the sodium ion. According to him, all
the water molecules in a single molecular water layer of Na montmoril-
lonite could be tied to the sodium ion and, therefore, could be expected
to yield a single endothermic peak.
Barshad's17 concept (see page 169) of a change in the configuration
of the water moJecuJes as the amount. of water sorbed increases, provides
another possible explanation for the mUltiple nature of the differential
thermal endothermic peaks resulting from the loss of adsorbed water.
Mering 19 has presented evidence suggesting that, at low relative humidi-
ties (below about 30 per cent relative humidity), a Ca montmorillonite
does not form a single water layer but instead a skeletal double layer,
corresponding to the octahedral coordination of hydration water about
the Ca++, and that the thickness of such a water layer is about 4.4 A.
At relative humidities between 30 and 80 per cent, two complete water
layers develop, and the thickness of the water layer increases to about
5.9 A. Mering has shown that saturation of about 30 per cent of the
base-exchange capacity with Ca++ is adequate to produce the hydration
characteristics of a Ca montmorillonite. Walker 37 has suggested a
similar hydration sequence for Mg vermiculite. It is clear from the fore-
going th,at further investigation is required before the detailed steps in the
sorption of water by montmorillonite with various adsorbed cations is
well understood.
Numerous investigations have brought out the interesting fact that
Na montmorillonite, on drying at room t{)mperature, tends to develop
a single water layer between the silicate layers and that Ca montmoril-
lonite under the same conditions tends to develop two water layers,
whereas, at high relative humidities and in the presence of an abun-
dance of water, Na montmorillonite sorbs by far the larger quantity of
water.

,', .
Clay-Water System 18]
Barshad 38 has studied the sorption of water by vermiculite. Accord
ing to him, vermiculite saturated with Mg++, Ca++, Ba++, H+, Li+, or
Na+ shows, when immersed in water, total sorption between the silicate
layer equal to only two molecular water layers. When vermiculite
saturated with K+, NHt, Rb+, or Cs+ is immersed in water, there is no
expansion of the lattice, indicating no water sorption between the silicate
layers.
Siefert 39 has presented the results, shown in Table 23, for the sorption
of water by kaolinite. These data show the relative influence of the
various cations on water sorption by this clay mineral.
TABLE 23. INFLUENCE OF CATIONS ON WATER SORPTION OF KAOLINITE
(After Siefert 3 ')
Relative humidity, per cent Order of decreasing sorption
10 Ca > H = Na = K
81. 5 H = Ca > Na K
99.9 H = Ca = Na K
Mackenzie 12 has recently suggested that a given ion like Na+ might
hydrate to a different degree on different clay minerals and even on the
same type of clay mineral if the bonding force holding the ion varied.
Thus, in montmorillonite the degree of hydration of Na+ might vary
depending on whether the charge holding it derived from a substitution
within the tetrahedral or the octahedral sheet.
STEPWISE HYDRATION OF MONTMORILLONITE
Careful study by Bradley, Clark, and Grim 40 of a hydrogen mont-

morillonite produced by electrodialysis of a Wyoming bentonite indicated
that the swelling of the lattice during hydration took place in a stepwise
fashion. A series of four apparently definite and discrete hydrates were
found, having cell heights of 12.4, 15.4, 18.4, and 21.4 A. Near the range
where any given hydrate is stable, successive orders of (OOl) reflections
appear for it alone. In intermediate ranges two suites of (DOl) reflections
appear simultaneously, one to be identified with the hydrate next higher
in the sequence and one with the hydrate next lower.
In their study, Bradley et al. 40 used oriented aggregates in which the
basal cleavages were substantially parallel. Hofmann,41 working with
38 Barshad, 1., Vermiculite and Its Relation to Biotite as Revealed by Base-Exchange
Reactions, X-ray Analyses, Differential Thermal Curves, and Water Content, Am .

Mineral., 33, 655-678 (1948).
3' Siefert, A. C., Studies on the Hydration of Clays, Ph.D. thesis, Pennsylvania
State College (1942).
40 Bradley, W. F., G. F. Clark, and R. E. Grim, A Study of the Behavior of Mont-
morillonite on Wetting, Z. Krist., 97, 216-222 (1937).
41 Hofmann, D., Neues aus der Chemie des Tons, Die ']hemie, 56, 283-294 (1942).
182 Clay Mineralogy
powders, failed to check the conclusions of Bradley et al., but later
Hofmann and Hausdorf,l1 using oriented aggregates, also concluded that
the hydration of montmorillonite took place by the formation of succes-
sive monomolecullu layers of water. The latter investigators, and later
Hendricks and Jefferson,lo have shown theoretically that powder diffrac-
tion should indicate that the c-axis-dimension spacing varies continuously
but not uniformly with water content, and that this apparently would be
a result of an averaging effect from a lattice that contains various numbers
of water layers in different parts. That is, the apparently continuous
change in the c dimension results from a random alternation of successive
discrete hydrates.
Mering 19 and others have shown that the development of discrete
monomolecular water layers probably does not hold precisely at low
moisture contents when the clay carries an adsorbed ion that hydrates,
e.g., Ca++ or Mg++. In these cases, the hydration net around the cation
prevents the orderly development of the initial hydration layers.
STABILITY OF MONTMORILLONITE HYDRATION
A considerable body of evidence, largely unpublished, shows that

montmorillonite clays, when subjected for long periods of time to sub-
stantially uniform moisture conditions, develop hydration characteristics
of considerable stability. Such materials resist change in the degree of
hydration, but when the hydration is changed even to a very slight degree,
the stability may be abruptly and completely lost. Thus ahigh-swelling
bentonite sample carefully collected to preserve its natural hydration can
be placed in water without any slaking, eveIl when moderately st,irred.
However, if such a bentonite is dried only a very small amount, it will
slake and swell immediately when placed in water. Apparently, if
hydration layers develop with a high degree of uniformity of thickness
and distribution, they have considerable stability, which is not present
if there is even a very minor variation in thickness from layer to layer.
This matter of stability of hydration is of great practical importance.
In soil mechanics, for example, tests run on samples with the natural
mQisture carefully retained (as is the usual procedure) are likely to give
results completely misleading if applied to the materials as used in the
field, where some loss of water may take place.
A further point in regard to the stability of montmorillonite hydration
is worthy of note. When the dried mineral is placed in water, it may
rapidly adsorb water up to a certain point and then become a thick pasty
mass. No more water is sorbed unless the mass is stirred vigorously.
Adsorption tends to stop at a given point, and an equilibrium is devel-
oped at a very high moisture content. This characteristic is shown
particularly by the Na montmorillonite in the bentonite from the Wyo--
ming area. It is of considerable commercial utility in that it permits the
mineral to be used for water impedance.
INFLUENCE OF ADSORBED ORGANIC MOLECULES
Gieseking 42 reported that montmorillonite clays lost their tendency to

swell by water sorption when saturated with a variety of organic cations.
These cations are adsorbed on the basal plane surfaces of the montmoril-
lonite (see Chap. 10). Hendricks 43 has shown that the amount of water
adsorbed by montmorillonite carrying certain adsorbed amine ions
approaches closely the quantity expected from the difference between the
total extent of the surface and the probable part of the surface covered
by the amine ions. Hendricks has also pointed out that the reduction in
water sorption should not be exactly correlative with the size of the
organic ion, since the shape of the organic ion may be such as to destroy
the configuration of water molecules in the adsorbed water layer.
Grim, Allaway, and Cuthbert 44 have shown that water adsorption
of kaolinite clays is reduced after their treatment with organic ions and
that the decrease in water adsorption is relatively less for kaolinite than
for montmorillonite. It appears certain that similar results would be
obtained for the other clay minerals and that the values would be inter-
mediate between those for montmorillonite and kaolinite.
HEAT OF WETTING
A dry clay material evolves heat when placed in water, and this phe-
nomenon is known as the heat of wetting. Heat is also evolved when
clay materials are placed in some liquids other than water, as, e.g.,
alcohols and various organic liquids. Janert4 5 has shown that heat of
wetting may be higher for organic liquids, and that values generally are
higher for polar than for nonpolar liquids. Janert has also shown that
there is no direct correlation between polarity and heat of wetting, so that
polarity is not the sole determining factor. Figure 63, giving data for
the heat of wetting for water and for carbon tetrachloride as developed
by a "brick clay," illustrates the kind of difference in values for water

lite-Nontronite Type of Clay Mineral, Soil Sci., 47, 1-14 (1939).
43 Hendricks, S. B., Base Exchange of the Clay Mineral Montmorillonite for Organic
Cations and Its Dependence upon Adsorption Due to van der Waals Forces, J. Phys.
Chem., 46, 65-81 (1941).
44 Grim, R. E., W. H. Allaway, and F. L. Cuthbert, Reaction of Different Clay
Minerals with Some Organic Cations, J. Am. Ceram. Soc., 30, 137-142 (1947),
45 Janert, H., The Application of Heat of Wetting Measurements to Soil Research
Problems, J. Agr. Sci., 24, 136-150 (1934).

184 Clay Mineralogy
and some other liquids. The heat of wetting decreases as the moisture
content of the clay increases at the time of making the determination
(Fig. 63). Miller et al. 46 present heat-of-wetting values for attapulgite
obtained with water and several organic liquids (Table 24). The organic
TABLE 24. HEAT-OF-WETTING VALUES FOR ATTAPULGITE, IN CALORIES PER GRAM
(After Miller, Heinemann, and McCarter 46)
Clay dried at
Liquid Dried raw clay
400C 450C 550C 900C
Water .. ................... 10.6 17.7 22.0 25.6 10.9

Benzene ................... 4.2 ." . 4.2 4.3 2.4
Ethyl alcohol ............... ... . 11.6 ., .. 17.1
liquids used by them also gave lower values than water, and for a series
of normal alcohols the values decreased markedly as the chain length of
the alcohol increased.
12
E
E
(.!)
~
Q)
a.
-=>,
00
un
.!:
'0
O>.~
c ....
+:0
+-
ClJ_
4
-
:;;:0
0ClJ 2
J:
0
14 16
Initial Moisture
FIG. 63. Heat of wetting for water and carbon tetrachloride in relation to initial
moisture content, from Marshall,46 after data by Janert.4.
Beat of wetting is measured in calories per gram of dried clay, with the
drying being done usually at nOec.
Values for Heat of Wetting. Heat-of-wetting values for kaolinite and
montmorillonite saturated with various cations and for an illite sample are
given in Table 25. Values for attapulgite, given in Table 24, are only
slightly lower than those of montmorillonite.
46 Miller, J. G., H. Heinemann, and W. S. W. McCarter, Heat of Wetting of Acti-

,'ated Bauxite and Attapulgus Clay, Ind. Eng. Chern., 42, 151-153 (1950).
TABLE 25. HEAT-OF-WETTING VALUES FOR KAOLINITE, MONTMORILLONITE, AND
ILLITE, IN CALORIES PER GRAM
Kaolinite Montmorillonite Illite

Reference
Nat- Ca++ H+ Na+ K+ Nat- Ca++ H+ Na+ K+ Natural
ural ural
--- -- - - -- -- --- -- --
Siefert 39 .. . 1.45 1.40 1.30 1.22 .... 22.1 20.1 12.1 9.1
Endell, Loos, et I
alY ........ 2.2 .. , . . , .. .... ... . ... . 22 . ... 16.1 I
Parmelee and
Frechette 4s 1.9 ... . ... . .. , . ... . 11.8 ... . ... . .... . .. 4.0
Values for the other clay minerals are not yet available, but they will
undoubtedly range between those for montmorillonite and kaolinite.
Chlorite will probably yield values closer to those for illite, and vermicu-
lite closer to those for montmorillonite.
Pate,49 Anderson,50 Janert,51 and others 39 have shown that the heat
of wetting varies with the adsorbed cation and that it is generally
higher for divalent than for univalent cations. Janert gives the order
Ca > Mg > H > N a > K.
Harmon and Fraulini 52 have presented data showing an increase in
heat of wetting for kaolinite as the particle size of the kaolinite decreases
(Table 26). This is an expected relationship, since heat of wetting
TABLE 26. HEAT OF WETTING OF KAOLINITE IN RELATION TO PARTICLE SIZE AND

CATION-EXCHANGE CAPACITY
(After Harmon and Fraulini 52)
Particle size, microns ...... 10-20 .5-10 .2-4 .1-0.5 0.5-0.25 0.25-0.10 0.10-0.05
Heat of wetting, cal/g ..... 0.95 0.99 1.15 1.38 1.42 1.87
Exchange capacity, meq/100
g ...................... 2.40 2.60 3.58 3.76 3.88 5.43 9.50
'7 Endell, K., W. Loos, H. Meischeider, and V. Berg, Ueber Zusammenhange

zwischen Wasseraushalt der Tonminerale und Boden physikalischen Eigenschalten,
Inst. deut. Forsh. Bodenmechanik, 5, (1938).
48 Parmelee, C. W., and D. Frechette, Heat of "Vetting Values of Fired and Unfired
Clays, J. Am. Ceram. Soc., 25, 108-112 (1942).
49 Pate, W. N., The Influences of the Amount and Nature of the Replaceable Bases
upon the Heat of Wetting of Soils and Soil Colloids, Soil Sci., 20, 32\)-375 (1925).
50 Anderson, M. S., The Influence of Substituted Cations on the Properties of Soil
Colloids, J. Agr. Research, 38, 565-584 (1929).

51 Janert, H., Cation Exchange and 'Water Adsorption in Soils, Z. Pjlanzenernahr.
Dungung u. Bodenk. A, 34, 100-108 (1934).
52 Harmon, C. G., and F. Fraulini, Properties of Kaolinite as a Function of Its
Particle Size, J. Am. Ceram. Soc., 23, 252-258 (1940).

186 Clay Mineralogy
should increase as the surface area increases with decreasing particle size.
It should apply equally well to the other nonexpanding clay minerals.
In the case of expanding-lattice minerals, such as montmorillonite and
vermiculite, there should be no strict correlation with particle size, since
the total surface is theoretically available to water sorption regardless
of the size of the particles. However, in the case of large units, as in
vermiculite, there would probably be some variation in the ease of wetting
of the particle surfaces, so that some variation of heat of wetting with
particle size probably would be found. Also, the surfaces of mont-
morillonites carrying certain adsorbed ions, e.g., K+, are not completely
available for water adsorption, and some correlation with particle size
would be expected for such montmorillonites.
Anderson and Mattson,53 Baver,54 and others have pointed out a
relation between cation-exchange capacity and heat of wetting for a
series of natural soils. Montmorillonites with a very high cation-
exchange capacity yield high values for heat of wetting; kaolinites with
low cation-exchange capacity yield low heat-of-wetting values; and
illites give intermediate values. Soils composed of one or a mixture
of these common clay minerals would show some correlation between
cation-exchange capacity and heat of wetting. It is doubtful if a close
correlation would exist between these two characteristics in a series of
natural soils-certainly it would not be a straight-line relationship-since
other factors, such as particle size and nature of adsorbed cation, would
influence the heat evolved on wetting.
Cause of Heat of Wetting. Behrends 55 has narrated early ideas offered
to explain heat of wetting, such as the compression of water at the
adsorbing surface, capillary condensation, etc. It is now generally
agreed that the phenomenon is due to two factors: (1) a change in state
of the water directly adjacent to the adsorbing surface, and (2) the
possible hydration of adsorbed ions. Baver and Winterkorn 56 have
emphasized the importance of the development of orientation of the water
molecules in the adsorbed water as a cause of heat of wetting.
Janert 45 has shown that the heat of wetting, calculated per equivalent,
is only a fraction of the values usually assigned to the total heat of
hydration of various cations in dilute solutions. This is explained on
53 Anderson, M. S., and S. Mattson, Properties of Soil Colloidal Material, U.S. Dept.
Agr. Bull. 1452 (1926).
54 Baver, L. D., The Effect of the Amount and Nature of Exchangeable Cations on
the Structure of a Colloidal Clay, Missouri Agr. Expt. Sta., Research Bull. 129 (1929).
55 Behrends, W. U., The Relation between the Surface, Hygroscopicity, and Heat
of Wetting of Soils, Z. Pjlanzenerndhr. Dungung u. Bodenk., 40, 255-309 (1935).
56 Baver, L. D., and H. Winterkorn, Sorption of Liquids by Soil Colloids, II, Soil
Sci., 40, 403-418 (1936).
the basis that the ion is not completely free and is, therefore, capable of
only partial hydration. According to Janert, the ratios of the heat of
wetting to the heat of hydration for various ions are as follows: for
H clays 1:11.5, for Mg clays 1:9, for Ca clay 1:7, for K clays 1:5.1, and
for Na clays 1 :4.9. Janert was of the opinion that ion hydration is a
major cause of the heat of wetting.
Siefert 39 computed the heat evolved per milliequivalent of cation and
found that a given cation adsorbed by kaolinite evolves more heat than
when adsorbed by montmorillonite. Since, in montmorillonite, there
is little surface aside from that occupied by the cation, Siefert concluded
that the heat of wetting due to the surface is greater than that due to
cation hydration.
It seems likely that the relative importance of surface and cation
hydration would vary for different clay minerals and for various adsorbed
cations. It is probably incorrect to assume that either factor is always
the more important cause.
Heat of Wetting in Solutions of Electrolytes. Siefert3 9 has shown
that when hydrogen kaolinite is wetted in 0.1 N NaOH or 0.015 N
Ca(OHh solutions, there is an increase in heat evolved of 0.2 cal/g over
that of pure water; this is probably due in part to the neutralization of
the clay acid by the base. However, Na kaolinite wetted in NaOH and
K kaolinite wetted in KOH showed a similar increase in heat evolved in
comparison to that developed in water. Somewhat similar results were
obtained by Siefert for montmorillonite clays. The explanation for this
phenomenon is not dear, but it may be merely the result of the better
dispersion of the clay in the alkaline solution.
Effect of Firing. The data given in Fig. 64, after Parmelee and
Frechette,48 show the relation of heat of wetting to firing temperature for
montmorillonite, kaolinite, and illite. In the case of the montmoril-
lonite, there is a sharp drop in heat of wetting after firing at the tempera-
ture where hydroxyl water is lost from the lattice. The illite shows a
slight increase in heat of wetting on firing up to about 450C and a reduc-
tion in heat of wetting after temperatures somewhat above that required
for the loss of hydroxyl water. The kaolinite shows a slight reduction
in the amount of heat evolved when heated to temperatures somewhat in
excess of that required for the loss of hydroxyl water. Data given by
Miller et al. 46 (Table 24) show the heat of wetting of attapulgite to
increase greatly on firing up to about 550C and then decrease, so that
after firing at 900C the same value as that for the unfired mineral is
obtained. The reason for the changes in heat of wetting of the various
clay minerals when heated to these particular temperatures is not clear.
Additional experimental data are needed before heat-of-wetting values
for fired clays can be understood.
188 Clay Mineralogy
12
f'-.,
i'-...
"""- r---:\
10
\ \ Monmori//onite
8
I~\
\\
6
\
E
-g
\
0
::'4
\\
0'>
\
:
QJ
2
- - - Autoc/oved \\
:s:
~
-
0
+-
0 0
- - Unoutocloved
~ ~
QJ
-
J: I..- ~
4
_.".. V ...., l\. Illite
,~
2
"\
\
~
0
:I TTIJo"Xe
100 300 500
Temperature of Firing Prior to DeterminatIOn
t IIitO 700 900 1100 c
of Heat of Wetting
FIG. 64. Heat of wetting of fired clay minerals, from Parmelee and Frechette."
Alexander, L. T., and T. M. Shaw, Determination of Ice-Water Relationships by

Measurement of Dielectric Constant Changes, J. Phys. Chem, 41, 955-960 (1937).
Bangham, D. H., and S. Mosallam, The Adsorption of Vapours at the Plane Surfaces
of Mica, II, Heats of Adsorption and the Structure of Multimolecular Films,
Proc. Roy. Soc. (London) A, 156, 558-571 (1938).
Bouyoucos, G. J., State in Which the Hygroscopic Moisture Exists in Soils as Indi-
cated by Its Determination with Alcohol, Soil Sci., 41, 443-447 (1936).
Brasseur, H., Water of Hydration in Crystals, Assoc. franr;. avance. sci. Liege, pp. 339-
342 (1939).
l.
Buergers, J. M., Introductory Remarks on Recent Investigations concerning the
Structure of Liquids, "Second Report on Viscosity and Plasticity," Noord Hol-
landsche uitgevers maatschaappi, Amsterdam (1938).
Carlson, H., and G. Mueller, On the Behavior of Ethereal Oils and Their Constituents
on Clay and Related Substances, Ber. deut. chem. Ges., 71,863-870 (1938).
Denisov, P. I., Nonfreezing Clay Solutions, Azerbatdzhanskoe N eftyanoe Khoz., pp.
31-32 (1941); Khim. Referat. Zhur., 4, 110 (1941).
Dumanski, A. V., Heats of Wetting and Hydrophily of Disperse Systems, Kolloid
Zhur., 12, 319-325 (1950).
East, W. H., Fundamental Study of Clays, X, Water Films in Monodisperse Kaolinite
Fractions, J. Am. Ceram. Soc., 33,211-218 (1950).
Endell, K., and P. Vageler, Der Kationen- und IVasseraushalt keramischer Tone im
rohen Zustand, Ber. deut. keram. Ges., 13, 377-411 (1932).
Freundlich, H., U. Schmidt, and G. Lindau, Ueber die Thixotropie von Bentonit-
Suspensions, Kolloid-Beihefte, 36, 43-81 (1932).
Hofmann, U., Ueber die Grundlagen der Plastizit1it der Kaoline und Tone, Ber.
/ deut. keram. Ges., 29,21-32 (1949).
i Kapp, L. C., The Approximate Size of Soil Particles at Which the Heat of Wetting is
Manifest, Soil Sci., 401-412 (1930).
Keppeler, G., Neuere Untersuchungen tiber Tonmineralien, Ber. deut. keram. Ges.,
19,159-176 (1938).
Kiefe, C., Sur les possibilities de liaisons et d'orientation des molecules d'un liquide
sur un solide. Epitaxie entre la kaolinite et l'eau, mimeographed, Paris (1947).
Kister, E. G., The Swelling of Clays, Neftyanoe Khoz., 25,23-27 (1948).
Longuet-Escard, J., The Effect of Progressive Dehydration on the Area of the Surface
of Montmorillonites, J. chim. phys., 47, pp. 113-117 (1950).
Makower, B., T. Shaw, and L. T. Alexander, The Specific Surface and Density of
Some Soils and Their Colloids, Soil Sci. Soc. Am. Proc., 2, 101-108 (1937).
Krist., 93, 481-487 (1936).
Norton, F. H., and A. L. Johnson, Fundamental Study of Clays, V, Nature of Water
Films in Plastic Clays, J. Am. Ceram. Soc., 27,77-80 (1944).
Puri, A. N., and R. C. Hoon, Physical Characteristics of Soils, III, Heat of Wetting,
Soil Sci., 47, 415-423 (1939).
Rideal, E. K., How Crystals Grow, Nature, 164,303-305 (1949).
Rios, E. G., and J. L. Vivaldi, Silicates of Laminar Structure, I, Hydration, Anales fis.
y quim. (Madrid) B, 45,291-342 (1949).
Schwarz, F., Heat of Wetting of Clays, Sprechsaal, 82, 24-25 (1949).
Sonders, L. R., D. P. Enright, and W. A. Weyl, Wettability, a Function of the Polar-
izability of the Surface Ions, Penn. State Coli. Min. Ind. Expt. Sta. Tech. Rept. 12,
N.R.032-265 (1949).
Vageler, P., and F. Alten, Boden des Nil und Gash Z. Pjlanzenerniihr. Diingung u.
Bodenk., 21, 335-345 (1931).
CHAPTER 9
Dehydration, Rehydration, and the Changes Taking Place
on Heating
Dehydration involves the loss of any water, adsorbed, interlayer, or

lattice OR water, held by the clay minerals. A study of dehydration is
concerned with the amount of water lost, rate of water loss, temperature
of dehydration, and energy involved. Dehydration frequently involves
significant changes in the structure of the clay minerals, and conse-
quently changes taking place during the heating of the clay minerals can-
not be considered apart from dehydration. Changes taking place in the
clay minerals on heating to relatively high temperatures are not neces-
sarily related to dehydration reactions, but for the sake of uniformity and
continuity, such changes will also be considered in this chapter.
Certain of the clay minerals take up water again on cooling after being
heated to moderately high temperatures. Such rehydration character-
istics will also be considered here.
METHODS OF STUDY
Various methods are available for studying the hydration properties

of the clay minerals. Since some of the methods supplement each other,
it is frequently desirable to use more than one procedure. This is par-
ticularly true where phase changes are important. For such studies, a
combination of X-ray diffraction and some other method is necessary.
Vapor-pressure-Water-content Determinations. This method has
been used by Thomas,1 Puri,2 Kuron,s Alexander and Haring,4 and many
other investigators of soils. Air-dried samples are allowed to absorb
.
moisture at various relative humidities, and curves are plotted of water
1 Thomas, M. D., Aqueous Vapor Pressure of Soils, Soil Sci., 25, 409-418, 485-493
(1928).
2 Puri, A. N., E. M. Crowther, and B. A. Keen, The Relation between the Vapour
Pressure and Water Content of Soil, J. Agr. Sci., 15,68-76 (1925).

3 Kuron, H., Adsorption von Dampfen und Gasen an Boden und Tonen und ihre
Verwendung, Kolloid-Beihejte, 36, 178-256 (1932).

4 Alexander, L. T., and M. M. Haring, Vapor Pressure-Water Relations for CertaiTl
Typical Soil Colloids, J. Phys. Chern., 40, 195-205 (1936).
190
Dehydration 191
content versus vapor pressure. It is likely, as N agelschmidt 6 has pointed
out, that the exchangeable ions are of greater importance than the clay
minerals themselves in causing the variations in moisture content as
determined. Only very limited application of the method has been made
to monomineral samples, and its contribution to the knowledge of clay
minerals remains for future studies.
L20
.~
18
16
- ~:::::: >-.:;::::_- -_
B
14
......... C ~, I'-.." "
II
\~ 12 ~ I
\ I
1: I
.2' 1.10
~ II
8
6
\!
II
4
I
2 ~
1.00 i~ ~
o 100 200 300 400 500 600 700 800
Temperature C
FIG. 65. Dehydration curves, from Nutting. s The weight at 800C is taken as the
base weight in plotting the curves. (A) Halloysite, Liege, Belgium; (B) kaolinite,
lone, Califomia; (C) anauxite, Mokelumne River, California.
Dehydration Curves. In this method the loss in weight of the material

upon heating to higher and higher temperatures is recorded and plotted
against the temperature. There are several variations of the method as
follows:
The sample may be heated at a given temperature until no loss in
weight occurs; it is then heated to a higher temperature and held at that
, temperature until no loss in weight takes place. This procedure is
repeated until a temperature is reached at which there is no further
loss in weight. This procedure has been used by Nutting,6 Ross and
6 Nagelschmidt, G., The Identification of Minerals in Soil Colloids, J. Agr. Sci., 29,
477-501 (1939).
6 Nutting, P. G., Some Standard Thermal Dehydration Curves of Minerals, U.S.
Geol. Survey Profess. Paper 197E, pp. 197-216 (1943).

lIBRARY
,'-'. p ,,/> "
t. . . . . . . .l!. ... L:"
)f p, ~~ ,ri cuI un.
\,"'_
). " nHIlo ~ra.IA5;1
\,(31 University
, .~: __ ,~ '~"a_ 30
f
192 Clay Mineralogy
Kerr,7,8 Kelley, 9 and many others. 10 ,11 Results for the various clay min-
erals are given in Figs. 65 to 69. Various investigators have plotted
their data somewhat differently, as illustrated in Figs. 65 to 69.
1.20
18 \
16 \\
14 1-~ r-A- --
-~ -_ -_-_
-fL
12 ,,'\ 1\ ~1
-
.<::
.2'1.10
\E 1"\ \ \\
I
~
8
)\\ ~ J2.._
\\
---;\; 1--- f---:::
\ ~ r---~ [\__ -~
6 ~"
4
'\~Q_
-_f:\- ~-:- N::-:
......
'- '-
'~ -'.
....
..., ....
'\~
~
I\:,~ ~ .... ' ....

\
~ --~:--.
~ l~ 1\::\ "-
2
1.00
o 100 200 300 400 "
Temperature C
500 600 700 800 900
FIG. 66. Dehydration curves, after Nutting. 6 The weight at 700 to 900C is taken
as the base weight in plotting the curves. (A) Penninite, Paradise Range, Nev.tda;
(B) chlorite, Danville, Virginia; (C) palygorskite (mountain leather), Montana; (D)
sericite, Prince Rupert, British Columbia; (E) vermiculite, North Carolina; (F) illite,
Fithian, Illinois; (G) glauconite, Lyons Wharf, Maryland; (H) &epiolite, Asia Minor.
The sample may be heated continuously at a constant rate, which has

varied in the experiments of different investigators from 5C/hr to
lOoC/min, and the loss in weight is recorded at successively higher
temperatures. The samples are weighed while they are hot. Migeon 12
7 Ross, C. S., and P. F. Kerr, The Kaolin Minerals, U.S. Geol. Survey Profess. Paper
165E, pp. 151-175 (1931).

8 Ross, C. S., and P. F. Kerr, Halloysite and Allophane, U.S. Geol. Survey Profess.
Paper 185G, pp. 135-148 (1934).

9 Kelley,.W. P., H. Jenny, and S. M. Brown, Hydration of Minerals and Soil Col-
loids in Relation to Crystal Structure, Soil Sci., 41,259-274 (1936).

10 Ross, C. S., and S. B. Hendricks, Minerals of the Montmorillonite Group. U.S.
11 Grim, R. E., R. H. Bray, and W. F. Bradley, The Mica in Argillaceous Sediments,
Am. Mineral., 22, 813-829 (1937).
12 Migeon, G., Contribution a l'etude de la definition des sepiolites, Bull. soc. franf.
mineral., 59, 6-133 (1936).
Dehydration 193
and LongchambonI3 particularly have used this procedure, and results for
some of the clay minerals are shown in Fig. 70. In this method equilib-
rium conditions are not reached at any given temperature.
III both the above procedures, the actual humidity of the air in contact
with the sample should be kept constant, since variations will affect the
results. N agelschmidt 5 has pointed out that it is generally assumed that
2e
26 .- L
1---.....
/
/
/ _- _-1.- -
( -
U
C
<LI
20
/ /
",'
_- _._
... i'1/-
V c
~ 16 J
.S
... / A.v- P
-
<LI
~
I tJ / ~-
...Q_
-
0
IS>
(/)
I~
10
;1
( (
.5 I
il ;
.6
~~1 ;!
Ei '"
~= f:"_:::;;-"
1/
f'
/ 1/
II
l,_....-- ~
~
f..--""
o100 200
- 300
--- >-~/
400 500 600 700 800 900
Temperature c
FIG. 67. Dehydration curves, from Ross and Kerr. 7,8 (A) Halloysite, Liege, Be\-
gium; (B) halloysite, Adams County, Ohio; (C) halloysite, Hickory, North Carolina;
(D) kaolinite, lone, California; (E) allophane, Moorefield, Kentucky.
the loss in weight of one mineral upon heating is not affected by the pres-
ence of a second mineral and that the dehydration curve of a mixture
would be equal to the superposition of curves of the separate minerals in
the mixture. This assumption has not been proved, and, indeed, data
from differentittl thermal analyses (see page 210) show that it is likely
to be true only in some mixtures and not in others.
In the case of natural clays which are not pure clay minerals, the loss
13 Longchambon, H., Sur les characteristiques des pa:Iygorskites, Compt. rend.
204, 55-58 (1937).

194 Clay Mineralogy
in weight which is measured is not necessarily due entirely to a loss of
water. It may be due to the loss of CO 2 from carbonates and to the loss
of volatile constituents which may be present in noncIay-mineral com-
ponents. Also, if divalent iron and/or manganese are present, their
6
3 "
I'l
2
t
!\'t,
N ~,
~ t---
~ r-&._
~
+- t ".
.c:
~
CT>
2
1\\\ I~\ ,,
'0 3
~\
i \
~
[\D 1\
\ :s.~\ \
"";E
.3'"
U) F,',
4
-~ ;t
\ "<..... ,
6
,6 "
"-..,-
7 \
8
" '- '....___
9
100 200 300 400 500 600 700 800 900 1000
Temperature C
FIG. 68. Dehydration curves, from Ross and Hendricks. 10 (Al Hectorite, Hector,
California; (B) montmorillonite, Belle Fourche, South Dakota; (C) montmorillonite,
Tatatilla, Mexico; (D) montmorillonite, Montrnorillon, France; (E) nontronite,
Spokane, Washington; (F) nontronite, Sandy Ridge, South Carolina; (G) montmoril-
lonite, Pontotoc, Mississippi.
oxidation would tend to cause an increase in weight and thereby to

reduce the apparent water loss. Variations in the specific characteristics
of the minerals being studied, such as grain size, crystallinity, nature of
adsorbed ions, etc., affect dehydration results. These factors also
influence differential thermal results, and they will be considered there
in some detail, as well as in the discussion of the dehydration propertiee
of the individual minerals.
Differential Thermal Analyses. The method of differential thermal
analysis determines, by suitable apparatus, the temperature at which
Dehydration 195
thermal reactions take place in a material when it is heated continuously
to an elevated temperature, and also the intensity and general character
of such reactions. In the case of the clay minerals, differential thermal
10 B
I
V~ --
/'
~
c
~A
~ v-
h / VD
~ I /
~
7" /'
/ /
/'
2
/ /
V
o L- v-
o
--
200
- /
400 600 800 1000
Temperature, C
FIG. 69. Dehydration curves, from Grim, Bray, and BradleyY (A) Illite, Gilead,
Calhoun County, Illinois; (E) illite, Fithian, Vermilion County, Illinois; (C) muscovite,
very finely ground; (D) muscovite, 100 mesh (coarser than C).
<J)
E
~ 80
.'?
~ 60
200 400 600 800

Temperature C
FIG. 70. Dehydration-weight-loss curves for (A) sepiolite, after ;\1igeon,12 and (E)
palygorskite, after Longchambon. 13
analyses show characteristic endothermic reactions due to dehydration

and to loss of crystal structure and exothermic reactions due to the forma-
tion of new phases at elevated temperatures. The method is, therefore,
useful for clay-mineral researches as a means of studying high-tempera-
ture reactions, in addition to its value in the investigation of hydration
phenomena.
r
196 Clay Mineralogy
The method is not restricted to
clays and, in fact, has been applied
to a wide variety of materials, such
as carbonates, hydrates, sulfides,
organics, etc. It is applicable to
any material which experiences
thermal reactions on heating which
begin abruptly and are completed
in a relatively short temperature
interval.
Differential thermal results are
plotted in the form of a continu-
ous curve in which the thermal
reactions are plotted against fur-
nace temperatures, with endother-
mic reactions conventionally shown
as downward deflections and exo-
thermic reactions as upward deflec-
tions from a horizontal base line.
The amount of divergence of the
difference curve from the base line
reflects the difference in tempera-
ture between the sample and the
furnace at any given temperature
and is, therefore, a measure of the
intensity of the thermal reaction.
Differential curves for the various
clay minerals are given in Figs. 71
to 76.
History of Method. Le Chatelier 14
in 1887, in an investigation of the
components of clays, appears to
have been the first person to set up
simple apparatus to measure and
0000008000000 record thermal reactions taking
900000 000000
C\Jr<)<j'IO(!)r--Cl)CJ>Q=~!:2
place in a material when it is heated
and to have used such data as ana-
FIG. 71. Differential thermal curves: lytical criteria to study the mate-
Scale B. (A) Kaolinite, Georgia, well-
crystallized; (B) kaolinite, Illinois, poorly
rial. Little use was made of the
crystallized; (C) hydrated halloysite, method by mineralogists un til about
Indiana; (D) anauxite, lone, California; 14 Le Chi1telier, H., De l'action de la
(E) allophane, Bedford, Indiana; (F) chaleur sur les argiles, Bull. soc. franr;.
allophane, Iyo, Japan. mineral., 10, 204-211 (1887).
Dehydration 197
8888888888888
-C\Jt<)<tLO(!)I'-COCJlQ=:!t<) 888~88gg88gC\J88
C\Jt<)" LOlOl'-COCJlQ 12
C C
FIG. 72. Differential thermal curves: FIG. 73. Differential thermal curves:
Scale A. (A) Montmorillonite, Otay, Scale A. (A) Biotite, University of
California; (B) montmorillonite, Tatatilla, Illinois collecti~Ils; (B) muscovite, Uni-
Mexico; (C) montmorillonite, Upton, versity of Illinois collections; (C) musco-
Wyoming; (D) montmorillonite, Cheto, vite, Bryman, California (minus one mi-
Arizona; (E) montmorillonite, Pontotoc, cron fraction); (D) illite, Fithian, Illinois;
Mississippi; (F) montmorillonite, Palmer, (E) illite, Grundy County, Illinois; (F)
Arkansas; (G) nontronite, Howard illite, Thebes, Illinois; (G) glauconite,
County, Arkansas; (H) hectorite, Hector, New Jersey; (H) glauconite, Washington.
California.
198 Clay Mineralogy
1910, when several investigators used it in a study of clays. In these
early studies the procedure generally followed was to place the material
in a small platinum crucible with a single thermocouple junction in the
center of the material. The whole mass was placed in a furnace and
heated at a rapid and relatively uniform rate.
The thermocouple was attached to a galva-
nometer which was read periodically at
short intervals or recorded photographically.
Thermal reactions in the material caused
variations in the galvanometer record as
compared to that obtained when the furnace
contained no sample. The record obtained
showed the thermal reactions of the material
superimposed on the heating curve of the
furnace. Examples of such records are
shown in Fig. 77.
8 2 2 C~ ~ - g In 1899 Roberts-Austen 15 devised the dif-
FIG. 74. Differential thermal ferential-thermocouple method for measuring
curves: Scale A. Illite asso- temperature differences between a material
ciated with chlorite: (A) ma- and a reference unit, and in 1908 Burgess 16
quoketa shale, Illinois; (B) suggested simple, adaptable, and usable dif-
Minford silt, Ohio. Illite
with mixed layers of mont-
ferential-thermocouple circuits. After the
morillonite: (A) Bear River, work of Roberts-Austen, the difference ther-
Wyoming; (B) Sarospatak, mocouple was widely used by metallurgists,
Hungary (sarospatite); (C) but Fenner 17 in 1913, in his study of the
Noyant Allier, France (bra- stability relations of the silica minerals,
vaisite); (D) Decorah shale,
vVisconsin. appears to have been the first to apply it
outside of the field of metallurgy. The tech-
nique devised by Fenner is substantially that used today in differential
thermal analyses with some elaboration and refinement. Following
Fenner's work, the method was used sporadically in the study of high-
temperature phase changes by Kracek 18 and others 19 and frequently
in the study of the constituents of clays by various investiga-
H<> Wohlin, R., Beitrage zur Kenntnis cler Thermischen Analyse von Tonen,
Bauxiten und einigen verwandten Korper, Sprechsaal, 46, 719-721, 733-735, 749-
751, 767-769, (1913).
15 Roberts-Austen, J., Fifth Report of the Alloys Hesearch Committee, Proe. Insf.
Meeh. Engrs. (London), 35, (1899).
16 Burgess, 'G. K., On Methods of Obtaining Cooling Curves, II/all. Bur. Standards
(U.S.) Tech. News Bull., 5, 199-225 (1908-0\).
17 Fenner, C. N., The Statility Relations of the Silica Minerals, Am. J. Sci., ser. 4,
36, il31-384 (1913).
18 Kracek, F. C., The Polymorphism of Sodium Sulphate, I, Thermal Analysis. J.
Phys. Chem., 33, 1281-1303 (1929).
19 Kracek, F. C., N. L. Bowen, and G. W. Morey, The System Potassium Metasili-
cate-Silica, J. Phys. Chem., 33, 1857-1879 (1929).
Dehydration 199
H
H
i
00 00 0 00 0 000 0 0 00 00 0000 00 0 00
0 0 00 0
-(\j r0<;t
0
lD ~f2
000 0
to OlQ= (\j
0
~
00 00
- (\J r0<;t gg12&5
00 0
(J)Q
00
~!:2
C C
FIG. 75. Differential thermal curves: FIG. 76. Differential thermal curves:
(A) pro chlorite, Chester, Vermont; (B) Scale A. (A) Attapulgite, Attapulgus,
clinochlore, Brewster, New York; (C) Georgia; (B) sepiolite, Salinelles, France;
vermiculite and chlorite, Lenni, Pennsyl- (C) palygorskite, North Africa; CD)
vania; (D) vermiculite, North Carolina, palygorskite, from Caillere;84 (E) ver-
natural; (E) vermiculite, North Carolina, miculite, Arizona; (F) pyrophyllite, North
Ca++; (F) vermiculite, North Carolina, Carolina; (G) talc, Vermont; (H) quartz,
Na+; (G) vermiculite, North Carolina, University of Illinois collections.
NHt; (H) vermiculite, North Carolina,
K+. Curves C to H from Barshad;82
vertical scale slightly less than for other
curves. Scale B for curves A and B.
200 Clay Mineralogy
tors. 20-22 In 1933 and 1935 excellent reports by Orcel 23 and Orcel and
Caillere 24 emphasized clearly the applicability of the method to clay-
mineral researches, and since that time it has been used in the study of
clays by many different investigators in many different countries. More
recently, the method has become popular with investigators of other
materials such as carbonates, sulfates, hydrates, etc. 25 ,26
Description of Method. In the method as generally used today, the
sample to be studied is placed in one hole of a specimen holder, and an
inert material that experiences no thermal reaction when heated to the
oooe
12
10
8
6
2
O~~~LJ~~~~L7~~~ ___________
5 J5
FIG. 77. Thermal reactions in clay superimposed on furnace heating rate, as recorded
by Wohlin."a Curve A is furnace heating curve. Curves B to D show furnace
heating rate with thermal reactions for clays superimposed.
temperature of the experiment, usually calcined aluminum oxide

(a-AI 2 0 3), is placed in another hole of the specimen holder. One junction
of the difference thermocouple (Fig. 78A) is placed in the center of the
sample and the other junction in the center of the inert material. The
holder and thermocouples are placed in a furnace so controlled as to
20 Houldsworth, H. S., and J. W. Cobb, The Behavior of Fireclays, Bauxites, etc., OIl
Heating, Trans. Ceram. Soc. (Engl.), 22, 111-137 (1922-1923).
21 Satoh, S., Heat Effects on Fireclays and Their Mixtures, Science Repts. T6hoku
Imp. Univ., ser. 3, 1, 157-203 (1923).
22 Orcel, J., Analyse thermique des chlorites, Bull. soc. fran~. mineral., 50, 278-322
(1927).
23 Orcel, J., UEmploi de I'analyse thermique differentielle dans la determination des
constituants des argiles, des laterites, et des bauxites, Congr. intern. mines met. et
geol. appl., 1,359-373 (1935).
240reel, J., and S. CaiIIere, L' Analyse thermique differentielle des argiles a mont-
morillonite (bentonites), Camp. rend., 197,774-777 (1933).
25 Grim, R. E., Method and Application of Differential Thermal Analysis, Ann.
N.Y. Acad. Sci., 53,1031-1053 (1951).
26 Berg, L. G., V. 1. Nikolaiev, and E. Y. Rode, Thermographia, Bull. 25 acado sci.
URSS (1944).
Dehydration 201
produce a uniform rate of temperature increase. The temperature of the
inert material increases regularly as the temperature of the furnace
increases. When a thermal reaction takes place in the sample, the tem-
perature of the sample is greater or less than that of the inert material,
depending on whether the reaction is exothermic or endothermic, for an
interval of time until the reaction is completed and the temperature of the
sample again comes to the temperature of the furnace. Consequently,
for an interval of time the temperature of one junction of the difference
couple is different from that of the other, and an emf is set up in the
differential-thermocouple circuit which is recorded as a function of time
or of the temperature of the furnace. The record may be made manually
b b
Difference
bob a Recorder
Difference
Temperature Proqrom
Recorder
Recorder Controller
to) (b)
FIG. 78. (a) Thermocouple setup as used today; (b) thermocouple setup as used in
early work and in some current work. (a) Platinum-lO per cent rhodium wire; (b)
platinum wire.
with a potentiometer or galvanometer, photographically with a galva-

nometer, or automatically with some electronic device. When no thermal
reaction is taking place in the specimen, the temperatures at both junc-
tions of the difference couple are the same, and no emf is set up. The
direction of the emf in the circuit depends on whether the temperature
of the sample is above or below that of the inert material, and conse-
quently the recording mechanism moves in opposite directions for endo-
thermic and exothermic reactions.
Figure 79 shows an idealized equilibrium dehydration curve and
differential thermal curve for kaolinite. The endothermic reaction
between about 500 and 700C corresponds obviously to the dehydration
of the mineral. A comparison of the curves illustrates that the differen-
tial method is a dynamic rather than a static one. The thermal reactions
are not instantaneous, and they are recorded as functions of time or as
202 Clay Mineralogy
functions of the furnace temperature, which is continuously increasing
as the reaction takes place. The temperature at which the dehydration
begins corresponds to the start of the endothermic reaction. The tem-
perature of the peak of the endothermic deflection varies, depending on
the details of the procedure followed, the character of the reaction
involved, and the material being studied.
....
.r:
12
10
.~+-
(1) C 8
3:(1)
U 6 A
.~ \._
(1) 4
~Q_ 2
.3 0
c:
<lJ
(1)
I~
(1) .....
. 0 "(1)
C
-
B
(1)-
u ~
c"O C>
C
~
r
(1)
~ 0
:::: c
.- (1)
Of ~
Q)'u ~
~
"- (1)
::on.
aU)
... i5
(l)
a. ~
E
~
o 100 200 300 400 500 600 700 800 900 1000
Temperature C
FIG. 79. Idealized dehydration and differential thermal eurves for kaolinite, after
Spiel et al. 27
Attempts have been made by Spiel,27 Kerr and Kulp,28 and Arens 2D
to analyze mathematically the differential thermal method. It will be
obvious from the discussion of the equipment used and the factors affect-
ing the results that the method has certain quantitative limitations which
restrict rigorous mathematical treatment. Currently an effort is being
made by a subcommittee of the International Committee for the Study
of Clays under the chairmanship of Dr. R. C. Mackenzie of Aberdeen,
27 Spiel, S.,. L. H. Berkelheimer, J. A. Pask, and B. Davies, Differential Thermal ~
Analysis-Its Application to Clays and Other Aluminous Minerals, U.S. Bur. Mines
Tech. Paper 664 (1945).
28 Kerr, P. F., and J. L. Kulp, Multiple Differential Thermal Analyses, Am. Mineral.,
33, 387-419 (1948).

29 Arens, P. L., "A Study of the Differential Thermal Analyses of Clays and Clay
Minerals,;' Wageningen, Netherlands (1951).
/t' !'
Dehydration 203
Scotland, to determine the inherent limitations of the method and to
standardize the procedure and equipment. The work of this committee
should permit a better analysis of the method. 30
Equipment Used. The specimen holder most widely used in the United
States is a block of nickel drilled with holes to accommodate about 0.5 g
of sample. Experience has shown that such blocks are serviceable and
provide peaks on the thermal curves of adequate sharpness and intensity.
Ceramic holders have been widely used in England, and Grimshaw 31
claims that they are desirable, because they cause the peaks for the
thermal reactions to be large and well defined as a result of the slow
transfer of heat between the test sample and surrounding material due to
the low heat conductivity of the ceramic material. Grimshaw used a
holder composed of recrystallized alumina, bonded with a small amount
of clay and fired to 1600C. In the early work of Le Chatelier 14 and
others, platinum crucibles were used to hold the specimen, and recently
Gruver 32 has claimed advantages for thin-walled platinum containers,
because the high heat conductivity of the metal permits the material to
reach furnace temperature quickly and because the thin walls provide
little heat capacity. In disagreement with Grimshaw, Gruver states that
thick-walled holders with higher heat capacity act as heat reservoirs and
tend to lower the intensity and sharpness of some reactions. Arens 29
claims that ceramic holders give relatively sharper endothermic peaks
and smaller exothermic peaks and that nickel holders give relatively
smaller endothermic peaks and sharper exothermic peaks. It may,
therefore, be desirable to use different types of sample holders for different
materials.
The thermocouple circuit most widely used at the present time is shown
in Fig. 78A, with an additional separate thermocouple to measure furnace
temperature for calibration. The circuit first used by Fenner 17 and still
used by some investigators, particularly in Europe, is shown in Fig. 78B.
In the latter circuit the sample temperature rather than the furnace tem-
perature is taken as the reference temperature.
Mackenzie 30 and later Arens 29 claimed that measurement of furnace
temperature in the clay sample has the advantage of providing more
comparable peak temperatures and facilitating calculations. As will be
shown, many things may cause a shift of the temperature of a reaction
peak so that no advantage will necessarily be gained. It seems probable
30 Mackenzie, R. C., Differential Thermal Analyses of Clay Minerals, Trans. 4th

Intern. Congr. Soil Sci., 11, 55-59 (1950).
31 Grimshaw, R. W., E. Heaton, and A. L. Roberts, Constitution of Refractory
Clays, II, Thermal Analysis Methods, Trans. Ceram. Soc. (Engl.), 44, 76-92 (1945).
32 Gruver, R. M., Precision Method of Thermal Analysis, J. Am. Ceram. Soc., 31,
323-328 (1948).
204 Clay Mineralogy
that greater advantages will accrue from calibrations with the actual
furnace temperature.
Early investigators used a platinum-platinum 10 per cent rhodium
thermocouple, and such couples are widely used today. Chromel-alurrtel
couples have been used satisfactorily for temperatures up to 1000C for
many materials, and they have the advantage over noble-metal CG1J.}\J.e.os.
of developing higher emf and thereby providing greater sensitivity.
Kracek 18 has u;::;ed gold-palladium-platinum-rhodium thermocouples,
which also have the advantage of developing high emf at elevated
temperatures.
In order to mltintain the horizontal position of the base line of the
differential temperature curve, it is necessary to maintain thermocouple
beads of constant size and to keep the beads in positions in the centers
of the sample and reference material. The thermocouple wire should be
relatively fine gauge, about 22, in order to reduce loss of heat by corl-
ductance through the wire.
The intensity of the thermal reactions varies greatly in differellt
materials, and it is essential to have some means (e.g., a variable resist-
ance in the difference-thermocouple circuit) of permitting some variation
and control of the sensitivity of the apparatus.
The furnace itflelf must have sufficient capacity to permit the desired
heating rate up to the maximum temperature and must be of sufficiellt
size to provide a uniform hot zone for the specimen holder. A wide
variety of furnaces, both horizontal and vertical, have proved satisfactory,
the selection being largely a matter of preference of the operator.
A wide variet1 of homemade and automatic tYges of 9rogram cotl-
trollers have been used to obtain a uniform heating rate. The use of I;1n
autotransformer, driven by a motor through a speed reducer in such a
way as to increase the voltage into the furnace gradually and continu-
ously, has been found quite satisfactory for furnaces with alloy windings.
Various manufacturers of temperature-control equipment have supplied
a variety of automatic controllers which have also proved satisfactory.
In general the controller should be tailor-made for the furnace, and care
must be taken with the on-off type that the on-off impulse does not get
into the differential-thermocouple record.
The differencwthermocouple record may be obtained by manual
reading of a galvanometer or potentiometer. However, it is desirable
to obtain a continuous difference record, and that can be done photo-
graphically with a reflecting galvanometer or with various automatic
recorders of commercial manufacturers. For use with noble-metal
couples, it is necessary to obtain automatic recorders of adequate and
variable sensitivity which at the same time have sufficient stability to
produce good records. This is important because the temperature
Dehydration 205
differences measured are very small, and the emf produced loy noble-
metal couples is therefore very small. In several laboratories the
photopen recorder used with a reflecting galvanometer has proved very
satisfactory.
Grim and Rowland 33 have constructed curves for their apparatus
correlating the height of the thermal deflections with the temperature
difference at the time of the reaction
(Fig. 80). These curves were based o
8
on measurements of the swing of the B
16
galvanometer for known temperature
24
differences. Use of these curves with ~
~
the differential curves will measure the o 32
temperature difference represented by I
various peak heights.
Experimental Factors and Variations l' 4
A
in Material Which Influence Results.
As has just been shown, variations in E 12
8 i
the equipment used may influence dif- ~ 16
ferential thermal results. Variations 20
in the materials being studied and in t-

o 200 400 600 800 1000
experimental technique may also cause Temperature,OC
significant variations in differential FIG. 80. Scales for determining

thermal curves. the temperature differences shown
The temperatures at the junctions by the peaks of the thermal curves
of the difference thermocouple depend for the apparatus of Grim and Row-
land. 33 Scale (A) 100 ohms series
to some extent on the coefficient of resistance; scale (B) 200 ohms series
thermal diffusivity of the material in resistance.
which they are embedded. The rate at
which a point in a hot body will cool under definite surface conditions
is known as thermal diffusivity and is equivalent to K/dc p , when K
is conductivity, d the density, and Cp the specific heat. The diffusivity
of the specimen may be different from that of the inert material, and,
furthermore, the diffusivity of the specimen may change as it is heated
because of the formation of new phases at high temperatures or because
of shrinkage of the sample. The effect of such a change on the differen-
tial curve will be to produce a sloping base line, or a sharp offset in the
curve, depending on the suddenness of the change. In the curve for
quartz (Fig. 76) the base line is shifted upward after the inversion of a
to iJ-quartz because of a difference in diffusivity.
The inert material used should have substantially the same specific
heat, heat conductivity, and thermal diffusivity as the sample studied.
33 Grim, R. E., and R. A. Rowland, Differential Thermal Analyses of Clay Minerals
and Other Hydrous Materials, Am. Mineral., 27, 746-761, 801-818 (1942).
2(;6 Clay Mineralogy
Also it should not experience any thermal reaction in the course of
analysis. Calcined alumina (a-Alz0 3 ) has proved quite satisfactory as
inert material. In an attempt to maintain uniformity between sample
I I I I ,
5-20 Microns
I I I I
/-5 Microns
..I-I-I--t-t-t"J ._
1\ I
v
I 1 I I
0.5- / Micron
L-I--+-+-H-r r-
I\J
I I I I I I
0.2-0.5 Micron
-r-- 1
!
I t I I I I
0./-0.2 Micron
-1-.1-'
I I t I I I
0.05-0./ Micron
,...... /' .._

-r-- l"-
V "~
200 400 600 800 1.000
Temperature C
FIG. 81. Variation of differential thermal curves for kaolinite with rartiele size. after
Spiel et al. 27
and inert material, calcined clay has been used occasionally as the inert
material. However, calcined clay may well have different diffusivity
from natural clay; hence little is gained by this procedure. Also, in some
cases, the calcined clay may show thermal reactions due to phase changes.
It is general practice to pack the sample in the holder under moderate
p!essure, and it follows from the foregoing statements that the sample
Ii
Dehydration 207
and inert material should be packed or prepared in the same way. The
importance of packing varies with the kind of material. It is particu-
larly important with light, fluffy material and, as shown by Gruver,32 is
of very little importance for granular, relatively fine grained material with
a considerable range of particle-size distribution so that it will pack
naturally to a fairly dense mass.
So far as packing of the sample is concerned, the particle-size distribu-
tion of the sample is generally of no importance, unless the material is all
relatively coarse (+60 mesh) or extremely fine grained (-2 microns).
A range of particle size permitting dense packing is advantageous.
Spiel et al. 27 has shown that for some materials the thermal curve varies
with the particle size of the component minerals, particularly when the
maximum size is about 2 microns. In general (Fig. 81), the size of the
thermal reaction and the temperature of the peak decrease as the particle
size decreases. In some materials (see page 213) the decrease in particle
size may be accompanied by a decrease in crystallinity, which is reflected
in the differential thermal curve by lower intensity of reactions and a
decrease in the temperature of the reaction peaks. According to Arens,29
if the size of particle is larger than about 20 microns, the surface area is
too small for dehydration reactions to occur rapidly enough to yield
pronounced effects on differential thermal records. Reactions due to
phase changes would be largely independent of particle size.
Arens 29 has investigated the effect on the differential thermal curve of the
shape and size of the sample hole and the position of the thermocouple
bead in the hole. According to him, the size and shape of the hole greatly
affect the peaks for reactions accompanied by a change of weight, e.g.,
dehydration, and have little effect on peaks due to phase changes. For
example, a deep, narrow hole will enhance the endothermic reaction for
the dehydration of illite. In the low-temperature region, heat transfer
occurs largely by conduction; in the high-temperature region, largely by
radiation. Placement of the thermocouple junction deep in the sample
yields curves with relatively strong endothermic peaks in the low-tem-
perature range and flat endothermic peaks in the high-temperature range.
Exothermic reactions are recorded best with deep placement of the
thermocouple junction.
Norton 34 and Spiel et alY have shown that, in general, the slower the
heating rate, the broader the peak and the lower the temperature of the
.I peak (Fig. 82). Increasing the heating rate delays the attainment of
the temperatures of both endothermic and exothermic peaks and increases
the height of the peak and the temperature interval during which the
reaction takes place. According to Spiel, the area under the curve for a
3' Norton, F. II., Critical Study of the Differential Thermal Method for the Identi-
fication of the Clay Minerals, J. Am. Ceram. Soc., 22, 54-63 (1939).
208 Clay Mineralo~y
given reaction and the temperature at the start of the reaction seem to be
independent of the rate of heating. ArenS 29 has not checked this con-
clusion, having found some variation in tIle area under curve with the
heating rate. The experiences of many investigators have shown that a
heating rate in the range of 10 to 15C/min is most satisfactory. A
slower heating rate unduly reduces the sharpness of the reactions, and
600 C
FIG. 82. Effect of variation in heating rate on the endothermic reaction for the
dehydration of kaolinite, after Spiel ei aZ.2 7
a faster heating rate causes too much overlapping of reactions, particu-

larly in samples which are mixtures. The heating rate must be uniform,
for even small changes in the rate of heating are likely to be reflected in
the differential curve.
In order to obtain reproducible results for many materials, the furnace
atmosphere must be kept constant, and for IJ].aterials subject to oxidation
it is important that the samples be prepared and placed in the furnace
in such a way that they always have the same ease of oxidation. Row-
Dehydration 209
35
land and Jonas have shown how variations in particle size, packing,
covering of block, and dilution of sample, in addition to furnace atmos-
phere, influence the ease of oxidation in siderite and hence the differential
curves obtainable from this mineral (Fig. 83).
I
;"
Curve 'A"
1000 2000 600 0 700 0 800 0 9000 10000
Curve ''8''
1000 2000 3000 4000
Curve 't'''
1000 2000 3000
++c._u_r_ve+I_~_'_''___-r__~_~~Ir-_'__
0 0 0 0 0 0
-4__-+
100 200 300 400 500 600 700 0
800 0
9000 10000
FIG. 83. Differential thermal curves for siderite, after Rowland and Jonas.a 6 (A)
Firmly packed without cover on sample holder; (B) firmly packed without cover and
run in nitrogen atmosphere; (C) firmly packed with cover on specimen holder; (D)
very finely ground, loosely packed, and without cover.
Arens 29 has pointed out that an increase in the partial vapor pressure
in the furnace may cause a significant delay in the occurrence of a dehy-
dration reaction. Rowland and Lewis 36 have shown the great effect
of an atmosphere of CO 2 in changing the reaction temperatures of the
carbonates.
In studying carbonaceous clays, it is common practice to use an inert
gas in the furnace to suppress the oxidation reaction, which otherwise
35 Rowland, R. A., and E. C. Jonas, Variations in Differential Thermal Analysis
Curves of Siderite, Am. Mineral., 34, 550-558 (1949).

36 Rowland, R. A., and D. R. Lewis, Furnace Atmosphere Control in Differential
Thermal Analyses, Am. Mineral., 36,80-91 (1951).
210 Clay Mineralogy
frequently produces an exothermic effect large enough and of sufficient
duration to mask all other thermal reactions.
Spiel et al. 27 have presented a series of curves for kaolinite diluted by
various quantities of an inert material, showing that the size of the reac-
tion and the temperature of the peak decrease as the amount of material
decreases. The temperature of the peak, therefore, is not an absolute
value but depends among other things on the amount of material used or
present in a mixture.
Grim 37 has shown that the characteristics of curves for mixtures of some
minerals depend somewhat on the intimacy of mixing of the components,
and that in clay minerals there is considerable difference between the
curves for mixtures of discrete particles several microns or more in
diameter and those for mixtures consisting of an intimate interlayering
of considerably smaller units, such as are frequently found in nature.
Arbitrary curves set up on the basis of dry mixtures of clay minerals,
therefore, are frequently of no value in evaluating curves for natural
clays. In general, the thermal reactions for the individual components
decrease in intensity and sharpness of reaction as the intimacy of mixing
Increases.
Care in the preparation of material for analysis and in making the
analysis will provide reproducible curves, which frequently permit
quantitative interpretations. It is obvious from the foregoing discussion
of the method that its accuracy for quantitative analysis is limited. The
accuracy varies with the kind of material being analyzed, but for many
substances values within 2 to 5 per cent are obtainable. In general, the
quantitative evaluations are based on the area under the peaks that
correspond to the reactions of the individual components. In some
instances it is difficult to measure the area under the curve, because the
reaction does not start or end abruptly. Berkelheimer 38 and Dean 39
have presented methods for measuring the areas that are workable for
reactions that are at least moderately sharp and intense.
The sensitivity of the method varies for different materials depending
on the intensity of their thermal reactions. For hydrates, such as
gibbsite, which yield intense abrupt thermal reactions, less than 5 per
cent can be detected, whereas in the case of the micas at least 10 to 15 per
cent is necessary for detection, because of the low intensity and lack of
abruptness in their thermal reactions. ."
37 Grim, R. E., Differential Thermal Curves of Prepared Mixtures of Clay Minerals,

Am. Mineral., 32, 493-501 (1947).
38 Berkelheimer, L. H., Differential Thermal Analyses of Quartz, U.S. Bur. Mines
Rept. Invest. 3763 (1944).
39 Dean, L. A., Differential Thermal Analyses of Hawaiian Soils, Soil Sci., 63, 95-105
(1947).
r
Dehydration 211
Identification of High-temperature Phases. The identification of
changes in phase on heating the clay minerals has been made almost
entirely by means of X-ray-diffraction analyses. In early investigations
the samples were heated to various temperatures, quenched or allowed
to cool, and then studied. Currently the use of high-temperature X-ray-
diffraction cameras has permitted the obtaining of diffraction patterns
while the minerals are at the high temperatures. Also, in the author's
laboratory, a small furnace has been mounted on the diffraction unit, so
that spectrometer traces are obtained while the minerals are at the
elevated temperature. The latter techniques not only save a tremendous
amount of time but remove the possible difficulties resulting from phase
inversions on cooling. It seems likely that the high-temperature X-ray-
diffraction technique will provide information of great value in studies
of the formation of minerals at elevated temperatures and of transforma-
tions at high temperatures. Except for the use of the petrographic
microscope, other possible methods of identifying high-temperature trans-
formations have been given little consideration. An exception is the
work of Comefero, Fischer, and Bradley,40 using the electron microscope
to reveal the development of mullite in heated kaolinite. The petro-
graphic microscope is adequate for identifying a new phase, if it is well
developed and relatively abundant. If the phase forms slowly, the
initially formed material is frequently not sufficiently abundant nor in
large enough units for identification optically; hence the point of the
beginning of the transformation cannot be fixed by microscopic study.
ALLOPHANE
Dehydration curves published by Ross and Kerr S (Fig. 67) show a con-
tinuous loss of water, with no breaks or very prominent flexures but a
gradual decrease in amount as the temperature is increased. Water of
the order of 10 to 15 per cent is lost below about HOC, and a roughly
similar amount is lost at higher temperatures up to about lOOOC.
These authors conclude that the dehydration curves of allophane are
characteristic of a material in which the water is held in solution and not
in chemical combination. Nutting 6 has published curves for two
allophane samples, one of which shows a gradual water loss, whereas the
other shows a distinct flexure between about 550 and 600C.
Grim and Rowland 33 have published differential thermal curves for
two allophane samples (Fig. 71), both of which show distinct endothermic
reactions at low temperatures, corresponding to the water loss below
llOC shown by the dehydration data. One of the differential thermal
40 Comefero, J. E., R. B. Fischer, and W. F. Bradley, Mullitization of Kaolinite,
J. Am. Ceram. Soc., 31,254-259 (1948).

212 Clay Mineralogy
curves also shows an endothermic reaction at a temperature suggestive
of loss of OR lattice water. The significance of this reaction and the
distinct flexure shown by Nutting's6 data might reflect some structural
organization in this allophane, or possibly only the presence of a crystal-
line impurity. The curve for the other allophane sample shows only very
slight reactions above 300C.
No investigation appears to have been made of the high-temperature
phases formed on heating allophane. Since the composition of allophane
can vary widely, great latitude in the high-temperature phases developed
in different samples would be expected. No information IS available
regarding the possible rehydration of allophane.
KAOLINITE
Dehydration and Phase Changes on Heating. Dehydration curves for

kaolinite are given in Figs. 65 and 67, and differential thermal curves are
given in Fig. 71. Ross and Kerr 7 have pointed out that the dehydration
curves are almost flat up to about 400C, showing little loss of water at
low temperatures, which is in accord with the absence of any thermal
reactions in the low-temperature region of the differential thermal curves
of many kaolinites. Differential curves for some poorly crystalline
kaolinites (Fig. 71) show a small initial endothermic reaction, indicating
that, when there is irregularity in the arrangement of the kaolinite units,
a small amount of water may be present between the layers. This is in
accord with the slightly greater c-axis spacing of poorly crystalline as
compared to well-crystallized kaolinites.
A series of dehydration curves presented by Ross and Kerr 7 show that
most of the dehydration takes place between about 400 and 525C.
The precise temperature for the loss of this OR lattice water varies from
kaolinite to kaolinite, and this variation may be explained by variations
in particle size, since the dehydration temperature is known to decrease
with decreasing particle size. It may also be explained by variations in
crystallinity, since the poorly crystalline material loses its hydroxyl water
somewhat more readily than well-crystallized kaolinite. A small amount
(2 to 3 per cent) of water is retained at about 525C, and this moisture is
lost gradually up to about 750 to 800C, where dehydration is essentially
complete.
Ross and Kerr 7 found no difference in the dehydration characteristics
of kaolinite and anauxite, but they did find that the dehydration tem-
perature for nacrite and dickite was slightly higher than that required for
kaolinite. Nutting 6 has published a few curves for kaolinite, showing a
slightly higher dehydration temperature than those published by Ross
and Kerr. Nutting also indicates that the dehydration temperature of

I
Dehydration 213
anauxite is slightly lower than that of kaolinite. The apparent difference
in the findings of these investigators may be explained by differences in
the particle size of their samples and possibly also in degree of crystallinity.
DeKeyser 41 has stated that kaolinite can be dehydrated completely at
temperatures as low as 3500 with long heating times (200 hr), and
indeed this is about the temperature of the very beginning of the break
in the dehydration curves for the mineral. In general, very long time
intervals are necessary to reach equilibrium for the clay minerals at tem-
peratures where the loss of OR water is beginning.
Differential thermal curves of kaolinite show an intense, sharp endo-
thermic reaction corresponding to the loss of OR water (Fig. 71). The
reaction begins at about 4000, and for well-crystallized kaolinite the
peak is at about 6000. The intensity of the reaction and hence the size
of the peak, as well as the peak temperature, are reduced slightly as the
particle size decreases and as the crystallinity decreases. The variation
seems to be greater for the crystallinity factor than for the particle-size
factor. According to Grimshaw et al.,31 the peak temperature is 20 to
300 lower for poorly crystalline kaolinite than for the well-crystallized
variety.
There seems to be no difference in the differential curves for kaolinite
and anauxite. The temperature of the peaks for the reaction for loss of
OR water is higher (about 1000) for nacrite and dickite than for
kaolinite.
The sections of the differential curves for poorly crystallized kaolinite
lying between the peak for loss of hydroxyl water and the intense exo-
thermic reaction at about 9500 are relatively fiat, whereas those for well-
crystallized kaolinite in this temperature range tend to have an upward
slope, with a slight endothermic reaction just before the exothermic
reaction. Grim and Bradley 42 have concluded that the loss of OR water
in poorly crystallized kaolinite is accompanied by a fairly complete loss
of structure, but that in well-crystallized kaolinites some structural
remnants persist along with some water, and that this remnant is lost
at the time of the higher-temperature endothermic reaction.
Numerous investigators have debated the nature of the material
immediately following the loss of OR lattice water from kaolinite. Insley
and Ewe1l 43 showed that the dehydrated kaolinite did not have the samE
chemical attributes, solubility in acid, etc., as mixtures of amorphom
4I DeKeyser, W., A Study of Kaolin and Some Belgian Clays, Ann. mines Belg., 40
711-806 (1939).
42 Grim, R. E., and W. F. Bradley, Rehydration and Dehydration of the Clay
Minerals, Am. Mineral., 33, 50-59 (1948).

<3 Insley, H., and R. H. Ewell, Thermal Behavior of Kaolin Minerals, J. Research
Nall. Bur. Standards, 14,615-627 (1935).

214 Clay Mineralogy
silica and alumina heated to similar temperatures. Oomefero, Fischer,
and Bradley 40 showed that the diffraction bands of dehydrated kaolinite
were more distinct than such bands from strictly amorphous solids; also,
on the basis of electron micrographs, they showed that the hexagonal
shape of the kaolinite was preserved above 6000. All this work appears
to have been done with well-crystallized kaolinite, and the conclusion
seems definite that, on dehydration, such material is not strictly amor-
phous but still retains some degree of order. It also appears certain that,
for poorly crystalline kaolinite, the degree of order is relatively less and
the loss of structure is fairly complete.
The exothermic reaction shown by the differential thermal curves
between about 900 and 10000 is somewhat less intense and takes place
over a slightly wider temperature interval for poorly crystallized kaolinite
than for the well-crystallized mineral. The explanation for this exo-
thermic reaction has been a matter of much dispute. Insley and EweH43
and many others have attributed it to the formation of 'Y-A1203. When
pure amorphous alumina is heated, 'Y-Ab03 crystallizes over a long
temperature interval and at a lower temperature. Furthermore,
'Y-A1 20 3 has a structure unlike one which would be anticipated from the
sharpness and high intensity of this exothermic reaction. Insley and
Ewell 43 attempt to explain these difficulties on the basis of some struc-
tural order in the dehydrated kaolinite which delays the formation of
'Y-A1203 until a relatively high temperature, with a consequent sudden
reaction releasing much energy. Oomefero, Fischer, and Bradley 40
present cogent reasons for attributing the exothermic reaction to the
formation of mullite in the sample of well-crystallized kaolinite that they
studied. These investigators, using a combination of electron micros-
copy and X-ray diffraction, found mullite needles developing at first
while the hexagonal outline of the kaolinite was preserved, and they con-
clude that, after the loss of OR water, there exists a uniquely organized
compound which collapses into muHite nuclei.
Several investigators 44 ,45 have reported both mullite and 'Y-A1203 in
kaolinite heated to 10000, so that the exothermic reaction could be
attributed to the formation of either or both of these compounds in some
samples.
Richardson 46 has recently published the results of X-ray-diffraction
44 McVay, T. M., and C. L. Thompson, X-Ray Investigation of the Effect of Heat
on China Clay, J. Am. Ceram. Soc., 11, 829-841 (1928).

45 Wilm, D., U. Hofmann, and K. Endell, Ueber die Bedeutung von rontgeninter-
ferenz Untersuchungen bei hohen Temperatures fur keramisch Forschung, Sprechsaal,

38, (1934).
46 Richardson, H. M., Phase Changes Which Occur on Heating Kaolin Clays,
"X-ray Identification and Structure of the Clay Minerals," Chap. III, pp. 76-85,
Mineralogical Society of Great Britain Monograph, (1951).
~',
I'
Dehydration 215
analyses of a series of kaolinites heated for 20 hr at each 50C temperature
interval from 800 to 1350C. Ris results indicate the presence of
'Y-alumina from 950 to above 1350C. Richardson also investigated
halloysite and found no mullite forming below nooc, and since hal-
loysite shows the same exothermic peak as kaolinite, the author concludes
that the exothermic reaction in kaolinite is probably due to the formation
of 'Y-alumina.
Recent unpublished work by Glass 47 suggests that the apparent
discrepancy in results may be due to variations in the crystallinity of the
material investigated and to the experimental conditions. According
to Glass, well-crystallized kaolinite at lOOOC yields strong diffraction
lines for 'Y-AhOa and weak lines for mullite. On further heating to
higher temperatures, or on long-continued heating at lOOOC, mullite
continues to develop. If the heating is rapid, as it is in a differential
thermal analysis, little further mullite develops until about 1250C, when
it abruptly appears in abundance and is accompanied by an exothermic
reaction. Cristobalite appears in abundance at about 1300C. In the
case of poorly crystallized kaolinite, at least in the samples studied by
Glass, only 'Y-A1203 appears at lOOOC. Mullite appears abruptly at
about 1200C, accompanied by an exothermic reaction, and aistobalite
appears in abundance at about 1300C. Johns 47 a; has recently shown
that the high temperature exothermic reaction for kaolinite can be
explained by a mullite nucleation which may be revealed by thermal data
before it is clearly shown by diffraction data. Johns has also shown that
the nucleation would vary somewhat with the crystallinity of the
kaolinite.
Bertorelli and Williams 48 have claimed that mullite formation is
enhanced by heating in the presence of inert gases. Caillere, Renin, and
Ture 49 have shown that there is a great variation in the intensity of the
exothermic reaction at about 950C as a consequence of variations in the
presence of a very small amount of impurities. Parmelee and Rodriguez 50
have shown that zinc, lithium, magnesium, iron, manganese, cerium, and
molybdenum markedly enhance the formation of mullite from kaolinite;
boron and calcium enhance it slightly; sodium, potassium, titanium, and
tin retard its formation; and the relative influence of these elements varies
47 Glass, H. D., Illinois State Geological Survey. Personal communication.
47a Johns, W. D., The Mineralogy of Flint Clays and associated Fireclays, Ph.D.
I thesis, Illinois, 1952.
's Bertorelli, O. L., and 1. Williams, Preparation of Mullite, U.S. Patent 2,536,122
(1949).
49 Caillere, S., S. Henin, and S. Ture, Investigation of the Differential Thermal
Analysis of Clays, Compt. rend., 223, 383-384 (1946).

50 Parmelee, C. W., and A. R. Rodriguez, Catalytic Mullitization of Kaolinite by
Metallic Oxides, J. Am. Ceram. Soc., 26, 1-10 (1942).

216 Clay Mineralogy
with temperature. Crookston 51 has found that the presence of potas-
sium reduces the formation of cristobalite from kaolinite markedly;
sodium reduces it slightly; and calcium, magnesium, and hydrogen
have little effect. These data make it
clear that the effect of small amounts
of impurities is undoubtedly a signifi-
E _v v- ~
_"..-
cant factor in explaining the apparent
variation in results obtained by differ-
V I 'r-O ent investigators on heating kaolinite.
0 ~
Hyslop and McMurdo 52 have pre-
C sented expansion-contraction curves for
the various clay minerals (Figure 84).
~~
8
- Kaolinite shows an initial slight expan-
A
"
"1"-r---,.\
sion up to about 500C, followed by
~ c contraction to the limits of the experi-
ments at 1100C. The beginning of
the contraction accompanies the loss
,
\ "- B
'E
of OH water. There is a flexure at
about 900C, corresponding to an inter-
val of relatively reduced contraction;
I~ this is the temperature of the exother-
mic reaction that is attributed to the
o 200 400 600 800 1000 1200
formation of I'-Ah03 or mullite.
Temperature C
Harmon and Parmelee 53 have stated
FIG. 84. Thermal expansion of the
clay minerals, after Hyslop and that
McMurdo. 52 One vertical division . . . only minor changes in true specific
is 0.80 per cent expansion or con- gravity take place up to the dehydration
traction. (A) Kaolinite, china clay,
of the kaolinite. At this point the specific
Cornwall, England; (B) halloysite,
Missouri; (C) beidellite (Putnam gravity drops suddenly as would be expected
soil), Missouri; (D) illite, Fithian, from the theory that dehydrated kaolinite
Illinois; (E) illite, Gilead, Illinois.is a mixture of amorphous alumina and
silica. The specific gravity then slowly
increases apparently indicating that a small amount of amorphous oxides are recom-
bining to form a compound. The increase is suddenly increased at about 900 o e.
Harmon and Parmelee did their work with ball clays in which the kao-
linite crystallinity was probably low. Rieke and Mauve 54 have presented
51 Crookston, J. A., The Effect of Exchangeable Bases on the Fired Properties of
Fireclays, Ph.D. thesis, University of Illinois (1949). ~
52 Hyslop, J. F., and A. McMurdo, The Thermal Expansion of Some Clay Minerals,
Trans. Ceram. Soc. (Engl.), 37, 180-182 (1938).

53 Harmon, C. G., and C. 'IV. Parmelee, Testing and Classification of Ball Clays-
Thermal History, Bull. Am. Ceram. Soc., 21, 280-286 (1942).

54 Rieke, R., and L. Mauve, Zur Frage des N achweises des mineralischen Bestand-
teile der Kaoline, Ber. deut. keram. Ges., 23, 119-150 (1942).
Dehydration 217
similar data (Table 27) for a kaolinite from Zettlitz in Czechoslovakia,
which is well crystallized; their data are in essential agreement with those
of Harmon and Parmelee.
TABLE 27. SPECIFIC GRAVITY OF CLAY ]\1"rNERALS AFTER HEATING TO VARIOUS
TEMPERATURES FOR 12 HR
(After Rieke and Mauve)H
Raw
Mineral 400C 450C 500C 600C 700C BOooe \loooe \l50oe lOoOoe
clay
- - - - --- - - - - - - - - - - - - - -
Kaolinite, Zettlitz ........ 2.64 2.64 2.49 2.47 2.50 2.53 2.60 2.62 2.69 2.74
Mica, Sarospatak ........ 2.B1 2.BO 2.71 2.63 2.58 2.56 2.51 2.52 2.53 2.55
Halloysite ............... 2.62 2 57 2.46 2.45 2.48 2.52 2.58 2.65 2.70 2.58
Muscovite ............ '. 2.86 2.89 2.91 2.92 2.84 2.81 2.71 2.64 2.59 2.58
Montmorillonite, Wyo- Could not be determined be- 2.59 2.58 2.49 2.52 2.55 2.52
ming cause of swelling
According to Norton, 55 kaolinites fuse from about 1650 to 1775C.

The fusion point is preceded by a vitrification interval, which may
extend over several hundred degrees, during which a glassy component
is an increasingly abundant phase. Small amounts of associated impuri-
ties may cause large changes in the determined fusion point and vitrifica-
tion range.
Rehydration. Grim and Bradley 42 have shown that well-crystallized
kaolinite, after heating to 600C slowly (70 hr), picks up a small but
appreciable amount of OH water at room temperature, and that poorly
crystallized kaolinite similarly heated apparently does not regain any
OR water under the same conditions. They interpret this as evidence
that, on dehydration of the well-crystallized mineral at 600C, some struc-
ture remains which has sufficient regularity to take up at least minor
amounts of hydroxyl water.
Van Nieuwenberg and Pieters56 have shown that after heating to
850C kaolinite can be completely rehydrated under steam pressure of
100 atm. Schachtschabe15 7 states that after heating to 800C kaolinite
rehydrates slowly at noc, but that at 175 0 to 205C under pressure it
completely rehydrates in about 100 hr.
HALLOYSITE
Dehydration and Phase Changes on Heating. Dehydration curves

published by Ross and Kerr S and N utting 6 (Figs. 65 and 67) show
55 Norton, F. H., "Refractories," McGraw-Hill, Ncw York (1949).
56 Van Nieuwenberg, C. J., and H. A. J. Pieters, Rehydration of Metakaolin and
the Synthesis of Kaolin, Ber. deut. keram. Ges., 10, 260-263 (1929).
57 Schachtschabel, P., Dehydration and Rehydration of Kaolin, Chem. Erde, 4, 395-
419 (1930).
218 Clay Mineralogy
significant loss of water below 100 0e for some halloysites and not for
others, depending, as is now known, on whether or not the halloysite is
the hydrated form (4H20). The curves show that at temperatures
between 1000 and 400 0e there is a very slight gradual loss of water, with
the amount increasing somewhat between about 300 0 and 400 0e. From
4000-430 0e to 500 0e the hydroxyl water is lost rapidly and abruptly.
0
Above about 500 e there is again a gradual loss of water up to about
800 0e, where dehydration is essentially complete. Ross and Kerr point
out that halloysite loses its OH lattice water at a temperature 600 to 80 0e
lower than that for kaolinite.
The water lost below 100 0e is pore water and that occurring between
basal plane surfaces of adjacent unit layers. Most of this interlayer
water can be lost at room temperature under conditions of low humidity
and over moderate periods of time. Brindley and Goodyear 58 have
shown that drying at room temperature does not completely remove all
interlayer water, and that the formula for air-dried halloysite is Al 20 v
2Si0 2.nH 20 with n = 274 to 2%, which is equivalent to about one layer
of water to every four silicate layers. Drying to about 400 0e is required
to remove the interlayer water completely. The moisture gradually lost
between 100 0e and 400 0 e undoubtedly corresponds to the residue of
interlayer water, which is driven off with difficulty. Brindley and
Goodyear have also shown that the complete loss of the interlayer water
is not accompanied by a reorganization of the randomly displaced silicate
layers.
Differential thermal curves for halloysites are given in Fig. 71. The
hydrated form shows an initial V-shaped endothermic peak resulting
from the loss of the interlayer water. This peak is, of course, small or
completely absent in the lower-hydration forms of the mineral. The
shape of the initial endothermic peak is unlike the one sometimes yielded
by poorly crystalline kaolinite at about the same temperature. The
peak given by kaolinite is less intense and less ahrupt than that of
halloysite.
The differential thermal curves above about 200 0e are essentially like
those for kaolinite. The sharp endothermic reaction due to loss of OH
water tends to develop a peak at a slightly lower temperature than that
in the case of kaolinite. Grimshaw, Heaton, and Roberts 31 found in their
differential thermal analyses that the endothermic peak for halloysite
was at 500 0e, for poorly crystalline kaolinite at 550 0 to 562e, and for
well-crystallized kaolinite at 583e. Kerr, Kulp, and Hamilton 59 state
58 Brindley, G. W., and J. Goodyear, X-Ray Studies of Halloysite and Metahalloy-
site, II, The Transition of Halloysite to Metahalloysite in Relation to Relative

Humidity, Mineralog. Mag., 28, 203-215 (1948).
69 Kerr, P. F., J. L. Kulp, and P. K. Hamilton, "Differential Thermal Analyses of
Dehydration 219
that this peak for halloysite is asymmetrical, being more abrupt on the
high-temperature side than the similar peak for kaolinite.
For the temperature interval above the dehydration reaction for OH
lattice water, the differential thermal curves of some halloysites are like
that of well-crystallized kaolinite, suggesting that some structure is
retained after dehydration and up to the temperature of the slight endo-
thermic reaction at about 900C. It is not known whether or not this is
a characteristic of all halloysites.
Glass 47 has investigated the phases developing on heating halloysite,
and the samples studied by him developed 'Y-A1 20 3 at the temperature of
the exothermic reaction of the differential thermal curve (950C );
mullite developed at 1250C; and cristobalite at 1300C. Like poorly
crystallized kaolinite, and unlike well-crystallized kaolinite, X-ray data
for halloysite did not indicate mullite in the temperature region of the
exothermic reaction at about 950C. Richardson 46 found on heating
halloysite that "I-alumina was present from 900 to 1150C, cristobalite
from 1100 to 1300C, and mullite from 1100 to 1350C. It would seem
from Richardson's data that for halloysite the sharp exothermic reaction
at about 950C must be explained by the formation of "I-alumina rather
than by mullite. However, J ohns 47a recent findings suggest that nuclea-
tion of mullite is a more likely explanation. .
It seems likely that the development of high-temperature phases would
be influenced by the presence of traces of various elements in a manner
similar to that shown previously for kaolinite.
Hyslop and McMurdo 52 have presented an expansion-contraction
curve for halloysite (Figure 84). The curve shows the halloysite to
expand slightly up to the dehydration of the mineral at about 500C.
Above this temperature there is a gradual contraction up to about 900C,
followed by a steep contraction up to about 1000C. From 1000C to
the end of the experiments at 1100C there is a further small amount of
contraction. The difference in the character of the curves for halloysite
and kaolinite may reflect differences in the high-temperature phases noted
above. When halloysite is fired to temperatures just short of complete
vitrification, masses of the mineral crack into small pieces, whereas
masses of kaolinite remain intact. The difference in physical character
must also reflect a difference in high-temperature phase reactions.
Although halloysite is very refrac~, it is difficult to use in ceramic
products because of this high-temperature characteristic. The tendency
of halloysite to crack into pieces at high temperatures makes it difficult
to determine its vitrification range and fusion point. It seems that the
Reference Clay Mineral Specimens," Rept. 3, American Petroleum Institute Project

49, Columbia University, New York (1949).
220 Clay Mineralogy
mineral fuses at a slightly higher temperature than kaolinite and that the
fusion point is preceded by a very short vitrification interval.
Rehydration. The interlayer water lost at room temperature is not
ordinarily regained; i.e., the reaction is not reversible. Bradley60 has
shown that certain organic molecules, e.g., glycol, will penetrate between
the layers of the dehydrated mineral and that, when the glycol is removed
by a solvent such as alcohol, mixed with water, water remains between
the layers. According to MacEwan 61 and Brindley,62 this reintroduction
of interlayer water is effective only if the original interlayer water has not
been completely removed. Brindley suggests that the glycol molecules
cannot enter between the layers unless some water is present to keep them
partly open.
Grim and Bradley42 have shown that halloysite, after being heated to
600C for 1 hr, regains a very small amount of OR lattice water in the
course of standing at room temperature for 70 days. A similar result
was found for well-crystallized kaolinite, but not for poorly crystalline
kaolinite. They interpret this to mean that some structural organization
is left in halloysite after the loss of OR lattice water. The similarity in
this attribute between halloysite and well-crystallized kaolinite rather
than poorly crystalline kaolinite suggests that the difference between the
kaolinite and halloysite structures is based on more than the variation in
the stacking of adjacent silicate sheets. Bates et al. 63 have emphasized
this point in considering the rolled tube-like shape of the halloysite
minerals revealed in electron micrographs.
MONTMORILLONITE
Dehydration and Phase Changes on Heating. Dehydration curves

after Ross and Hendricks 10 are given in Fig. 68. The curves show con-
siderable water loss at low temperatures (100 to 200C), and, as has been
shown (pages 17.5-180), the amount of this water, which occurs mostly as
interlayer water between the silicate sheets, is contingent upon the nature
of the adsorbed ions and the pretreatment of the sample, e.g., amount of
drying, relative humidity, etc. The dehydration curves are S-shaped,
60 Bradley, W. F., Diagnostic Criteria for Clay Minerals, Am. Mineral., 30, 704-713
(1945).
61 MacEwan, D. M. C., Complexes of Clays with Organic Compounds, I, Complex
Formation between Montmorillonite and Halloysite and Certain Organic Liquids, "
62 Brindley, G. W., The Kaolin Minerals, "X-ray Identification and Structure of the
Clay Minerals," Chap. II, pp. 32-75, Mineralogical Society of Great Britain Mono-
graph (1951).
Endellite and Halloysite, Am. Mineral., 35, 463-484 (1950).
Dehydration 221
usually showing no distinct break between the temperature of loss of the
last interlayer water and the beginning of loss of OH lattice water. Ross
and Hendricks have arbitrarily taken the temperature of 3000 as the
break point, and this is probably the best single value that can be named.
The loss of the interlayer water is accompanied by a reduction in the c
dimension to 9.4 to 10 A, with the exact value depending on the size of
the interlamellar ions.
The curves of various montmorillonites show wide variation of tem-
perature for the loss of OH lattice water. For normal montmorillonites,
i.e., those with moderate to low substitutions of iron and magnesium for
aluminum, rapid loss of OH water begins at about 5000, and the dehy-
dration is practically complete at 8000. For nontronites the rapid loss
of hydroxyl water begins at about 4000, and the dehydration is sub-
stantially complete at 8000. Rectorite is unusual, in that the rapid loss
does not begin until about 7000 and the loss is still not complete at
9300. In hectorite the hydroxyls are partially replaced by fluorines,
which are possibly more tightly held within the lattice structure than the
hydroxyls.
The loss of (OR) water in montmorillonites, unlike that in kaolinite and
halloysite, does not begin or end abruptly, except in the case of hectorite.
Another point of difference from kaolinite and halloysite, as Oaldwell and
Marshall 64 and Kelley et al. 9 have shown, is the small variation in the
dehydration characteristics of montmorillonites with particle size. This
is to be expected, since a reduction in particle size of the montmorillonites
is essentially a reduction of aggregates and not of primary components;
i.e., it is a separation of flakes along basal planes of easy separation rather
than a cross breaking of flakes.
The dehydration curves of some of the montmorillonites show flexures
in the temperature interval of the loss of OR water; the significance of
these will be discussed when the differential curves are considered.
Differential thermal curves for a series of montmorillonites are given
in Ifig. 72. At low temperatures they show variable endothermic peaks
due primarily to the loss of interlayer water. The size and character of
these peaks are contingent upon the nature of the adsorbed cation and
on the pretreatment of the sample (see pages 178-179).
Most normal montmorillonites with relatively small amounts of iron
and magnesium replacing aluminum show an endothermic reaction due
to loss of OR lattice water, beginning rather gradually at about 450 to
5000, ending at about 7500, and with a peak temperature of about
7000. Large replacement of aluminum by iron causes a reduction in the
Properties of the Clay Minerals N ontronite, Attapulgite and Saponite, ColI. Agr. Univ.
Missouri Research Bull. 354 (1942). /
Clay Mineralogy
temperature of the reaction, and for nontronite the peak temperature is
550 to 600C. Some montmorillonites which have a moderately high
magnesium content, e.g., the sample from Otay, California, also show a
lowering of the dehydration temperature. On the other hand, the mag-
nesium-rich hectorite dehydrates at a higher temperature than the
aluminous montmorillonites.
Some montmorillonites of relatively low iron and magnesium contents
show unusually low temperatures for the hydroxyl-loss endothermic
reaction, and some samples are known (e.g., montmorillonite from Palmer,
Arkansas) that show a dual endothermic peak for hydroxyl water loss.
In some cases the dual peak can be explained tentatively by a mixing of
montmorillonites, e.g., high-iron and aluminous varieties. In some other
examples the possibility of small amounts of kaolinite or halloysite as
impurities cannot be eliminated. There are, however, some samples for
which there is no satisfactory explanation. Kerr et al. 59 have reasoned
that the explanation may reside in the geometry of the replacement and
population of the octahedral positions in the structure. According to
these authors, variation in the position of the cations in the octahedral
packing and hence variation in the nearness to hydroxyl ions should
affect the strength of the bonding of the OR and hence the energy neces-
sary for its release.
Thilo and Schunemann 65 and Grim and Bradley66 have shown that
the general layer type of structure illustrated by montmorillonite and the
micas is able to survive the elimination of OR water with only moderate
readjustments. According to Bradley and Grim,67 the removal of
hydroxyl water is correlative with increases of 0.1 to 0.3 A in c-axis
periodicity and involves the expulsion of about one-sixth of the oxygens
of the octahedrally coordinated portion of the structure. This correlation
is apparently confined to the heptaphyllite types of montmorillonites.
An idealized rearrangement of the octahedral layer which adequately
accounts for the meager data observable is illustrated in Fig. 85. It
represents merely the lifting of the adjacent oxygen layer out of the
packed position, with the immediate consequences that the vertical
height is increased and the intensity of the 4.5 A diffraction line is aug-
mented. A diagram of a typical dehydration specimen is compared with
its normal dry state in Fig. 85. Diffraction data for the rearranged
material, after Bradley and Grim, are given in Table 28. The idealized
66 Thilo, E., and H. Schunemann, Chemical Studies of Silicates, IV, Behavior of
Pyrophyllite on Heating and the Existence of a "'Vater-free" Pyrophyllite, Z. anorg.
u. allgem. Chem., 230, 321-325 (1937).
66 Grim, R. E., and W. F. Bradley, Investigation of the Effect of Heat on the Clay
Minerals Illite and Montmorillonite, J. Am. Ceram. Soc., 23, 242-248 (1940).
67 Bradley, W. F., and R. E. Grim, High Temperature Thermal Effects of Clay and
Related Materials, Am. Mineral., 36, 182-201 (1951) .
.1\ ..
,'

Dehydration 223
b
FIG. 85. Probable rearrangement in heptaphyllite-type clay minerals upon loss of
hydroxyl water. The diffraction diagrams illustrate the Tatatilla montmorillonite
after heating to 550 0 (upper) and 725C (lower). The schematic sketches represent
the arrangement within an octahedral layer before and after dehydroxylation, after
Bradley and Grim. 67
scheme shown in Fig. 85 involves shared pairs of octahedral coordination

faces for a heptaphyllite and is probably unstable. Applications of the
same scheme to an octaphyllite type would demand sharing of three such
pairs of faces, which is presumably impossible. No rearranged anhydrite
was observed for hectorite and talc, which were the only two certain
octaphyllite structures studied by Bradley and Grim. 67
The structure of many montmorillonites persists to temperatures of the
order of 800 to 900C, and Grim and Bradley 66 have correlated the
destruction of the lattice with the endothermic reaction frequently shown
at about this temperature interval. This correlation probably applies
only to heptaphyllite montmorillonites, for, as has just been shown, an
anhydride structure should not persist in the octaphyllite members.
Some montmorillonites which have been considered to be heptaphyllites,
e.g., some nontronites, do not produce this high-temperature endothermic
peak, and therefore it appears that the anhydrite structure may not
persist in all heptaphyllites.
221 Clay Mineralogy
TABLE 28. DIFFRACTION DATA FOR MONTMORILLONITE (TATATILLA, MEXICO)
725C
IGNITED TO
(From Bradley and Grim 67 )
Calculated relative
Indices as Observed rela- intensities for
d, A
orthohexagonal tive intensities idealized Scheme
of Fig. 85
001 9.7 m 2
002 4.8.'5 m 2
110,020 4.48 88 18
003 3.22 8 10
200, 130 2.59 m 4
2.5 Diffuse m
220,040 2.23 m 2
2.1 Diffuse w
005 1. 93 w 1
240, 150 1. 72 w 2
1.67 Diffuse m
330,060 1.51 m 2
1.49 Diffuse w
400,260 1.31 w 3
420,350 1.25 w 1
88 = very strong; 8 = strong; m = moderate, W = weak.

Page has suggested another explanation for the third endothermic
68
peak of the montmorillonites. According to him, it is due to the loss of

OR water which is bonded with magnesiums in octahedral coordination,
rather than to lattice destruction. It has not been established that there
is a correlation between the magJ'lesimll content and the presence or size
of this peak. McConnell 69 has recently attributed this high-temperature
endothermic reaction to the loss of hydroxyls which are in the silica layer
in tetrahedral configuration. McConnell attempts to correlate the peak
with the amount of water lost at high temperatures, but the data are not
completely convincing. An amount of water of the order of 0.5 to 1 per
cent does persist beyond the temperature of the major loss of hydroxyl
water at 500 to 700C. The amount of water to be lost at high tem-
peratures is small in comparison with the size of the endothermic peak,
and furthermore the loss of these hydroxyls does not seem to take place
abruptly, as would be required for the development of a peak of the
character frequently shown on the thermal curves.
Following the third endothermic reaction, the differential thermal
68 Page, J. B., Differential Thermal Analyses of Montmorillonite, Soil Sci., 56,273-
283 (1943).
166-172 (1950).
Dehydration
curves show a wide variety of thermal reactions, indicating that there is
considerable variation in the phases formed when various montmoril-
lonites are heated to elevated temperatures (Fig. 72). Data given in
Table 29 from Bradley and Grim 67 show that this is the case.
TABLE 29. HIGH-TEMPERATURE PHASES DEVELOPED ON FIRING MONTMORILLONITES

(From Bradley and Grim 67)
Sample location 900C 1000C 1100C 1200C 13000 C
Otay, California ... .......... ii-Quartz (1) Cristobalite (1) Cristobalite (1) Cristobalite (1)
Enstatite (3) ii-Quartz (1) Cordierite (1) Cordierite (1)
Enstatite (2) Periclase (3)
Tatatilla, Vera .......... ii-Quartz (2) ............... . .............. Cristobalite (1)

Cruz, Mexico Mullite (2)
Cordierite (2)
Upton, Wyoming .. Spinel (1) Spinel (1) Spinel (2) ............... Mullite (2)
a-Quartz (2) Cristobalite (1)
Cheto, Arizona . ... . . . . .. . . . . ii-Quartz (1) ii-Quartz (1) ............... Cristobalite (1)
Anorthite (?) (3) Cristobalite (3) Cordierite (1)
Anorthite (?) (3)
Palmer. Arkansas .. . . . . . . . . . . Spinel (1) Rpinel (1) Cristobalite (1)

Quartz (3) Spinel (1)
Mullite (1)
Sierra de Guada- . . . . . . . . . . Spinel (1) Cristobalite (1) Cristobalite (1) Cristobalite (1)
lupe, Mexico Spinel (1) Spinel (1) Cordierite (1)
Cordierite (2)
Harris County, .......... Spinel (2) Cristobalite (1) Cristobalite (1) Cristobalite (1)
Texas Cristobalite (3) Spinel (1) Spinel Mullite (2)
Mullite (2)
Pontotoc, Missis- .......... Spinel (1) Cristobalite (1) . .............. Cristobalite (2)
sippi a-Quartz (2) Spinel (1) Cordierite (2)
Fairview, Utah .. .. .......... ............... . .............. Mullite (1) Mullite (1)

Cristobalite (1)
Wagon Wheel Gap, Spinel (2) Cristobalite (1) ............... ............... Cristobalite
Colorado Mullite (3)
Spinel (2)
Nashville, Arkan- .......... ............. " ............... ............... Mullite

sas (woody non- Cristobalite
tronite) Spinel
Parenthetic numbers signif.y (1) important, (2) moderate, (3) minor.
Considerably more experimentation is necessary before all the general

factors controlling the formation of high-temperature phases for mont-
morillonites are well known and before a complete correlation of thermal
reactions and specific phase formation can be made. However, Bradley
226 Clay Mineralogy
67
and Grim have arrived at some general conclusions, based on a careful
study of many samples, and the following discussion is taken from their
work.
The high-temperature portions of the thermal curves of montmoril-
lonites tend to differentiate into two separate types: those which present
an S-shaped feature, and those which exhibit a distinct shoulder directly
above the third endothermic reaction. Clays yielding the S-shaped
curves contain spinel following the first exothermic reaction; it is propor-
tional in amount to the magnitude of the thermal effect. In the second
group the first prominent new phase appears to be quartz. The new
phase is thus seen to develop in the one case from the octahedral region
in the clay and in the other case from the tetrahedral region.
From the point of view of the composition of the mineral, montmoril-
lonite seems to develop quartz if there is substantially no suhstitution
in the tetrahedral sheet, and spinel if there is suhstantial suhstitution of
AIH in tetrahedral positions. The regularity in the composition of the
tetrahedral scheme appears to endow it with ability to maintain its unity
under conditions which disorganize substituted layers. Furthermore the
montmorillonites developing the second type of curve, with the delayed
exothermic reaction, are relatively low in iron. In these low-iron mont-
morillonites, enstatite, mullite, or anorthite develops at about 1000C, in
addition to the .B-quartz, probably depending on the amount of mag-
nesium present and possibly also on the amount of exchangeable Ca++.
Clays of either type, which have undergone either of the foregoing
syntheses, may show an additional, fairly prominent reaction near
1200C which seems to be due to either mullite or cordierite with about
equal frequency. Bradley and Grim 67 point out that the first high-
temperature phases appear to be largely determined by the structural
attributes of the montmorillonite itself. The initial phases are not the
ones which necessarily would be expe'cted from phase-diagram considera-
tions and bulk chemical composition but are due primarily to "inherit-
ance" from the original structure. The later phases developed at higher
temperatures are dependent primarily on bulk composition, and here
"inheritance" is of little importance. Thus in the case of the develop-
ment of mullite or cordierite, the controlling factor is probably the amount
of magnesium present.
The spinel-forming specimens all develop more or less cristobalite
0
around 1000 or 1100C without going through the quartz intermediate
stage. No separate features appear in the differential thermal curves to
indicate any energy effect associated with the growth of the cristobalite.
Since this is true also of opal fired alone, it is assumed that any exchange
involved is not picked up by the thermal method. Cristobalite is
actually effectively somewhat less dense than the condensed-layer con-
Dehydration 227
figuration derived from the articulation of the silica layers (see page 228)
and might even require an energy input under the conditions of the
analyses.
The spinel tends to disappear at temperatures above about 12000 in
all samples studied except the nontronite sample.
Bradley and Grim 67 have considered the structural changes involved
in the development of high-temperature phases when montmorillonites
are heated, and the following statements are also from their work.
As the spinel phases first appear, X-ray-diffraction lines are notably
diffuse, and probably the phase does not have any particular composition.
Rather, it is merely an irregularly constituted assemblage of small cations,
some octahedrally coordinated and others tetrahedrally coordinated,
with only the oxygen packing actually approaching crystalline regularity.
The spinels grow rapidly to a clearly crystalline powder affording a clean
diffraction diagram. No noticeable difference in relative intensities is
observed between the diffuse and the sharp patterns, although cube-edge
parameters may change somewhat (with changing composition), and no
erratic appearance of new lines has been observed.
The cryptocrystalline quartz produced on firing many montmoril-
lonites at about 10000 exhibits many vagaries with respect to its
apparent relation to high-low inversion. Synthesized quartz, when
observed by Bradley and Grim at room temperature, was found in some
cases as the a, in others as the {3 modification, and in one case apparently
intermediate. Both the a and {3 forms make the normal slow transition
to cristobalite at more elevated temperatures. Bradley and Grim 67 sug-
gest that the failure to invert is caused by the trapping of occasional
extraneous ions in the channels of the structure.
The synthesized quartz develops suddenly with a correlative sharp
exothermic effect, and the crystallites are subject to moderate grain
growth in soaking periods of a few hours before eventually transforming
into cristobalite. The original concept of the layer silicates was drawn
from observations of the dimensional compatibility of brucite and the
cristobalite-tridymite type of tetrahedral layer. One might postulate
that a quick crystallization of silica would lead to one of these forms, but
that does not seem to be the case. Bradley and Grim 67 picture the
synthesis of the quartz as illustrated in Fig. 86, which shows two sets
of linked tetrahedrons, representing adjacent montmorillonite surfaces,
in which oxygens are partitioned into two kinds. Suitable articulation
of these pairs actually does simulate the quartz arrangement of height
equivalent to one complete unit cell, and the new grouping effects a
localized gain in density of the order of about 10 per cent. This postula-
tion of the articulation of montmorillonite surfaces into the quartz
arrangement is a consequence of extended clean clay-mineral surfaces.
228 Clay Mineralogy
(a) (b) (c)

FIG. 86. The articulation of tetrahedrally arranged montmorillonite surfaces to form
quartz, after Bradley and Grim. 67 (a) A montmorillonite surface; (b) two distorted
arrangements of a; (c) a superposition of the two levels b.
To articulate two such surfaces into a segment of the cristobalite structure

would require the introduction of additional oxygen between surfaces to
complete the coordination of any silicons which might provide valence
bonds between layers.
The fine-grained cristobalite resulting from syntheses in heterogeneous
matrices is noticeably variable and of uncertain inversion character.
Such material gradually becomes typical cristobalite upon ignition to
sufficiently high temperatures, apparently without any sudden energy
effects.
The MgSi0 3 formed in the firing of some montmorillonites is com-
parable with the" mesoenstatite" of Thilo and Rogge. 7o This material
is probably to be looked upon as a rudimentary nucleation rather than
as a normal crystallization. The structural relations of the layer silicates
and of the pyroxenes are such that projections onto the respective ortho-
hexagonal axes are approximate equivalents. In the imperfectly
developed diffraction diagrams of the fired montmorillonites, the lines
which appear are only those which relate to the pseudohexagonal nature.
In fact they constitute an identification of MgSi0 3 only in the sense
that they correspond to equivalent lines which appear in better and
better crystallized materials of higher magnesium content for gradations
up to and including the typical enstatite crystallization obtainable on
firing tremolite asbestos.
Bradley and Grim 67 have discussed the development of forsterite from
chrysolite and clinochlore. They conclude that the transition from the
old to the new phase requires no chaotic disintegration followed by the
70 Thilo, E., and G. Rogge, Chemische Untersuchungen von Silikaten, VIII, Ber.
deut. chem. Ges. B., 72,341-362 (1939).
I
1
Dehydration 229
reconstitution of a new assemblage, but simply the freedom of a few Mg
ions to migrate to the nearest equivalent interstices. In other words
the oxygen packing remains substantially the same, and the transforma-
tion is a matter of the shift of a few magnesium ions. They conclude
further that the development of mullite from pyrophyllite, which is
probably similar to its development from montmorillonite, is carried
forward to a considerable extent along the same lines as the development
of the forsterite.
Steger 71 has investigated the expansion-contraction of montmoril-
lonite. The specimen examined by him showed a very slight gradual
expansion from 200 to about 700 0 e. Between 700 and 800 e there
0
was a rapid expansion possibly to be

correlated with the development of +2
the anhydride form. This expansion
- +1
was immediately followed by a large c
contraction up to about 950 0 e, when U... 0
Q)
~\
the sample again expanded (Fig. 87). ~ \ A L...-'" /
Rieke and Mauve&4 have shown .S-I
(Table 27) that the specific gravity of -0-
a sample of montmorillonite decreased ..J ~ -2
on heating from 600 to 800 e, fol- c
0 I\J
lowed by a slight increase when it was ~-3
0-
heated to 950C. There was a further oc V\
slight decrease on firing to 1000 0 e, '0-
4
+-
,...8
the maximum temperature of the

-5100 300 500 700 900 1100
experiments.
Temperature C
As would be expected from the large
variations in the chemical composition FIG. 87. Expansion and contraction
of montmorillonite (Wyoming benton-
of the mon tmorilloni tes, the fusion ite) on firing, after Steger. 71 (A)
point of montmorillonite minerals Preheated to 1100; (B) without
exhibit wide variations. Iron-rich preheating.
members of the group may fuse below
1000C, whereas iron-poor members may not fuse until about 1200 to
1300C. The presence of alkalies and alkaline earths renders mont-
morillonites nonrefractory. The vitrification range of montmorillonite
minerals is frequently not large.
Rehydration. Hofmann and Endell 72 have shown that the property of
.. montmorillonite to regain interlayer water and expand is lost for Li
71 Steger, W., Die Langenanderungen von Kaolinen und einigen anderen Ton-
miner alien beim Brennen bis 1l0000, Ber. deut. keram. Ges., 23, 46-92, 157-174 (1942).
72 Hofmann, U., and J. Endell, Die Abhangigkeit des Kationaustausches und der
Quellung bei Montmorillonit von der Vorerhitzung, Ver. deut. Chemiker, Beiheft, 36,
10 (1939).
230 Clay Mineralogy
montmorillonite after heating to 105 to 1250, for H or Oa montmoril-
lonite after heating to 300 to 3900, and for Na montmorillonite after
heating to 390 to 4900. Hofmann and Endell carried their rehydration
tests over short periods of time, and somewhat different results might be
found for longer intervals of rehydration.
Grim and Bradley42 have shown that N a
montmorillonite, after being heated for 1 hr
at 6000, very slowly picked up interlayer
water, so that a small amount is shown at
the end of 268 days. Samples heated to
8000 showed no pickup of interlayer
water after the same length of time. The
same authors present data for a Oa mont-
morillonite showing that it also regained
interlayer water after heating to 6000 and
that the rehydration was slightly slower
than for the N a montmorillonite.
According to Grim and Bradley,42 sam-
H
ples heated to 6000 for 1 hr regained
about one-fourth of the original OR lattice
8 8N 8 8<::;t 8 8<..0 8l"-
!'() L() water in 268 days. As shown in Fig. 88,
cc the endothermic reaction corresponding to
FIG. 88. Differential thermal the removal of this rehydration water yields
curves of sodium montmoril-
a double peak, suggesting that the regained
lonite (bentonite), Clay Spur,
Wyoming, after Grim and water is of two kinds or forms in the lattice.
Bradley. 42 (A) ~ot heated; Some of the rehydration water is removed
(B) heated, 500C for 1 hr, at about the same temperature as the origi-
curve run after standing 13 nal hydroxyl water, and the remainder at a
days; (C) heated, 600C for 1
temperature about 1500 lower. Both
hr, curve run after standing 11
days; (D) heated, 600C for 1 kinds or 'forms of water are regained at
hr, curve run after standing 68 about the same rate. The rehydration of
days; (E) heated, 600C for 1 the OH lattice water does not afford any
hr, curve run after standing 146 clear-cut X-ray-diffraction interference, but
days; (F) heated, 600C for 1
hr, curve run after standing 268
in the light of the anhydrous configura-
days; (G) heated, 800C for 1 tion illustrated (Fig. 85), it seems reason-
hr, curve run after standing 76 able to assume that only a small part is
days; (H) heated, 800C for 1 recombined into the original configuration,
hr, curve run after standing 268 with an additional significant amount attain-
days.
ing some other bonding of lower nergy.
Experience in drying bentonite commercially has shown that, as a
practical matter, regaining of interlayer water is difficult if the last trace
of interlayer water is removed. So long as some water remains between
the layers, swelling is generally relatively easy ..
Dehydration 231
VERMICULITE
Dehydration and Phase Changes on Heating. A dehydration curve

published by Nutting 6 (Fig. 66) shows a large water loss below about
lOOoe, and a gradual loss from this temperature to about 850 0 e, where
dehydration is substantially complete. The curve shows flexures indi-
cating relatively more rapid dehydration in the vicinity of 400 0 and e
above 700 0 e. Walker 73 has also published dehydration curves for ver-
miculite. His data show the water to be lost mostly in three steps: below
lOOC, from 250 to about 400C, and from 600 to 850C. According
to Walker, the water lost in each of these steps expressed as percentage
of total water content of the mineral is about 48, 25, and 27 per cent,
respectively.
Barshad 74 and Walker 73 have published differential thermal curves for
the mineral (Fig. 75). Natural vermiculites show a large, dual, initial
0
endothermic reaction with a peak at about 150 to 200C, immediately
followed by a smaller peak at about 250 to 275C. This reaction results
from the loss of interlayer water.
Walker 73 explains the character of the initial endothermic peaks as
follows: The interlayer water in natural vermiculites consists of a hydra-
tion shell around the Mg ions made up of six water molecules and also
includes "unbound" water in the space between the hydrated Mg++.
The initial endothermic peak corresponds to the loss of the unbound
water, which is driven off on heating to about 100C. Accompanying
the loss of the unbound water is the collapse of the c dimension to 11.8 A,
which provides room for only a single molecular layer of water between
the silicate sheets. The collavse of the structure causes the hydration
envelope of the Mg++ to be spread out flat, so that two water molecules
become unbound and the Mg++ is surrounded by only four water mole-
cules. The two water molecules becoming unbound in the collapse of the
structure are also lost as part of the initial endothermic reaction. The
second low-temperature endothermic reaction corresponds to the loss of
the remaining four water molecules around the Mg++.
As shown in Fig. 75, the part of the differential curves corresponding
to the loss of interlayer water varies with the exchangeable cation carried
by the vermiculite. This variation is to be expected, since Barshad 74 has
shown that, when Ca++ and Mg++ are the exchangeable ions, there is a
double layer of water; when Ba+, Li+, and Na+ are the exchangeable ions,

of Crystal Lattice, Am. Mineral., 35, 225-238 (1()50).
232 Clay Mineralogy
there is a single layer of water; and when NRt, K+, Rb+, and Os+ are the
exchangeable ions, the mineral has no interlayer water. The sizes and
complexities of the initial peaks for vermiculite with any of these ions
vary in accordance with the hydration characteristics of the particular
ion. Thus K vermiculite shows only a very slight low-temperature
endothermic reaction, and Na vermiculite shows only a single initial
endothermic peak.
It appears from the data given by Nutting,6 Barshad,14 and Walker 73
that the OR lattice water from the silicate part of the structure is lost
gradually from about 500 to 8500 and hence does not tend to cause a
sharp corresponding thermal reaction. The endothermic reaction
between 700 and 8000 is probably due to the destruction of the silicate
structure, accompanied by the loss of the last hydroxyl water. The endo-
thermic peak at 6000 for the sample from Pennsylvania (Fig. 75) is
probably due to some chlorite in mixed layers with the vermiculite.
According to Barshad,74 the initial stages of dehydration are attended
by a contraction of the lattice, with the development of diffraction char-
acteristics like those of biotite. When vermiculite is heated abruptlY to
about 3000, the mineral exfoliates, with the production of a material of
low bulk density of considerable economic importance. The exfolia-
tion does not take place on heating to about 1500, from which fact it
may be concluded that all interlayer water must be driven off for it
to occur.
The high-temperature phases formed on the ignition of vermiculite have
not been studied in detail. Walker 73 has shown that enstatite is formed,
corresponding to the sharp exothermic reaction at about 8000 in samples
studied by him. The high-temperature phases could be expected to
show considerable variation, depending on lattice substitutions and
exchangeable ions, and probably would develop the general composition
and high-temperature phase relations' shown by montmorillonites (see
pages 225-227). Indeed, in the high-temperature parts of the thermal
curves shown by Barshad for vermiculite prepared with various exchange-
able cations (Fig. 75), the variations must mean a considerable variety in
high-temperature phases.
No information is available concerning the expansion-contraction of
vermiculite as it is heated to elevated temperatures. Also there are no
data on specific-gravity changes with temperature. The temperature of
fusion of the mineral and its vitrification range would be similar to those
of montmorillonite.
Rehydration. Walker 73 has shown that the unbound water driven off
at about lOOoO is regained almost instantly on cooling to room tempera-
ture, with an expansion of the lattice to its original dimensions. Bar-
shad 74 has shown that, after all the interlayer water is driven off at
Dehydration 233
temperatures up to 550C, the mineral will still rehydrate rapidly and
regain its original lattice dimensions and dehydration characteristics.
The ability to rehydrate disappears gradually above 550C, and
according to Walker it is completely lost at about 700C. Apparently
the loss of the rehydmtion ability accompanies the loss of hydroxyl water.
Since the silicate unit is octaphyllitic, a loss of structure would probably
accompany the loss of hydroxyls, and the ability to rehydrate after
heating to the higher temperatures would not be expected ..
ILLITE
Dehydration and Phase Changes on Heating. As currently defined,

the illite group includes clay minerals having either the muscovite type
20
...
C
<I>
(.)
...
<I> 10
a.
E
en
Muscovite /"
<J)
0
Biotite) ___
-
-l 5
...:
~
..... , ..
..... -'~
"
_--::~~~-'-- Pl7loqopite
~~-
0
200 400 600 800 1000
Temperature "c
FIG. 89. Dehydration curves of biotite, phlogopite, and muscovite, after Roy.75
Samples held at respective temperatures for 24 hr.
or the biotite type of structure, and it is necessary to consider the dehy-

dration properties of muscovite and biotite, as well as of the clay-mineral
micas. Dehydration curves for muscovite, biotite, and phlogopite are
given in Fig. 89, after Roy.75 The dehydration curve shows that mus-
covite has a slow, relatively gradual loss of water up to about 800C
without any restricted interval of large water loss. Phlogopite shows
virtually the same dehydration characteristics as muscovite. Biotite
loses water gradually up to about 400C; it shows very little loss between
0
about 400 and 850C and considerable loss between 850 and 1000C .
Walker 76 has also presented a dehydration curve for biotite, and, unlike
Roy's curve (Fig. 89), it shows an almost uniform rate of water loss up to
75 Roy, R., Decomposition and Resynthesis of the Micas, J. Am. Ceram. Soc., 32,
202-210 (1949).
76 Walker, G. W., The Decomposition of Biotite in Soils, Mineralog. Mag., 28,693-
703 (1940).
234 Clay Mineralogy
about 8500, where the dehydration is essentially complete. Kelley,
Jenny, and Brown 9 present dehydration curves for muscovite which show
an interval of relatively rapid dehydration between about 450 and
8500. The variation in dehydration data obtained by various investi-
gators for the micas undoubtedly results to a considerable extent from the
factor of time. Very great differences in water loss are obtained by
heating at a given temperature for varying intervals of time.
Roy75 has investigated the length of time required for dehydration of
muscovite to be completed at various temperatures, and his results are
given in Fig. 90. His data show that
about 10 hr is required for complete
dehydration at 3000, i.e., for re-
moval of all the water which can pos-
sibly be removed at this temperature
under ordinary pressure. At 450
and 6500 some water was still being
lost at the end of 100 hr, and at
Q) 6500 the loss of water in 1 -hr is
S:I only 20 per cent of the total lost in
100 hr. Obviously, tremendously
I 3 5 10 25 50 100 variable dehydration results can be
Heating Time in Hours (Log.) obtained for muscovite micas unless
FIG. 90. Dehydration curves for the time factor is carefully con-
muscovite, after Roy,75 showing rela-
tion of water loss to time of heating. sidered. The same conclusion prob-
ably holds also for the biotite micas.
There may well be some variation in dehydration properties within par-
ticular types of micas due to variations in composition and structure,
and some of the differences found in the literature may be due to this
cause.
Dehydration curves for illites published by Grim, Bray, and Bradley!!
(Fig. 69) show a considerable water loss below 1000, a gradual loss from
1000 to about 3500, a relatively abrupt, large loss from 350 to about
6000, and a gradual loss above 6000. These illites have a muscovite
type of crystallization. A sample of weathered biotite which probably
is somewhat representative of an illite with biotite attributes is shown by
Walker 76 to have a small loss of water below 1000, a gradual, very slight
loss from 100 to about 3000, a rather sharp but gradual loss from 300
to 6000, and a slow, gradual loss above 6000. Thus, the dehydration
properties of the illites with muscovite and biotite attributes for which
data are available are about the same. The illites differ from muscovite
and biotite by having a water loss below 1000 and a more abrupt loss of
much of the OR lattice water between about 300 and 6000. The low-
temperature water loss is to be expected, since the clay-mineral micas
", .
Dehydration 235
differ from the well-crystallized micas in having some interlayer water,
which is a consequence of fewer interlayer cations, less bond between
layers, less uniform orientation of successive layers, smaller particle size,
and/or a difference in composition within the silicate layer itself. The
difference in loss of hydroxyl water cannot be so completely explained
but is probably largely the result of the difference in particle size, as shown
in the following discussion of differential thermal curves.
Differential thermal curves (Fig. 73), after Grim and Bradley,77 for
coarse-grained muscovite (10 to 20 microns) show only a single dehydra-
tion reaction, beginning at about 800C with a peak at about 900C. In
the rapid heating of the differential thermal procedure there would be
little loss of hydroxyls below about 800C, and then, in accordance with
Roy's75 data, an abrupt loss of the OR water should occur. Grim and
Bradley show that a decrease in the particle size of the muscovite is
accompanied by a reduction in the temperature of the beginning of the
endothermic dehydration reaction and by an increase in the temperature
interval during which it takes place. Differential thermal analyses of
biotite yield a curve without any pronounced deflections up to lOOOC,
indicating an absence of any abrupt dehydration in this temperature
range under the conditions of the analyses (heating rate, lOoC/min).
Differential thermal curves for illite presented by Grim and Bradley77
show an initial endothermic reaction corresponding to the loss of inter-
layer water, a second endothermic reaction beginning at about 450 to
500C with a peak at about 550 to 650C, a third, slight endothermic
reaction between about 850 and 950C, and frequently an exothermic
reaction between about 900 and lOOOC (Fig. 73). The size and tem-
perature interval of the second endothermic peak, which corresponds
to the loss of OR lattice water, vary in different samples, and the final
part of the curve also shows considerable variation. Information is not
yet available on the precise variation of thermal characteristics of illites
with differences in their structural and composition attributes. Grim
and Bradley (Fig. 73G) have shown one octaphyllitic illite that yields a
differential thermal curve like those of heptaphyllitic illites. On the
other hand, another sample of glauconite showed a relatively smaller
initial endothermic reaction and a relatively higher temperature for the
reaction accompanying the loss of hydroxyl water than the heptaphyllitic
illites. Mackenzie, Walker, and Rart7 8 have recently described an illite
showing a double endothermic reaction in the hydroxyl dehydration
71 Grim, R. E., and W. F. Bradley, The Mica Clay Minerals, "X-ray Identification
and Structure of the Clay Minerals," Chap. V, pp. 138-172, Mineralogical Society of
78 Mackenzie, R. C., G. W. Walker, and R. Hart, Illite in Decomposed Granite at
I Ballater, Aberdeenshire, Mineralog. Mag., 28,704-713 (1949).

236 Clay Mineralogy
interval, with peaks at 550 to 600C and 713C, and for which there is as
yet no satisfactory explanation.
According to Roy75 muscovite shows no marked structural change on
heating to 940C, and the loss of the structure takes place between 940
and 980C. The expulsion of at least most of the OR lattice water is
accompanied by a slight expansion of the lattice in the c dimension and
not by complete destruction of the lattice. On heating, phlogopite
exhibits the same characteristics as muscovite, except that its structure is
lost at a temperature about 50C lower. The biotite structure, according
to Roy, persists to about 1100C, and the biotite structure, therefore, is
also not lost simultaneously with the expulsion of the OR lattice water.
This is surprising in view of the reasoning of Bradley and Grim 77 that
octaphyllite anhydrite-mica structures are unlikely (see page 223).
Grim and Bradley 66 have shown that the loss of OR water in the illite
studied by them is not accompanied by a loss of structure. It is accom-
panied by a slight structural change of the kind described formontmoril-
lonite (see page 223). According to these investigators, the structure
of these illites is not destroyed until at least 850C, and Grim and Bradley
would correlate its destruction with the third endothermic reaction in the
differential thermal curves. The illites studied by them had the musco-
vite type of crystallization. Maegdefrau and Rofmann 79 have also
reported the continuation of an illite-like structure beyond the loss of OR
water, and, in at least one of their samples, it is retained to 1000C.
Roy75 reports that muscovite on firing at 1000C develops I'-Al 2 0.
and/or spinel (the X-ray data were too poor for a positive determination);
at 1200C the presence of 'Y-A1 2 0 s is definitely established, and a-A1 2 0 3
appears; and above 1400C corundum (a-AI 20 3 ) and glass are the only
phases present. Zwetsch 80 has reported that muscovite ignites at 1050C
to 'Y-A1 20 3, a-AI 20., and leucite, and finally at 1300C the crystalline
phases are leucite and a-Al 20 3. On firing phlogopite, according to
Roy,15 spinel is formed at lOOOC, and this persists as the only crystalline
phase to 1550C. On firing biotite to about 1100C, the phases develop-
ing are a high-iron magnetic spinel, leucite, and mullite; at about 1300C
the phases are th~ high-iron spinel and leucite; and at 1500C only olivine
and glass are to be found. In a biotite studied by Grim and Bradley,
leu cite, 'Y-Fe203, and a spinel were identified after firing to 1200C.
On heating several illites, Grim and Bradley 66 found spinel appearing
in all of them at about 850C, and the spinel increased in amount and
particle size up to about 1200C. These investigators suggested that the
79 Maegdefrau, E., and U. Hofmann, Glimmerartige Mineralien als Tonsubstanzen,
Z. Krist., 98,31-59 (1937).

80 Zwetsch, A., Rontgenuntersuchungen in der Keramik, Ber. deut. keram. Ges., 14,
2-14 (1934).
Dehydration 237
octahedral sheet of the illite lattice, carrying the alumina, magnesia, and
iron, goes into the formation of the spinel, and that the alkalies and the
silica from the tetrahedral layers yield an amorphous glass. They found
that at 1300C the spinel had dissolved in the glass, and that mullite
became apparent at llOOC and persisted to at least 1400C. The
amounts of spinel and of mullite appeared to be complementary, and the
specimens richer in spinel were subject to fusion at a somewhat lower
temperature. These investigators found no clear evidence of a quartz
phase developing from their illites at elevated temperatures.
It is reasonable to expect that there would be considerable variation in
the high-temperature phases developed from illites of different structural
and composition attributes, and indeed the variations in the high-tem-
perature portions of the differential thermal curves show that this must
be true. Considerably more study of ignited illites is necessary before
general conclusions can be reached regarding high-temperature phase
transformations in these minerals.
Hyslop and McMurd0 52 have published expansion-contraction curves
for two illite samples (Fig. S4). Both samples show considerable expan-
sion (1.1 and O.S per cent, respectively) up to SOOC. Above SOOC one
of the samples showed an abrupt contraction, which was continuous up
to the limit of the experiments at llOOC. The other sample showed
slight contraction between SOOO and 950C, followed by a secondary
expansion from 950 to 1000C. The secondary expansion could well be
explained by a reaction producing gas from an impurity in the sample and
probably is not a characteristic of the mineral.
The fusion point of the illites varies between wide limits. Some
glauconites fuse below 1000C, whereas iron-poor varieties relatively
low in alkalies do not fuse until temperatures of the order of 1300C.
In general, illites with the higher fusion point have a moderately long
vitrification range (200C).
Rehydration. Grim and Bradley 42 have studied the rehydration prop-
erties of three illites. In one of them (Fig. 91), after heating to 600C
there was a rapid pickup of adsorbed water and hydroxyl lattice water,
so that a considerable part (25 per cent) of both types of water was
regained within a few hours. Additional amounts of adsorbed and
hydroxyl water were regained slowly for a period of many months. The
loss of rehydration hydroxyl water on heating again began at a lower
temperature (400C) than that of the original water (475). After
heating to SOOC there was also a gradual regaining of adsorbed water and
OH lattice water. The data suggest that, on heating to either 600 or
SOOC, there was an initial, relatively rapid rehydration followed by a
very slow and gradual rehydration. A second sample of illite showed
nearly the same rehydration characteristics as those just described, except
238 Clay Mineralogy
that the rehydration was slower and there was no lowering of the tem-
perature required to remove the water regained on rehydration. The
third sample studied by Grim and Bradley contained some chlorite mixed
with the illite. After heating to 6000, this sample showed a slow gradual
pickup of adsorbed water, and after 217
days a larger aIDount was adsorbed than
had been present in the original sample.
Also in this sample, some hydroxyl lattice
water was regained in material heated to
8000. Roughly the same amount of hy-
droxyl water waS taken up by samples
heated to both 500 and 8000. The total
amount of OH lnttice water taken up by
each of the illites in about 9 months was
close to one-third of the original amount.
The nature of the probable pairing of the
layers in the illites investigated is such that
data similar to that obtained for montmoril-
Ionites cannot be obtained (see page 223).
8 0(\) ~ 8 8 8W 8f'- 888
co
'" '<;j"
C
l() (j) 0
- However, it seemS likely that the structural
FIG. 91. Differential ther- changes are much the same for both minerals.
mal curves of illite, Fithian, Roy75 succeeded in re-forming muscovite
Illinois from Grim and Brad- and phlogopite from material decomposed at
ley: 42 (A) not heated (B) 10500 by heating the latter to 400 to 6500
heated, 600 e for 3 hr, curve
0
for at least 12 hr under pressures varying be-

run after standing 2 hr; (C)
heated, 600 e for 1 hr, curve tween 1,000 and 10,000 psi. No success was
0
run after standing 1 days; (D) attained in a similat treatment of decom-

heated, 600 e for 1 hr, curve
0
posed biotite.
run after standing 70days; (E)
heated, 800 0 e for 1 hr, curve CHLORITE
run after standing 13 days; (F)
Dehydration and Phase Changes on Heat-
heated, 800 0 e for 1 hr, curve
run after standing 70 days; ing. Nutting 6 has presented dehydration
(G) heated 800 0 e for 1 hr,curves for several chlorites (Fig. 66). All of
curve run after standing 147Nutting's curves for the chlorite minerals
days; (H) heated, 'soooe for
show very little water loss prior to about
1 hr, curve run after standing
275 days. 5000 and a sharp loss of a good deal of
water between about 500 and 5500. The
curves above about 6000 show considerable variation; some of them
show a gradual continuous loss of weight from about 600 to 8500,
where dehydration is essentially complete, whereas others show a rela-
tively slight loss between 600 and about 750 to 800C, and then a
relatively abrupt loss from 800 to essentially complete dehydration at
850 to 900C.
Dehydration 239
Orcel,22 Orcel and Renaud,81 Barshad,82 and the writer have obtained
differential thermal curves for a large number of chlorite samples (Fig.
75). The samples that have been studied show little or no low-tempera-
ture endothermic reaction, indicating an absence of interlayer or adsorbed
water. Further work may well reveal chlorites yielding such a reaction,
as it seems likely that very fine grained chlorites with some interlayer
water would occur in some clay materials. All the samples exhibited a
distinct, sharp endothermic reaction between 500 and about 700C.
Most of them showed a second endothermic reaction at about 800C, and
this peak was frequently followed immediately by a sharp exothermic
peak. Orcel concluded from his work that the hydroxyl water of the
chlorites is driven off in two stages, and when a second endothermic peak
is not shown, the two stages of dehydration occur within such a nar-
row temperature interval as to be unresolved by differential analyses.
Unpublished work by the writer and studies by Barshad 82 show that
extremely sensitive differential analyses of a series of chlorites (Fig. 75)
do tend to resolve the endothermic reaction into two parts. Nutting's6
dehydration curves show definitely that the chlorites he investigated lost
their OR water in two stages, but that the second stage may immediately
follow the first stage without an interval during which no water is lost, and
further that the second stage may take place gradually. Differential
curves of chlorites in which the second-stage loss of water was very
gradual would be expected to show an overlap of thermal reactions or
even the complete absence of a second endothermic reaction.
All the available dehydration and differential thermal data are for
well-crystallized chlorites, and the variations that would be found in
clay-mineral chlorites are a matter of surmise. Reasoning from analogies
with illites as compared to muscovite and biotite, it appears likely that
clay-mineral chlorites would show the same thermal reaction as the well-
crystallized chlorites, but perhaps at somewhat lower temperatures.
The smaller particle size of the chlorite clay minerals and possibly some
defects of orientation of layers would perhaps permit the loss of some of
the OR water of the brucite sheet at a relatively lower temperature.
Also, the clay-mineral chlorites could well have some interlayer water,
providing a small initial endothermic reaction as in many illites.
Brindley and Ali 83 have investigated the changes taking place in some
81 Orcel, J., and P. Renaud, Dehydration of Ferromagnesian Chlorites, Compt. rend.,

212, 918~920 (1941).
82 Barshad, 1., Vermiculite and Its Relation to Biotite as Revealed by Base-Exchange
Reactions, X-ray Analyses, Differential Thermal Curves, and Water Content, Am.
Mineral., 33, 655~678 (1948).
83 Brindley, G. V{., and S. Z. Ali, Thermal Transformations in Magnesian Chlorites,
Acta Crystal/og., 3, 25~30 (1950).

240 Clay Mineralogy
chlorites (penninite, clinochlore, and sheridanite) after heating to suc-
cessively higher temperatures. According to these investigators, in the
0
temperature range from 400 to 700 e, the intensities of the COOl) reflec-
tions progressively change; in particular, the (001) reflection becomes
several times more intense, and the (002) and (003) become several times
weaker. Also the lattice constants change, Co showing a small contraction
and ao and bo showing small expansions. These changes are attributed
to the gradual breakdown of the brucite layer, while the mica layer remains
unaffected. The dehydration of the brucite layer corresponds to the
first endothermic reaction of the differential thermal curves. At 700.e
about two-thirds of the hydroxyls of the brucite layer have been removed,
and Fourier syntheses of (OOl) intensities show that the magnesium atoms
move away from their original position in a plane midway between the
mica sheets to positions directly adjacent to the mica-unit surface. This
change precedes the dehydration of the mica-layer portion of the chlorite
structures. In some chlorites the loss of water by the mica layer is
abrupt, whereas in others it is gradual. In either case it is accompanied
by the total destruction of the chlorite structure, which in turn is imme-
diately followed by the crystallization of olivine. The endothermic
reaction frequently shown at 750 to 800 0 e is correlated with the dehy-
dration of the mica layer and the loss of its structure. The exothermic
reaction at about 800 to 900 0 e is correlated with the development of the
olivine.
Brindley and Ali 83 have shown that the newly formed olivine phase has
a high degree of preferential orientation relative to the structure of the
original chlorite, and they have considered in detail the probable shifts
in the atomic lattice accompanying the formation of the olivine. They
have shown that the temperature of lattice destruction and the tempera-
ture of the formation of olivine vary for different chlorites. They have
also shown that other phases in addition to olivine form when various
chlorites are heated, and that the temperature of formation of the addi-
tional phases is also variable. Thus for penninite, olivine first appears at
about 800 0 e, for clinochlore at 850C, and for sheridanite at 950 0 e.
According to these workers, penninite at 950C shows olivine and spinel,
and at 1180 0 e olivine, spinel, and enstatite; clinchlore at 850C shows
0
olivine and spinel; and sheridanite at 1100 e shows olivine, spinel, and
enstatite. Brindley and Ali 83 also present differential thermal curves for
these same chlorites carried to high temperatures, and they are unable to
correlate the thermal reactions satisfactorily in each case with the phase
developed as shown by X-ray-diffraetion analysis.
The explanation for the apparent lack of correlation between the high-
temperature phases revealed by X-ray analysis and the reactions shown
by differential thermal curves for chlorites and for some other clay
Dehydration 241
minerals is probably as follows: In some cases the new phase may develop
slowly and/or involve only a very slight structural shift, so that it would
not be accompanied by an intense or sudden energy change, and conse-
quently no thermal reaction would be expected on the differential thermal
curve. In other cases, the reaction may begin abruptly by a somewhat
crude nucleation of the new phase, which would involve a structural
change of considerable magnitude and a definite thermal reaction but
would not yield discernible diffraction evidence of the new phase. Fur-
ther heating or more time would be required to develop the new phase
to the point where distinct diffraction effects become obvious. In the
writer's experience, the thermal reaction for a new phase frequently
precedes its distinct appearance in X-ray reflections, and the X-ray
reflections further precede its appearance in optical measurements.
No data are available for the expansion-contraction properties of
chlorite on heating or for the changes in specific gravity of the mineral
on firing. Also no precise data are available for the fusion temperature
and vitrification range of these minerals. It appears likely that they
would be closely similar to the biotite type of mica in fusion and vitrifica-
tion characteristics.
Rehydration. No experimental data could be found on the possible
rehydration of the chlorite minerals. Reasoning from structural
analogies with the micas, it is likely that the chlorite clay minerals would
show a slow, perhaps relatively slight pickUp of hydroxyl water, at least
in the mica-layer part of the structure, after heating to temperatures of
the order of 700C.
SEPIOLITE-ATTAPULGITE-PALYGORSKITE
Dehydration and Phase Changes on Heating. The dehydration and

high-temperature characteristics of only a few samples of these minerals
have been studied, and there is no assurance that the results obtained
have wide application to the group as a whole.
A dehydration curve for a sepiolite given by Nutting 6 (Fig. 66) shows
a large water loss below about lOOC and a gradual continuous loss above
that temperature to complete dehydration at about 800C. There is a
relatively slight increase in water loss between about 200C and 400C
and above 675C. Migeon 12 has presented (Fig. 70) a continuous, i.e.,
" not equilibrium, weight-loss curve for sepiolite which is generally similar
to Nutting's curve in that it shows a large loss below 100C and a gradual
loss above that temperature, with flexures indicating relatively rapid loss
at about 275 to 325C, 550 to 600C, and slightly above 800C.
A differential thermal curve for a sepiolite is given in Fig. 76. It shows
a large initial endothermic peak beginning below 100C, followed by a
242 Clay Mineralogy
series of slight endothermic peaks between about 200 and about 700"C.
The exact temperature of these peaks is known to vary from sample to
sample. Between 700 and 800C there is generally a fairly sharp endo-
thermic reaction, followed immediately by a sharp exothermic reaction.
Caillere 84 has published a series of differential thermal curves for many
sepiolites, and they show the same general characteristics and variations
noted above.
Some variation is to be expected in the dehydration characteristics of
the sepiolites because of probable variations of composition. Also, in a
mineral such as sepiolite in which the dehydration is gradual rather than
abrupt, the corresponding thermal reactions would be slight, with the
consequence that small variations in experimental techniques would be
expected to give somewhat different results.
According to Migeon 12 and Longchambon 85 the water driven off below
about 350C is both hygroscopic and zeolitic water. The hygroscopic
water is considered to be surface-adsorbed water and the zeolitic water
to be present in the structural channels. The zeolitic water would be lost
at the higher temperatures, and many differential thermal curves of
sepiolites show a double endothermic reaction below about 350C, the
lower-temperature peak corresponding to the loss of hygroscopic water
and the higher-temperature peak to zeolitic water. There is no change
in the structure of the mineral on heating to 350C. Above this tem-
perature, the loss of water is accompanied by slight change in the lateral
dimensions of the fibers. The a dimension, according to Migeon and
Longchambon, is increased, and the b dimension is decreased. There is
no change in the length along the fiber. These investigators designate
this modification of the mineral as sepiolite II.. The modified form of
sepiolite persists to about 700C, when it passes into an amorphous phase,
according to Caillere. 84 Apparently the OR water is retained initially
in the modified phase, sepiolite II. As firing proceeds, it is lost gradually
and later abruptly at the time of the loss of structure. The high-tem-
perature endothermic reaction at 700 to 800C corresponds to this final
dehydration and the accompanying loss of structure.
The exothermic reaction just above 800C is probably due to the
development of a magnesium silicate phase. At 1000C Longchambon85
records the presence of enstatite and at 1300C a mixture of enstatite and
cristobalite.
Longchambon85 has shown, in a series of dilation curves for sepiolite, "
84 Caillere, S., Etude de quelques silicates magnesiens a facies asbestiforme ou

papyrace n'appartenant pas du groupe de l'antigorite, Bull. soc. fran~. mineral., 59,
353-386 (1936).
86 Longcham bon, H., Sur certaines caracteristiques de Ill, sepiolite d' Ampandan-
dravll, et la formule des sepiolites, Bull. aoe. fran!;. mineral., 60, 232-276 (1937).
Dehydration 243
that there are two intervals of contraction perpendicular to the fiber
axis, one interval of about 1 per cent at about 3500, and another of about
2 per cent taking place abruptly at about 8000. Parallel to the fiber
axis, there is a slow gradual contraction of about 0.5 per cent up to
8000. The contraction during this temperature interval becomes more
prominent as the temperature increases. At 8000, a further contrac-
tion, exceeding 2 per cent in the direction parallel to the fiber axis, takes
place suddenly.
Nutting 6 has presented a dehydration curve for a palygorskite (Fig. 66),
showing a sharp water loss below 1000, abrupt water losses at 150 to
2000 and 375 to 4250, and a gradual loss above 4250, with complete
dehydration at about 7000. A continuous weight-loss curve (Fig. 70),
given by Longchambon 13 shows similar results.
Differential curves for a palygorskite and an attapulgite (Figs. 76A
and C) show an initial endothermic reaction below about 2000 and
additional endothermic reactions at about 225 to 3500 and 400 to
5250, corresponding approximately to the intervals of abrupt water
losses shown by the dehydration curves. There is no definite-higher-
temperature endothermic peak, and apparently the small water loss above
about 5500 causes no thermal reaction adequate for recording.
Oaillere and Henin 86 have presented a series of differential thermal
curves for palygorskites, and they in general have similar characteristics
(Fig. 76D), except for larger initial endothermic reactions. The various
palygorskites that have been analyzed show additional variations in
detail, particularly in the high-temperature part of the curves, and this
is to be expected because of the probable differences in composition, with
resulting variations in high-temperature phases.
Bradley 87 has stated that the water molecules in the channel-like
interstices of the structure are lost below 1000, a temperature which
would correspond to the initial endothermic reaction. The water
coordinated about the magnesium ions is lost at a slightly higher tem-
perature, and this dehydration is probably correlated with the second
endothermic reaction. The loss of the hydroxyl water from the silicate
unit is probably correlated with the third endothermic reaction. Appar-
ently most of the hydroxyl water is lost at about 4000., with the remain-
der being lost slowly between 450 and about 7000.
The palygorskites-attapulgites differ in dehydration from the sepiolites
" in the loss of the hydroxyl water more abruptly and generally at a lower
temperature. Also in at least some sepiolites, there is little or no indica-
86 Caillere, S., and S. Henin, Application de l'analyse thermique differentielle ft,
l'etude des argiles des sols, Ann. agron., 17, 23-72 (1947).
87 Bradley, \V. F., The Structural Scheme of Attapulgite, Am. Mineral., 25, 405-
413 (1940).
244 Clay Mineralogy
tion of so-called zeolitic water, which in attapulgite is that coordinated
with the magnesiums.
According to Longchambon,88,89 the loss of water up to about 400C is
not accompanied by any structural changes. At about 400C, there is a
rapid contraction of the crystal lattice normal to the length of the fiber.
The modified structure persists to about 775C, when the structure is
destroyed and further shrinkage takes place.
According to Caillere and Henin,90 a comparison of the results of Long-
chambon and Bradley indicates that above 400C the (hkO) reflections
are replaced by halos. This is explained by dislocations in the octa-
hedral units of the amphibole chains, which lead to a closing of the
cavities or channels with the subsequent disappearance of the water-
adsorbing property. No satisfactory structural explanation has been
offered for the contraction of the lattice normal to the fiber axis described
by Longchambon.
Caillere and Henin 90 state that the high-temperature phases of some
attapulgites-palygorskites are probably enstatite, sillimanite, and cristo-
balite. The high-temperature parts of the differential thermal curves
show large variations, which indicate that there is significant difference
in the high-temperature phases formed when these minerals are heated.
This is to be expected, by analogy with the montmorillonite minerals,
because of probable large variations in their chemical composition and
possibly also in their structure. The precise nature of these high-
temperature phases and their relation to composition and structure
remain for future research.
According to Grim,91 attapulgite fuses at approximately the same tem-
perature as montmorillonite. Data are not avaiJable for the fusion point
of other members of this group, but probably they would be similar to
attapulgite.
Rehydration. Very little data are available regarding the rehydra-
tion of the sepiolite-attapulgite-palygorskite minerals. According to
Migeon,12 the water lost below 350C by sepiolite is easily regained,
whereas water lost at higher temperatures is not taken up again. Atta-
pulgite-palygorskite minerals probably have about the same rehydration
characteristics as sepiolite.
88 Longchambon, H., On the Characteristic Properties of Palygorskite, Com pt.
rend., 203, 672-674 (1936).
89 Longchambon, H., Recueil des communications du congres technique de l'indus-
trie de ceramique, pp. 75-141 (1938).
90 Caillere, S., and S. Henin, Palygorskite-Attapulgite, "X-ray Identification and
Structure of the Clay Minerals," Chap. IX, pp. 234-243, Mineralogical Society of
91 Grim, R. E., Relation of the Composition to the Properties of Clays, J. Am.
Ceram. Soc., 22, 141-151 (1939).
Dehydration 245
CLAY-MINERAL MIXTURES
Many clay materials are made up of mixtures of several different clay
minerals. The mixtures may be of two types: a combination of discrete
particles of the separate minerals or an interstratification of layers of the
I I I I I
100% Kaolin
1\
........ ",\ V
80% Koolin-
" 1'- -
20% Bentonite
-r--.
~ r-
........
1\
60% Koolin-
-r--. mnl,,"'" ........ V'-
f" v
i"- r--
-r-
40% KoLJ
60 % Bentonite
- \ - ~~
1f1lJ.l
l-
"'\ 1/"V V
-- 1'J:1MIOdl' ~ r- r- r- I-'V'V\
1\ --~ I
~
100% BeLl1
-~ 1-\ ,_, r- r- r- 1-'/
i'. J
V
200 400 600 800 1,000

Temperature C
FIG. 92. Differential thermal curves of discrete mixtures of kaolinite and mont-
morillonite, from Spiel et al. 27
separate minerals in which the thickness of the layers is of the order of

one or a few unit layers.
In mixtures of discrete particles the dehydration characteristics tend to
reflect those of the individual components. Numerous investigators 27 33
have published differential therm!1l curves of prepared mixtures of
various clay minerals. As shown in Fig. 92, such mixtures may show
the thermal effects of the separate clay minerals with intensities propor-
tionate to their relative abundance.
246 Clay Mineralogy
37
Grim has published a large series of differential thermal curves of
clay-mineral mixtures prepared in different ways. He has shown that
the thermal characteristics depend on the intimacy of mixing and on the
nature of the clay minerals themselves. In
mixtures prepared by combining dry pow-
ders without subsequent wetting, the dehy-
dration characteristics of the individual
components are maintained in proportions
roughly equal to their relative abundance.
If the mixing is done by stirring together
wet suspensions of the separate minerals,
and particularly if the mixture is allowed
to dry and then rewetted and remixed sev-
eral times, the dehydration characteristics
of the individual components may be modi-
fied significantly. As shown in Fig. 93, the
modifications may be such as to mask com-
pletely the presence of relatively minor com-
ponents. Amounts of the order of 25 per
cent could easily be hidden by such modifi-
o a 0 0 0 0 000 0
000 000 0 0 0 0 cations of dehydration characteristics.
- N
~ 7~ ill K ro m 0
C - In such mixtures of discrete particles the
FIG. 93. Differential thermal modification of dehydration properties takes
curves of prepared mixtures, place more readily and to a greater degree as
after GrimY (A) Illite, the crystallinity decreases. Thus, the effect
Grundy County, Illinois; (B)
illite, 0.504 g, and kaolinite
is more pronounced for poorly crystalline
K 0.252g; (C) illite, 0.371 g, kaolinite than for tpe well-crystallized
and" kaolinite K 0.371g; (D) mineral.
illite, 0.518 g, "and kaolinite In general, it seems that, as the intimacy
K" 0.259g; (E) illite, 0.371 g, of mixing increases, the temperature for the
and kaolinite K 3 , 0.371g; (F)
illite, 0.502 g, and kaolinite
loss of hydroxyl water is reduced and the
K", 0.251g; (G) illite, 0.363 g, loss of this water is more gradual, so that its
and kaolinite K 3 , 0.363g. thermal effect is spread out over a wider tem-
Kaolinite K, and K, rela- perature interval.
tively poorly crystalline; K3 With regard to interlayer mixtures, the
well crystallized.
published curves show that in some cases
the individual components are registered in the differential thermal
data and that in other cases they are not. Grim and Rowland 33 have
shown that a series of samples listed as beidellite provided differen-
tial thermal curves indicating that they are mixtures. Grim and
Bradley 77 have shown (Fig. 74) that the so-called sarospatite and bra-
vaisite from the type localities are mixtures of mica and montmorillonite
and that both components are indicated by the thermal curves. These
,', .
Dehydration 247
same investigators point out that other samples composed of mixtures of
illite and montmorillonite provide thermal data suggestive of a single
mineral (Fig. 74) and also that mixtures of illite and chlorite may give no
definite indication of more than one component.
In unpublished work by the author and his colleagues, it was found that
some bentonites from Texas composed of mixtures of kaolinite and a low-
iron montmorillonite in nearly equal amounts, yielded a single endo-
thermic peak at about 600C for the loss of hydroxyl water. In contrast
to this finding, some of the kaolins from the Georgia area contain 15 to
20 per cent of montmorillonite, and differential thermal curves reveal dis-
tinct thermal reactions for both the kaolinite and the montmorillonite.
The factors determining whether or not the individual components are
revealed by dehydration data are not completely known. The tendency
for a separate reaction decreases as the relative abundance of the com-
ponent decreases. Also a further factor probably is the nature of the
interlayering; e.g., a component present in layers of nearly unit-cell
thickness would not be revealed, whereas if the same mineral were present
in the same amount but concentrated in layers several unit cells thick, its
presence might be shown. Variations in composition may be a further
influencing factor. Thus, in mixtures of illites and montmorillonites, a
montmorillonite of one composition may be more readily concealed than
one with a different composition.
Differential thermal data are used widely for clay-mineral identifica-
tion, and in some cases quantitative estimates have been made solely on
such data. The foregoing discussion illustrates the fact that such
identifications must be made with caution. Differential thermal data
are extremely valuable for clay-mineral identifications, but they c::mnot
be used blindly, basing identifications solely on data derived from a few
pure clay minerals. Frequently, trustworthy identifications can be made
only on the basis of several types of analytical data, e.g., differential
thermal plus X-ray diffraction plus chemical data.
Substantially no specific information is available on the high-tempera-
ture phases formed in clay-mineral mixtures. It seems likely that the
initial high-temperature phases of interlayer mixtures, in some cases at
least, would be considerably different from those of the individual com-
ponents. This surmise is based on variations shown in the high-tem-
perature portion of the thermal curves for such mixtures.
Allaway, W. H., Differential Thermal Analyses of Clays Treated with Organic

Cations as an Aid in the Study of Soil Colloids, Soil Sci. Soc. Am. Proc., 13, 183-
188 (1948).
248 Clay Mineralogy
Augustnik, A. 1., P. Z. Tandura, and L. 1. Sverchkova, Mechanism of Reactions in
Talc upon Heating, J. Applied Chem. (USSR), 22, 1156-1159 (1949).
Belyankin, D. S., and V. P. Ivanova, On the Behavior of Montmorillonite at High
Temperatures, Compt. rend. acado sci. URSS, 18,279-282 (l938).
Dubois, P., Balance-analyseur thermique it enregistrement photographique, Bull.
soc. chim. France, 3, 1178-1181 (1936).
Eitel, W., H. O. Muller, and O. E. Rodzewski, Uebermikroscopische Untersuchungen
an Tonmineralien, Ber. deut. keram. Ges., 20, 165-180 (1939).
Endell, K., and M. Ardenne, Demonstration of Sinterin~ and Fusion of Ceramic Raw
Material in a Heated Electron Microscope, Kolloid,Z., 104,223-231 (1943).
Faust, G. T., Thermal Analysis of Quartz and Its Use in Calibration in Thermal
Analysis Studies, Am. Mineral., 33,337-345 (1948).
Freberg, A. K., Interpretation of the Second Exothermal Effect in the Thermal
Analysis of Refractory Clays, Compt. rend. acado sci. URSS, 28,90-92 (1940).
Garcia, F. G., Effect of Heat Treatment on the Hydration Capacity, of Montmoril-
lonite, Anales fis. y quim. B, 46, 1183-1210 (1949).
Gilard, P., Autoclave Treatment of Kaolin and Metakaolin, Verre .~ilicates ind., 13,
57-59 (1948).
Gor/:iuna'1-, N. 1., and E. A. Slrurygina, Tlrermal Cun-e'! of Minerals Encountered in
Soils and Rocks, Pedalogy (USSR), pp. 367-373 (H)50).
Grim, R. E., Method and Application of Differential 'thermal Analysis, Ann. New
York Acad. Sci., 63, 1031-1053 (1951).
Grim, R. E., and R. A. Rowland, Differential Thermal Analyses of Clay and Shales-
A Control and Prospecting Method, J. Am. Ceram. Soc., 27, 65-76 (1944).
Gruver, R. M., E. C. Henry, and H. Heystek, Suppres~ion of Thermal Reactions of
Kaolinite, Am. Mineral., 34, 769-873 (1949).
Hendricks, S. B., S. S. Goldich, and B. A. Nelson, A Portable Differential Thermal
Analysis Unit for Bauxite Exploration, Econ. Geol., 41, 64-76 (1946).
Honus, A., The Thermal Decomposition of Kaolin and the Compounds Which Form
from Kaolin at Higher Temperatures, Stavivo 387 (1936), Chern. Obzor, 12,
Abstracts 96 (1937).
Hyslop, J. F., Decomposition of Clay by Heat, Trans. Geram. Soc. (Engl.), 43, 49-51
(1944).
Jourdain, M. A., Recherches sur les constituents des argiles refractaires it l'aide de
I'analyse thermique. Ceramique 40, pp. 135-141 (1937).
Jung, U., Chemical and Physical Nature of the DehYdration and Rehydration of
Kaolin, Chem. Erde, 7, 113-120 (1932).
Keller, W. D., and J. F. Westcott, Differential Thermal Analyses of Some Missouri
Fireclays, J. Am. Ceram. Soc., 31, 100-105 (1948).
Kelley, W. P., and J. B. Page, Criteria for the Identifi(~ation of the Constituents of
Soil Colloids, Soil Sci. Soc. Am. Proc., 7, 175-181 (1942).
Kiefer, C., Thermal Dehydration of Phyllitic Minerals, Compt. rend., 229, 1021-1022
(1949).
Larsen, E. S., and G. Stieger, Dehydration and Opticltl Studies of Alunogen, Non-
tronite, and Griffithite, Am. J. Sci. ser. 15, 6, 1-19 (1928).
McCarter, W. S. W., K. A. Krieger, and H. Heinemann, Thermal Activation of Atta-
pulgus Clay, Ind. Eng. Chem., 42,529-533 (1950).
Menshutkin, B. M., History of Thermal Analysis, Ann. 4ecteur anal. phys.-chim. Inst.
chim. gen. (USSR), 8,373-406 (1936).
Mervielli, J., and A. Boureille, Identifications des argilEls ceramique par la thermo-
balance, Bull. soc. franr;. ceram., 7, 18-27 (1950).
Munier, P., Prediction of the Time-Temperature Curve for the Shrinkage of Ceramic
Products, Chaleur & Ind., 26,87-93, 120-128 (1945).
"
Dehydration 249
Norin, R., The Decomposition Products of Kaolinite, Geol. Foren. i. Stockholm Forh.,
66, 15-18 (1944).
Norton, F. H., and W. G. Lawrence, Notes on Anomalous Heat Adsorption of Kao-
linite, J. Am. Ceram. Soc., 26,388-:389 (1\)43).
Rios, E. G., and J. L. Vivaldi, Silicates of Laminar Structure, I, Hydration, Anales lis.
y quim. B, 46, 291-342 (1949).
Rode, E. Y., and T. V. Rode, Physicochemical Nature of Nontronite, Akad. Nauk
SSSR, 16, 273-278 (1947).
Siefert, A. C., Studies on the Hydration of Clays, Ph.D. thesis, Pennsylvania State
College (1942).
Stringham, B., and A. O. Taylor, Nontronite at Bingham, Utah, Am. Mineral.,
36, 1060-1066 (1950).
Tscheiswili, L., W. Bussem, and W. Weyl, Metakaolin, Ber. deut. keram. Ges., 20,
249-276 (1939).
Urbain, P., Introduction a I' etude petrographique et g60chemique des roches argile-
uses, Actualities sci. et ind., 60, (1935).
Zvanut, F. J., and L. J. Wood, X-ray Investigation of the Pyrochemical Changes in
Missouri Halloysite, J. Am. Ceram. Soc., 20, 251-255 (1937).
CHAPTER 10
Clay-mineral~Organic Reactions
INTRODUCTION
Investigations of clay materials in many different fields for many years

have produced evidence of some kind of a reaction between the fundamen-
tal components of clay and organic material. For example, early lit-
erature l on decolorizing oil with clays refers to the adsorption of some
components of the oil during the process and to changes in the nature of
the oil as a consequence of decolorization.
Students of soils in the field of agriculture have presented evidence for
some sort of inorganic-organic combination in many soils. For example,
Demolin and Barbier 2 showed a definite fixation of humic acid and protein
by clay, and Mattson 3 demonstrated a reduction in base-exchange capac-
ity by complexing clay with protein. Sedletsky 4 investigated the matter
in considerable detail and concluded that many soils contain clay-
mineral-organic complexes.
Further evidence for such reactions is provided by the color reactions
produced when certain clays and organic compounds are mixed (see pages
274-276). The development of color or a change in color requires some
kind of reaction between the clay and the organic co~pound added. The
fact that montmorillonite and halloysite experience changes in their optical
properties when immersed in certain oils (see pages 286-288) also indi-
cates a reaction between these clay minerals and some organic compounds.
Prior to the determination of the exact nature of the fundamental con-
stituents of clays and soils, there could be no precise concept of the reac-
tion between the components of clay and organic compounds and of the
nature of clay-organic complexes. The finding that clay materials are
composed of small crystalline particles and the determination of the
structure of the particles provided a new basis for studying clay-organic
1 Gilpin, J. E., and M. P. Cram, Fractionation of Crude Oil by Capillary Diffusion,
U.S. Geol. Survey Bull. 365 (1908).

2 Demolin, A., and G. Barbier, Conditions de formation et constitution du complexe
argilo-humique des sols, Compl. rend., 188, 654-656 (1929).
3 Mattson, S., The Laws of Soil Colloidal Behavior, VII, Proteins and Proteinated
Complexes, Soil Sci., 23, 41-72 (1932).
4 Sedletsky, I. D., "X-ray Study of Soils," Moscow (1939).
250
Clay-Mineral-Organic Reactions 251
reaction", and since the early 1930'" a large amount of work has been
done on this problem, and the character of clay-mineral-organic reactions
has now been fairly well established. Thus, in 1934, Smith" reacted
organic bases and their salts with montmorillonites and presented evi-
dence that the reaction was one of ion exchange. Gieseking and J enny 6
in 1936 showed that methylene blue replaced sorbed cations in certain
clays, and later Gieseking 7 and his colleagues 8 ,9 showed definitely that
organic ions enter into cation-exchange reactions with the clay minerals,
particularly with montmorillonite.
Hofmann and his colleagues lO in 1934 showed that the c-axis dimension
of montmorillonite varied following treatment with alcohol, acetone, and
ether. In 1939 Ensminger and Gieseking 8 shmved definitely that organic
ions could be sorbed on the basal plane surfaces of montmorillonite.
The early work of Hofmann et al. Io in 1934, and that of Myers l l and
Sharov 12 in 1937, suggested that organic compounds with polar active
groups could be adsorbed by the clay minerals. Bradley 13,14 in 1945, and
MacEwan 15 ,16 at about the same time quite independently, showed def-
initely that the nonionic organic molecules of polar character could be
adsorbed by the clay minerals. These investigators also considered the
nature of the bond between the nonionic organic molecules and the clay
minerals.
6 Smith, C. R., Base Exchange Reactions of Bentonites and Salts of Organic Bases,
J. Am. Chem. Soc., 56, 1561-1563 (1934).
6 Gieseking, J. E., and H. Jenny, Behavior of Polyvalent Cations in Base Exchange,
Soil Sci., 42, 273-280 (1936).
lite-Kontronite Type of Clay Minerals, Soil Sci., 47, 1-14 (1939).

8 Ensminger, L. E., and J. E. Gieseking, The Adsorption of Proteins by Montmoril-
lonite Clay, Soil Sci., 48, 467-473 (1939).

9 Ensminger, L. E., and J. E. Gieseking, Adsorption of Proteins by Montmorillonite
Clays and Its Effect on Base-Exchange Capacity, Soil Sci., 51, 125-1:"\2 (1941).
10 Hofmann, U., K. Ende!!, and D. Wilm, Rontgenographische und kolloidchemischc
Untersuchungen tiber Ton, Angew. Chem., 47, 539-547 (1934).
11 Myers, H. E., Physicochemical Reactions between Organic and Inorganic Soil
Colloids as Related to Aggregate Formation, Soil Sci., 44, 331-359 (1937).

12 Sharov, V. S., Action of Substances having a Hydrophobic and a Gel-forming
Effect on Clays, Grozhen8ki~ N eftyanik, 6, 10-15 (1936); Chem. Abstracts, 31, 2764
(1937).
13 Bradley, W. F., Molecular Associations between Montmorillonite and Some Poly-
functional Organic Liquids, J. Am. Chem. Soc., 67,975-981 (1945).

14 Bradley, W. F., Diagnostic Criteria for Clay Minerals, Am. Mineral., 30, 704-713
(1945).
15 MacEwan, D. M. C., Identification of the Montmorillonite Group of Minerals by
X-Rays, Nature, 154,577-578 (1944).

16 MacEwan, D. M. C., The Identification and Estimation of the Montmorillonite
Group of Clay Minerals, with Special Reference to Soil Clays, J. Soc. Chem. Ind. (Lon-
don), 65, 298-305 (1946).
252 Clay Mineralogy
REACTIONS WITH MONTMORILLONITE AND HALLOYSITE
The reactions of montmorillonite and halloysite with organic com-

pounds have been studied more thoroughly than those of other clay min-
erals, and they illustrate well the nature of the reactions and their effect
on certain properties of the clay minerals themselves; hence they will be
discussed first and in considerable detail.
Ionic Reactions. Many common organic cations contain basic amino
groups, and it is the reaction between these compounds and the clay min-
erals that has been studied in most detail. These cations are ammonium
ions in which one or more of the hy-
95
drogens have been substituted by or-
ganic groups.
Because of its low cation-exchange
capacity, such reactions with halloysite
<IJ t are relatively slight and unimportant.
' ~ 55 The organic ions are held around the
.s:::-
~ ~ 45 edges of the particles by the same bond-
wo
c'5 35 ing mechanism as that operating for
,g 2" montmorillonite.
0::: 25
U ~ The data in Fig. 94, after Ensminger
15 !:-----:':!-::-_ _-,'-:_ _.....J and Gieseking,9 show that the adsorp-
o 50 100 150 tion of gelatin may have a very
Grams of Gelatin per
100 Grams of Clay marked effect on the cation-exchange
FIG. 94. Effect of gelatin adsorp- capacity of montmorillonite and a mont-
tion on the cation-exchange capacity morillonitic soil. The gelatin-clay com-
of montmorillonite, after Ensminger plexes were prep~red by mixing alkaline
and Gieseking. 9 Montmorillonite suspensions of gelatin and clay together
( - - - - - ) ; colloid fraction of
Hartsburg soil (- - - - -).
and then acidifying to pH 2.6 with
acetic acid. These authors showed
that the protein did not reduce the cation-exchange capacity in an alkaline
medium. As the hydrogen-ion concentration is increased, however, the
cation-exchange capacity decreases. They conclude, from the fact that
an increase in hydrogen-ion concentration increases the basic properties
of the protein, that it is adsorbed as cations.
The data in Table 30, after Hendricks,17 show that the cation-exchange
capacities determined with relatively small organic molecules are nearly
the same as that found by displacement of the hydrogen ion with the
barium ion. Hendricks showed also that extremely weak bases like the
0- and m-nitroanilines do not form salts with montmorillonite, and that
17 Hendricks, S. B., Base-Exchange of the Clay Mineral Montmorillonite for Organic

Cations and Its Dependence upon Adsorption Due to van der Waals Force~, J. Phys.
Chern., 45, 65-81 (1941).
large molecules like the alkaloids brucine (strychnos group) and codeine
(morphine group) neutralize less of the hydrogen on the clay, even though
they are stronger bases than aniline and benzidine. Thus codeine, in
concentrations of 4 meq to 0.5 meq of acid clay in 25 cc of water, neutral-
ized 0.63 meq of hydrogen per gram of clay, and brucine similarly neu-
tralized 0.65 meq of hydrogen per gram of montmorillonite. The differ-
ence, namely, about 0.30 meq per gram, between these amounts and the
exchange capacity of the montmorillonite (see Table 30) represents the
amount of hydrogen so covered by the large alkaloid molecules as to be
unavailable for neutralizing other molecules (cover-up effect). Organic
ions larger than the area per exchange position, i.e., about 80 A2, would
be required to produce the cover-up effect. Hendricks has shown that
both groups of an organic diamine are effective in neutralizing an acid
clay.
TABLE 30. CATION-EXCHANGE CAPACITY OF A MONTMORII,LONITE DETERMINED WITH

CERTAIN ORGANIC CATIONS
(After Hendricks 17 )
Base Cation-exchange capacity, meg/100 g
Benzidine ...................... :. 91
p-Aminodimethylaniline. . . . . . . . . . 90
p-Phenylenediamine. . . . . . . . . . . . . . 86
a-Napthylamine. . . . . . . . . . . . . . . . . 85
2,7-Diaminofluorene...... .. .. .. . . 95
Piperidine. . . . . . . . . . . . . . . . . . . . . . . 90
Barium. . . . . . . . . . . . . . . . . . . . . . . . . 90-94
Hendricks 17 has indicated that the organic ions are held by Van der
Waals forces in addition to the coulombic force. In general, the larger
ions are more strongly adsorbed because of the greater Van der Waals
forces, and larger organic ions are difficult or impossible to replace with
smaller ions. Grim, Allaway, and Cuthbert 18 have shown that small
ions are adsorbed only up to the cation-exchange capacity, whereas larger
ions may be adsorbed in excess, and that these excess molecules are not
dissociated and probably are adsorbed by Van der Waals forces. Bradley
and Grim 19 believe that the coulombic forces are supplemented by
C-H' . 0 bonds (see pages 259-261) between the organic molecule and
the clay-mineral surface.
The adsorption of the organic ions on the basal plane surface of mont-
morillonites is shown conclusively by the shift in the c-axis spacing of the
montmorillonite units accompanying the adsorption (Fig. 95). Hen-
18 Grim, R. E., W. H. Allaway, and F. L. Cuthbert, Reaction of Different Clay
Minerals with Organic Cations, J. Am. Ceram. Soc., 30, 137-142 (1947).
19 Bradley, \V. F., and R. E. Grim, Colloid Properties of Layer Lattices, J. Phys.
Chem., 62, 1404-1413 (1947).
254 Clay Mineralogy
dricks 17 found that the organic ions he studied were oriented between the
montmorillonite layers so as to cause the minimum expansion of the
layers. Thus, flat-shaped ions were found to lie with their flat surfaces
parallel to the basal surfaces of the montmorillonite flake (Fig. 96).
Hendricks,17 Bradley and Grim,19 and Jordan 20 have shown that for
some organics mOre than one molecular layer may be adsorbed on the
basal surfaces of the montmorillonite planes. Jordan has determined the
c-axis spacing of montmorillonite from Wyoming bentonite treated with
aliphatic amines of increasing chain lengths. Calculating the total inter-
nal surface area per cation-exchange position as 165 A2, the c-axis spacings
Per Cent of Basal Plane Area

Covered by Amine E
0
.~
19 26.5 34 41.5 49 56.5 64 71.5 79 '8
I 9. 6 r---T--"''i-''-_'''--"''F--ic'---''--T''---''r---'-T'~ 10 ==
0> 17.6 8 ~
. " ~o
g_ 15.6 I 6 ~.~
OJ II
-(/)
Q)
c
13.6 ,"'"""Cr----:;---' 4 02
/ cO
0 /
0::: 11.6 / 2 :8 0:::
a / o
(/) 9.6 I o ;:;
0
(]J 0~~--7-~--~8~~'O~~'2~~'~4~'~6~'8
2 4 6 fr
Number of Carbon Atoms in Amine Chain OJ
FIG. 95. Effect of amine chain length on montmorillonite basal spacing, aftpr
Jordan.'o
are plotted in Fig. 95 as a function of amine chain length. The stepwise

separation of the flakes in increments of 4 A, which is about the Van der
Waals thickness of a methyl group, indicates that the chains lie flat along
the clay flake surfaces with the planes of the zigzag carbon chains parallel
to the plane of the mineral. On this basis the areas covered by the amine
chains have been computed from atomic dimensions and included in Fig.
95. When the organic chains occupy no more than half of the available
area per exchange position, the organic molecules on the top surface of one
layer fit into the gaps between those on the bottom surface of the layer
directly above it, so that the resulting separation of the two layers is
4 A, or the thickness of one hydrocarbon chain. When the chains occupy
more than 50 per cent of the surface area per exchange position, adjacent
laminae are unable to approach more closely than 8 A, which is the thick- ~
ness of two hydrocarbon chains.
The water-adsorhing properties of montmorillonite are gradually
reduced as the basal surfaces of the mineral are coated with the organic
20 Jordan, J. W., Organophilic Bentonites, I, Swelling in Organic Liquids, J. Phys.

& Colloid Chem., 53, 294-306 (1949).
Clay-MineraI-Organic Reactions 255
FIG. 96. Schematic drawing showing position of organic cations between mont-
morillonite Layers, after Hendricks."
256 Clay Mineralogy
ions. In general, the larger the organic ion, the greater the reduction in
the water-adsorbing capacity. Figure 97 shows differential curves for
untreated montmorillonite from Wyoming bentonite and for butyl-
dodecyl- and octadecylamine complexes of the same clay after exposure to
an atmosphere of 75 per cent relative humidity for 4 days. These curves
Differential Thermal Curves
~ ;-------~7n-t~re-a-t-ed~B-e-nt~o-m7re-----------r.-"-----,
~ - - - - Bulylammonium Bentonite , /
I
g ----
&. -----
Dodecylommonium Bentonite
Octadecylommonium Bentonite I
,/'
, I
I
//. /-~
/,
/' .-.\1.
\
'----
1*'/'
_...-' ,'/' Oxidotion
'/ I \\ of Orgonic
, / \\ Moffer
/J~/ \" ___1:

\ \ ,i
1\ "
\ \ /:
(/)
c
~
\ \ I :I
o
a
<lJ '\ I
a::
.~
E
<lJ
.c Adsorbed
,
\
\
I I
I I
I
'0 Moisture I I I ( 'Pyrolysis
\\I of Organic
"0
c Maffer
W
H! .
\1\/
v,.
o 10 200 300 400 500 600
Temperature, C
FIG. 97. Differential thermal curves of amine-treated and untreated montmorillonite,
from Jordan. 21
show a progressive decrease in the size of the low-temperature endother-

mic peak; this decrease results from the loss of adsorbed water with an
increase in the size of the aliphatic chain. Gieseking,7 Hendricks,17 and
Grim, Allaway, and Cuthbert1 8 have presented data on the relation of
water adsorption to amine adsorption which indicate the same conclusion.
Jordan 21 has shown that the viscosity of aqueous bentonite suspensions
21 Jordan, J. W., Alteration of the Properties of Bentonite by Reaction with Amines,
Mineralog. Mag., 28, 598-605 (1949).
Clay -Mineral-Organic Reactions 257
varies with the amount of treating amine, and that the maximum affect is
obtained in the region of complete cation exchange. As shown in Fig. 98,
maximum flocculation with maximum viscosity is reached in an octadec-
ylamine-montmorillonite when the amount of the added amine is nearly
equal to the cation-exchange capacity.
24
If)
~ 20
'0
0..
2 16
c
Q)
u 12
.~
i:' 8
'in
o
u
(J)
;;
O~-L__~__~~~~__~__~~L_-L__~
o 20 40 60 80 100 120 140 160 180
Milliequivalents of Amine per 100 Grams of Cloy
FIG. 98. Effect of octadecylamine acetate on the viscosity of montmorillonite
(vVyoming bentonite), after Jordan. 2l Concentration of suspension, 1.9 per cent.
Adsorption of Polar Molecules. Debye 22 has shown that many non-

ionic and organic molecules are dipoles as a result of the lack of symmetry
of electron distribution within individual molecules. Such molecules act
as if they carried both centers of positive charges and centers of negative
charges. Since the clay-mineral structures are also polar, when they are
in contact with liquids, solutions, or suspensions containing other polar
substances, the negative centers on the clay minerals attract the positive
centers on polar substances in surrounding liquids, and positive centers
on clay minerals attract negative centers on surrounding polar particles.
The foregoing notion of dipole character is valuable as a general concept.
However, it should be borne in mind that more than one polar group can
enter many molecules, and some quite stable associations occur with
molecules which have no dipoles.
The water molecule is polar, and it is by far the most important polar
compound sorbed by the clay minerals. Bradley 13 in 1945 showed that
glycols, polyglycols, and polyglycol ethers were adsorbed on the basal
plane surfaces of montmorillonite, displacing water from these surfaces.
About contemporaneously and quite independently, MacEwan 15 16
showed that a large variety of polar organic molecules can be adsorbed
on the basal surfaces of both montmorillonite and halloysite (Table 31).
The c-axis spacing of both minerals varies with the organic molecule
adsorbed. The inorganic cations present on the surfaces of the clay min-
22 Debye, P., "Polar Molecules," Reinhold, New York (1929).
258 Clay Mineralogy
TABLE 31. DATA FOR MONTMORILLONITE- AND HALLOYSITE-ORGANIC COMPLEXES
(After MacEwan 23)
Montmorillonite HaUoysite
Complex-forming substances Meth- Base Meth-

od of No. No.
satur- od of Do obs. of
Do obs. Do calc. of ~ calc.
forma-
layers ation forma- layers
tion of clay tion
--- --- --- ----- ------ --- --
Normal monohydric alcohols
MethanoL ..... a 7.4 6.9 2 NH,+ a 3.4
b 7.2 Ca++
Ethanol. ....... a 7.9 8.1 2 NH.+ a 2.8d
I-Propanol .... b 4.5 4.5 1 NH,+ a 0.3 0
I-Butanol. .... b 4.6 4.5 1 NH.+ a 0.3 0
1-Pentanol. .... b 4.6 4.5 1 NH.+
I-Hexanal ...... b 4. 1 4.5 1 NH.+
1-Heptanol. ..... b 4. 1 4.5 1 NH,+
1-Hexadecanol. .... K 4. 1d 4.5 1 NH,+
Other monohydric alcohols
4-HeptanoL. b 3.1 4.5 NH,+
2-0ctanol. b 4.1d 4.5 Ca++
2-Ethyl-1-bu tanol. b 4.2 4.5 NH,+
2-Methyl-2-butanol. b 5.4 5.4 NH.+
Cyclohexanol. . . . . . . . . . . . . b 5.6 5.5 NH.+
Polyhydric alcohols
Ethylene glycol. .......... a 1.6 1.3 2 NH.+, a, e 3.1
Ca++
1,3-Propanediol. .......... a 8.6 8.6 2 NH,+, a, e 4.4 4.5
Ca++
1,3-Butanediol. ........... i. . . . a 4 Od 4.5
GlyceroL ... a 8.3 8.3 2 ! NH,+', a, e 3 8 3.9
Alcohol ethers and ethers
1,4-Dinxane .. a, b 5.6 5.9 Ca++ a d
Methyl cellosol ve ..... a 3.5d
Ethyl cellosolve ... a 3.6d
Butyl cellosolve. a 3.1d
Diethylene glycol ... a 3.6
Triethylene glycoL a 3.5
Hydrocarbons
n-Hexane ........ b 1-2d 0 Ca++
n-Heptane ........ .. b 0.6d 0 NH,'"
Benzene. ..... , ..... b, c 4 6d 2 NH.+
Naphthalene. b 3.5d 1 NH.+
Tetrahydronaphthalene .... b 4.6d 2 NH,+
Decahydronaphthalene .... b 0.6 0 NH.+
Other compounds
Chloroethanol. ........... b 8.1 1.8 2 NH.+ a, e 3.6
Ethanediamine .. , b 4.3 4.4 1 Ca++ a 4.5 4.4
1.2-Propanediatnine ..... a d
Diethylamine ........ ... a 0.3 0
Acetone ..... ...... a, b 8.2 8.1 2 Ca++ a,1 3.9 1
Acetaldehyde .... a ,3.6 1
Acetonitrile . ..... a, b 10.2 10.6 3 Ca++ a 3.4 1
Nitromethane . .... b 10.4 10.5 3 Ca++ a, a, h 3.3 1
Nitrobenzene ...... .. , .. b .5.7 2 NH,+ d
A = "clear space" between mineral sheets measured from the surface of the 0 or OH layers,
in kX
a; = treated with large excess of cold liquid
b = boiled down to half volume with excess of liquid
c = dehydrated at 100C over P,O, in vacuum, then covered with liquid
e = covered with excess of liquid, and heated to 100C for 1 to 2 hr
I. g, h, i = like a, starting from comple" with ethanediamine (f), methyl cellosolve (g), methanol (h).
or chloroethanol (i)
K = treated with the molten alcohol at about 20DoC
d = diffuse or double line.
Clay -Mineral-Organic Reactions 259
erals are not necessarily displaced by the adsorption of the organic
molecules.
On the basis of one-dimensional Fourier syntheses of diffraction data,
Bradley 13 concluded that the polar molecules are held to the clay surface
through a C-H . 0 (clay-mineral surface) bond. MacEwan 23 has
come to the same conclusion regarding the nature of the bond. In gen-
eral, the sorption energy for the organic dipolar molecules is greater than
0
for water molecules, for montmorillonite dehydrated at 400 to 500C
will not rehydrate in water but will take up glycol and expand. The
organic dipolar molecules are oriented between the basal surfaces of the
clay minerals in positions as flat as possible.
Bradley showed that some organic dipoles were adsorbed in multiple
molecular layers by montmorillonite. MacEwan has confirmed this find-
ing for montmorillonite and has indicated that a single molecular layer
is the general result for all organic polar molecules with halloysite. In
the case of montmorillonite, three molecular layers is the greatest number
that has been observed.
MacEwan explains the multiple-layer adsorption by montmorillonite
as compared to the single-layer adsorption by halloysite as follows: Each
of the surfaces of the montmorillonite sheets is formed by a layer of
oxygen ions, and thus they behave rather like sheets with a uniform
diffuse negative charge. Each sheet of montmorillonite will tend to col-
lect a layer of dipoles on both its surfaces, so that between such sheets
there will be two layers. In the case of nonpolar molecules or groups,
only the Van der Waals force of attraction between them and the mont-
morillonite sheets need be considered. Since this force is nondirectional,
the question of orientation does not arise and only a single molecular
layer could be held in common by two neighboring montmorillonite
sheets. Thus markedly polar molecules would form a two-layer struc-
ture, and less polar molecules a one-layer structure. Very strongly polar
liquids may form even more than two layers (e.g., acetonitrile and nitro-
methane). According to MacEwan,24 the number of layers in montmoril-
lonites in contact with excess of the liquid appears to increase with the
function M/[P] where M is the molecular dipole moment and [P] is
molecular size (parachor). Some polar groups appear to be necessary for
adsorption, for saturated hydrocarbons do not form complexes (see Table
31). Presumably, therefore, as the hydrocarbon chain of a monohydric
J alcohol is made longer and longer, the adsorption would gradually dimin-
23 MacEwan, D. M. C., Complexes of Clays with Organic Compounds, I, Trans.
Faraday Soc., 44, 349-367 (1948).

24 MacEwan, D. lVr. C., The Montmorillonite Minerals, "X-ray Identification and
Structures of the Clay Minerals," Chap. IV, pp. 86-137, Mineralogical Society of
260 Clay Mineralogy
ish. The necessity of polar groups for adsorption is no doubt bound up
with the fact that the montmorillonite sheets are charged and are held
together by positive ions between the sheets. The energy of formation
of a purely Van der Waals adsorption complex is presumably insufficient
to cause the breakup of the charged-sheet-ion-
charged-sheet complex. But if a few polar groups
are present, these will tend to congregate about
the charged spots on the sheets and about the
positive ions. The result will be an effective
increase of the radius of the ions and a conse-
quent diminution in electrical stress. The re-
sulting structure may be more stable than the
dehydrated montmorillonite structure, in which
the base cations occur between the structural
sheets.
With halloysite only highly polar molecules
are adsorbed between the unit layers. In this
mineral the adjacent sheets of successive layers
carry positive and negative charges, respectively,
and there is no tendency for two layers of polar
molecules to form. Figure 99, after MacEwan,23
shows schematically the process of complex
formation.
On the basis of observed distances between
FIG. 99. The process of
successive clay-mineral sheets in the complexes
complex formation in
montmorillonite and hal- from c-axis measurements and computed values
loysite, after MacEwan. 23 based on Van der Waals dimensions (Table 31),
(a) Polar molecules with MacEwan concluded that each layer of dipoles
montmorillonite; (b) occupies its full" clearance space" in the c-axis
mainly nonpolar groups
direction, i.e., the space necessary to allow its
with some polar groups
with montmorillonite; molecules to move freely over a plane parallel to
(c) polar molecules with the surface of the structural sheets. This sug-
halloysite. The specific gests that the adsorbed layers are in the state
role of hydrogen-bond of a two-dimensional liquid, firmly adhering to
formation, which is
the clay surface but able to move over it in a ran-
important, has been
neglected. dom manner. This conclusion is supported by
the extremely labile nature of the complexes,
which is such that one complex is changed to another by washing
with excess of another miscible liquid, and by the absence of a simple
integral number of molecules per unit cell, which would be expected for a
crystalline arrangement of the adsorbed molecules on the clay surface.
MacEwan has indicated that, when glycol complexes are made in the
presence of water, less glycol is needed to form the complex. He stated
that regardless of the water present the c-axis spacing is the same, and
concluded that the glycol molecules must set the pattern into which the
water molecules fit. Mackenzie 25 has found that the c-axis spacing of a
montmorillonite-ethylene glycol complex can be maintained by adding
water if the glycol content is low. The additional water is proportional
to the glycol deficiency in the molecular ratio of about 1: 6. Recently
Glaeser26 has shown that this conclusion does not hold for all complexes.
For example, in the case of acetone-montmorillonite complexes, when the
complex is formed in the absence of water, the distance between the mont-
morillonite layers is 3.7 to 3.9 A; at 5 per cent relative humidity, 8.0 to
8.2 A; and at 100 per cent relative humidity, 12.5 to 15.1 A. Apparently
a water-acetone complex develops between the montmorillonite layers.
The data given in Table 31, therefore, may be subject to variations in
some cases depending on the water present during the preparation of the
complexes. Glaeser 27 has shown also that the character of the adsorbed
cation is only of slight importance in the formation of dipole complexes
if they are formed in the absence of water or at low relative humidities.
However, if they are formed at high relative humidities, the thickness of
the adsorbed layer may vary greatly according to the cation present.
Thus for acetone-montmorillonite com-
plexes formed at 100 per cent relative
humidity, the distance between the
0
/8
v
units is about 12.5 A, if Na+ is the S /6 1\
adsorbed ion, and 15.1 A, if it is Ca++.
The explanation probably resides in u
the hydration tendencies of the cation. Cf)
0'
c
8. /4
\ V
---.
Comparison of Ionic and Polar Com- g /2
plexes. Figures 95, 100, and 101, ~
from Jordan,20 Bradley,13 and Mac-
/0
Ewan,23 after MacEwan,24 show the 2 4 6 8 fO
variation in the c-axis spacing of Number of Carbon Atoms
montmorillonite with the number of FIG. 100. Variation of basal spac-
carbon atoms in the chain, in com- ing of montmorillonite complexes
with dihydric alcohols, from data by
plexes with the straight-chain amines, Bradley.1s
alcohols, glycols, and nitriles. In each
case the spacing always corresponds to the presence of a whole number
of molecular sheets between the montmorillonite layers. The spacing
25 Mackenzie, R. C., Complexes of Clays with Organic Compounds, II, Trans.
Faraday Soc., 47, 368-375 (1948).

26 Glaeser, R., Sur Ie mecanisme de formation des complexes montmorillonite-ace-
tone, Compt. rend. 226, 935-937 (1948).

27 Glaeser, R., On the Mechanism of Formation of Montmorillonite-Acetone Com-
plexes, Mi;,eralog, Soc. Gr. Britain Clay Mineral Bull., 3, pp. 88-90 (1949).
262 Clay Mineralogy
alters in a homologous series in jumps of one sheet at a time. The
figures illustrate the difference between ionic and polar types of com-
plexes. In the ionic type the spacing increases with increasing number
of carbon atoms, whereas in the polar type the spacing decreases with
increasing number of carbon atoms. The explanation is that, in the
---- Three
0 Layers
1\
8
\
i<i V ---- Two
\ 1\
Q)
::i5 Layers
o
I- 6
Q)
Q)
(/) \ \
0
I
4 :> -._, ,__ - ~.- One
.'= ,, \
\ Layer
<J ~\
2
i\
\
o
o 2
Number of Carbon A toms
\
---
4
--- -
6 16
.. - No
Complex
o Montmorillonite Complexes with Monohydric Alcohol

o Montmorillonite Complexes with Nitriles
Halloysite Complex with Monohydri'c Alcohol
FIG. 101. Variation of basal spacing with chain length for montmorillonite com-
plexes with monohydric alcohols 0, with nitriles 0, for halloysite complexes with
monohydric alcohols e, from MacEwan. 24 '
case of the ionic type, enough organic molecules must be present to

neutralize the charge on the montmorillonite, and if the molecules are
large, this will require more than two layers. In polar adsorption the
number of layers is determined by the energy of adsorption, 'which
decreases as the nonpolar part of the molecule becomes larger.
REACTIONS WITH CLAY MINERALS OTHER THAN MONTMORILLONITE

AND HALLOYSITE
Kaolinite. Grim, Allaway, and Cuthbert 18 have shown that kaolinite

enters into ion-exchange reactions with organic ions and that the water-
sorption properties are reduced by the replacement of inorganic cations
with organic cations. Because of the low cation-exchange capacity of
these minerals, the quantity of organic cations sorbed is small. Bradl ey 13
has pointed out that the adsorption of polar molecules would take place
on the surfaces bf kaolinite. The organic molecules do not penetrate
between the unit layers of kaolinite as they do in halloysite and mont-
morillonite and, therefore, would be retained only around the edges and
exterior surfaces of the particles.
Illite. It has been shown 18 that illite enters ion-exchange reactions
with organic cations and that the water-sorption properties are reduced
after reaction with such ions. The quantity of organic ions sorbed is
somewhat greater than for kaolinite because of the higher cation-exchange
capacity. The sorption of polar molecules undoubtedly takes place in a
manner similar to that previously described for kaolinite.
Chlorite. No data are available on the reaction of organic molecules
with the chloritic clay minerals, but it seems certain that they would
react similarly to illite and kaolinite. Because of the increase in ion-
exchange capacity and surface area with decreasing particle size, the
sorption of both ionic and polar organic molecules would increase
greatly with decreasing particle size for chlorite as well as for kaolinite
and illite.
Vermiculite. Bradley14 pointed out that the ideal structure of vermic-
ulite suggests that the interlayer water could be displaced by organic
molecules. Walker28 has shown that complexes are formed with organic
ions and that the amount of displacement of the vermiculite basal reflec-
tion depends on the concentration as well as the nature of the organic ion.
Barshad 29 has shown that Mg, Ca, Ea, H, Li, and N a vermiculite take up
a single layer of glycol molecules, with a c-axis spacing of 14.3 A, when
immersed in hot anhydrous glycol. Under similar conditions montmoril-
lonite takes up two molecular layers of glycol, with a c-axis spacing of
17.7 A. Also under similar conditions NH 4 , K Rb, and Cs vermiculite
remain contracted, taking up no glycol molecules between unit layers.
Walker 28 has shown that ethylene glycol, methane, cyclohexanol, acetone,
and acetonitrile show limited penetration of Mg vermiculite, or none at
all, but that penetration is easy if Li vermiculites are used. Probably
because of the generally larger size of the vermiculite flakes and the higher
perfection of stacking orientation, penetration of organic molecules is
more difficult in vermiculites than in montmorillonites.
Walker28 has emphasized that the c-axis spacing of glycol-vermiculite
(14.28 A) is very close to that of natural magnesium-saturated vermic-
ulite (14.36 A), and therefore the penetration of the glycol molecule is
difficult to observe. Walker also points out that in the montmorillonite
complex the glycol molecule orients itself with a height of 3.9 A, whereas
in vermiculite the space available is 5 A; hence it is concluded that in
Walker, G. F., Vermiculite-Organic Complexes, Nature, 166, 695-697 (1950).
28
Barshad, I., The Effect of the Interlayer Cations on the Expansion of the Mica
29
Type of Crystal Lattice, Am. Mineral., 35, 225-238 (1950).
264 Clay Mineralogy
vermiculite the organic molecules either (1) attach themselves to the clay-
mineral surfaces by a weaker bonding force than that in a montmorillonite
complex or (2) orient themselves differently with respect to the adjacent
silicate surfaces. In vermiculite, single layers of glycol are interleaved
with the aluminosilicate layers, whereas in montmorillonite there are
multilayers. Walker states that vermiculite exerts a catalytic effect on
the decomposition of glycerol. Vermiculite boiled a few minutes in
glycerol acquires a black metallic coating on interlamellar as well as
boundary surfaces. The basal spacing of such a vermiculite complex is
14.06 A.
Attapulgite. The fibrous clay minerals belonging to the sepiolite-
attapulgite-palygorskite group should adsorb organic ions and polar mol-
ecules around their edges like kaolinite, illite, and chlorite. Bradley 14
pointed out that the dimensions of the cross section of the channels in
attapulgite are about 3.7 by 6.0 A, which is the approximate cross-
sectional dimension for single strings of ethylene glycol molecules. The
presence of organic molecules in the channels would be difficult to detect
by X-ray diffraction, because only minor intensity variations in the
reflections would be expected. However, careful determinations of the
change in the optical properties of attapulgite on treatment with glycol
suggest that a few of the organic molecules have penetrated the channels.
Nederbragt 30 has presented evidence, based on the preferential adsorption
of some paraffins, indicating that some aliphatic chains are adsorbed in
the open channels or in the gutters around the edges of particles and per-
haps also in the interior channels.
RESISTANCE OF ADSORBED ORGANIC MOLECULES TO BIOLOGICAL

DECOMPOSITION
The work of several investigators suggests that the resistance of some

organic materials to biological decomposition may be increased when they
are adsorbed by the clay minerals. Thus, W aksman 31 reached this con-
clusion because it is easier to build up the humus content in a clay-rich
soil than in a clay-poor soil. Mattson 3 believed that adsorbed proteins
might be more resistant to microbial decomposition than free proteins.
Ensminger and Gieseking 32 and Erickson 33 have investigated the matter
30 Nederbragt, G. W., and J. J. deJong, The Separation of Long-Chain and Compact
Molecules by Adsorption, Rec. trav. chim., 66,831-834 (1946).

31 Waksman, S. A., "Humus; Origin, Chemical Composition and Importance in
Nature," Williams & Wilkins, Baltimore (1936).

32 Ensminger, L. E., and J. E. Gieseking, Resistance of Clay-Adsorbed Proteins to
Protolytic Hydrolysis, Soil Sci., 63,205-209 (1942).

33 Erickson, A. E., Ph.D. thesis, University of Illinois (1948). Reference from J. E.
Gieseking, The Clay Minerals of Soils, Advances in Agron., 1, 598-605 (1949).
of the decay of adsorbed organics and have shown that the decay char-
acteristics are often quite different in free as compared to clay-mineral-
adsorbed organic material. Ensminger and Gieseking observed that
protein-montmorillonite complexes prepared by ion exchange in which
relatively large amounts of protein were adsorbed as cations, showed only
a slight tendency to putrefy. These same authors showed that the
adsorption of albumin and hemoglobin by ion exchange in montmoril-
lonite interfered with the enzymatic hydrolysis of these proteins in both
acid pepsin suspensions and alkaline pancreatin suspensions. In kaolin-
ite complexes there was no significant effect on the hydrolysis of albumin
and hemoglobin. Apparently sorption between basal units, as in the
case of montmorillonite, rather than around particle edges, as in kaolinite,
is necessary to retard chemical change. Ensminger and Gieseking 32 con-
clude that the hydrolysis is reduced because the protein molecules, when
adsorbed by the clay, are oriented in such a way that the active groups
are inaccessible to the enzyme. Another possible explanation is that the
enzyme is partly adsorbed and thereby rendered partly inactive.
Erickson 33 has shown that the stability of amine-montmorillonite com-
plexes, or their tendency to decomposition, depends on the enzyme and on
the amine used in the complex. Actually in some cases, e.g., in tyrosine-
montmorillonite complexes, after an initial lag 111 oxidation, the clay
seems to catalyze the enzymatic reaction.
ORGANOPHILIC CLAY-MINERAL COMPLEXES
Hauser 34 appears to have discovered that certain organic compounds

of montmorillonite have the property of swelling and dispersing in organic
liquids. Jordan 2o ,21 has investigated this matter in great detail and the
following statements are based on his work.
The gel volume of a series of normal primary aliphatic amine-mont-
morillonite complexes plotted as a function of the number of carbon atoms
are given in Fig. 102. The gel volumes were determined in nitrobenzene,
benzene, and isoamyl alcohol. The data show that the organophilic
properties are negligible until an amine with a chain of 10 carbon atoms
is reached, and that 12 carbons are required for maximum swelling. An
amine with a 12-carbon chain covers slightly more than half of the avail-
able basal surface of the montmorillonite units per exchange position, and
initiates the interlayer distance of about 8 A, which is the thickness of
two molecular layers, (Fig. 95). Apparently this condition is necessary
for maximum swelling.
The swelling behavior of dodecylamine complexes in a wide variety of
liquids is given in Table 32. These data suggest that swelling is low in
34 Hauser, E. A. Reference in J. W. Jordan, Mineralog. Mag., 28, ,598-605 (1949).
266 Clay Mineralogy
liquids of a nonpolar nature, such as the aliphatic and aromatic hydro-
carbons, and that generally the gel volume increases with the dielectric
constant of the liquid, although the correlation is not perfect. It appears
that the most effective liquids are those which combine highly polar with
__ IOO~--~--~--~--~----~--~--~--~--~
2
0. o Nitrobenzene
80 o Benzene
U) C:. Isoomyl Alcohol
d
N 60
-=
~ 40
Qi
E
:::l
20
~
a:; 0
<.::> 4
20 6 B 10 12 14 16 18
Number of Carbon Atoms in Amine Choin
FIG. 102. Effect of amine chain length on swelling of organic-ammonium-montmoril-
lonite in organic liquids, after Jordan. 20
O~ __ ~ __ ~ ____ ~ __ ~ __ ~ __ ~~ __ ~ __ ~ ____ L_~
o ~ ~ ~ ~ W w ro 00 ~ 00
Per Cent Alcohol in Alcohol- Toluene Mixture
FIG. 103. Swelling of octadeeylammonium-montmorillonite in binary mixtures, after
Jordan. 20
highly Qrganophlli~ ~hara~teristi~s, good examples being nitrobenzene

and benzonitrile. It is thought that, in the case of an incompletely clad
amine-montmorillonite complex, adsorption of a highly polar organic
liquid on the uncoated portion of the clay surface would greatly enhance
solvation of the micelle in the remainder of the liquid.
Jordan found that, for complexes prepared with single-chain primary
amines of less than sufficient size to coat the montmorillonite particles
TABLE 32. GEL VOLUMES OF 2-G SAMPLES OF DODECYLAMMONIUM-BENTONITE IN
VARIOUS LIQUIDS
(After Jordan 20 )
Gel Gel
volume, volume,
Liquid ml Liquid ml
Water (untreated bentonite) ....... . 31 Butyl Carbitol ............ . 12.5
Water (dodecylammonium-bentonite) 2.0 n-Butyl phthalate ......... . 12.5
Petroleum oil, Gulfpride SAE 10 ... . 2.5 Isophorone ............... . 12.5
Petroleum oil, Gulfpride SAE 40 ... . 2.5 Benzyl alcohol ............ . 13.0
Dow Corning Fluid 200 ............ . 2.5 Bromoform ............... . 13.0
Petroleum ether .................. . 3.0 Ethyl acetate ....... . 13.0
Piperidine ........................ . 3.0 TricresyI phosphate ....... . 13.0
Naphtha ......................... . 3.5 Acetone .......... . 13.5
Carbon disulfide .................. . 4.0 Ethanol (95 %) ........... . 13.5
Carbon tetrachloride .............. . 4.0 Nitroethane .............. . 13.5
Dibutylamine .................... . 4.0 Acetonitrile .............. . 14.0
Glycerol ......................... . 4.5 Isoamyl acetate ........... . 14.0
Tributylamine .................... . 4.5 Castor oil ................ . 14.0
Amyl nitrate ..................... . 6.0 Linseed oil ............... . 14.0
a-Butylene hromide ............... . 6.0 Oleic acid ................ . 14.5
Eucalyptol. ...................... . 6.5 n-Butylaldehyde .......... . 15.0
Styrene .......................... . 6.5 Cyclohexanone ........... . 15.0
Toluene ......................... . 6.5 Dodecylamine ............ . 15.5
Bromobenzene .................... . 7.0 Ethyl bromide ... ' ....... . 15.5
Linoleic acid ..................... . 7.0 n-Butyl tartrate.. . ....... . 16.5
Cymene ......................... . 7.5 n-Heptaldehyde .. 18.0
Aniline .......................... . 8.0 Methyl iodide. . . ...... . 18.0
Cyclohexanol ..................... . 8.0 y- Picoline ................ . 18.0
Ethylene diehloride ............... . 8.0 Acetophenone ... . 19.0
Benzene... . . ............ . 9.0 Tetraethyl orthosilicate .. 19.0
Paraldehyde. . . . . . . .. . ........... . 9.0 Coconut oil. ....... . 20.0
Aeetic anhydride .................. . 10.0 Dodecyl alcohol. .......... . 20.0
Chloroform ...................... . 10.0 Methyl ethyl ketone ...... . 20.0
o-Cresol . ........................ . 10.0 Diethyl ketone .. 21.0
Ethyl malonate ................... . 10.0 Hexadienal .. 21.0
Formamide.. . .......... . 10.0 Pyridine .... 28.0
Furfuryl alcohol .................. . 10.0 Benzaldehyde ... 31.0
Toluidine .................... " .. . 10.0 Benzoyl chloride .. . 33.0
Phenol .......................... . 10.5 Crotonaldehyde ......... . 34.0
Butyl stearate .................... . 11.0 Ethyl ether ............. . 35.0
2-Nitropropane ................... . 11.0 Furfural ...... , . , ........ . 35.0
Acetic acid, glacial. ............... . 12.0 Benzonitrile. . . .. . ....... . 50.0
Isoamyl alcohol. . . . . . . . . . . . .. . ... . 12.0 Nitrobenzene ... , ......... . 88.0
1-Nitropropane ................... . 12.0
completely with a layer of hydrocarbon chains, swelling can be developed

in nonpolar hydrocarbons if a second organic liquid that is highly polar,
e.g., an alcohol, ester, ketone, or aldehyde, is added to the system. Figure
103 shows the great influence of small additions of alcohols on the swelling
characteristics of octadecylammonium-montmorillonite in toluene.
268 Clay Mineralogy
It is postulated that the highly polar additive is adsorbed on the
uncoated surface of the montmorillonite, thereby rendering the individual
flakes entirely organophilic and compatible with the hydrocarbon portion
of the solvating liquid. In Fig. 104, data are presented for toluene-ester
systems with octadecylammonium-montmorillonite; these data indicate
that the acid portion of the ester is more important than the alcohol por-
tion. For example, ethyl formate is more effective than methyl acetate;
methyl, ethyl, and butyl acetate are approximately equal in effect; and
butyl stearate is practically without influence.
o Ethyl Formate
-6_ 50 o Methyl Acetate
E '" Ethyl Acetate
o Butyl Acetate
(J)
40 Butyl Stearate
0I
~
:g 30
O~~~ __~__~___ L_ __ L_ _~_ _~~_ _L __ _~~

o 10 20 30 40 50 60 70 BO 90 100
Per Cent Ester in Mixture
FIG. 104. Swelling of octadecylammonium-montmorillonite in toluene-ester mix-
tures, after Jordan. 20
Jordan has shown that montmorillonite complexes witb quaternary

ammonium salts having two long aliphatic chains are organophilic in
single-hydrocarbon liquid systems. Dispen;ion in toluene takes place
much more readily and completely with the double-chain complexes than
with the single-chain complexes, probably because the completely coated
particles are more compatible with the hydrocarbon. In the case of
double-chain complexes, there is a tendency for an initially greater sep-
aration of the montmorillonite units than in single-chain complexes.
Jordan concludes that the swelling process involves two factors for an
incompletely clad clay complex: (1) high adsorption energy of liquid for
uncoated clay, and (2) high solvation energy of liquid for the organic
coating. Both conditions are fulfilled by a single liquid like nitrobenzene
or by a binary liquid system of the toluene-alcohol type. Adsorption of
polar molecules by the clay must tend to separate the plates, thereby
rendering the micelle more organophilic and the organic coating more
accessible. The dialkyl quaternary complexes are solvated by toluene
alone because of the complete coating of organic matter and the initially
Clay-Mineral-0rganic Reactions 269
higher separation of the plates. Data for this conclusion are given in
Fig. 105, which shows gel volumes for several organic-ammonium-
montmorillonite complexes in several primary aliphatic alcohols. In
agreement with this conclusion, octylammonium-montmorillonite sol-
vates most in the most polar alcohol, while dimethyllaurylcetylammo-
nium montmorillonite solvates least in methanol and to a progressively
greater extent in the less polar members of the series.
27
24
/'
Dimethy/dodecy/-
hexodecy/ Ammonium
Bentonite
y' / '
"*-E 21
0
(f) "............ / ' Dodecy/omine
E 18 / ('., '-o--..a....~/' /Bentonite
0
'- / ,/ . . ,..... 7---1(_
(!) I '" ' -_
/ ~;~~~~~~;~;;;~~~~T~=--::::~
I /
N
15
..j
~ ,/ Bentonite
.!:: 12
CI.>
E
::J
g 9
Q;
(!)
:3
1 2345678
Carbon Chain Length of Alcohol
FIG. 105. Effect of chain length on swelling of organic-ammonium-montmorillonite
in alcohol, after Jordan.20
STRUCTURAL IMPLICATIONS OF MONTMORILLONITE-

ORGANIC COMPLEXES
Studies of the reactions between montmorillonite and organic molecules

have led some investigators to certain conclusions regarding the atomic
. structure of the mineral. Thus, Berger 35 concluded that methylation
takes place when dry hydrogen montmorillonite is treated with diazo-
methane, and that this requires the presence of acidic hydroxyls on the
exterior of the montmorillonite structure. Berger further concluded from
quantitative determinations of the methylation that the Hofmann struc-
86 Berger, G., The Structure of Montmorillonite, Chem. Weekblad, 38, 42-43 (1941).
270 Ciay Mineralogy
ture would not provide an adequate number of such hydroxyls. Since
the Edelman-Hofmann structure (see page 61) postulates an abundance
of such hydroxyls, the methylation was believed to be evidence for this
structure. Edelman 36 has pointed out that the capacity for methylation
of montmorillonite determined by Berger requires the inversion of only
about one-third of the number of silica tetrahedrons (see page 62) orig-
inally postulated as inverted by Edelman and Favejee. 37
Gieseking 38 has expressed the opinion that the reaction of acetyl chlo-
ride with air-dried montmorillonite, producing a lyophobic material which
no longer swells in water, affords evidence for the occurrence of OH groups
on the basal surfaces of montmorillonite.
Deuel et aZ. 39 have studied the esterification of hydrogen montmoril-
lonite by 1,2-epoxides and the reaction of sodium montmorillonite and
mustard gas, with the production of hydrophobic clay esters of largely
reduced cation-exchange capacities. The clay esters produced in the
reactions can be saponified. These investigators also found that hydro-
gen clays can be transformed by reaction with 80Cb into clay chlorides,
which in turn will react with alcohols to form esters. Deuel et aZ. con-
clude that these reactions require the presence of a sufficiently large num-
ber of surface hydroxyls to provide evidence for the Edelman-Favejee
structure.
Numerous ir.vestigators have pointed out that the Edelman-Favejee
structure is not supported by diffraction data for montmorillonite. For
example, the value of d(OOl) in completely dried alkali and alkaline-earth
montmorillonites is less than 10 A, which is too low for the inverted silica
tetrahedron structure of Edelman and Favejee. Projecting silica tetra-
hedrons from the basal planes of the mineral would requir~ a d(OOl) value
of at least 11 A in the dried condition.
The problem of the structural implications of montmorillonite-organic
reactions cannot now be resolved, but in the face of the X-ray-diffraction
data, it seems that some explanations other than the Edelman-Favejee
structure must be found for the reactions described by Berger, Gieseking,
and Deuel. An alternative explanation, which cannot yet be evaluated
adequately, is the structure suggested by McConne1l 40 (see page 63),
86 Edelman, C. H., Relation entre les proprietes et la structure de quelques mincraux

argileux, Verre silicates ind., 21, 3-6 (1947).
37 Edelman, C. H., and J. L. Favejce, On the Crystal Structure of Montmorillonite
and Halloysite, Z. Krist., 102, 417-431 (1940).

38 Gieseking, J. E., The Clay Minerals in Soils, Advances in Agron., 1, 59-204
(1949).
39 Deuel, H., G. Huber, and R. Iberg, Organische Derivate von Tonmineralien,
Helv. Chim. Acta, 33, 1229-1232 (1950).

166-172 (1950).
which would provide some hydroxyls in the basal montmorillonite sur-
faces without projecting silica tetrahedrons.
ANALYTICAL TECHNIQUES BASED ON CLAY-

MINERAL-ORGANIC REACTIONS
The reactions of clay minerals and organic molecules have been used as
the basis for analytical techniques for the identification of the clay min-
erals, for the determination of certain properties of the clay minerals
themselves, and for the determination of the geometry and properties of
the organic molecules.
X-ray Techniques. The montmorillonite minerals commonly give
diffuse reflections and a nonintegral series of basal reflections. Bradley 14
and MacEwan 15 have pointed out that organic-montmorillonite com-
plexes tend to have a high degree of regularity in the c spacing, giving an
integral series of very sharp reflections. Thus, the identification of mont-
morillonite by X-ray diffraction is greatly simplified by preliminary treat-
ment with certain organic reagents. It is desirable to use an organic
liquid of low volatility which is readily miscible with water, so that the
natural hydrated montmorillonite may readily be changed into an organic
complex. Glycerol and ethylene glycol are used extensively. The
hydration state of the natural mineral does not alter appreciably the
position or intensity of the reflections given by the organic-montmoril-
lonite complex. It is common practice to add enough glycerol to air-dried
material to make it visibly moist. Any montmorillonite present will
give rise to a periodicity along c of about 17.7 A, with a series of higher
orders, most of which will be adquately separated from reflections given
by the other clay minerals. By the use of the organic treatment tech-
nique, relatively small amounts (5 per cent) of montmorillonite can be
identified. Prior to the development of this technique, such small
amounts of montmorillonite would not have been detected.
Glycerol treatment of halloysite (4H 2 0) causes a shift in the basal
reflections to 11 A, and first, third, and fifth orders from this spacing are
apparent. A similar shift takes place for the dehydrated form of this
mineral if the dehydration is not essentially complete. Thus, air-dried
samples will show the shift, whereas samples dried at 75C for about 12 h
or at higher temperatures for shorter intervals do not form the complex
and exhibit no shift.
Glycerol forms complexes with vermiculite with a spacing of 14 A,
which is so close to the spacing of the fully hydrated natural mineral that
it has little diagnostic value.
The glycol or glycerol treatment technique is of tremendous value in
resolving very intimate interlayer mixtures of some of the clay minerals.
272 Clay Mineralogy
For example, the reflections from a mixture of montmorillonite and illite
are likely to be diffused and complex because of the variable c spacing of
the former mineral. Treatment with glycol or glycerol develops a uni-
form spacing of the montmorillonite and thereby sharpens the reflections
so that they Can be interpreted more readily. Use of the organic treat-
ment technique has shown that many materials described as single species
are actually interlayer mixtures. Thus, the type bravaisite and many
materials described as beidellite were found to be mixtures of illite and
montmorillonite (see page 36). For a detailed discussion of the glycol-
glycerol diffraction technique the works of MacEwan 15,24 and Bradley 14
should be consulted.
Differential Thermal Techniques. When clay-mineraI-organic com-
plexes are heated, the organic material is oxidized, and an exothermic
reaction takes place. Allaway, 41 working with piperidine-montmorillonite
complexes, showed that the detailed features of the exothermic reactions
varied with the composition of the montmoriIIonite. Triethanolamine,
monoethanolamine, and n-butylamine complexes gave results similar to
those for piperidine complexes. Allaway found that the exothermic
reaction was always multiple and that a peak at 7000 suggested a high-
magnesium montmorillonite, a peak between 450 and 5000 suggested
an iron-rich montmorillonite, and a peak at about 6000 suggested a
member of the montmorillonite group containing some aluminum in tetra-
hedral coordination. Similar variations in the exothermic reactions were
found for several illites, but they were of relatively lower intensity than
for the montmorillonites. Piperidine treatment had little effect on the
differential thermal curves of kaolinites.
Allaway concluded that the piperidine held by the montmorillonite
breaks down on firing with a loss of hydrogen, leaving a Qoating of carbon
on the external surfaces and between the unit layers of the clay mineral.
The carbon is burned off rather slowly as long as the clay-mineral struc-
ture remains intact, but when the structure breaks down, the carbon is
quite rapidly oxidized. The use of organic cations is essentially a tech-
nique for magnifying the thermal effects of the breakdown of such struc-
tures as can be coated with carbon, with strong exothermic reactions
replacing weaker endothermic reactions. Certain lattice changes which
appear in the ordinary differential thermal procedure to be slow gradual
breakdown of the clay are resolved into two separate peaks of thermal
activity by the use of the piperidine treatment. Thus, the treatment
procedure may resolve mixtures or permit the identification of certain
mineral components which would not be revealed by the usual method of
thermal study.
41 Allaway, W. H., Differential Thermal Analy~es of Clays Treated with Organic
Cations as an Aid in the Study of Soil Colloids, Soil Sci. Soc. Am. Proc., 13, 183-188
(1948).
Clay-Mineral-Organic Reactions i
Optical Methods. The adsorption of organic molecules between the

basal plane surfaces of montmorillonite and to a lesser extent of halloysite
causes a slight but definitely determinable change in their indices of
refraction (see pages 286-288). Therefore, when these clay minerals are
mounted in certain liquids for optical determinations, the indices of refrac-
tion can be seen to change as the liquid is adsorbed by the clay mineral.
This characteristic can be used as a diagnostic criterion for the identifica-
tion of montmorillonite and halloysite. No specific data are avail-
able for vermiculite, but its indices of refraction should also change,
since it possesses the property of interlamellar adsorption of organic
molecules.
Cation-exchange Capacity. Robertson and Ward 42 have recently
described a rapid method for estimating the cation-exchange capacity of
clay materials based on the adsorption of methylene blue. The change
in color of a standard methylene blue solution produced by the adsorption
of the clay material permits a determination of the amount of the organic
molecule adsorbed, from which the cation-exchange capacity can be calcu-
lated. For most clays the results are in good agreement with those
obtained by other procedures.
Surface-area Determination. Dyal and Hendricks 43 have pointed out
that the total surface, the external surface, and, by difference, the internal
surface of clays can be determined from adsorption reactions with polar
molecules, e.g., ethylene glycol. Total surface can be determined by the
retention of ethylene glycol in an evacuated system. After the clay has
been heated to about lOOoe for halloysite and 600 0 e for montmorillonite,
the organic molecule does not penetrate between the unit layers, and
values for retention on external surfaces alone are obtained. The differ-
ence between total retention and external retention expresses the amount
of interlayer surface, and from this value the amount of halloysite or
montmorillonite can be estimated.
Geometry and Properties of Organic Molecules. The study of clay-
mineral-organic reactions and the resulting complexes provides an
approach to the investigation of the shape and certain bonding character-
istics of the adsorbed organic molecules themselves. Thus the space
required for the organic molecules between the montmorillonite layers, as
revealed by determination of d(OOI) spacings, provides information on
the thickness, areal dimensions, and in some cases the bond distribution
in the organic molecules. This subject is aside from the field of this
volume, and for information concerning it, the works of Hendricks, 17
MacEwan,23 and Bradley 13 should be consulted.
42 Robertson, R. H. S., and R. M. Ward, The Assay of Pharmaceutical Clays, J.

Pharm. Pharmacal., 3, 27-35 (1951).
43 Dyal, R. S., and S. B. Hendricks, The Surface of Clays in Polar Liquids as a
Characteristic Index, Soil Sci., 69, 421-432 (1950).
274 Clay Mineralogy
STAINING TESTS FOR CLAY MINERALS
According to Faust,44 who has investigated the history of staining tech-

niques in general and has summarized much of the early work, Behrens 45
in 1871 appears to have first applied staining methods to mineralogy.
Staining techniques were used for biological material beginning about
1838. An extensive literature has grown up regarding the use of staining
methods, and many persons have attempted to apply them to clays.
The adsorption of various organic substances by natural and chemically
altered or heat-treated clays can produce color changes in the clay. Fre-
quently such changes vary, depending on the identity of the clay mineral
and its composition. The color changes, therefore, provide a possible
basis for identifying the clay-mineral components of clay materials. A
staining test has the advantage of being rapid and simple to perform, even
in the field.
In general, the reaction and resulting color change are slight or absent
for clay minerals of low adsorptive capacity, such as kaolinite, but typ-
ically are pronounced for highly adsorptive clay minerals, such as mont-
morillonites. Two mechanisms have been proposed to explain the color
reactions obtained with clay minerals and various reagents: (1) an acid-
base reaction in which the natural or acid-treated clay reacts as an acid,
and (2) an oxidation-reduction phenomenon in which certain ions, mainly
ferric iron contained in the clay-mineral lattice, cause an oxidation of the
reagent. Substances probably causing color changes by the acid-base
reaction include triphenylmethane, azine, and azo dyes. Substances
probably changing the color of clays by the oxidation-reduction reaction
include benzidine and other aromatic amines.
The general applicability of staining tests for the identification of the
clay minerals is restricted, because the development of characteristic
color reactions can be inhibited, augmented, or masked by several com-
mon ingredients of clay materials other than the clay minerals. Man-
ganese dioxide may cause the oxidation-reduction reaction with benzidine,
even in the absence of adsorptive clays, and ferrous iron and other reduc-
ing agents may prevent the color-forming reaction from taking place.
In techniques that require the preparation of hydrogen clays for the stain-
ing reaction, the acid treatment may destroy the structure of the clay
minerals. Thus iron- and magnesium-rich montmorillonites, and conse-
C\_uently some nontronites and hectorite, might not appear to react with ~
the staining solutions like other members of the montmorillonite group.
44 Faust, G. F., Staining of Clay Minerals as a Rapid Means of Identification in
Natural and Beneficiation Products. U.S. Bur. Mines Rept. Invest. 3522 (1940).
46 Behrens, H., Mikrochemische Methods fur Mineralanalyse, Verslag. M ededeel.
Koninkl. Akad. Wetenschap. Amsterdam, 1, 17 (1881).
Clay-Mineral-Organic Reactions
TABLE 33. CHAItACTERISTIC STAINING OF CLAY MINERALS
(After Mielenz et al. 46 )
Untreated
Acid-treated clay
clay
Clay mineral
Benzidine Safranine Y Malachite green
Kaolinite No reaction Red-purple red. * Strong Blue-green blue and blue-

to weak pleochroism green. * Strong to weak
from reddish purple par- pleochroism from yel-
allel to cleavage to yellowish green parallel to
lowish red perpendicular cleavage to blue per-
to cleavage pendicular to cleavage
Halloysite No reaction Blotchy stain: purple, Blotchy stain: yellow
purple-blue purple, and green-yellow, blue-
red-purple red. Not green, and green-yel-
pleochroic low. Not pleochroic
Dickite No reaction Crystals not stained. Crystals not stained.
Very, weak pleochroism Very weak pleochroism
from reddish purple or from yellowish green or
purple parallel to cleav- colorless parallel to
age to reddish yellow cleavage to light blue
perpendicular to cleav- perpendicular to cleav-
age age
Nacrite No reaction Crystals not stained. Crystals not stained.
Weak pleochroism from Weak pleochroism from
reddish purple parallel yellowish green parallel
to cleavage to yellowish to cleavage to blue per-
red perpendicular to pendicular to cleavage
cleavage
Montmorillonite Purple-blue Purple-blue Yellow-red yellow
N ontronite Blue-green Red-purple red * Green blue-green and
blue-green blue *
Hectorite Purple-blue Red-purple red * Blue-green blue *
Illite No reaction Red-purple red * Green blue-green *
Attapulgite No reaction Red-purple red * Blue-green and blue-
green blue *
Pyrophyllite No reaction Not stained Not stained
* Dye absorbed without change in color. Samples of nontronite included in these

tests did not change the color of the dyes in these tests, but specimens of nontronite
reacting in a manner similar to montmorillonite have been examined.
Ii Also, the strong original color of a sample particularly rich in iron might
tend to mask the color change which would result from the adsorption of
the staining reagent.
Because of the above factors caution must be used in applying any
staining test indiscriminately to clay materials. Under restricted con-
ditions, when the samples to be studied are relatively pure clay-mineral
276 Clay Mineralogy
materials, or where the range of composition is restricted and known in
general terms, staining techniques may be entirely satisfactory. For
example, some of the sedimentary kaolins in the Georgia area contain
montmorillonite in addition to kaolinite. There are no other constit-
uents of these kaolins that might cause a color reaction, and a staini~g
test quickly shows the presence of montmorillonite.
A very large number of reagents have been suggested and tried for
clay-mineral staining tests. In recent years the tendency has been to
use several reagents separately in a test, so that the deficiencies of one can
be compensated by the value of another for a given material. Such a
test has been described by Mielenz, King, and Schieltz 46 and appears to
give quite satisfactory results. In this procedure a pulverized portion of
the natural clay is treated with benzidine, and portions of the clay, after
treatment with strong hydrochloric acid and washing to remove the
excess chloride, are treated separately with a saturated solution of safra-
nine Y in nitrobenzene and with a saturated solution of malachite green
in nitrobenzene. In addition to the visual determination of the color
developed, the samples are examined with the petrographic microscope
to determine the development of any pleochroic colors and other changes
in optical characteristics. The results of the use of these reagents are
summarized in Table 33 taken from the work of Mielenz, King, and
Schieltz. 46 Even with such an elaborate staining technique, caution
must be used; anomalous results are sometimes obtained because various
nonclay substances may cause or inhibit staining reactions; and some
clays fail to react in a characteristic manner for as yet inexplicable
reasons.
Bangham, D. H., and S. Mosallam, The Adsorption of Vapours at Plane Surfaces of
Mica, Proc. Roy. Soc. (London) A, 166,558-571 (1938).
Bossazza, V. L., On the Adsorption of Some Organic Dyes by Clays and Clay Minerals,
Am. Mineral., 29,235-241 (1944).
Caillere, S., R. Glaeser, and J. Esquerin, Preparation of Halloysite with Interplanar
Distances of 14 A and 17 .it, Compt. rend., 230, 308-310 (1950).
Emodii, B. S., The Adsorption of Dyestuffs by Montmorillonite, Mineralog Soc. Gr.
Erbring, H., and H. Lehman, Exchange Reactions with Voluminous Organic Colloid
Ions on Sodium Bentonites, Kolloid-Z., 107, 201-205 (1944).
Grandjean, M. F., Coloration des argiles par les couleurs d'analine, Bull. soc. fran!;.
mineral., 32, 408-419 (1909).
Hauser, E. A., Modified Gel Forming Clay and Process of Producing Same, U.S.
Patent 2,531,429 (1950).
Hauser, E. A., and M. B. Leggett, Color Reactions between Clays and Amines, J.
Am. Chem. Soc., 62, 1811-1814 (1940).
46 Mielenz, R. C., M. E. King, and M. C. Schieltz, "Staining Tests," Rept. 7,
American Petroleum Institute Project 49, Columbia University, New York (1950).
Hendricks, S. B., and L. T. Alexander, A Qualitative Color Test for the Montmoril-
lonite Type of Clay Minerals, J. Am. Soc. Agron., 32,455-458 (1940).
Jackson, M. L., and N. M. Hellman, X-Ray Diffraction Procedure for the Positive
Identification of Montmorillonite, Soil Sci. Soc. Am. Proc., 6, 133-145 (1941).
Jordan, J. W., B. J. Hook, and C. M. Finlayson, Organophilic Bentonites, II, Organic
Liquid Gels, J. Phys. & Colloid Chem., 64, 1196-1208 (1950).
Lloyd, J. U., Discovery of the Alkaloidal Affinities of Hydrous Aluminum Silicate.
J. Am. Pharm. Assoc., 6,381-390, 490-495 (1916).
Miller, J. G., H. Heinemann, and W. S. McCarter, The Effect of Solvent on the
Adsorption of Liquids by Bauxite and Attapulgus Clays, J. Phys. & Colloid
Chem., 66, 515-523 (1951).
Page, J. B., Unreliability of the Benzidine Color Reaction as a Test for Montmoril-
lonite, Soil Sci., 61, 133-140 (1941).
CHAPTER 11
Optical Properties
The optical properties of the clay minerals as reported in the literature

are summarized in Table 34. In general it is necessary to give a range of
optical values instead of a single value, because of inherent variations in
the clay minerals themselves and because of difficulties in measuring
precisely the optical properties of the extremely small clay-mineral par-
ticles. Also, it is usually possible to determine only some of the optical
characteristics.
Inherent variations in the clay minerals which cause changes in optical
properties are replacements within the lattice. Also, variations in the
amount of any interlayer water in the expanding-lattice minerals are
reflected by changes in optical properties. In addition, the clay minerals
from time to time may exhibit strain or warping within the lattice or
deviations from perfection of crystallinity, which would cause some altera-
tion in optical characteristics. The variation due to strain or imperfec-
tions in crystallinity is likely to be small. Further, the optical values of
some clay minerals are different in different liquids which may be used as
index liquids to measure them (see pages 286~288).
KAOLINITE
As shown in Table 34, there is a measurable and diagnostic difference in

the optical properties of kaolinite, dickite, and nacrite. The optical
characteristics of anauxite are substantially the same as those of kaolinite.
Because of the absence of replacements within the kaolinite lattice and
the low adsorptive properties of the mineral, kaolinite exhibits narrowly
restricted optical values. Kaolinite frequently occurs in particles of suffi-
cient size to permit fairly complete and precise optical measurements. As
the degree of crystallinity decreases, the poorly crystallized variety has
slightly decreasing indices of refraction and birefringence.
The data in Fig. 106, after Mehmel,l show that the mean index of
refraction of kaolinite increases slightly on heating up to the point of loss
of OR lattice water slightly above 400C. Accompanying this dehydra-
1 Mehmel, M., Beitrage zur Frage des Wasserhaltes der Minerale Kaolinit, Halloysit
und Montmorilionit, Chem. Erde, 11, 1-16 (1937).
278
Optical Properties 279
s,....,
S
o ~
~~
If., ,. . ,
+1 ,....,~
ISle
CO +8
1J d
II liS ,....,
o
,....; 0-1 S
~~
d

~~
"P.-1
i~ /I /I
o ISlISl
---~------------------------------~~--~------I
1 Q)
1 Q)
i;:; i;:;
~. ~ ~ ;::,
VAV
RRR
e
e.>0
;::;00
- I
d 0
;>C'l
1O
fo I I I I I I I I I +1 I I I I
m I',
0> .-<0 1O0 C'l1001Ot-0 1OC'l1O

g 0";'
00 8;;g; ";'lOC':)lOOCO
000000
..... co co
000
~8 ,,1O ...... ~~ d~~

1 9 111 99 9
;:;;MCn'1~8
0
828 88 00.0
.0.0
. 02 g22
00 000 00 0000000 000
1'"""11"""'1 ~1'"""11'"""I1'"""I1'"""I_'
~~~:6~~~~~~
'';''';'101O1O1O1O1O1O1O
280 Clay Mineralogy
14 Kaolinite __
12 /" 1.58
10 __ -_........v-wt. Loss 1.57
8 ---r:Jndex 1, /.56
6 J""_"-.. 1.55
4
2
.....-,---
...-/.
--- 1.54
1.53
o !:-...JoC::-!-::-...J_~,..-L~...J_....L.__J___J___J___j_-L-L_J 1.52
o 100 200 300 400 500 600 700
Temperature C
_
c:
24
22
Holloysite __ -- ~
c:
<3 20 tcWt. Loss u

'- 18
3:. 16 Index~_~
I 1.57 ~
Qj
1.56 0:::
.S 14 ------- f\. 1.55 '0
'"
2
12
I
r_---J '\...__ 1.54
a '-L-~..J-...,,-l::-:,-L--~::-L~,,,--L_l-~:;L.-__J_---L-::-L..,,--J 53 ~
><
_.j 0 100 200 300 400 500 600 700

f.
.s
Temperature C
c:
o
<1J
~
24 Man tmorl/lonite
(---
-
22
20 _,..-/
18
16 --<-wt. Loss
14 /
12 I
1.57
10 / _---{Index
1.56
: I/~--- " 1.55
4 I ' 1.54
~ri "-..,_,~ 1.53

\.52
o 100 200 300 400 500 600 700
Temperature C
FIG. 106. Relation of mean index of refraction to weight loss and firing temperature
for kaolinite, halloysite and montmorillonite, after Mehme1. 1
tion reaction, there is a sharp reduction in the mean index from about
1.575 to about 1.525 and probably also a complete loss of birefringence.
RALLOYSITE
Usually only the mean indices of refraction can be determined for the
halloysite miner~ls b~cause of their low degree of crystallinity. As shown
in Fig. 107, the transition from the 4H 20 to 2H 20 form is accompanied,
according to Mehmel 1 and Correns and Mehmel,2 by an increase in index
from about 1.532 to almost 1.550. Alexander and colleagues 3 have
computed the theoretical value of the 4H 20 form at 1.490, assuming the
structure for the mineral suggested by Rendricks 4 and using the molecular
'.
16
14 Ho//oysite 1.55
.-.-
12 I
I
I
10 I
I
I
8 I, 1.54
I
6
Index:J_. __ . __:
4
~~s} . - - - - - - - '
2 1.53
c:
a ~
CQ)
10 9 8 7 6 5 4 3 2 I 0 105 a e (j
u
... Water Vapor Pressure in mm. Od, .......
0
-
Q)
Q)
(l.. a::
.S 16 1.56 0
on
on
0
..J
14 Mon/morl/ /onite ,/--; 1.55 '0
oS
x
Q)
....
.c
12
./ /
/ 1.54 c
0
/ /'
0>
.~ 10 1.53 ~
. /A-WI. Loss
8 lndexy 1.52
6
. I 1.5J
4 ./ /
I
tJio,
/ /
0
2
--_../ 1.49
1.48
7 6 5 4 3 2 I 0 105e
Water Vapor Pressure in mm, Od.
FIG. 107. Relation of mean index of refraction, weight loss, water-vapor pressure,
and drying at 105C for halloysite and montmorillonite, after MehmeJ.!
refractivity of kaolinite and water. No indices nearly as low as 1.490

appear to have been reported for the 4R 20 form, which may be taken as
lack of support for this structural concept of the mineral.
On heating to temperatures above 100C, halloysite shows a slight
gradual increase in index of refraction, until the OR water is driven off at
2 Correns, C. W., and M. Mehmel, Ueber den opt is chen und r6ntgenographischen
Nachweis von Kaolinit, Halloysit, und Montmorillonit, Z. Krist., 94, 337-348 (1936).
3 Alexander, L. T., G. T. Faust, S. B. Hendricks, H. Insley, and H. F. McMurdie,
(1943).
4 Hendricks, S. B., On the Structure of the Clay Minerals: Dickite, Halloysite, and
Hydrated Halloysite, Am. Mineral., 23, 295-301 (1938).

282 Clay Mineralogy'
about 400C. The dehydration of the mineral is accompanied by a sharp
reduction in the index of refraction from about 1.560 to 1.535, as shown
in Fig. 106.
MONTMORILLONITE
The montmorillonite clay minerals show large variations in optical

properties as a consequence of substitutions within the lattice. The data
in Fig. 108, after Ross and Hen-
dricks,5 show that an increase in
1.660
the Fe 2 0a content from 0 to 32 per
c 1.640 v cent is accompanied by an in-
.Q
t; 1.620
L)-- crease in the l' index from about
~
,/
'Q; 1.600 ./ 1.545 to 1.645 and by an increase
0::
... V ~
v in the birefringence from about
'01.560 0.025 to 0.045. Mehmel 1 has
x /v V shown that the indices increase
~ 1.560
c
- 1.540
/" VV as the content of magnesium in-
1.520
/ v creases. In the samples of mont-
o 4 8 12 16 20 24 28 32 morillonite studied by him, the
Ferric Iron in Per Cent mean index increased from about
FIG. 108. Indices of refraction and bire- 1.530 to 1.548 as the MgO content
fringence of montmorillonite in relation to increased from about 2 to 6 per cent.
content of ferric iron, after Ross and
Hendricks. 5 As shown in Fig. 107, the mean
index of refraction increases in the
range from about 1.49 to 1.55 as the thickness of the interlayer water
decreases. The identities of the excha,ngeable cations affect the indices
indirectly under certain conditions of humidity because of their control
of the thickness of the interlayer water. Thus air-dried Na montmoril-
lonite tends to have interlayer water One molecular layer thick, whereas
Ca montmorillonite under the same conditions tends to have interlayer
water two molecular layers thick. Under such conditions the Ca mont-
morillonite will have the lower indices. There is no positive evidence
that the exchangeable cations themselves directly influence the optical
properties, but the data by Marshall 6 on birefringence in an electrical
field (see page 290) suggest that such an influence does exist.
Mehmel, as shown in Fig. 106, and Grim and Bradley 7 have independ-
ently shown that the mean index of refraction of a montmorillonite rel-

6 Marshall, C. E., The Orientation of Ani~otropic Particles in an Electric Field,

7 Grim, R. E., and W. F. Bradley, The Eff()ct of Heat on the Clay Minerals Illite
and Montmorillonite, J. Am. Ceram. Soc., 23, 242-248 (1940).
OptiCal Properties 23~
atively low in iron increases from about 1.54 to about 1.570 when all the
interlayer water is eliminated on heating to 250 to 300C. At higher
temperatures there is no appreciable change in indices until about 450 0 to
500C, when the OR lattice water is lost, accompanied by a sharp drop
in mean index to about 1.52. This drop in index is accompanied by a
1.700
c:
0
'"5 1.660 )'
t'
ill
0: 1.620
CIt
'0
x 1.580
(J)
"0
E 1.540
0 100 200 300400 500 600 700 800
Temperature C
1.62
c
0
.~
1.58 50ponite
_g
QJ
0: 1.54
'0
x 1.50
QJ
"0
E 146
0
Temperature C
FIG. 109. Effect of heating on the refractive indices and birefringence of nontronite
and saponite, after Caldwell and Marshall. 8
reduction in birefringence. Caldwell and Mar:ohall s present data for sap-

onite (Fig. 109) showing somewhat the same effect of heating, except that
the decrease in indices accompanying the loss of OR water is more grad-
ual. These investigators show also that at higher temperatures, above
that at which the hydroxyl water is lost, there is an increase in indices and
birefringence. A sample of nontronite investigated by Caldw~ll and
MarshallS showed (Fig. 109) a gradual inerease in indices on heating to
about 850C, with only a slight flattening of the curve accompanying loss
of OR water. These authors also show that the birefringence of non-
tronite increases as the mineral is dehydrated, up to about 700C. At
temperatures above 700C, the birefringence decreases moderately.
8 Caldwell, O. G., and C. E. Marshall, A Study of Some Physical and Chemical
Properties of the Clay Minerals Nontronite, Attapulgite, and Saponite, Coil. Agr.
Univ. Missouri Research Bull. 354 (1942).
284 Clay Mineralogy
ILLITE
The exact variations of the optical properties of the illites accompany-

ing substitutions within the lattice are not known, but they are probably
of the same order of magnitude as those for the montmorillonite minerals.
The illites with the muscovite type of structure have somewhat lower
indices of refraction and lower birefringence than muscovite itself, and
the same relation apparently holds for the illites with the biotite type of
structure. Thus, in the case of illite with the muscovite structure, the 'Y
value is frequently of the order of 1.570 to 1.580, as compared to 1.595 for
1.60
c
o
't; 1.58
o
.t:
<lJ
0::: 1.56
'0
x
~ 1.54
.
I. 5 2 ;:----::::':c::------:--=-=---=-=:::-----:-.':-::---c::'::-:--~".__---J
o 200 400 600 800 1000 1200 1400
Temperature C
FIG. 110. Effect of heating on the refractive indices and birefringence of illite, after
Grim and Bradley. 7
muscovite itself. For biotite-type illite the 'Y value is frequently 1.62 to
1.63, as compared to 1.64 to 1.66 for biotite. This difference must result,
at least in part, from the lesser population of interlayer cations and from
the presence of some water molecules between the silicate layers of the
illites. It probably also is somewhat dependent on the crystallinity of
the clay minerals, i.e., on defects in the stacking of the silicate layers and
on some variations in the population of the silicate units.
Grim and Bradley 7 have shown (Fig. 110) that on heating there is fre-
quently a slight increase in indices at very low temperatures (200C); e.g.,
the 'Y value increases from about 1.575 to 1.585. Above about 200C
there is little change until about 400C, when there is a drop in indices
accompanying the loss of OR water. At this temperature the 'Y index
decreases from about 1.585 to 1.560. At still higher temperatures, there
is a further decrease in the 'Y index, and the birefringence also decreases.
CHLORITE:
There is substantially no information available in the literature on the

optical properties of the chlorite clay minerals. Unpublished data
I .
uptical Properties 285
obtained by the author indicate that the chlorite clay minerals have some-
what lower indices of refraction than the well-crystallized chlorite min-
erals, and that the difference is of about the same order of magnitude as
that for the illites as compared to muscovite and biotite. There are no
experimental data on the changes in optical properties on heating the
chlorite clay minerals, but it is probable, by analogy with the illite min-
erals, that they would show an increase in indices accompanying any
loss of adsorbed water and a decrease in indices at higher temperatures,
accompanying the loss of hydroxyl water.
VERMICULITE
There are no precise data available on the optical properties of the

vermiculite clay minerals. However, they probably have substantially
the same optical characteristics as those of coarse grain vermiculite.
Because of the structural similarity of vermiculite and montmorillonite,
it would be expected that the optical properties of vermiculite would vary
with the degree of hydration and with substitutions within the lattice in a
similar manner to that discussed for the montmorillonite clay minerals.
Also, it would be expected that heating vermiculite would cause changes
in optical characteristics similar to those found for the montmorillonites.
SEPIOLITE-ATTAPULGITE-PALYGORSKITE
The optical properties of only a few specimens of these clay minerals

have been measured, so that the possible variations are not definitely
known. It appears probable that the 'Y index varies from about 1.50 to
1.555 and that the birefringence varies from about 0.01 to 0.035. Accord-
ing to Von Englehardt,9 the indices increase in the palygorskites as the
magnesium content increases. The similarity between the indices and
birefringence of some of these clay minerals and those of the montmoril-
lonites is noteworthy and has led sometimes to their erroneous identifica-
tion as montmorillonites.
Caillere 10 has stated that the indices of sepiolite increase from a mean
value of 1.49 to 1.54 when the mineral is dehydrated above about 465C.
At still higher temperatures the mean index increases further to about
1.61 on calcination.
Caldwell and Marsha1l 8 showed that attapulgite behaves differently
9 Von Englehardt, W., Ueber silikatische Tonminerale, Fortschr. Mineral. Krist. u.
Petrog., 21, 276-337 (1937).

10 Caillere, S., Etude de quelques silicates magncsiens it facies asbestiforme ou
papyrace n'appartenant pas du groupe de l'antigorite, Bull. soc. fran~. mineral., 59,
353-386 (1936).
286 Clay Mineralogy
from the other clay minerals on heating (Fig. 111). It shows an increase
in indices up to about lOOC, the 'Y value increasing from 1.52 to 1.55;
there follows a decrease to about 200C, 'Y decreasing to 1.53. Above
200C there is a slight gradual increase up to about 850C, with 'Y attain-
ing a value of about 1.56.
c
..g
(.)
.g '7
(J)
0::
'0
'571
1.530 : fZ:
x 1.490 I I !
(J)
"0
0 500 600 700 800
.E Temperature C
FIG. 111. Effect of heating on the refractive indices and birefringence of attapulgite,
after Caldwell and Marshall. 8
INFLUENCE OF IMMERSION MEDIA ON OPTICAL PROPERTIES
Larsen and Wherryll in 1917 found that the indices of refraction of a

clay material from Beidell, Colorado, now known to contain a montmoril-
lonite mineral, increased gradually when the material was immersed in a
mixture of cinnamon oil and oil of clove. This appears to be the first
mention in the literature of variations in the optical properties of the clay
minerals resulting from the media in which they are immersed. Correns
and MehmeP showed that the indices of montmorillonite and halloysite
(4H 2 0) varied in different liquids. In general, the indices of the mineral
increased with time over a period of about 2 hr, and it is now known that
much longer time, perhaps several days, would be necessary for equilib-
rium to be reached. In the case of montmorillonite, changes in optical
values were noted for both air-dried samples and material dried at 105C.
Correns and Mehmel believed that the effect was due to the penetration
of the organic molecules between the silicate layers and thought that
no effect would be produced by organic molecules too big for such
penetration.
Van Baren 12 investigated the influence of a variety of organic liquids
on the indices of refraction of kaolinite, halloysite (2H 20), and montmoril-
lonite (Table 35). He found no changes in the indices of kaolinite or
halloysite (2H 20) but significant increases for montmorillonite for some
immersion liquids. It is probable that under somewhat different condi-
tions and perhaps with different liquids, halloysite material with only
11 Larsen, E. S., and E. T. Wherry, Leverrierite from Colorado, J. Wash. Acad. Sci.,
7,208-217 (1917).
12 Van Baren, F. A., Ueber den Einfiuss verschiedener Flussigkeiten auf den Brech-
ungsindex von Tonmineralien, Z. Krist., 95, 464-499 (1936).

TABLE 35. INDEX OF REFEAC'l'ION IN RELATION TO IMMEESION LIQ1TID
(After Van Baren 12)
Mont-
Mont- moril-
Kao- Hal- Non-
Liquid and index of liquid linite loysite tronite morillonite
lonite (HCI
treated)
- - - - - - - - - - - - - - - - - 1 - - - - - - - - - ---- - - -
Mean index of mineral., ,,", ..... . 1.57 1.575 1.565 1. 50 1.50
Chlorobenzene, 1.524 ........... . + + +
Ethylenebromide and ehlorobenzene, 1.535 .. + + +
Benzylchloride, 15.41 .............. . + + +
Benzylamine, 1.546 ................. . + + + + +
o-Nitrotoluene, 1.547 ................... . + + +
Nitrobenzene, 1.552. . ........ . + + +
Monobromobenzene, 1.560 ........ . + + +
Xylidine, 1.560 ................... . + + + + +
o-Anisol, 1.562 ................... . + + +
o-Toluidine, 1.572 ........... : ..... . + x + +
Anisidine, 1.576 .................. . + x +
Aniline, 1.585 ................... . + +
Bromoform, 1.596 .......... .
Iodobenzene and bromobenzene, 1.605 ..
Iodobenzene and bromobenzene, ] .616 .
Chinoline, 1.624 .................... , ' + x x
cc-Tetrahydro-ar-napthylamine, 1.631, , , ?
+ =index of mineral higher than liquid

=index of mineral lower than liquid
= index of mineral and liquid the same
x = index of mineral increased but not to a value higher than liquid
slightly more than 2H 2 0 would show some variation in optical properties.

Van Baren concluded that organic molecules with NH2 radicals are most
likely to cause significant variations in optical characteristics and that the
time necessary for the liquid to obtain its maximum influence increased
as the size of the organic molecule increased. He showed further that the
indices of a montmorillonite, when immersed in a liquid of higher index
than the highest index of the mineral, sometimes increase to even higher
values than that of the liquid itself. That is, the indices of the mineral
may not only increase to equal that of the liquid but may become higher
I than that of the liquid. Thus a montmorillonite with a mean index
of 1.50 increased to a value above that of benzylamine (1.546) when
immersed in this liquid.
VendeP3 has recently studied in detail the characteristics which are
13 Vendel, M., Zur Bestimmung der Liehtbrechung silikatischer Tonminerale,

Chem. Erde, 15,325-370 (1945).
1
288 Clay Mineralogy
likely to be found in liquids causing a variation in indices of the clay min
erals. According to him, a liquid should have the following properties tc
permit a determination of the true optical values of the clay minerals:
1. The liquid must not chemically attack the clay mineral.
2. It must not be soluble in water, and water must not be soluble in it,
otherwise a change in hydration state might develop with an attendant
change in optical values.
3. It must not enter into any ion-exchange reaction with the clay.
4. It must not permit the dissociation of any of the adsorbed ions of the
clay.
5. It must be nonpolar.
6. Its index must be approximately that of the mineral.
7. If a mixture is used, the components must be completely miscible
and must have nearly the same boiling point.
8. The liquids should be well-defined compounds.
Montmorillonite, halloysite, and possibly vermiculite to a limited
extent (see Chap. 10) have the property of adsorbing certain organic mol-
ecules on their basal plane surfaces. The types of organic molecules
adsorbed are also those which influence the optical properties. The var-
iation in optical properties is, therefore, undoubtedly largely a conse-
quence of this adsorption. The optical properties of these minerals could
also be altered by certain liquids as a consequence of an effect on the
extent of interlayer hydration water. Illite, kaolinite, and chlorite min-
erals would be expected to show no variation with different immersion
liquids, because of their very low adsorptive power for organic molecules
and because of their lack of interlayer water. For the same reasons the
attapulgite-sepiolite minerals also should show no effect from immersion
liquids except in the rare cases when an organic molecule is of the correct
size and shape to enter the channels of their structure.
This attribute of the clay minerals requires that the index liquids used
in their study must be carefully chosen, and, in general, when values are
published, the liquids in which they were determined should be recorded.
Obviously, the influence of immersion liquids on certain clay minerals and
not on others can be of diagnostic value in the identification of the clay
minerals.
ORIENTED-AGGREGATE TECHNIQUE
When the flake-shaped clay-mineral particles settle from a deflocculated

clay-water-suspension on a horizontal surface, they tend to come to rest
with one flake on top of another so that their basal plane surfaces are sub-
stantially parallel. Frequently successive flakes assume positions in
which there is a considerable degree of parallelism of the a and b crystal-
lographic axes. In other words, an aggregate orientation closely approx-
"' ..
imating crystal growth may take place in the accumulation of such mate-
rial under the foregoing conditions.
Grim 14 in 1934 described a technique taking advantage of the above
phenomenon for the preparation of oriented aggregates for the determina-
tion of the optical properties of the clay minerals. In this technique, a
clay-water suspension is allowed to stand until some of the clay mineral
has accumulated on a flat surface, either the bottom of a vessel or a glass
plate hung in a horizontal position in the suspension. The flat surface
with the accumulated clay is then removed from the suspension and dried.
Aggregate particles are sliced or cut from the dried film of clay for mount-
ing in index-of-refraction liquids.
The Ci and 'Y indices and the birefringence of such aggregates can usually
be determined with considerable accuracy. Frequently the aggregates
yield reasonably good interference figures, from which the sign of the
mineral and the order of magnitude of 2V can be determined. Measure-
ment of the latter value implies considerable orientation paral1el to the
a and b crystallographic axes.
Without the use of oriented aggregates, it is often possible to determine
only a mean index of refraction from the more or less randomly oriented
clay-mineral particles in the crude clay. 111e birefringence and other
optical values frequently cannot be even approximated in the crude clay.
INTERLAYER MIXTURES
It has been pointed out (see pages 79-80) that clay materials may be
composed of several clay minerals mixed primarily by interstratification
on substantially unit-cell dimensions. Such materials may be composed
of aggregates made up of alternations of one or a few unit layers of the
component clay minerals. Also in some clay materials the clay-mineral
units are interlayered with ferric hydroxide units.
Optical data from such intimate interlayered mixtures can easily be
mistaken for that of a single mineral. Thus, Grim and Rowland 15 showed
that a sample from N amiquipa, Mexico, previously described as beidellite
on the basis of optical values, was actually a mixture of halloysite and
limonite. X-ray and differential thermal data left no doubt regarding
this interpretation, whereas the optical values were suggestive of a single
clay mineral of the montmorillonite type.
An example such as that above emphasizes the fact that clay-mineral
determinations based solely on optical data must be made with great
14 Grim, R. E., The Petrographic Study of Clay Minerals-A Laboratory Note, J.
Sediment. Petrol., 4, 45-46 (1934).
15 Grim, R. E., and R. A. Rowland, Differential Thermal Analysis of Clay Minerals
and Other Hydrous Material, Am. Mineral., 27, 746-761 (1942).
290 Clay Mineralogy
caution to avoid gross errors. Frequently the character of the interfer-
ence figure reveals whether a clay-mineral aggregate is made up of a single
mineral or a mixture. If the interference figure is not good and the
isogyres are not reasonably complete, there is probably more than one
mineral present. Homogeneity and parallelism in the a and b directions
are required for fairly complete figures of the layered clay minerals.
ORIENTATION IN AN ELECTRICAL FIELD
Marshall 6- 16 has studied in considerable detail the tendency of clay-

mineral particles to orient themselves in an electrical field, and the fol-
lowing statements are taken largely from his publications.
Freundlich, Kruyt, and their coworkers 17 showed that extremely non-
equidimensional particles, such as needle-shaped particles, are oriented
in a water suspension in an electrical field relative to the electrical lines of
force. Procopiu 18 showed that the prime cause of orientation in suspen-
sions of various minerals was the electrical anistropy of the particles them-
selves, rather than hydrodynamic effects. Thus, irrespective of the shape
of the particles or the hydrodynamic conditions of flow, a turning moment
operates when an electric field is applied, until the direction of greatest
dielectric constant becomes parallel to the lines of force.
Marshall 6 ,16 has measured the birefringence developed in suspensions of
various clays when subjected to an electrical field, and some typical results
for fractions of a soil listed as beidellite, containing a large proportion of
montmorillonite, are given in Table 36. It can be seen from the data in
Table 36 that the electrical birefringence is affected by the exchangeable
cation. No such variation was found for kaolinite, which is evidence, as
Marshall pointed out, for the exterior position of the exchange cations in
kaolinite and their internal position in montmorillonite.
Marshall 16 found that dilute suspensions of lithium and sodium mont-
morillonites sometimes gave a negative instead of a positive birefringence.
Recently Muller 19 ,2o and N orton 21 have found that, at low frequencies,
negative values exist for montmorillonite suspensions and, for high fre-
16 Marshall, C. E., "The Colloidal Chemistry of the Silicate Minerals," Academic

Press, New York (1949).
17 Freundlich, H., H. R. Kruyt, et al., See C. E. Marshall, "The Colloidal Chemistry
of the Silicate Minerals," Academic Press, New York (1949).
18 Procopiu, S., Sur la birefringence 6lectrique et magn6tique des suspensions, Ann.
phys., 1, 213-223 (1924).
19 Muller, H., Electro-Optical Effects in Colloids, Phys. Rev., 55, 508 (1939).
20 Muller, H., and B. W. Sakmann, A New Electro-Optical Effect in Colloids, Phys.
Rev., 66, 615-616 (1939).
21 Norton, F. J., Influence of Frequency on the Electro-Optical Effects in Colloids,
Phys. Rev., 66, 668-669 (1939).
quencies, the values are positive. The lower the concentration of the
clay, the lower the range of frequencies for which negative values were
observed. These authors conclude that the negative values are due to
the presence of long-range interparticle forces, the positive ones being
normal for an orientation like that of the Kerr effect (see Marshall 16 ).
Whiteside and Marsha1l 22 have shown that, at 4,000 volts/cm with 60-
cycle a-c current, the saturation values for Na and K clays can be
obtained, and, for the Putnam ma.terial (beidellite), the Na and K clays
are very significantly lower in birefringence than the hydrogen clay.
Their data indicate that the optical and electrical ellipsoids are similar in
character.
TABI,E 36. ELECTRICAL BIREFRINGENCE OF PUTNAM (BEIDEI,LITE) CLAY FRACTIONS
(After Marshall'6)
Birefringence at 800 volts/em, 60 cycles a-c

Exchange
cations
500-200 mIL 200-] 00 mIL 100 mIL
~----- - - - - - - - - - _ - - -------_-
H+ 0.023 0.029 0.026
Ca++ 0.026 0.0:31 0.031
K+ 0.023 0.020 0.011
FORM BIREFRINGENCE
Wiener23 showed that hirefringence ("form birefringence") may arise

in a system in which there is a parallel arrangement of plate- or rod-shaped
particles when these particles are separated by a medium of different
refractive index. There is no form birefringence when the solid and the
liquid have the same index of refraction. The birefringence increases as
the difference between the indices of the liquid and solid increases.
Wiener developed his concepts for isotropic particles, but Ambronn and
Frey,24 who have studied this phenomenon in great detail, have shown
that the effect can be superimposed on systems which contain birefrin-
gent components. Thus, the inherent birefringence of the clay-mineral
particles would augment the form birefringence developed in suspensions.
In the case of the clay minerals, form birefringence would arise pri-
marily because of the penetration of water or an organic liquid between
22 Whiteside, E. P., and C. E. Marshall, Mineralogical and Chemical Studies of the
Putnam Silt Loam Soil, Call. Agr. Univ. Missouri Research Bull. 386 (1944).
23 Wiener, 0., Theory of Reaction Constants, Abhandl. math.-phys. Klasse sachs.
Akad. Wiss. (Leipzig), 32,256-276 (1912).
24 Ambronn, H., and A. Frey, "Das Polarizationsnikroskop," Akl1demische Verlags-
gesellschaft m.b. H., Leipzig (1926).

292 Clay Mineralogy
the unit layers of the expanding clay minerals. Thus, under the usual
conditions for the measurement of optical properties, it might be signif-
icant for montmorillonite, vermiculite, and halloysite but probably not
for the other clay minerals. Where the ratio of volume of liquid to solid
is small, the effect would be small, but when considerable fluid has been
adsorbed between layers, it might be significant. Undoubtedly part of
the variation of the optical characteristics of montmorillonite and hal-
loysite minerals reported in the literature is due to form birefringence.
Obviously the phenomenon of form birefringence must be kept in mind
in any optical study of fine-grained particles in a liquid medium. Unfor-
tunately this has not always been the case. /
(
DISCUSSION OF THE APPLICATION OF OPTICAL METHODS IN

CLAY-MINERAL STUDIES
The value of optical methods in studying the clay minerals has been the
subject of much discussion. Some students of clays assert that optical
methods are worthless for such fine-grained materials, whereas others
plead for their universal use and base detailed clay-mineral determina-
tions solely on optical studies. In the writer's opinion, optical methods
are very frequently of value, but they must be used with caution, and it
must be realized that they are worthless for some types of materials. An
optical examination can be made in a very short time, and the chance of
obtaining valuable data usually justifies the expenditure of the required
time. Also, after some experience has been gained in the study of clay
minerals under the microscope, more can be done with them than would
be supposed at first. It is usually possible to determine with fair accuracy
at least the indices of refraction and the birefringence of clay-mineral
samples. The optical properties of the various clay minerals are suffi-
ciently diverse so that an idea of the identity of the clay minerals usually
can be gained from such data. As noted previously (page 289), caution
must be used because of anomalous results possible from interlayer
mixtures.
Certain of the clay minerals frequently have distinctive appearances
under the microscope which aid in their identification. Thus, aggregates
of the montmorillonite clay minerals often appear as a large single crystal
that has been strained. The individual particles of montmorillonite
either cannot be seen or else are indistinct and tend to merge into each c-
other. The attapulgite-sepiolite minerals sometimes have a somewhat
similar appearance. Aggregates of the other clay minerals have the
appearance of being made up of discrete particles, even though the indi-
vidual particles are less than 1 micron in diameter, and the particles have
a high degree of uniform orientation. With a little experience one can
usually recognize relatively pure montmorillonite quite easily under the
microscope by its appearance without reference to optical values.
The presence of nonclay minerals, such as quartz, calcite, and feldspar,
may cause considerable difficulty if these minerals are present in particles
of about the same size as the clay minerals. An abundance of extremely
fine quartz or calcite makes optical methods valueless in many cases.
Sometimes the presence of such nonclay minerals will produce an unusual
appearance or strange optical characteristics which suggest that some
such material might be present. However, in many cases, the nonclay
material would go undetected by optical methods.
The presence of even relatively small amounts of extremely fine-grained
carbonates makes optical study of the clay minerals impossible, because
of the high birefringence of these nonclay minerals. Further, it is not
possible to remove the carbonates in all cases by dissolving them, because
of the susceptibility of some of the clay minerals to acid attack. Certain
of the clay minerals (see pages 296~298) are relatively insoluble in acid, but
others are quite soluble. Unfortunately many calcareous materials also
contain the more soluble clay minerals (see page 359), so that in a study
of such material it is necessary to determine the clay-mineral composition
by nonoptical methods.
Organic material and ferric iron compounds may mask the optical char-
acteristics of the clay minerals. The organic material can frequently be
removed by solvents or by mild oxidants, such as hydrogen peroxide,
without altering the clay minerals. Some organic material, particularly
that in the older sediments, is not affected by such reagents and can be
removed only by vigorous oxidizing agents or by heating to elevated tem-
peratures. Unfortunately, if there is iron present, it may be oxidized in
the process, so that it may then conceal the characteristics of the clay
minerals.
Sometimes the free iron compounds can be removed by reducing, e.g.,
by hydrogen sulfide, and then dissolving the ferrous iron. Again caution
must be used because of the danger of attack on the clay minerals, partic-
ularly those of the montmorillonite and attapulgite-palygorskite groups,
by the reducing agents.
Optical methods, of course, are of great value in determining the non-
clay minerals present in a clay material. In general, minerals found in
particles as small as 5 microns in diameter can be identified. If the min-
eral has particularly distinctive optical properties, considerably finer par-
ticles can be identified. Thus with a little experience carbonates can be
spotted in particles at least as small as 1 micron. Quantitative deter-
minations of such nonclay minerals are, however, very difficult for par-
ticles smaller than about 5 microns.
The petrographic microscope is, of course, useful in studying the texture
294 Clay Mineralogy
of clays as revealed by thin sections. Examples are the shards which are
characteristic of many bentonites and the parallel orientation of the clay
minerals in many shales. The textures of clay materials are not included
in the objective of this text and will not be considered. The identification
of the clay minerals in thin sections is usually difficult and often impos-
sible. Optical methods of identification can be applied best by the use
of immersion techniques and prepared, oriented aggregates.
Caillere, S., S. Henin, and S. Meriaux, Xylotite, a Ferrous Variety of Sepiolite, Compt.
rend., 227, 855-856 (1948).
Edsall, J. T., Streaming Birefringence and Its Relation to Particle Size and Shape,
Advances in Colloid Sci., 1, 269-316 (1941).
Grim, R. E., Properties of Clays, "Recent Marine Sediments," pp. 466-496, American
Association of Petroleum Geologists (1939).
Hendricks, S. B., and C. S. Ross, Chemical Composition and Genesis of Glauconite
and Celadonite, Am. Mineral., 26, 683-708 (1941).
Hofmann, U., and W. Bilke, Ueber die innerkristalline Quellung und das Basenaus-
tauschvermiigen des Montmorillonits, Kolloid-Z., 77, 238-251 (1936).
Marshall, C. E., Mineralogical Methods for the Study of Silts and Clays, Z. Krist.,
90,8-34 (1935).
Ross, C. S., and P. F. Kerr, The Kaolin Minerals, U.S. Geol. Survey Profess. Paper
165E, pp. 151-180 (1931).
Ross, C. S., and P. F. Kerr, Halloysite and Allophane, U.S. Geol. Survey Profess.
Paper 185G, pp. 135-148 (1934).
CHAPTER 12
Miscellaneous Properties
SOLUBILITY OF THE CLAY MINERALS
General Statement. The solubility of the clay minerals in acids and

alkalies is important fundamentally because it reveals certain attributes
of the clay minerals and certain differences between them which are not
obvious from other methods of study. From a practical standpoint, sol-
ubility characteristics are important in determining the utility of various
clay materials as sources of metallic aluminum and for the manufacture
of alum. Such data are also of practical value in numerous other ways,
e.g., in the bonding of refractories by add media and in the fixation of
certain materials such as phosphates in soils. The solubility character-
istics of the clay minerals are important in determining the limitations of
acid treatment and electrodialysis as auxiliary tools in the analyses of com-
plex clay materials containing carbonate, and in determining their ion-
exchange capacity.
Nutting,l who has studied the solubility of the clay minerals in great
detail, arrived at the following general conclusions:
The solubility of a number of clays in acids and alkalies of a wide range of con-
centrations indicates that the solution process is of at least three kinds for every
clay, dependent on the concentration of solvent.
At equilibrium, a fract,ion of the acid or alkali remains free, and a fraction of
each clay remains undissolved regardless of the proportions present. Hence the
reactions appear to be reversible as indicated by varying equilibrium conditions.
Over a range of low acid concentrations, 0.05 to 0.2 normal, the" solution" is
essentially a silicate hydrosol similar in composition to the clay dissolved. At
higher concentrations it contains also salts in solution; at lower concentrations
excess silica. At acid concentrations of 20 per cent and over, bases but no silica
go into solution and no hydrosols are formed.
Over a range of low alkali concentrations, 0.002 to 0.005 normal (0.02 to 0.05
per cent) sodium carbonate, the solution essentially is an alkali silicate hydrosol,
only silica being removed from the clay. At concentrations of alkali below the
critical range, no carbonate is left in solution.
With montmorillonite clays the critical anion concentration is around 0.1 mole
1 Nutting, P. G., The Action of Some Aqueous Solutions on Clays of the Montmoril-
lonite Group, U.S. Geol. Survey, Profess. Paper 197F, pp. 219-235 (1943).
295
296 Clay Minera]ogy
per liter while the critical cation (Na+) concentration is around 0.001 mole per
liter. Within the critical acid range, the anions may be considered as competing
on substantially equal terms with silicic acid for the bases of the clay. Within
the critical alkali range, the alkali in solution is competing on equal terms with
the bases of the clay for the silica.
In general, above certain minimum concentrations, acids remove alkali
metals, alkaline earths, iron, and aluminum from the clay minerals, and
alkalies dissolve the silica. The literature contains a large amount of
information on acid solubility, because of the general economic importance
of the acid-clay reaction, as noted previously, but relatively little data on
solubility in alkalies.
Solubility of Clay Minerals in Acid. The solubility of the clay min-
erals in acids varies with the nature of the acid, the acid concentration,
the acid-to-clay ratio, the temperature, and the duration of treatment.
Also the solubility of the various clay-mineral groups is quite different,
and there is great variation in solubility characteristics of members of
some individual groups. Thus, in general, a magnesium-rich montmoril-
lonite is much more soluble than one that is rich in aluminum, with an
iron-rich member somewhere in between. In the case of the clay min-
erals showing variations in the degree of crystallinity, such as kaolinite,
the solubility increases as the degree of crystallinity decreases. The sol-
ubility would, of course, also increase as the particle size decreases.
Calcining the clay minerals changes their solubility characteristics and
their relative solubility with respect to each other.
Pask and Davies 2 (Table 37) have shown, using 0.5 g of clay dried at
130C in 30 cc of 20 per cent solution of H 2S0 4 and boiling for Y2 hr, that
only 3 per cent of the total alumina is dissolved from kaolinite and only
9 per cent from anauxite. Halloysite is moderately soluble, showing solu-
tion of from 50 to 90 per cent of the total alumina under similar condi-
tions. Illite is slightly soluble, with 11 per cent of total alumina going
into solution in one sample. Montmorillonite ranges from low to high
solubility, showing 33 to 87 per cent of total alumina dissolved. The
attapulgite-sepiolite clay minerals would probably have characteristics
similar to those of the montmorillonite minerals.
With acid of the same concentration and a similar acid-to-clay ratio,
but with digestion under pressure at 155C, the solubility of all the clay
minerals increased (Table 37). All the alumina of halloysite became sol-
uble, and almost all of it was dissolved from the montmorillonites (85 to
93 per cent). Even the ordinarily relatively insoluble kaolinite and illite
lost the major part of their alumina, 70 and 87 per cent, respectively,
going into solution. '
2 Pask, J. A., and B. Davies, Thermal Analysis of Clays and Acid Extraction of
Alumina from Clays, U.S. Bur. Mines Tech. Paper 664, pp. 56-78 (1945).
Miscellaneous Properties 297
TABLE 37. EXTRACTION OE' ALUMINA FROM VARIOUS CLAY MINERALS WITH SULFURIC
ACID
(From Pask and Davies')
Aluminum extracted, per cent total aluminum present
Clay mineral
Clay dried Clay calcined Pressure digestion
at 130C* at 800C* at 155ct
Kaolinite ............... .. 3 100 70

Anauxite .... .......... ... . 9 95
Halloysite .............. ... 63 100 98
50 96 I
90 I
I
78 100 100
Illite ..................... 11 52 87 i
Montmorillonite ........... 87 28 85 !
33 19 93
Muscovite ... . . . . . . . . . . . . . 17 29
* Determinations made after boiling 0.5 g of sample in 35 cc of 20 per cent solution

of sulfuric acid for 1 hr.
t Clay dried at 130a C, then digestion of 0.5-g sample in 35 cc of 20 per cent solution
of sulfuric acid for one hour.
Under similar conditions of extraction, but after calcining the clay

minerals to 800C, all or substantially all the alumina of the kaolinite,
anauxite, and halloysite is soluble, whereas the solubility of the alumina
in the illite is increased only moderately, and for the montmorillonites it
actually decreases (Table 37). The explanation resides in the fact that
the kaolinite and halloysite lattices are probably destroyed or at least con-
siderably disrupted 'at this temperature, and no new high-temperature
phases have yet formed. In the case of illites and montmorillonites
the lattice has been changed slightly but not destroyed; hence the sol-
ubility of the alumina has been changed, but not so that it can be com-
pletely removed. When kaolinite is heated to still higher temperatures
(975C), new crystalline phases develop, and the solubility of the alumina
decreases. Similarly, when the other clay minerals are heated to near
this temperature (975C), new high-temperature phases form, which
would affect the solubility in acid. In the case of some clay minerals,
.. such as the montmorillonites, the high-temperature phase forming (see

page 225) varies considerably with substitutions in the original mont-
morillonite lattice, and correlative variations in solubility in acid would
be expected.
A comparison of the data obtained by Thiebaut 3 with those of Pask
3 Thiebaut, J. L., "Sediments argilo-calcaires du bassin de Paris," Nancy, France
(1925).
298 Clay Mineralogy
2
and Davies show the great variation in results obtained under different
treatment conditions. Using 50 per cent hydrochloric acid, a treating
temperature of 80 to 85C, a treating time of 2 hr, and clay dried at 105
to 108C, Thiebaut found that the solubility of biotite was 100 per cent,
muscovite 5 to 32 per cent, kaolinite 10 per cent, halloysite 6 to 15 per
cent, and montmorillonite 62 per cent, based on cation solubility. Using
sulfuric acid of similar concentration and the same minerals but evaporat-
ing to dryness, Thiebaut found that the solubility of all the above minerals
was complete, again based on their cation solubility. The work of
Thiebaut 3 indicates the generality that the clay minerals are more soluble
in sulfuric than in hydrochloric acid, and also that for hydrochloric acid,
both kaolinite and halloysite are less soluble than the three-layer clay
minerals. Biotite occupies an anomalous position in being about equally
soluble in both acids.
The solubility of the clay minerals in other acids is not well known.
However, investigations of anion exchange indicate that the solubility is
likely to be considerable, especially for the acids with an anion having a
size and geometry approximating that of the component parts of the clay-
mineral lattice. As a consequence, some relatively weak acids may
strongly attack certain clay minerals. Murray4 has shown, for example,
that phosphoric acid attacks the kaolinite lattice under some conditions
with greater vigor than sulfuric acid.
Wolf5 has presented interesting data showing the effect of concentration
of acid on solubility. Using 10 g of Zettlitz kaolinite and cooking for
2 hr, he found that 0.02 N HCI dissolved 3.1 mg of Al 2 0 3 in 100 cc of acid;
that in similar amounts of acid 0.5 N HCI dissolved 53.5 mg of AbOa;
and that 5 N HCI dissolved 124.4 mg of AI 20 a.
Nature of the Acid Reaction. In the attack of acid on the mont-
morillonite clay minerals and probably also on the illite and sepiolite-
attapulgite clay minerals, it appears that the alkalies and alkaline earths
are removed relatively more rapidly than the aluminum or iron, and that
the iron is removed more rapidly than the aluminum. Hofmann and
Endell,6 Glaeser,7 and Mering 8 have shown that 75 to 85 per cent of the
total alumina must be removed from the montmorillonite lattice before
it is completely destroyed. Acid attack begins around the edges of the
4 Murray, H., The Structure of Kaolinite and Its Relation to Acid Treatment,
Ph.D. thesis, University of Illinois (1951).

5 Wolf, L., Zur Chemie des Kaolins, Ber. deut. keram. Ges., 14,393-403 (1933).
6 Hofmann, U., and K. Endell, Mitteilung tiber die Aktivierung der rohen Bleich-
erde, Angell'. Chem., 48, 187-192 (1935).

7 Glaeser, R., Effect of Acid Treatment on the Base-Exchange Capacity of Mont-
morillonite, Campt. rend., 222, 1241-1242 (1946).

8 Mering, J., Reactions of Montmorillonite, Bull. soc. chim. France, pp. 218-223
(1949).
particles and works inward. In relatively dilute alkaline solutions, the
solubility of the silica residue of the structure left around the edges of the
particles is increased as a consequence of the attack by acids.
In the case of aluminum it appears that its removal is stepwise; i.e., it
moves first from octahedral positions to exchange positions and then to
'complete solubility. If the sample is dried before all the aluminum is
removed, at least some of it appears to move back from exchange positions
to octahedral positions.
. Brindley and Youell 9 have recently presented data for a chlorite (pen-
ninite) indicating that the aluminum in octahedral coordination is more
soluble in HCI than the aluminum in tetrahedral coordination. These
authors do not suggest that relative solubility can be used as a method
for the general determination of octahedral versus tetrahedral aluminum
in the clay minerals.
Murray4 has shown that under similar conditions of treatment the
actual sulfate or phosphate compound resulting from treatment with
sulfuric and phosphoric acid, respectively, is almost the same for all
kaolinites and halloysite, regardless of variations in degree of crystallinity.
There were, however, some differences which suggested to him that the
structural attributes of the parent clay mineral did exert some influence
on the nature of the resulting reaction product.
Decomposition by Electrodialysis. Numerous investigators have
indicated that electrodialysis of certain of the clay minerals may cause
their decomposition. Thus Kelleylo has shown that, as the cations are
replaced by H+, aluminum moves from octahedral positions to exchange
positions, and Hofmann and Giese l l have shown that, in general, unex-
changeable cations are lost from within the lattice before all the exchange-
able cations are replaced by H+.
The amount of loss of unexchangeable cations during electrodialysis
varies with the clay mineral. Magnesium-rich montmorillonite minerals
seem to be most susceptible, and Nutting 12 has shown that all the magne-
sium can be removed from hectorite by electrodialysis with complete
breakdown of the structure. Caldwell and Marshall 13 have shown that
9 Brindley, G., and R. F. Youell, A Chemical Determination of Tetrahedral and
Octahedral Aluminum Ions in a Silicate, Acta Cryslallog., 4, 495-496 (1951).

10 Kelley, W. P., The Agronomic Importance of Calcium, Soil Sci., 40, 103-109
(1935).
11 Hofmann, U., and K. Giese, Ueber den Kationenaustausch an Tonmineralien,
Kolloid-Z., 87, 21-36 (939).
12 Nutting, P. G., A Study of the Association of Magnesia with Silica in a Pure
Magnesium Clay, J. Wash. Acad. Sci., 30,233-237 (1940).

Properties of the Clay Minerals Nontronite, Attapulgite, and Saponite, Call. AliT.
Univ. Missouri Research Bull. 354 (1942).
300 Clay Mineralogy
the same results are obtained with saponite. Other montmorillonites are
quite susceptible to such alteration, but less so than saponite and hector-
ite. The iron-rich montmorillonites are more susceptible than those high
in aluminum.
Attapulgite-sepiolite, vermiculite, chlorite, and some of the biotite
micas are quite susceptible to disintegration by electrodialysis. Roy14
has shown that biotite suffers appreciable loss of cations after only a few
hours of electrodialysis and that 80 to 90 per cent of all of them are lost at
the end of 28 days, without, however, complete destruction of the biotite
structure. After a similar length of time, muscovite and phlogopite
micas lost only a small amount of their iron and alkalies and substantially
no aluminum. It appears that kaolinite is relatively little affected by
electrodialysis, and data are not at hand for halloysite.
Electrodialysis has frequently been used to prepare H clays for cation-
exchange studies and for use as the starting point for investigations of the
physical properties of clays carrying specific exchangeable cations.
Because of the likelihood of significant amounts of disintegration of the
clay mineral, the procedure must be used with caution. Many of the
investigations which have used this procedure are of little value, because
the clay minerals were altered, and the cation composition was not what
it was thought to be, as a result of very difficultly replaeeable aluminum
moving from the lattice to exchange positions.
Cation Liberation by Neutral Salts. M ukherj ee and his colleagues 15,16
have shown that AI3+ and Fe3+ may be removed from certain hydrogen
clays by repeated treatment with BaCb. They determined the amount
of H+, Aj3+, and Fe3+ in the leachate after leaching the clay minerals with
BaCh The clay minerals had been rendered acid by prior treatment
with 0.02 N HCl. After each leaching with BaCI 2 , the clay minerals were
rendered acid again before repeated leaching with BaCI 2 These inves-
tigators found that some AI3+ and Fe H continued to come out of the lat-
tice of montmorillonite through at least eight cycles of acid treatment
and leaching, whereas for kaolinite there was little effect after the first
leaching. The removal of the Aj3+ and Fe3+ from the montmorillonite
was accompanied by a reduction in cation-exchange capacity, whereas
the capacity of kaolinite was substantially unaffected. The work of
Mukherjee shows the extreme sensitivity of some of the clay minerals to
chemical treatment.
14 Roy, R., Decomposition and Resynthesis of the Micas, J. Am. Ceram. Soc., 32,
202-210 (1949).
15 Mukherjee, J. N., B. Chatterjee, and B. N. Baverjee, Liberation of H+, AJ+++
and Fe+++ from Hydrogen Clay by Neutral Salts, J. Colloid Sci., 2, 247-256 (1947).
16 Mukherjee, J. N., B. Chatterjee, and A. Roy, Liheration of 11+, Al+++ and Fe+++
from pure Clay Minerals on Repeated Soil Treatment and be saturation, J. Colloid
Sci., 3, 437-446 (1948).
,...
Rational Analysis. It is frequently desirable, particularly in the indus-
Grial use of clays, as in the ceramic industry, to have some rapid, simple
method of determining variations in the composition and properties of
clays. Before the development of X-ray-diffraction, differential thermal,
and improved optical methods about 1930, the only methods available
were chemical procedures based on the difference in the resistance of var-
ious clay substances to chemical attack. There developed, therefore,
so-called" rational methods of analysis" based primarily on the difference
in solubility of various clay substances in acids. Such methods were
widely used, e.g., in the ceramic industry, for raw-material control.
Most methods of rational analysis were based on the difference in sol-
ubility of clay components in hot concentrated hydrochloric and sulfuric
acids. The part of a cl~y soluble in hydrochloric acid was frequently
called allophaneton, or allophane clay, and the part insoluble in hydro-
chloric but soluble in sulfuric acid was called kaolinton, or kaolin clay.
The allophaneton often was largely the montmorillonite portion of a clay,
together with other components soluble because of extremely fine grain
size. The kaolinton was likely to include primarily the kaolinite and
illite components. Frequently the method was elaborated to obtain
some determination or estimation of the potash content as a basis for an
evaluation of a possible mica component. Also in some instances a value
for silica was obtained to indicate the quartz content. The wide varia-
tions in the methods of rational analysis that have been used are sum-
marized by Harkort and Harkort. 17
Washington 18 and later Correns 19 have pointed out the great fallibility
of rational analytical procedures because of the tremendous range in sol-
ubility of the components of clays as a consequence of particle-size varia-
tions. Since about 1930, with the development of other analytical meth-
ods, the use of rational analysis has gradually decreased. It should be
pointed out, however, that such methods may be quite satisfactory for a
specific limited purpose if used with caution and only for a limited range
of clays. Thus an industry using only a particular kind of clay can set
up an arbitrary set of chemical tests, which would be rapid and simple, to
check certain variations in the material supplied to them.
Solubility of Clay Minerals in Alkalies. Except for the investigations
of Nutting, 1 there has been very little study of the solubility of the clay
minerals in alkalies. Nutting has studied particularly the solubility of
17 Harkort, H., and H. J. Harkort, A Rapid Rational Analysis, Sprechsaal, 66,705-

707, 723-726, 739-741 (1942).
18 Washington, H. S., The Calculation of the "Rational Analysis" of Clays, J. Am.
Ceram. Soc., 1, 405-421 (1918).

19 Correns, C. W., Ueber die Bestandteile der Tone, Z. deut. geol. Ges., 86, 706-711
(1933).
302 Clay Mineralogy
montmorillonite and attapulgite in dilute solutions of sodium carbonate.
His results for montmorillonite, given in Table 38 and Fig. 112, show that
TABLE 38. SILICA DISSOLVED FROM MONTMORILLONITE (WYOMING BENTONITE) BY
SOLUTIONS CONTAINING SODIUM CARBONATE
(After Nutting!)
Per cent Na 2 CO a........... 0.0 0.005 0.01 0.02 0.05 0.1 0.2 0.5 1.0
Si0 2 in solution, g lliter* .... 0.0640.0670.1320.1780.133 0.2600.397 0.5980.700
* After digestion of 48 g of clay in 2-liter solutions for 48 hr at 90C.
for this clay mineral the amount of silica removed in solution increases as
the concentration of sodium carbonate increases to a maximum atabout
.8
-~
:.J
....
1/
2i .6
<n
E
o....
/
L?
.<:: .4
V
"0
<IJ
>
'0
/
v
<n
<n J
(5 .2
./
v- i'(:/
v
o
o .005,o1.Q2.05 .2 .5 1.0 2
Per Cent Sodium Corbonote in Two Litre Solutions
FIG. 112. Solubility of silica from montmorillonite (Wyoming bentonite) in sodium
carbonate solutions, after Nutting.!
0.025 per cent sodium carbonate concentration. At higher concentra-

tions, the solubility decreases to a minimum at about 0.05 per cent sodium
carbonate and then rises steadily to 0.70 g/liter at 1 per cent sodium car-
bonate concentration. According to Nutting, only silica is dissolved by
the alkali up to a concentration of 0.05 per cent sodium carbonate. He
points out that an extremely dilute solution of alkali is very effective in
removing silica from montmorillonites and, given enough time, probably
could remove all of it. ..
Additional data for the removal of silica from montmorillonite by
sodium carbonate are given in Table 39, together with comparable data
for attapulgite. The results for the two clay minerals are much the
same, except that the relative amount of silica dissolved is lower for the
attapulgite.
Miscellaneous Properties
.303
TABT,J<; 39. MILLIGRAMS OF SILICA DISSOLVED FROM 1 G OF CLAY IN VARIOUS
AMOUNTS OF WATER CONTAINING 1 G OF SODIUM CARBONATE
(After Nutting!)
Water, Montmorillonite Attapulgite

liters (Wyoming bentonite) (Quincy, Florida)
10 169 32
5 481 253
2 215 192
1 104 79
0.5 70 87
0.2 86 50
0.1 100 59
INFRARED SPECTRA OF THE CLAY MINERALS
General Statement. Atoms which are grouped together in molecules

do not remain at rest but are continuously in vibration. Such vibrations
produce periodic displacement of atoms with respect to one another, caus-
ing a simultaneous change in interatomic distances. The frequencies of
the vibrations fall within the range of 10 13 to 10 14 cycles per second, which
is of the same order of magnitude as the frequencies of infrared radiations.
Vibrations which are accompanied by a change in dipole moment give
rise to the absorption of radiations in the infrared region of the electro-
magnetic spectrum. Several modes of vibration may occur for a par-
ticular atomic group, each at a characteristic frequency and normally
independent of other modes.
If molecules of a substance whose vibrations are accompanied by a
change of dipole moment are irradiated by a succession of monochromatic
bands of infrared, those radiated frequencies which correspond to the
intramolecular vibrational frequencies may be absorbed wholly or in part.
If the per cent of radiation which is absorbed by a substance is plotted
against the incident wavelength (or frequency), the ensuing graph may be
interpreted in terms of the intramolecular vibration. The graph, there-
fore, will be characteristic of the material and can be used in its identifica-
tion. In addition, it should also provide data on the structure and
bonding characteristics within the molecule.
Infrared absorption has been used extensively and successfully in the
study of organic compounds. Its utility in mineralogy and particularly
in clay mineralogy has not been fully explored. For details regarding
infrared absorption in general, the text of Barnes et al. 20 should be con-
20 Barnes, R. B., R. C. Gore, U. Liddell, and V. Z. Williams, "Infrared Spectros-
copy," Reinhold, New York (1944).

304 Clay Mineralogy
suIted, and for the application to clay mineralogy, the pioneer work of
Keller and Pickett 21 ,22 and Kerr and collaborators (see Adler et al. 23) should
be referred to. The statements herein are taken largely from these
publications.
The range of infrared used in absorption studies in general is 4,000 to
400 cm- 1 in frequency and 2.5 to 25 microns in wavelength. In the case
of clay mineral investigations, wavelengths of 2 to 15 microns are used,
since the characteristic spectra occur in this range. Wavelength is con-
ventionally expressed in terms of microns and frequency in terms of wave
numbers or reciprocal centimeters. The relationship is shown by
10- 4
1 J.I = 10- 4 cm or 1 cm- 1
Infrared absorption by powdered material cannot be measured satis-
factorily unless scattering and reflection of the radiation from the tiny
particles is reduced to a low value. It has been shown that the use of
particles smaller than the minimum wavelength of radiation used, about
2 microns, reduces these factors to a very low level and generally improves
results. Early work was done by mounting the clay in an oil medium
(Nujol), which caused difficulty by adding its own spectrum to that of the
clay. Currently techniques are available for mounting the clay without
any additive.
Data for Clay Minerals. Kerr and collaborators (see Adler et al. 23 )
present infrared absorption spectra obtained by different investigators in
different laboratories on the same series of clay-mineral samples. The
curves obtained by different investigators for the same clay-mineral sam-
ples have the same general characteristics but show considerable variation
in detail. Further investigation is required to show which of these minor
features of the curves are significant and which are spurious and due to
factors of technique. In Figs. 113 to 116 and Table 40 infrared absorp-
tion data for a variety of type clay minerals are given. These data are
taken from the work of Hunt, one of the contributors to the Report of the
American Petroleum Institute, 23 and they illustrate the general character-
istics of the absorption spectra for the various clay minerals.
Buswell and Dudenbostep4 have shown that the adsorption at 2.75
21 Keller, W. D., and E. E. Pickett, Absorption of Infrared Radiation by Powdered
Silica Minerals, Am. Mineral., 34,855-868 (1949).
22 Keller, W. D., and E. E. Pickett, Absorption of Infrared Radiations by Clay
Minerals, Am. J. Sci. 248,264-273 (1950).
23 Adler, H., E. E. Bray, N. P. Stevens, J. M. Hunt, W. D. Keller, E. E. Pickett,
and P. F. Kerr, "Infrared Spectra of Reference Clay Minerals," Rept. 8, American
Petroleum Institute Project 49, Columbia Cnivcrsity, New York (1950).
24 Buswell, A. M., and B. F. Dudenbostcl, Spectroscopic Studies of Base Exchange
Materials, J. Am. Chem. Soc., 63, 2554-2558 (1941).

Wove Number in CM.- 1
00 00 0
00 00 0 0000 0 0 0 oo
fiS~ g ~ 2
,~1
fiS~ g
."1-.;;:
~ ~
~ -
8 g to
lOa I, I, ,/, ,L
../ t"'\
80
~ fV'1}.
l..- V \
60
40
20 'cI
L V \
\
III f
\
/
-- Kaolinite>
Murfree>sboro
Arkansas
o IV
8 '~~~llIImIft:~;M
rig IWij\M~~n
~'~g_~5e
'~!-fml_i;::~
2 3 4 5 6 7 8 9 10 II 12 13 14 15 16
Wove Length in Microns
FIG. 113. Infrared absorption curves for kaolinites and halloysites, after Hunt (see
Adler et al. 23).
microns is due to unbonded OR, and that bonded OR caused absorption

at 2.7 to 3.2 microns and at 6.15 and 7.55 microns. The absorption at
about 6.1 microns shown by some of the clay minerals is believed due to
adsorbed water, and the absorption at about 9 microns is believed due to
the Si-O linkage. The octahedral alumina sheet is probably responsible
for the absorption at about 10 microns. The causes of many of the
absorption bands shown by the clay minerals are not yet understood.
306 Clay Mineralogy
:t
o
rl
..oOtr.l~~l.C
~~g~ ;;~~B:SS8
~,...-jo),-j Ci)..-ICi)!"""'I0),.....!
00
~~~~
<:.O~<:.C.,....,
~ e.:> ~ ..;< ~e.:>.stt:>

e.:><oe.:>e.:>
t-<OOOlC ~~t-R
C'le.:>C'le.:> C'le.:>C'lM
o
1.':
000
o
00
1.':
00
,...;
,...; 0>
,...;
.,
lC
,...; lC
00
00>
,...;
C';l ~ IYJ M 0
C'1t-Ol1';lGr;l~
e<:>OlCOo>O
O::,.....,Ci.l~OJ~
~~1.':~Cl~
M....... ~,......, Ci.l""""
oo.,.....!oo,......,oo,......,
o
"'0
t-
oolC
C'l e<:>
..,
Q)
'2
o
~
I=i
o
Z
308 Clay Mineralogy
Wove Number in CM.- 1
8 88 8 000 0 8 a
2<t ;;;S l(3 @ ~~g _ 8
100 I .1 .1 .1
" "I "I "
80 l.--+---1'-\ h I-
/ \ ! Dickite
60
fl.lr """
40
20 r
/
\ If
v
" Son Juanita
Mexico
./
f--'"
1\ /": I-' "'-'
/"
V'" '\
\ IJV
"" '" pyrop!'yllite
Robbins
Ill. Corolina
c: 20
1\/
tJ 0
100
ao / \
V\ I \ V""
60
40 /'
../'
\. ~ '"'" "'- Hectorite
Hector
California
20 V
,./'
\ /
o "
2 :3 4 5 6 7 8 9 10 II 12 13 14 15 16
Wave Length in Microns
FIG. 114. Infrared absorption curves for dickite, pyrophyllite, nontronite, and hec-
torite, after Hunt (see Adler et aP').
For example, the causes of variations in the spectra within the mont-
morillonite group cannot yet be related to variations of composition or
structure.
SURFACE AREA
Many methods have been tried to determine the surface area of clay
materials. The most trustworthy results have been obtained with the
Wove Number in CM,-I
00 00 0 0000 0
00 00 0 0 0 0 0 It)
0000 0
00
lD<t
OlD
r<lN
0
N 1Qst ~ ~ ::?
- Q 0
())
0
a:l
0
I'-
C\J
~
,I ,I, I
100
80
/[\
"'" I/"" \ 1--- Montmori//oni'te
60
40 V \ v- I'-. Chambers
Arizona
20 V '-
/
o
'~!~4~m II UtEm~~i:::lm;"
'in
~ '~~JE~.~Ef%r$
~ '~Wlm.1rg~!tl'
'~ffn.ml~~i:;;:n;t.
2 3 4 5 6 7 8 9 10 II 12 13 14 15 16
Wave Length in Microns
FIG. 115. Infrared absorption curves for montmorillonite, after Hunt (see Adler
et al.23).
method developed by Brunauer, Emmett, and Teller,25,26 which depends

on the adsorption of simple molecules, such as nitrogen, at temperatures
in the neighborhood of their boiling point. A comprehensive discussion
25 Brunauer, S., and P. H. Emmett, The Use of Low Temperature van der \Vaals
Adsorption Isotherms in Determining the Surface Area of Various Adsorbents, J. Am.

Chern. Soc., 59, 2682-2689 (1937).
26 Brunauer, S., P. H. Emmett, and E. Teller, Adsorption of Gases in Multimole-
cular Layers, J. Am. Chem. Soc., 60,309-319 (1938).
310 Clay Mineralogy
Wove Number in CM.- I
gg g g g 0000 0 0 0 8 If)
8~ ~ ~ @ ~! ~ ~ 8 ~ <Xl
N
<D
100 I, ,I,,,, .1 ,I ,I ,I, ,I
f-"""I _'\.
80 ~ L r---."
60
V \ / Illite
40
V J F/th/an
1\ Illinois
20
o
Altopulqite
Attapulgus
Georgio
15 16
FIG. 116. Infrared absorption curves for illites, mixed-layer minerals, and attapul-
gites, after Hunt (see Adler et al.'3).
of the basic theory of physical adsorption and surface-area measurement

has been presented by Brunauer.27 Emmett, Brunauer, and Love 28 and
27 Brunauer, S., "The Adsorption of Gases and Vapours," Princeton University
Press, Princeton, N.J. (1943).
28 Emmett, P. H., S. Brunauer, and K. S. Love, The Measurement'of Surface Area
of Soil~ and Soil Colloids by the Use of Low Temperature van der Waals Adsorption,
Soil Sci., 40, 57-65 (l938).
Makower, Shaw, and Alexander29 first applied the method to soils and
clay materials.
Obviously, specific surface is not a characteristic of any particular clay
mineral, since it varies with particle-size distribution, particle shape, and
the presence of cracks and pores in the sample. The data given in Table
41, taken from Nelson and Hendricks,30 are intended to give comparative
TABLE 41. SPECIFIC SURFACE AREA OF SOME CLAY MINERALS
(From Nelson and Hendricks 30 )
Heat treat- Loss in Specific surface,

Mineral
ment,OC weight, % m 2 /g
-
Kaolinite, -0.3 micron .... 30 .... 15.5
200 0.2 15.3
500 12.7 18.1
700 14.3 16.5
900 14.6 1.5
Illite, -0.3 micron ......... 30 ... 97.1
200 0.4 92.2
500 5.8 91. 6
7QO 7.7 80.4
Montmorillonite ............ 30 .... 15.5
Halloysite .. ......... . " . 30 .... 43.2
Heated at temperature given until no further weight loss.
values of the order of magnitude of the surface area of some clay min-
erals. The low surface-area value obtained for montmorillonite is inter-
esting in view of the fine grain and the expanding-lattice characteristics
of the mineral. The low value for montmorillonite means that the major
part of the potential surface of the mineral could not be reached by the
nitrogen or other gases used in the determinations. The theoretical sur-
face of montmorillonite on dispersion to nearly unit-cell dimensions is
8 X 10 6 cm 2 /g.
The values in Table 41 show that only a small amount of surface is lost
when kaolinite and illite are dehydrated.
Recently Hendricks and Dya}31 have used the adsorption of ethylene
glycol for surface determinations. The organic is adsorbed on both
29 Makower, B., T. M. Shaw, and L. T. Alexander, The Specific Surface and Density
of Some Soils and Thcir Colloids, Soil Sci. Soc. Am. Proc., 2, 101-109 (1937).
3U Nelson, R. A., and S. B. Hendricks, Specific Surface of Some Cluy Minerals,
Soils, and Soil Colloids, Soil Sci., 56, 285-296 (1944).

81 Hendricks, S. B., and R. S. Dyal, Surface Mcasurement by Ethylene Glycol
Retention of Clays and Its Application to Potassium Fixation, 'Trans. 4th Intern.
Congr. Soil Sci., 2, 71-72 (1950).
312 Clay Mineralogy
external surfaces and inner surfaces accessible only by swelling. A com-
bination of the Brunauer, Emmett, and Teller method with that of Hen-
dricks and Dyal provides values for both external and swelling surfaces.
DENSITY
Determinations of the density of the clay minerals must be made on

aggregates of fine particles. The measurement is inherently difficult,
because of the likelihood of incomplete penetration of the liquid used into
the interstices of the aggregate mass and because of the possibility of a
chemical or physical change during the measurements as a consequence
of adsorption of the liquid or an exchange reaction between the liquid and
the clay mineral. It is particularly difficult to measure the density of the
clay minerals in their natural state with their adsorbed water, which, in
some of the clay minerals, such as montmorillonites and halloysite
(4H 2 0), is an integral part of their structure. For these clay minerals
the density of the dehydrated material can be determined, but it does not
represent the mineral as it occurs naturally.
In the case of the clay minerals which are subject to isomor-
phous substitution within the lattice, e.g., montmorillonite, illite, etc.,
the density varies from one specimen to another. For such clay min-
erals, there is a range of characteristic density values rather than a
single value.
Kaolinite. According to Gruner,32 the theoretical density of kaolinite
computed from the structure is 2.609. Values given in the literature
vary from 2.60 to 2.68, with 2.63 frequently quoted in mineralogical text-
books. Because of the low adsorptive capacities of kaolinite and the
general absence of isomorphous substitutions, the value for this mineral
is a fairly definite and easily determined quantity.
Halloysite. Values of 2.55 and 2.56 are given by Makower and col-
leagues 29 for completely evacuated material, which was probably close to
the 2H 2 0 form. The hydrated form of the mineral (4H zO) would have
a lower density, and values given in the literature 33 for halloysites in
general are as low as 2.0 to 2.2.
Illite. Values given by Dana 34 are 2.76 to 3.0 for muscovite and 2.7
to 3.1 for biotite. Illites would be expected to be in this range, but per-
haps on the low side because of deficiencies of interlayer ions and slight
32 Gruner, J. W., Densities and Structural Relationships of Kaolinites and Anauxites,

Am. Mineral., 22, 85.';-860 (1937).
33 Bosazza, V. L., The Specific Gravity of Clays, S. African J. Sci., 36, 155-157
(1939).
34 Dana, E. S., "Textbook of Mineralogy," W. E. Ford, ed., Wiley, New York
(1921).
replacement of such ions by hydration. DeWit and Arens 35 have
recently given values of 2.642 to 2.688 for materials with 0.0 per cent
adsorbed water. These authors have determined values for the decr'ease
in density of two illites as they pick up adsorbed water in increasing rela-
tive humidities (Table 42). A certain amount of such change is to be
TABLE 42. DENSITY DETERMINATIONS OF CLAY MINERALS AFTER WATER
ADSORPTION AT VARIOUS RELATIVE HUMIDITIES
(After de Wit and Arens')
Relative Moisture content,

Mineral Density
humidity oven-dry basis, %
Montmorillonite (thixotro'n). .......... . .... 0.0 2.348

0.00 . . . .1, 2.608
0.25 11.6 2.199
0.50 16.6 2.134
0.75 28.4 2.001
LOO 46.0 1.772
Illite (Holland). ...................... . ... 0.0 2.649
0.0 ..... 2.660
0.25 3.0 2.579
0.50 4.85 2.504
0.75 6.9 2.422
LOO 18.8 2.128
Illite (Maquoketa shale, Illinois) ........ .... 0.0 2.642
0.0 ..... 2.688
0.25 8.3 2.409
, o 50 12.0 2.310
0.75 24.8 2.02
LOO 76.0 1.48
Kaolinite "Brocades" ........ '" . . .. 0.0 2.667
0.0 2.682
0.75 0.4 2.663
0.50 0.8 2.651
0 75 0.9 2.642
1.06 6.5 2.427
expected, but a large amount of it would probably indicate the presence

of interlayer montmorillonite or vermiculite.
Montmorillonite. Makower and coworkers29 give a value of 2.53 for a
low-iron montmorillonite and 2.74 for one with a somewhat higher iron
content (3.6 per cent). DeWit and Arens 35 give a value of 2.348, and
Caldwell and Marshall 13 give values for nontronite of 2.2 to 2.7 and for
saponite of 2.24 to 2.30. These values are for essentially dehydrated
35 DeWit, C. P., and P. L. Arens, Moisture Content and Density of Some Clay
Minerals and Some Remarks on the Hydration Pattern of Clay, Trans. 4th Intern.
Congr. Soil Sci., 2, 59-62 (1950).
314 Clay Mineralogy
material, and it may be concluded that the density of such montmorillo-
nite ranges from about 2.2 to at least 2.7. Probably 2.7 is too low for the
substantially pure, iron-end member of the series.
The data in Table 42 after DeWit and Arens indicate the large change
in density of the mineral with the increase in adsorbed water as the
measurements were made following treatment at increasing relative
humidities.
Vermiculite. Precise density data are not available for clay-mineral
vermiculite, but the oven-dried material would be expected to have about
the density of biotite mica. Because of its hydration properties, natural
material would give a lower value than biotite mica, and the density
would decrease with increasing adsorbed water, as does that of mont-
morillonite. The minimum value would be higher for vermiculite than
for montmorillonite, since the amount of adsorbed water is restricted in
vermiculite.
Chlorite. Dana 34 gives values of 2.6 to 2.96 for the chlorite minerals.
The clay-mineral chlorites would be expected to give values in this range.
Like the illites as compared to museovite and biotite, they would prob-
ably be in the low part of this range. The very low water-adsorptive
properties of the mineral would mean that the density of air-dried mate- \
rial would differ little from that of natural material, and there would be
little change as determinations were made following treatment at varying
relative humidities.
Attapulgite-Sepiolite-Palygorskite. Caillere 36 gives a value of 2.08
for sepiolite and 2.29 and 2.36 for two samples of palygorskite. These
values indicate the probable range for these minerals, but more data are
necessary before their density characteristics are known in any detail.
Solubility of the Clay Minerals

Bujor, D. J., Electron Microscope Studies of the Formation of Montmorillonite
from Decomposed Biotite, Neues Jahrb. Mineral. Geol. A, pp. 25-28 (1943).
Granger, A., Hational Analyses, Ceramique, 30, 56-58 (1927).
Kallauner, 0., and J. Matejka, Rational Analyses of Kaolinitic Clays, Czechoslov.
Ceram. Soc., 69, 68-70 (1928).
Nagelschmidt, G., Hontgenographische Untersuchungen an Tonen, Z. Krist., 87, 120--
146 (1934).
Nutting, P. G., Solution and Dispersion of Minerals in Water, J. Wash. Acad. Sci., 22,
261-267 (1932).
Tammann, G., and C. F. Grevemeyer, Ueber die Gemenge von Kontronit und Quartz
von St. Andreasberg Tone, Z. anorg. u. allgem. Chem., 135, 197-200 (1924).
Wilson, H., "Ceramics-Clay Technology," .'\fcGraw-Hill, New York (1927).
36 Caillere, S., S. Henin, and S. Meril1ux, Xylotile, a Fibrous Variety of Sepiolite,
Compt. rend., 227, 855-856 (1948).
" .. '
Zalmazon, E. S., and E. S. Shishova, Method of Separation of the Colloid Fraction
from Carbonaceous Clays and Ooze, Bull. acado sci. URSS ser. geol., pp. 145-149
(1950).
I nfrared Absorption Spectra
Buswell, A. 1\1., K. Krebs, and W. H. Rodebush, Infrared Studies, III, Absorption
Band of Hydrogels between 2.5 and 3.5 Microns, J. Am. Chern. Soc., 59,2603-2605
(1937).
Drummond, D. G., Infrared Absorption of {luartz, Nature, 130, 928-929 (1932).
Williams, V. Z., Infrared Instrumentation and Techniques, Rev. Sci. Instruments, 19,
135-178 (1948). ~.
Swface Area
Emmett, P. H., The Measurement of the Surface Areas of Finely Divided or Porous
Solids by Low Temperature Adsorption Isotherms, Advances in Colloid Sci., 1,
1-36 (1941).
Longuet-Escard, J., The Effect of Progressive Dehydration on the Area of the Sur-
face of Montmorillonite, J. chim. phys., 47, 113-117 (1950).
McAuliffe, C. D., N. S. Hall, L. A. Dean, and S. n. Hendricks, Exchange Reactions
between Phosphates and Soils: Hydroxylic Surfaces of Soil Minerals, Soil Sci.
Soc. Am. Proc., 12, 119-123 (1948).
Ritter, H. L., and L. C. Erich, Pore Size Distribution in Porous Materials, Anal.
Chern., 20, 665-670 (1948) ..
Schofield, R. K., Calculation of Surface Area of Clays from Measurements of Negative
Adsorption, Trans. Cerarn. Soc., (Engl.) 48, 207-213 (1949).
Spiel, S., Effect of Adsorbed Electrolytes on Properties of Monodisperse Clay-Water
Systems, J. Am. Cerarn. Soc., 23, 33-38 (1940).
Teichner, S., Measurement of Specific Surfaces of Certain Clays, Compt. rend., 231,
1063-1064 (1950).
Density
Ralston, O. C., and A. G. Stern, "Report of Non Metallic Division," U.S. Bur. Mine:s
Rept. Invest. 3599 (1941).
CHAPTER 13
Origin and Occurrence of the Clay Minerals
SYNTHESIS OF THE CLAY MINERALS
Introduction. The synthesis of the clay minerals at elevated tempera-

tures and pressures from oxides and hydroxides and from various min-
erals, particularly the feldspars, in the presence of acids and alkalies has
been studied in considerable detail. Some investigation has been made
of the synthesis of the clay minerals at ordinary temperatures and pres-
sures, but the amount of such work is inadequate for general conclusions.
Consideration is given herein only to syntheses investigations of a sort
which throw light on problems of the occurrence and origin of the clay .
minerals in nature.
The identification of the crystalline phases formed in syntheses experi-
ments is often very difficult because of their very small particle size, their
poor crystallinity, and their contamination with large amounts of unre-
acted or unchanged material. Usually, X-ray diffraction provides the
only trustworthy method of identifying the new phases. The identifica-
tions made in the years before the development of X-ray-diffraction
techniques and before adequate diffraction data were available for the
pure clay minerals are, therefore, often open to serious question. Even
with the use of diffraction methods, there may be important uncertainties
in the identifications reported in the literature. Thus, some of the mate-
rial reported as kaolinite may actually be halloysite, since in few cases
have the identifications been checked by electron micrographs.
Syntheses from Mixtures of Oxides and Hydroxides at Elevated
Temperatures and Pressures. Information on this subject is largely the
result of the \vork of Noll,l-4 who has published a series of classic papers
on the synthesis of many of the clay minerals.
The reaction products obtained by Noll, for the system A1 2 0 3-Si0 2-H 2 0
1 Noll, W., Hydrothermal Synthese des Muscovits, Nachtr. Ges. Wiss. Gottingen, 20,
122-134 (1932).
2 Noll, W., Mineralbildung im System A1 2 0 3-Si0 2-H 2 0, Nelws Jahrb. Mineral. Geol.
Beilage Bd. A, 70, 65-115 (1935).

3 Noll, W., Synthese von Montmorilloniten, Chem. Erde, 10, 129-154 (1930).
4 Noll, W., Ueber die Bildungsbedingungen von Kaolin, Montmorillonit, Sericit,
Pyrophyllit, und Analeim, Mineralog. petrog. u. Mitt., 48, 210-246 (1936).

316
Origin and Occurrence of the Clay Minerals 317
at temperatures between 250 and 500C with varying amounts of
alumina and silica, the pressure bomb containing a constant percentage
of water, are given in Table 43. Silica gel and alumina gel were used in
TABLE 43. REACTION PRODUCTS IN THE SYSTEM AL 2 03-SI02-H 2 0

(After N oIl 4)
Molecular ratio Al,O,:SiO, (H20 constant)

Temp.,oC
(Pressure, atm.)
1 :0 > 1:2 1:2 1 :4 < 1:4 I 0: 1
5000 Corundum Pyrophyllite

(530-540) + ?
--------- --~--- -_-_-- - - - - - Pyrophyllite
Pyrophyllite Pyrophyllite + SiO,
4000 + (amorphous)
(300)
boehmite +
kaolinite
SiO,
?
(amorphous)
------- Kaolinite ----- ----- _----
3500 Boehrnite +
(168) boehmite
300 0 Kaolinite
(87) Kaolinite Kaolinite + SiO,
250 0 (amorphous)
I
(41)
the experiments; the latter crystallized rapidly to bayerite and boehmite

on aging the gel prior to the experimentation. Most of the alumina used
was in the form of a mixture of boehmite and bayerite. When the two
were used separately, the only observable difference in the results was a
difference in the rate of reaction, boehmite reacting more readily. The
rate of cooling and the degree of filling of the bomb had little, if any,
effect on the nature of the product.
The data in Table 43 indicate that in the absence of silica, boehmite is
formed up to 400C, and at 500C corundum is formed. According to
Roy and Osborne,s at very low temperatures gibbsite would form, with
a transition to boehmite at about 120C. Bayerite is a metastable phase
formed during the rapid crystallization of alumina hydrate from sodium
aluminate solutions. Also, according to these authors, diaspore instead
of boehmite would be the monohydrate phase above 280C under certain
conditions of elevated pressure. Roy and Osborne show that 'Y alumina,
a metastable phase, intervenes in the transition of boehmite to corundum
(ex alumina).
Noll's data show that, with molecular Al 2 0 a : Si0 2 ratios of greater than
Roy, R., and E. F. Osborn, Studies in the System AI,03-SiO z-H,O, presented at
American Institute of Mining and Metallurgical Engineers, St. Louis, 1)51.
318 Clay Mineralogy
1: 2, kaolinite and boehmite are formed up to 400C, and an unidentifiable
uhase develops at 500C. With a ratio of 1: 2, only kaolinite is formed
up to 350C; above that temperature kaolinite, pyrophyllite, and boeh-
mite are formed up to 400C; and at 500C, kaolinite and boehmite dis-
appear, with pyrophyllite and an unidentifiable phase the only compo-
nents present. With a ratio of 1 : 4, kaolinite is the only crystalline phase
up to 350C, and at 400 and 500C pyrophyllite is the only phase
developed. With molecular alumina-to-silica ratios less than 1: 4, the
results of the synthesis are the same as when the ratio is exactly 1: 4,
except that amorphous silica is present at all temperatures. In the
absence of alumina, no crystalline silica phase is formed under the condi-,
tions of Noll's experiments. Noll points out that the rate of reaction
varies greatly with the temperature; e.g., kaolinite formed at 300C in
1 hr, whereas it took 111 hr at 200C.
It would seem from the work of Roy and Osborne 5 that, at very low
temperatures and pressures, halloysite (2H 2 0) may be the resulting
phase in the system Ah03-Si02-H20, rather than kaolinite. Up to the
present time halloysite (4H 20) does not appear to have been synthesized.
The data in Table 44, after Noll, 4 show the reactions obtained in the
TABLE 44. REACTION PRODUCTS IN THE SYSTEM (CAMG)O-(K2NA2)O-AL20a-SI02-
H 20
(After N o1l4)
(CaMg)0-(K,Na,)0-Al,03-SiO, (H,O constant)
< 0.2:1:4 0.2:1:4 0.37:1:2 1: 1:4
300 a C Alkali solu- Kaolinite ( Montmorillonite ............... Montmorillonite

MgO tion Montmorillonite ( <)
---- ----- ---------1-------1-------- --------
300 a C Alkali solu- Kaolinite ( Montmorillonite _.............. ?
CaO tion Montmorillonite )
------1-----1------------ - - - - - - - - - - - - - - --------
300aC Alkali solu- Kaolinite ( Montmorillonite ............... Analcime
Na,O tion Montmorillonite ( <) (+ kaolinite)
-----I-----I--~-----------------------I-----------
Alkali solu- Kaolinite ( Montmorillonite Mica K feldspar* +?
tion Montmorillonite ) (+ kaolinite)
Acid solu- Kaolinite Kaolinite Kaolinite Kaolinite

tion
----1-----1---------1--------- - - - - - - -------
Alk'ali solu- ................ Mica
tion
----- - - - - - - - ------- ------- --------
Acid solu- ................... . Pyrophyllite +
tion boehmite and
kaolinite
* In pure condition at AhO,,6SiO,.

system (CaMg)O-(K2Na2)O-Ah03-Si02-H20 on heating to 300C and
400C for periods of 8 to 24 hr. In each case the bomb was 0.35 per cent
filled with water, and the pressure developed was 87 atm at 300C and 300
atm at 400C. As the alkalies or alkaline earths are added to a mixture
of composition A1 20 3 : 4Si0 2 , montmorillonite forms in increasing amounts
at the expense of kaolinite. A concentration of about 0.2 mole of
(CaMg)O or (Na2K2)O per mole of alumina with an AI 2 0 3 :Si0 2 ratio of
1:4 appears to be the optimum for the formation of montmorillonite. At
slightly higher concentrations with MgO as the alkaline earth, mont-
morillonite is still the phase formed. At higher concentrations with
Na 20, analcime is formed, and at moderately higher concentrations with
K 20, muscovite is formed. The reaction product with CaO at higher
concentrations could not be determined by Noll.
At very high concentrations of K 2 0 with the A1 2 0 3: Si0 2 ratio equal to
1 :4, a potash feldspar was formed. At very high concentrations of MgO
with the same A1 2 0 3 : Si0 2 ratio, according to Noll and also Stresse and
Hofmann,6 talc and then serpentine formed plus kaolinite or pyrophyllite
instead of montmorillonite. Pyrophyllite rather than montmorillonite
is the phase formed if the temperature used is above 400C.
All the above results are for alkaline systems. N oIl shows that for
acid systems with K 20, kaolinite rather than the mica is formed. Other
data suggest the generality that kaolinite is likely to be the phase formed
at 300C in an acid system containing any of the alkalies or alkaline
earths. Gruner 7 confirms the formation of kaolinite at 300C with K 2 0
in an acid system, even in the presence of a large excess of potash. How-
ever, Gruner shows that at 400C muscovite will form in an acid system
in the presence of potash, and also that kaolinite will be converted to
muscovite at this temperature in the presence of an excess of KCI.
Ewell and Insley 8 have reported the formation of nontronite from a
mixture of silica gel and ferric oxide heated at 350C for 6 days at a
pressure of 167 atm.
Syntheses from Mixtures of Crystalline Minerals and Chemical
Reagents at Elevated Temperatures and Pressures. A considerable
amount of work has been done on the alteration of ground feldspar sub-
jected to elevated temperatures and pressures in the presence of acids
and various other reagents. Unfortunately, little work has been done
with minerals other than the feldspars.
6 Stresse, H., and U. Hofmann, Synthesis of Magnesium Silicate Gels with Two
Dimensional Regular Structures, Z. anorg. u. all gem. Chern., 247, 65-95 (1941).
7 Gruner, J. W., Formation and Stability of Muscovite in Acid Solutions at Ele-
vated Temperatures, Am. Mineral., 24, 624-628 (1939).

8 Ewell, R. H., and H. Insley, Hydrothermal SynthesiA of Kaolinite, Dickite,
Beidellite, and Kontronite, J. Research Nail. Bur. Standards, 15, 173-186 (Ul35).
320 Clay Mineralogy
According to Gruner,9 kaolinite, pyrophyllite, muscovite, and boehmite
are the phases which may form when microcline and albite are subjected
to high temperatures and pressures in the presence of aluminum hydrox-
ide, silica, and potassium chloride under acid conditions. In Gruner's
experiments the concentration of the potassium ions and the Al 2 0 3 : Si0 2
ratio of the system were the factors determining which minerals formed
from the feldspars. Kaolinite will form from feldspars and is stable
below approximately 350C, regardless of the potassium-ion concentra-
tion provided the ratio of AI: Si is about 1: 1. This means that additional
easily available aluminum has to be in the system. If the additional
aluminum is not present, pyrophyllite formed instead of kaolinite below
about 350C.
In the temperature range from 350 to at least 530C, pyrophyllite
formed from the feldspars if the potassium-ion concentration was low
(it must not be much greater than the amount that the acid can remove
from the potash feldspars). Under these conditions boehmite formed
from any alumina in excess of that required for the pyrophyllite.
If the concentration of potassium and sodiuq! ions was high but no
additional alumina was available, the feldspars remained unaltered. No
true potassium feldspar was synthesized under the conditions of the
experiments. If alumina, silica, and potassium chloride are added to the
feldspar, muscovite and pyrophyllite are synthesized without any forma-
tion of feldspar.
Muscovite formed readily from feldspar, beginning at about 350 to
525C, the highest temperature tried, provided that the potassium and
aluminum were in sufficient concentration.
Schwarz and Trageser 10 stated that orthoclase and anorthite in the
presence of 0.5N HCI gave kaolinite as an alteration product at tempera-
tures below about 400C, pyrophyllite and boehmite at temperatures of
400 to 550C, and corundum at about 600C.
Badger and Allyl! produced kaolinite when a potash feldspar was
heated at 225C for 24 hr in the presence of 5 per cent hydrofluoric acid.
These same authors found no reaction when the same feldspar was heated
in the presence of carbonic acid for 156 hr at 60C and a pressure of
1,800 psi. '
The formation of secondary silicate compounds from solid phases like
feldspar and aluminum hydroxide in the foregoing type of experiments
9 Gruner, J. W., Hydrothermal Alteration of Feldspars in Acid Solutions between ..

300 0 and 400 0 e., Ecan. Geal., 29, 578~589 (1944).
10 Schwarz, R., and G. Trageser, Ueber die Synthese des Pyrophyllits, Z. anorg. u.
allgem. Chem., 225, 142-150 (1935).

11 Badger, A. E., and A. Ally, Note on the Formation of Kaolin Minerals from Feld-
spar, J. Geal., 40,745-747 (1935).
is little understood. Schwarz and Trageser 10 believe that the hydrolysis
of the original mineral is essentially complete and that the new insolublB
compounds are reaction products of the hydrolyzed components. Gruner 9
suggests that the degree of ionization, particularly with respect to Si0 4
groups, must be very small in such systems. Jander and Wuhrer 12 sug-
gest that there is only partial hydrolysis of the original mineral with the
formation of some kind of an amorphous intermediate product. Various
investigators have favored a structural transformation of the primary
minerals to the secondary ones with only very slight hydrolysis. Schwarz
and Trageser10 point out that kaolinite can form under similar conditions
from a variety of parent minerals (e.g, leucite, in addition to the feldspars),
which means that no single original structure is required if the change is
largely a structural transformation.
Syntheses from Mixtures of Oxides and Hydroxides at Ordinary
Temperatures and Pressures. Only a few investigations have been made
of the formation of the clay minerals from their constituent oxides at
ordinary conditions of temperature and pressure. The resulting data are
scant, and in some cases the identifications of the resulting products are
not very convincing.
Sedletsky 13 mixed sodium silicate and sodium aluminate and then
leached with N MgCl 2 until the pH of the leachate was 5.6. After
further washing to remove all the free chloride, the gel was held in a closed
vessel at the temperature of the laboratory for 4 years, at the end of
which X-ray diffraction showed the presence of a product similar to
montmorillonite.
Caillere and Henin 14 report the synthesis of clay minerals by the
electrolysis of a solution of silica or a mixture of solutions of silica and
alumina. Platinum was used as the cathode and various anodes of
aluminum, iron, nickel, and magnesium were tried. The product
depended on the anode used; e.g., with a magnesium anode an antigorite-
like mineral was formed, and with aluminum the product seemed to be
kaolinite.
Stresse and Hofmann 6 report that magnesium silicate gels obtained by
boiling together solutions of MgCl 2 and hydrated silica developed clay-
mineral structures when they were boiled further with potassium hydrox-
ide, calcium hydroxide, or sodium hydroxide. The structures in every
case were like those of montmorillonite, except when the concentration
12 Jander, W., and J. Wuhrer, Die Bildung von Magnesium-hydrosilikaten, Z.

anarg. u. allgem. Chem., 236, 282-287 (1938).
13 Sedletsky, I. D., Genesis of Minerals from Soil Colloids of the Montmorillonite
Group, Campi. rend. acado sci. URSS, 17,375-377 (1937).

14 Caillere, S., and S. Henin, Essais de synthese des mineraux argileux, Verre silicates
indus., 13, 63-64 (1948).

322 Clay Mineralogy
of potassium hydroxide became fairly high; then a mICa structure
d2veloped.
Teichner and Pernoux 15 have pointed out that some catalysts for the
Fischer process frequently are prepared by impregnating kieselguhr with
nickel and that in the impregnation of the silica the catalyst develops a
montmorillonite-like structure. Unpublished reports16 indicate that
nearly all the clay minerals have been synthesized with nickel proxying
for alumium. The conditions under which the syntheses were made,
however, have not been revealed.
Transformations of Clay Minerals at Ordinary Temperatures and
Pressures. Many investigators, including Caillere and Henin,17 Volk,18
Aleshin,19 and Barshad,2 have shown that a material substantially like'
illite is produced from montmorillonite when all the exchange positions
are occupied by potassium ions and the material is thoroughly dried at
about 110C. Such material does not again expand, even on treatment
with a polar organic liquid, and it has substantially the same X-ray-
diffraction characteristics as illite. Similar results are obtained under
the same conditions by treating montmorillonite with NIlt. Vermiculite
reacts with both K+ and NHt, as does montmorillonite, and with similar
results.
Caillere and Henin 21 and Barshad have shown also that the treatment
of either montmorillonite or vermiculite under certain conditions wit.h a
solution containing Mg++, so that all the exchange positions are occupied
by the magnesium ions, produces a product which does not expand and
has the X-ray-diffraction characteristics of chlori~e. The exact experi-
mental conditions of the treatment process necessary to obtain chlorite
rather than montmorillonite with exchangeable Mg++ are not known.
Caillere and Henin 22 report the formation of kaolinite from mont-
morillonite in the laboratory. The kaolinite was obtained by treating
montmorillonite with 20 per cent CaC12, 0.5 per cent Na3Al03, and 10
15 Teichner, S., and E. Pernoux, Texture and Surface Area of Kieselguhrs after
Certain Treatments, Mineralog. Soc. Gr. Britain Clay Mineral Group Bull. 1, 145-150
(1949). .
16 DeLange, ~ J., Notes, J. Sci. Instruments, 24, no. 1, p. 20 (1947).
17 Caillere, S:, S. Henin, and S. Meriaux, Transformation experimentale d'une

montmorillonit en une phyllit it 10 Atype illite, Compt. rend., 226, 680-681 (1948).
18 Yolk, G., Nature of Potash Fixation in Soils, Soil Sci., 45, 263-276 (1938).
19 Aleshin, S. N., Changes of Montmorillonite into Hydro-mica, Doklady Akad.
Nallk SSSR, 61, 693-695 (1048).

20 Barsh ad, 1., The Effect of Interlayer Cations on the Expansion of the Mica Type
of Crystal Structure, Am. Mineral., 35, 225-238 (1950).

21 Caillere, S., and S. Henin, Experimental Formation of Chlorites from Montmoril-
lonite, Mineralog. Mag., 28, 612-620 (1949).

22 Caillere, S., and S. Henin, Formation of a Phyllite of the Kaolinite Type by
Treatment of Montmorillonite, Compt. rend., 224, 53-55 (1947).
,', .
per cent AI(N0 3)3 for 3 to 4 days and then precipitating with HCI or
NH 4 0H.
General Conclusions from Synthesis Data. Some general conclusions
regarding the environmental conditions favorable for the formation of the
clay minerals can be reached from the available data from synthesis
experiments. At low temperatures and pressures, acid conditions appar-
ently favor the formation of the kaolinite type of mineral, whereas
alkaline conditions favor the formation of montmorillonite, or mica, if
potassium is the alkali metal and if it is present in concentration above
a certain level. From the work of Noll the minimum concentration of
K 2 0 necessary for the formation of the mica is between 0.2 and 0.37 mole
per mole of A1 2 0 3. The presence of magnesium particularly favors the
formation 6f montmorillonite.
At temperatures somewhat above about 350C and with moderate
pressures, pyrophyllite forms in place of kaolinite, with any excess A1 20 a
forming boehmite. At more elevated temperatures and pressures, other
alumina phases develop. At more elevated temperatures, the foregoing
generalities for results under acidic and alkaline conditions do not always
hold. Thus, mica can form under acid conditions, and kaolinite or
pyrophyllite can develop in the presence of an excess of K 20, depending
on the temperature, the amount of K 20, and the A1 20 3: Si0 2 ratio.
CLAY MINERALS OF HYDROTHERMAL ORIGIN
Introduction. It has been recognized for a long time that argillaceous

alteration products due to hydrothermal action are frequently to be
found as an aureole around metalliferous deposits. Such alteration prod-
ucts are also found associated with hot springs and geysers. The size of
the alteration envelope varies widely. At Butte, Montana,23 for example,
a width of 10 ft is common in the peripheral area of the ore body, whereas
the width may become 100 ft or more in the ZOne of intermediate minerali-
zation intensity. In fissure veins the alteration lies parallel to the walls
of the fissure and is of relatively uniform width, varying according to the
size of the vein. If the veins are closely spaced, the alteration halos may
merge, as in the central zone at Butte and in porphyry copper deposits.
Bateman 24 has pointed out that, with epithermal veins, the alteration
zone is likely to be narrow and the changes may be hardly discernible,
.. whereas, with mesothermal veins, the alteration halo is likely to be wide
and intense. The alteration products associated with hypothermal
deposits are generally not argillaceous, and the secondary products
23 Sales, R., and C. Meyers, Wall Rock Alteration at Butte, Montana, Am. lnst.
Mining Met. Engrs. Tech. Pub. 2400 (1948).

24 Bateman, A. M., "Economic Mineral Deposits," Wiley, New York (1942).
324 Clay Mineralogy
found with other types of metalliferous ore bodies are likely to contain
few or substantially no clay minerals.
Deseriptions of hydrothermal alteration products, before the develop-
ment of modern X-ray, thermal, and optieal methods of clay-mineral
identification, are of little value because of questionable clay-mineral
identifications. For example, a large mass of hydrothermal halloysite
at Eureka, Utah, was first described as talc. Unfortunately, even in very
recent years, some students of ore deposits have failed to realize the
complexities of clay-mineral identification and have published determina-
tions which are questionable because they are based on insufficient
analytical data.
The study of the alteration products associated with many ore bodies
is complicated by the difficulty of distinguishing between hypogene and
supergene products. In many cases, later supergene alteration is super-
imposed on earlier hypogene alteration, and the same clay mineral may
be developed in both processes. Kaolinite, for example, in some instances
may be formed as a hypogene product and later by supergene alteration
of some other clay minerals.
The probable economic importance of wall-rock alteration has been
recognized in recent years, because of the likelihood of significant rela-
tionships between kind of alteration and mineralization. As a conse-
quence, the problem has been studied by many investigators, but much
of the data have not yet been published, and the topic has not reached
the place where many broad generalizations are established.
Types of Clay Minerals in Hydrothermal Deposits. All the various
clay minerals, with the possible exception of attapulgite-palygorskite and
vermiculite, have been authentically reported in hydrothermal bodies.
The various forms of the kaolinite minerals are recorded, including
anauxite, dickite, and nacrite, as well as various montmorillonites,
including nontronite and the magnesium-rich members. The white
mica found in alteration halos is frequently described as sericite and
undoubtedly has similarities to illite. The beidellite which is frequently
recorded, particularly in the earlier literature, is probably often a mix-
ture 25 of clay minerals. Halloysite and allophane are often reported, and
although the identifications are sometimes established, some are based
solely on optical 'measurements and must be questioned. Material
appearing to be isotropic and therefore amorphous and called allophane,
may. be revealed by X-ray diffraction to be largely crystalline but
extrerrl.ely fine grained and an aggregate of randomly oriented particles.
In sbme instances, hydrothermal elay-mineral bodies are of large size
(of the order of 100 ft in diameter) and are composed substantially of a
25 Grim, R. E., and R. A. Rowland, Differential Thermal Analyses of Clay Minerals
and Other Hydrous Materials, Am. Mineral., 27, 746-761 (1942).
single clay mineral. An example is the Cornish kaolin, whose hydro-
thermal origin is now well established by the finding of kaolin masses
which are overlain by sills and which do not reach the surface. Other
examples are the halloysite body at Eureka, Utah, which is composed of
a single clay mineral. The so-called bentonite from the Island of Ponza,
Italy, may be an example of a huge mass of almost pure montmorillonite
formed from the alteration of an eruptive igneous rock.
In many, perhaps most, instances, the alteration product is a mixture
of several clay minerals. The association is frequently so intimate that
the identification of the individuals is very difficult. Mixed-layer clay
mineral specimens are extremely common in hydrothermal deposits.
Mode of Formation and Occurrence. In many hydrothermal clay
bodies a zonal arrangement of the clay minerals has been observed.
Often there is an inner halo of sericite (the term sericite will be usedforthe
white-mica alteration product, since that name is used in the literature
and since the precise relationship to illite is not established), an inter-
mediate zone of kaolinite, and an outer zone of montmorillonite and
chlorite, with chlorite being most abundant on the outermost fringe.
The boundaries of the zones are gradational. In some other bodies the
alteration products appear to be superimposed rather than zonal, as if
they had been developed in several fairly distinct stages.
Because of the seeming difference in the occurrence of the argillaceous
alteration products, two concepts of their formation have been suggested.
On the basis of their classic study of wall-rock alteration at Butte,
Montana, Sales and Myer 23 conclude that the alteration is essentially
contemporaneous with ore deposition and that different mineralogical
zones in the surrounding rock may be regarded as reaction rims, repre-
senting diminishing activity of the powerful solution that deposited the
ore minerals. The different mineralogical and chemical responses within
the wall rock are dependent not on a drastic change in the hydrothermal
fluid itself, or on a periodicity in its operation, but on continuously vary-
ing conditions of physicochemical environment within the wall rock
outward from the vein.
The second concept appears in the writings of Lovering,26 Kerr,27
Schwartz,28 and others 29 who have described wall-rock alteration at
numerous deposits which seems to have developed in stages, by differing
26 Lovering, T. S., "Rock Altcration as a Guide to Ore-East Tintic District,
Utah," Economic Geology Monograph I (1949).

27 Kerr, P. F., J. L. Kulp, C. M. Pattcrson, and R. J. Wright, Hydrothermal Alter
ation at Santa Rita, New Mexico, Bull. Geol. Soc. Am., 61,275-348 (1950).
28 Schwartz, G. M., Hydrothermal Alteration of the" Porphyry Coppcr" Deposits,
Eeon. Geol., 42, 319-352 (1947).

29 Anderson, A. L., Structural Control and Wall Rock Alteration at the Webber
Mine, Butte County, Idaho, Eeon. Geol., 42, 368-383 (1947).

326 Clay Mineralogy
solutions appearing at different periods separated by time intervals of
relative inactivity. In some cases the products developed during a given
stage' show some zoning. Often there appears to be an early stage of the
development of chlorite and sericitic mica, a second stage of development
of halloysite, kaolinite, and montmorillonite, and a late stage of develop-
ment of sericite. In the second stage, the halloysite and kaolinite are
likely to be close to the vein, and Lovering has recognized a zoning in
some deposits of dickite-kaolinite-halloysite outward from the vein.
It seems likely that both concepts are applicable at particular deposits.
Adequate information is not yet at hand to determine what controls the
operation of one or the other process at a given deposit. Also, adequate
data are not available for broad generalizations on the nature of the
zoning or on the sequence of clay minerals likely to develop in the
periodic process.
According to Sales and Meyer,23 the alteration of the quartz monzonite
at Butte, Montana, consists of an outermost fringe in which the horn-
blende and biotite are changed to chlorite. Progressing toward the vein,
montmorillonite becomes the dominant alteration product developing
from plagioclase, with the nontronite variety forming from hornblende.
A small amount of amorphous material is also formed with the mont-
morillonite. Progressing further toward the vein, kaolinite is found in
increasing abundance with the gradual reduction in the amount of
montmorillonite. In the kaolinite subzone there is always some mont-
morillonite present, and frequently there is a regeneration of the chlorite.
In the argillaceous alteration zone, the orthoclase is unaltered except for
a faint trace of kaolinization at the innermost edge. The argillaceous
zone grades veinward into the sericite zone, which is made up solely of
sericite, quartz, and pyrite. In this zone all the remaining silicate
minerals are altered.
Sales and Meyer show that the abundance of lime, soda, and silica .
declines considerably and magnesia moderately, from the unaltered rock
toward the vein. They conclude that there is a continual removal of
these components during the gradual outward movement of the altera-
tion; i.e., these components are carried outward as they are released by
the breakdown of the parent minerals. The result over a period of time
is a gradual pushing outward of the front of attack as the total envelope
of alteration grows wider while maintaining the identity and positions
relative to each other of its component zones. The montmorillonite sub-
zone, for example, prevailed, until the outward migration caused a
depletion of alkalies and alkaline earths, when kaolinite became the stable
phase.
, According to . Sales and Meyer, sericitization and argillization are
contemporaneous processes. The potash content is nearly constant
throughout the entire alteration halo. The orthoclase feldspars break
down only in the sericite zone, and in fact the beginning of the sericite
zone coincides with the destruction of the potash feldspar. Orthoclase
is broken down only close to the vein as a consequence of the higher
temperatures and the greater potency of hydrothermal acid water.
With the destruction of the orthoclase near the vein, the potash necessary
for the formation of sericite is available. Sericite is not formed at a dis-
tance from the vein because of the absence of available potash as a result
of the stability of the orthoclase. An alkaline environment, in addition
to the presence of potash, is usually required for the formation of white
mica, but Gruner 7 has shown that the mica can form in an acid environ-
ment at temperatures of the order of 350C in the presence of an excess
of potassium.
Lovering 26 describes wall-rock alteration in the Tintic district of Utah
as a periodic development. The first stage is chloritization of the
volcanic rocks and dolomitization of the limestones, both involving the
addition of magnesium. The second stage, or argillitic stage, witnessed
the development of the clay minerals other than chlorite. The clay
minerals are better developed in the volcanics than in the limestones, and
Lovering was able to detect a zoning within the argillic zone consisting
of dickite-kaolinite-halloysite (2H 20)-beidellite-halloysite (4H 20), read-
ing progressively away from the ore deposit. The final stage was exten-
sive sericitization and silicification.
Kerr and coUeagues 27 have described the hydrothermal alteration of
intrusives at Santa Rita, New Mexico, as occurring in several stages.
The initial stage of clay mineral development is chloritization of biotite
and hornblende. The next stage is one of "hydromica-argillic altera-
tion," with the development primarily of hydromica from the biotite and
chlorite, and of sericite and kaolinite from the plagioclase feldspars.
The final stage is one of sericitization and silicification, in which nearly
all the primary silicates are altered. The absence of montmorillonite at
Santa Rita is noteworthy.
Peterson and colleagues 30 describe three stages of alteration at Castle
Dome, Arizona. The earliest stage is the development of chlorite and
sericite with some nonclay minerals. This is followed by an argillic stage
in which montmorillonite and" hydromica" are the important constit-
uents. Kaolinite is also developed in the argillic stage, but it is of
minor abundance. The final phase is one of intense sericitization and
silicification.
Schwartz 28 has summarized the data for the alteration associated with
a series of porphyry coppers and has shown that there is great variation
30 Peterson, N. P., C. M. Gilbert, and G. L. Quick, Hydrothermal Alteration in the
Castle Dome Copper Deposit, Arizona, Eean. Geol., 41, 820-846 (1946). .'
328 Clay Mineralogy
in the kind and amount of alteration in various ore bodies. In some
d{'_:)osits, notably at Ajo, Arizona, there are substantially no clay-mineral
alteration products, whereas, in others, e.g., at Morenci and Miami,
Arizona, hypogene clay minerals are extensively developed. He has
shown also that the identity and relative abundance of the clay minerals
differ widely in different deposits. In most deposits, however, the
development of argillic material seems to be followed by a final period
of intense sericitization and silicification.
As stated earlier, the available data are not adequate to establish
general conclusions, but they do suggest a general tendency toward
sericite adjacent to the vein, an outer chloritic zone, and an intermediat~
zone of kaolinite, halloysite, allophane, and/or montmorillonite. In the
intermediate zone, montmorillonite is mostly in the outer part and the
kaolinite minerals in the inner part. In some cases the sericitization
seems to be later than the argillic development.
Nature of the Hydrothermal Solutions. Information concerning the
composition of the hydrothermal solutions is scant, but in many, per-
haps most, cases, such waters are thought to be acid to begin with, carry-
ing chlorine, sulfur, carbon dioxide, and/or silica. The composition of
the solutions must change as reactions take place with the host rock, and
in the case of silicate rocks the solutions would become alkaline as a con-
sequence of such reactions. The alkaline constituents would probably
be transported outward by the solutions, and the alkalinity of the solu-
tions would persist only so long as alkalies or alkaline earths were being
released by the breakdown of the parent rock. As the action of the acid
water proceeded, therefore, the acid-alkaline front would move outward
as the alkaline earths and alkalies were carried outward. As Sales and
Meyers have postulated, the later breakdown of a resistant mineral, such
as orthoclase, which had persisted through the beginning of the acid
encroachment until temperatures or solution concentration became suffi-
ciently high to break it down, can cause a reversal to alkaline conditions.
It is by no means established that hydrothermal solutions are always
acid initially. Andreatta,31 for example, has described a sequence of
hydrothermal clay minerals developed in the order micaceous clay
mineral-montmorillonite-halloysite-kaolinite; this sequence he believes
to have formed from warm water originally alkaline and rich in potassium
which became neutral and finally acid with progressive cooling.
Relation to Parent Materials. If the hydrothermal alteration is
intense, as a result of relatively high temperatures and concentrations,
and long continued, so that all the original minerals have been affected,
and if there has been considerable transportation of components, such
31 Andreatta, C., Study of the Hydrothermal Stratum of the Clay Minerals of
Capalhia. Italy, Ind. ceram. e,silicati, 2, 17-21 (1949).

as alkalies and alkaline earths released by the breakdown of the parent
minerals, the alteration products will tend to be the same regardless of.the
parent rocks, except for carbonate and quartzitic rocks. In the case of
limestones, the tendency is toward silicification. In the case of dolomitic
rocks, silicification may also be the result, but the reaction between the
magnesia and the silica sometimes forms a magnesium-rich clay mineral.
The deposit of the montmorillonite mineral hectorite, at Hector, Cali-
fornia, has probably had such an origin.
If the hydrothermal alteration is relatively slight, the character of the
primary mineral largely determines the alteration product. With very
slight alteration, the magnesium-rich minerals such as the hornblendes
and biotite tend to change to chlorites. With slight alteration and in
the presence of alkalies and alkaline earths, except potassium, the micas,
the ferromagnesium minerals, and the plagioclase feldspars are likely to
yield montmorillonite. The presence of magnesium particularly favors
the formation of montmorillonite. The presence of potash, either from
feldspars or primary micas, favors the development of secondary micas.
ZsolnY,32 for example, has described alteration in adjacent rhyolite and
andesite; the former shows illitic secondary products and the latter,
alteration to montmorillonite. The literature has many references to
montmorillonite types of alteration products in basic igneous rocks,
indicating the frequency of montmorillonitic alteration in this type of rock.
Kaolinite may form from any constituents, if the alkalies and alkaline
earths are removed approximately as fast as they are liberated from the
parent rock or if the environment is acid and the temperature is moderate.
When present in the environment, calcium particularly appears to retard
the formation of kaolinite.
Sales and Meyer,23 Ross and Hendricks,33 and others have shown that
the clay minerals themselves alter to other clay minerals if the hydro-
thermal environmental conditions change. Such alterations take place
in accord with the generalities just mentioned above.
Relation to Mineralization. The impetus for much of the recent study
of wall-rock alteration associated with ore bodies has been the hope that
correlations of type of alteration with mineralization would provide keys
for prospecting. Very little such data have been published, but the
meager published data suggest that in some cases, at least, the hopes
have been realized.
Lovering 26 has shown at Tintic, Utah, that ore values are not found
when the late stage of sericitization and silicification has not been opera-
32 Zsolny, L. M., Illite, Montmorillonite, Halloysite and Volcanic Ash as Whiteware
Body Ingredients, J. Am. Ceram. Soc., 29, 254-260 (1946).

330 Clay Mineralogy
tive. At Santa Rita, New Mexico, Kerr and colleagues 27 have shown I
that at least part of the mineralization is contemporaneous with the

argillic alteration. These authors also point out that the feldspars in the
area of copper mineralization show prominent kaolinite alteration,
whereas in the zinc area, montmorillonite is the dominant alteration
product. Sales and Meyer have observed a general close association of
sulfides and the sericite alteration at Butte, Montana.
Clay Minerals Associated with Hot Springs, Fumaroles, Etc. The
action of hot springs, fumaroles, etc., provides an excellent opportunity
to study hydrothermal alteration of which clay mineralogists have not
yet taken full advantage. Fenner,34 studying samples from boreholes
in the Yellowstone area, found that beidellite was the alteration product
at depths greater than about 95 ft when the solutions were alkaline. At
shallower depths kaolinite was found, and it was believed to have resulted
from descending meteoric acid waters. The acid resulted from the
oxidation of sulfides.
SOILS AND WEATHERING
Modern investigations by Kelley, Hendricks, MacEwan, Henin,

Marshall, Jackson, and many other students of soils have provided a
tremendous amount of information on all phases of clay mineralogy.
Each of these investigators and many others have extensive bibliographies
dealing with clay mineralogy, so that it is not feasible to attempt to list
even the major publications of the most active workers. Soil investiga-
tors were the first to apply modern methods of clay-mineral analysis to
their problems extensively. In this section, the distribution of clay
minerals in various soil types and the environmental conditions leading
to the formation of particular clay minerals during weathering processes
will be considered.
Factors Controlling Weathering Processes. No arrangement of the
factors controlling weathering processes in a general order of importance
is possible, since the factors are closely interdependent and since their
relative importance is not always the same.
f Parent Rock. The composition and texture of the parent rock are
.~ important in initial stages of weathering, but their importance decreases
as the duration of weathering increases. Under conditions of drastic
alteration under humid conditions, as in the development of podsolic and
lateritic soils, the influence of parent rock is relatively short-lived,
whereas in aridic zonal soils it prevails almost indefinitely. It has been
proved that soils containing kaolinite and soils containing montmoril-
lonite can both develop from the same parent rock under different condi-
34 Fenner, C. N., Bore Hole Investigations in Yellowstone Park, J. Geol., 44, pp. 225-
315 (1916).
."
tions of climate, topography, and time. 35 Also it has been shown 36 that
the same type of soil with a characteristic clay-mineral compositi<l can
develop after a relatively long period of time from rocks of widely differ-
ent composition and texture.
The most important factor of composition in parent rocks is the content
of alkalies and alkaline earths. Rocks containing no alkalies, such as
kaolin clays, can yield only kaolinite or lateritic weathering products
unless ground-water movement brings alkalies and/or alkaline earths to
the environment. Igneous rocks, shales, slates, schists, and argillaceous
carbonates can yield a variety of weathering products, at least in the
initial stages of weathering, because of their content of alkalies and
alkaline earths in addition to alumina, silica, etc.
Climate. The temperature and the rainfall, particularly its seasonal
distribution, are the significant climatic factors. Decay of the parent
minerals is most rapid in warm, humid climates. The direction of the
movement of water through the weathering zone and the leaching of
soluble salts from the material depend on the amount of rainfall and its
seasonal distribution. Thus, in a continuously wet climate the move-
ment of the water is downward, tending to remove downward the com-
ponents liberated by the decay of the parent minerals. In a continuously
dry climate or seasonal one with long dry seasons, the prevailing water
movement may be upward, and the decay components would not be
removed from the decay zone. The climate determines the amount and
kind of vegetation and further the amount and kind of organic alteration
products yielded by the decay of the organic material. Thus in a humid
cool climate an abundance of organic material decays slowly to produce
an abundance of active organic acids and other compounds to react with
the parent material, whereas in a hot humid climate with seasonal rainfall
the organic material is rapidly destroyed by oxidation and alteration.
As a consequence, the kind and amount of organic products available for
leaching are very different in the two environments.
Topography. Topography determines whether or not there is active
vertical movement of water though the weathering material. This is
particularly significant in regions of relatively high rainfall in which
low, flat areas may be saturated with water almost to the surface. The
prevalence of oxidizing or reducing conditions and the depth to which
active oxidizing conditions penetrate depend in part on the water satura-
35 Hosking, J. S., The Soil Clay Mineralogy of Some Australia Soils Developed on
Granitic and Basaltic Parent Material, J. Council Sci. Ind. Research, 13, 206-216
(1941).
36 Humbert, R. P., and C. E. Marshall, Mineralogical and Chemical Studies of Soil
Formation'from Acid and Basic Igneous Rocks of Missouri, Agr. Expt. Sta. Univ.
Missouri Research Bull. 359 (1943).
332 Clay Mineralogy
tion of the soils. Poorly drained areas saturated with water for long
perio~s of the year generally have only reducing or feebly oxidizing con-
ditions, or at best a thin surficial zone of oxidation.
Topography, by its control of the vertical movement of ground water,
also influences the leaching processes. Thus, in low, flat areas with little
movement of water through the soils, there would be relatively little
leaching.
Topography also influences the rate of erosion at the surface, and hence
the rate of removal of the products of weathering and the rate at which
fresh parent material is brought close to the surface into the most active
zone of alteration.
Vegetation. The significant factors with regard to vegetation are its
amount and kind and the amount and kind of resulting products of its
decay. In an arid area where vegetation is absent or in an area with long
hot, dry intervals where vegetation is quickly oxidized, few or no organic
acids are produced from the vegetation to enter into the alteration proc-
esses. Also in countries of eucalyptus rather than deciduous trees, there
would be scant accumulation of leaf litter and scant production of organic
decay products, even though conditions favored the abundant growth of
vegetation.
Soil investigations suggest that, under cool, humid conditions with
deciduous forest growth, organic decomposition products develop in the
soil which have very drastic effects on parent silicate minerals. Under
temperate conditions and prairie grass cover, the resulting vegetal decom-
position products are different and appear to have less drastic effects on
the parent minerals.
Time. Since weathering processes are relatively slow, time is a signifi-
cant factor in the development of alteration products. In the initial
stages of the weathering of some rocks, alkalies and alkaline earth may be
present in the weathering zone, and as a consequence a certain kind of
alteration product will develop. After a long time interval all the alkalies
may have been removed by leaching, and a different type of alteration
product will form. As a consequence; weathering products from a given
type of parent material may be quite different in an initial stage and in a
later stage. Bray 37 has, for example, shown the great difference in the
mineral composition of weathering products of loess in Illinois as a conse-
quence of time. Manifestly, the factor of time is most important when
weathering in general is moderately severe and when the parent rock has a
composition permitting the formation of a variety of alteration products.
Soil-profile Development. Weathering processes, especially where
there is vertical (downward or upward) movement of moisture, tend to
37 Bray, R. H., Chemical and Physical Changes in Soil Colloids with Advancing
Development in Illinois Soils, Soil Sci., 43, 1-14 (1937).
"' ..
develop a sequence of horizons within the zone of alteration. The hori-
zons frequently grade sharply into adjacent ones and have thicknesses
measured in inches or a few feet, although there are some very mature
..
soils with horizons many feet in thickness. The horizons show differences
in degree of breakdown of parent material, in kind and relative abundance
of secondary mineral components, in particle-size distribution, in organic
pH
L Litter 4.78
Portiolly Decoyed
F Organic Matter, 3.48
Reddish 8rown
H Row Humus, Black 3.54
Silicious Gray Loyer 4.20
Precipitated Humus,
3.93
Dark 8rown to 81ack
Compac~ Yellow to Reddish

8rown. Accumulahon of
B2 5" 4.50
Sesquio)(ide and Colloidal
Cloy
o
_.::_. 0 _. Friable Yellow
B.3 7"
o Sandy Loom
4.55
(_) . (:j '_'.

At 24" Friable
_ _ '_-(J 4.80
Yellow Sand Downward
.l:J '. o (l .
FIG. 117. 'Well-developed podsolic soil profile, from IJyons and Buckman."S
material, and/or in pH and character of alkali and alkaline-earth content.

Vertical movement of water causes the translocation of soil constituents,
thereby developing the sequence of varying horizons. A fairly typical
profile of a podsolic type of soil is given in Fig. 117, from Lyons and
J Buckman.38 As a consequence of the difference in composition and tex-
ture, the physical properties of plasticity, tilth, etc., vary from horizon to
horizon in the profile.
3S Lyons, T. L., and H. O. Buckman, "The Nature and Properties of Soils," Mac-
millan, New York (1943).
334 Clay Mineralogy
The soil horizons are designated from the surface downward, A, B, C,
D. The nomenclature of soil horizons outlined recently by the Soil Sur-
vey Staff of the U.S. Department of Agriculture, as summarized by
Winters and Simonson,39 is as follows: The A horizon is defined as a layer
of maximum accumulation of organic matter, and/or as the surface layer
that is lighter in color than the underlying horizon and has lost some of
its clay and sesquioxides. The B horizon is defined as the soil layer that
contains an accumulation of clay and sesquioxides with small amounts
of organic material, and/or has a blocky or prismatic structure and a
stronger color than the overlying A horizon. The C horizon is defined as
a layer of unconsolidated and partly weathered material relatively
unchanged by soil-forming processes. The D horizon is unaltered parent
material.
In description::; of weathering products it is obviously necessary to
describe the profile as a whole as well as the composition and properties
of the individual horizons.
Classification of Great Soil Groups of the Wodd. The various types
of soil are considered by student8 of 80il8 under three main headings as
follows:
1. Zonal Soils. After soil-forming processes have been at work for a
considerable time, soil8 over wide area8, underlain by rocks of varying
kinds, tend to become alike in general characteristic8 if the climatic influ-
ences are reasonably uniform and continuous and if erosion is not too
vigoroU8. Such 80il group8 are called zonal and con8titute the Great Soil
Groups of the World. In the8e 8oil8, climate, with its associated kind of
vegetation and organic decomposition material, is the major factor in
80il gene8is.
2. Intrazonal Soils. The8e 80ils are a8sociated with the zonal groups
and are th08e that reflect the influence of 80me local conditions 8uch as
poor drainage, alkali salt8, or 80me other unique characteri8tic super-
imposed on the general characteri8tics resulting from the climate and
vegetation factors. Such soil8 frequently cross zonal boundaries and
thus are called intrazonal.
3. Azonal Soils. These are soils without profile development, such as
alluvial and colluvial soil8. There i81ittle or no alteration of parent mate-
rial because of their youth or environmental setting.
The classification of the great soil groups with 80me of the more impor-
tant associated intrazonal soils suggested by Lyon and Buckman is given.
in Table 45. A map showing the world di8tribution of the great soil
groups is presented in Figure 118. A somewhat idealized diagrammatic
presentation of representative profiles of five zonal soils is given in
Figure 119.
39 Winters, E., and R. W. Simonson, The Subsoil, Advances in Agron., 3,2-92 (1951).
, ....
TABLE 45. CLASSIFICATION OF GREAT SOIL GROUPS O"F THE WORLD
(After Lyons and Buckman 38)
I. Zonal humid soils
Tundra
Dark gray peaty accumulation over gray mottled mineral horizons. Sub-
stratum ever frozen and often the subsoil as well
Podsolic
Typical podsol (gray forest soil)
Brown podsolic
Gray-brown podsolic
Red and yellow podsolic (lateritic)
Prairie dark brown and reddish brown
Lateritic
Typical and well-developed laterites
Yellowish-brown lateritic soils
Reddish-brown lateritic soils
Terra rossa
Associated intra zonal groups
Plano sols-strongly leached surface soils over clay pans
Rendzinas-dark-colored grassland soils from soft limy material
Brown forest soils, similar to gray-brown podsolic but neutral instead of acidic
II. Zonal aridic soils*
Chernozem (black earth)
j Chestnut and reddish chestnut

Brown and reddish-brown aridics
T Reddish desert soils
I Noncalcic brown soils
Associated Intrazonal Groups
Saline soils (Solonchak)-high content of soluble salts of Ca, Mg, and Na.
Mildly alkaline
Alkali soils (Solonctz)-moderate content of soluble salts but with those of
Na dominant. Strongly alkaline
* Direction of arrow indicates a progressive reduction in effective rainfall.
The general relation of the great soil groups to climate conditions is

shown in Fig. 120, after Millar and Turk. 40 A Pedalfer soil is one in
which there is a tendency for alumina and ferric iron to accumulate in the
profile. A Pedocal soil is one in which there is a tendency for calcium
carbonate to accumulate in the profile.
Description of Great Soil Groups. In zonal humid soil regions the
climate varies from cold to tropical with rainfall greater than about
25 in./year. The vegetation is forest and grasses in warmer regions,
and mosses, lichens, and shrubs in tundra areas. The soils are character-
ized generally by a concentration of iron and aluminum in certain hori-
zons. Ordinarily there is no concentration of carbonate when the soils
have reached a mature stage.
40 Millar, C. E., and F. M. Turk, "Fundamentals of Soil Science," Wiley, New York
(1943).
336 Clay Mineralogy
12 12 12 12
24 24 24 24
36 36 36 36
48 48 48 48
TUNDRA PODZOLIC CHERNOZEMIC DESE"RTIC LATOSOL

FIG. 119. Idealized profiles of five zonal soils, from ~Wintersand Simonson.a 9 Depth
in inches.
Dry Cold Wet Cold

Tundra
I -1- -I- I Degraded
,- l
I I I
I Pedocal ChcmDzem
I Pedalfer
I I I
I I I I
Chest- I I Podsol
I I
I nut chemozeml I I
I
I Gray I I I Brown Podsolic I
I
I
~
I .....
I
I ~
Desert
~
OJ
t
I
I
Prairie
IGray Brown Podsolic t
I
I ~
_j I
I I
I ~ ~ I
1
I
I l2
Qj I
I
~ Red and Yellow I
I Reddish I Podsolic I
I -<:::: Chestnut I ,- I
I ~
"tS
I
Reddish
I
I
I
I I I
I
I
Red
Desert
~ I Prairie I
I Yellow-brown I
I I
Lateritic I
I I I Red-brown I
I Shantunq I I Lctcritic
I Brawn I I I
L _j L
Laterile J
Dry Hot Wet Hot
FIG. 120. Classification of soils in relation to climatic conditions, after Millar and
Turk. 40
Tundra Soils. These soils are characterized by an accumulallbn of

organic matter giving a peaty surface overlying a bluish-gray sticky and
compact subsoil. The frozen subsoil causes poorly drained, boggy con-
ditions for most of the year. The clay-mineral content of tundra soils
is frequently rather low.
338 Clay Mineralogy
Podsolic Soils. Such soils are developed under cool and damp climatic
contiitions providing an abundant surface accumulation of organic mate-
rial. Forests are the most common vegetative cover, although podsoliza-
tion does occur with grass cover. The covering vegetation must be such
as to yield on decomposition organic acids and other compounds of great
potency, and there must be enough rainfall to provide active leaching
conditions. The environment is acid, 'and pH values as low as 3.5 are
common in the humus and 4.0 to 4.5 in the clay component.
Under such conditions carbonates are quickly dissolved and alkaline
and alkaline earths are removed from the profile. The base saturation
ranges from about 20 to 80 per cent and is commonly 40 to 60 per cent.
The organic and inorganic colloidal components are translocated from the
upper horizon downward to about the water table. In addition leaching
conditions are so intense that the inorganic colloidal complex becomes
unstable in the surface horizons, and the iron ~illd alumina are carried
away to be reprecipitated at about the water table. As a consequence
of the removal of aluminum and iron, a gray, highly siliceous, acid horizon
develops directly beneath the surface aC'cumulation of organic material.
Because of the gray color and ashy appearance of the highly siliceous hori-
zon, the term podsol (Russian pod, meaning under, and zola, ash) has
been applied to such soils. An example of a well~developed podsol profile
is shown in Fig. 117.
Brown podsolic soils are found in the Northern Hemisphere to the south
of the true podsols, and where leaching is less intense. The siliceous
gray A2 horizon (Fig. 117) is replaced by a friable light brown horizon
containing considerable organic material. There is little breakdown of
the colloidal silicates and little translocation of either the colloidal silicates
or the aluminum and iron.
Gray-brown podsolic soils lie to the south of the brown podsolic soils in
the Northern Hemisphere and show very mild leaching, primarily because
of a vegetative cover which yields less-potent organic decomposition
products and a climate which yields a smaller quantity of such products
because of more complete oxidation. ' The gray siliceous layer is replaced
by a gray or yellow-brown horizon, and considerable exchangeable cal-
cium remains on the clay minerals. There is some downward movement
of colloids, but the organic acids are too feeble to cause appreciable dis-
integration of the silicates.
Red and yellow podsolic soils are transitional with laterite soils in warm
areas and have formed under mild climate, abundant rainfall, and a forest
cover which is either mixed or largely deciduous. Leaching is marked,
but the breakdown of the silicates is small. There is some translocation
of colloidal silicates. The relatively rapid oxidation of the iron checks its
downward movement, so that it tends to remain in the upper part of the
profile.
Prairie soils are sometimes classed separately from the podsols. They
form under climatic conditions similar to the gray-brown podsols but
under grass instead of forest cover. Apparently the grass cover produces"
decomposition products of less potency than does the forest cover. The
A horizon is extremely rich in organic materilil, and downward leaching
is moderate. Long times are required to remove the exchangeable cal-
cium and other ions, and there is little breakdown of the silicates. There
is frequently considerable downward movement of colloidal silicates to
form impervious horizons (pans) in the profile.
Laterite Soils. These soils develop under tropical conditions of high
temperature and high rainfall that is often seasonal. The primary sil-
icates are quickly broken dowll, and the extensive rainfall causes the
Gray Chest- Chern-

Desert Brawn prairie Podsol
Desert nut ozem
I[
Hot Dry Cool Wet
~r:-- ~ 0/0/" ,<lcc{/,

CorOO/7.~V/o/;-OI7
r:Concen tration
oflOrqanic
IAIA2
""
Ol'~
Material B
FIG. 121. Zones of concentration of organic material and carbonate in relation to

soil types and climate, after Millar and Turk. 40
quick removal in solution of any alkalies and alkaline earths. Iron tends
to be oxidized, particularly under seasonal rainfall, and its movement
along with alumina is retarded. The organic material is very rapidly
oxidized so that it does not accumulate, and downward-seeping waters
carry little organic acid. The downward-moving waters are almost neu-
tral, or more often slightly on the alkaline side as a result of the presence
of alkalies or alkaline earths from the primary silicates. Under such
conditions of slight alkalinity the silica tends to be dissolved and removed,
with attendant concentration of the alumina and ferric oxide in the
upper part of the weathering zone.
Zonal Aridic Soils. In zonal aridic soil regions the rainfall is less than
about 25 in./year, and the vegetative cover is grass or desert plants.
There is little or no disintegration of the silicates or migration of the
colloidal components. These soils are characterized by the movement
of carbonate and by its concentration in some particular horizon. As the
climate becomes drier, the zone of carbonate concentration gradually
comes closer to the surface (Fig. 121).
340 Clay Mineralogy
Chernozem soils are found in areas of 12 to 25 in. of rainfall under grass
cwver and in a cool climate. Because of the cool climate, the organic
material is oxidized very little, and a thick black surface horizon rich in
organic material develops. There is little disintegration of the silicates,
and leaching is very mild. The leaching removes the carbonates from
the A2 horizon, and they are reprecipitated and concentrated in the B
horizon.
The chestnut-brown aridic soils are found where rainfall is 10 to 15
in./year and the climate is warm. The vegetation is scant, and as a
result the surface color is gray brown, instead of black as in Chernozems.
Also the low rainfall permits slight leaching, and the horizon of carbonate
concentration is close to the surface.
Gray or red desert soils develop under conditions of high aridity and in
the general absence of a cover of vegetation. Because of the scant rain-
fall there is no downward leaching. The dominant movement of moisture
through the soil is upward, owing to the dominance of evaporation, with
the accumulation of carbonates and other salts at the surface.
Clay-mineral Composition of Soils. An early study of the clay-mineral
composition of soils of various groups by Alexander, Hendricks, and
Nelson 41 indicated that kaolinite is the dominant clay-mineral component
of red and gray podsolic soils, with illite frequently present but in quan-
tities usually less than about 10 per cent. In Prairie soils both illite and
kaolinite are usually present, with the former slightly the more abundant.
Small amounts of montmorillonite are often present in the hardpan hori-
zon of Prairie soils. In the series of Chernozem soils studies by these
investigators, illite was the dominant component, and kaolinite and mont-
morillonite were also present in about equal amounts; and in a series of
Desert soils from California, illite and montmorillonite were the dominant
clay minerals.
Knox: 42 has recently made a compilation of the clay-mineral determina-
tions reported in the literature for about 100 soils. Kaolinite was the
only clay mineral reported in mOJ,'e than half of a total of over 20 red and
yellow Podsols. In the remainder kaolinite was the dominant compo-
nent, but small amounts of illite and montmorillonite were also reported.
Kaolinite was reported as the dominant constituent in 6 out of 12 Gray
Brown Podsols that have been studied. The remaining 6 Gray Brown
Podsols are reported to contain illite and montmorillonite in varying
amounts, as well as kaolinite. The analyses of 6 Planosols were reported
in the literature, and they all contained montmorillonite. In all but one
of the Planosols, illite was also present in about equal abundance with the
41 Alexander, L. T., S. B. Hendricks, and R. A. Nelson, Minerals Present in Soil
Colloids, II, Estimation in Some Representative Soils, Soil Sci., 48, 273-279 (1939).
42 Knox, E., personal communication.
montmorillonite, and in 2 of them small amounts of kaolinite were present
also. A series of 3 Rendzina soils all showed illite, kaolinite, and mont-
morillonite, .but in varying amounts. Kaolinite was the component of
all the laterites whose clay-mineral composition was recorded.
The clay-mineral analyses of 15 Prairie soils were compiled by Knox,
and illite is reported in 11, montmorillonite in 12, and kaolinite in only 4
of them. It seems clear from Knox's data that illite and montmorillonite
are the characteristic clay minerals in Prairie soils.
Knox's work shows definitely the predominance of montmorillonite and
illite in the soils of arid regions. Montmorillonite was recorded in all 7
Chernozem soils that have been reported. Illite is reported in 4 of the
Chernozems, with small amounts of kaolinite in only 2 of them. Illite is
reported as the only clay-mineral component in 5 out of 10 Solonetz soils
that have been analyzed. Montmorillonite is reported in 5 of these soils
and is the dominant component in 2 of them. No kaolinite was found in
any of the Solonetz soils. In a series of 5 soils classed as White Alkali and
Black Alkali, illite was present in all of them and was the only clay mineral
in 3 of them. Montmorillonite was present in 2 of these alkali soils, and
kaolinite was completely absent in all of them.
Recently Winters and Simonson s9 have published compilations of data
on the clay-mineral composition of the B horizon of the great soil groups
which agree with the findings of Knox. According to them the composi-
tion of the B horizon of various soil groups is as follows: Tundra soils
contain mainly mica 43 and/or montmorillonite types of clay minerals;
podsolic soils contain dominantly illite, with lesser amounts of mont-
morillonite and occasional minor amounts of kaolinite; lateritic soils are
dominantly kaolinitic with minor amounts or none of the other clay min-
erals; chernozemic soils are largely composed of illite and montmorillonite;
and Desert soils contain largely montmorillonite and illite clay minerals.
Investigations by Millot 44 and by the author indicate that the paly-
gorskite-sepiolite clay minerals are likely to be significant components of
Desert soils.
A further generality of importance regarding soils is reported by
Beavers.45 According to him, the organic material in many soils accu-
mulates only up to a certain maximum amount. Additional quantities
in excess of the equilibrium amount do not tend to persist. He has found
that the attainable maximum may be relatively high for montmorillonite
soils and that it is low for kaolinitic soils.
43 The authors refer to a 2: 1 type of clay mineral, which would include chlorite,
illite, vermiculite, and montmorillonite clay minerals.

44 Millot, G., Relations entre la constitution et la genese des roches sedimentaires
argileuses, "Geol. appliq. et prosp. min." vol. II, Nancy, France (1949).
4& Beavers, A. H., personal communication.
342 Clay Mineralogy
(DiScuSSion of Weathering Products Formed from Various Types of
Rock llnder Varying Conditions. Adequate information on the composi-
tion of soils, the specific influence of the various factors of weathering,
and the environmental conditions favorable for the formation and occur-
rence of particular types of clay minerals is not available for complete and
positive generalizations on the formation and occurrence of the clay min-
erals during weathering. However, the data are adequate for some gen-
eralizations which appear reasonably well established. These may be
summarized as follows:
~When a basic igneous rock contains considerable magnesium and
weathers under conditions which, because of poor drainage or low rainfall,
permit the magnesium to remain in the weathering zone after it is released
by the breakdown of the parent mineral, montmorillonite will be the
alteration product. If, however, because of high rainfall and good drain-
age, the magnesium is removed as soon as it is released from the parent
mineral, kaolinite will be the weathering product. If the leaching of the
magnesium is relatively rapid, montmorillonite will be an initial stage of
weathering and kaolinite a later stage. Under long-continued weathering
by processes removing magnesium, where such processes evolve in a cold,
wet climate producing an abundance of organic acids (e.g., podsolic altera-
tion) and hence a potent, acid, leaching environment, the clay minerals
will ultimately be broken down, with the aluminum and iron carried
downward and the silica concentrated near the surface. Under long-
continued weathering by processes removing magnesium, but evolving
in a wet, hot climate producing no organic acids and hence a neutral or
slightly alkaline environment (lateritic alteration), the silica is carried
away, and iron and alumina are concentrated at or near the surface.
An acid igneous rock containing considerable quantities of potassium
as well as magnesium, under weathering conditions permitting the potash
and magnesia to remain in the weathering environment after the break-
down of the parent minerals, will yield illite and montmorillonite as the
alteration products. If the content of magnesium is low, illite will be the
only product, and if the content of potash is low, montmorillonite will be
the only product. Rapid removal of the potash and magnesia leads to
the formation of kaolinite. The same sequence of events noted for long-
continued weathering of the basic igneous rocks will apply to the acid
igneous rocks. The other alkalies and alkaline earths which may be
present appear to playa relatively minor role, except that there is a sug-
gestion that calcium favors the formation of montmorillonite and that
kaolinite does not form so long as there is any appreciable calcium in the
environment.
It will be shown later that the clay-mineral components, and thus per-
haps the chief silicate components, present in calcareous sediments
(including dolomites) are illites and montmorillonites. In the weather-
ing of calcareous sediments, there is substantially no alteration of the sj.l-
icates until the carbonate is completely broken down and the calcium
removed from the environment. After the removal of the calcium,
alterations may be podsolic, with the development of a silica-rich surficial
material, or lateritic, with the development of a zone of aluminum and
iron concentration.
The weathering of a shale will progress about like that of the acidic
igneous rocks. In the case of a calcareous shale, the calcium must be
removed before there is any appreciable alteration of the primary silicates.
In the weathering of kaolin clays in which there is no appreciable con-
tent of alkalies or alkaline earths, there can be no formation of illite or
montmorillonite, since potash and magnesia are necessary for their forma-
tion, unless some alkalies or alkaline earths are brought into the weather-
ing zone, as, e.g., by the upward movement of moisture under hot, dry
climatic c0nditions.
In summary, then, a significant condition in weathering is the presence
of alkalies and alkaline earths, particularly potassium and magnesium,
in the environment of alteration, and the length of time they remain in the
environment after their liberation from the parent minerals. The kind
of alkali or alkaline earth is also important, since potash leads to the for-
mation of illite, magnesium to the formation of montmorillonite, and cal-
cium probably to the formation of montmorillonite, with an added tend-
ency to block the formation of kaolinite. Carbonate tends to retard the
disintegration of the primary silicates.
The pH of the water moving through the soils is also important, par-
ticularly under conditions of active leaching and long periods of time.
Under the latter conditions, acidity favors the removal of iron and alumi-
num and the concentration of silica, whereas neutral or alkaline conditions
favor the accumulation of iron and aluminum near the surface.
Reversion of Weathering Cycle. A drastic change in the environment
of alteration can lead to the regeneration of clay minerals or to the forma-
tion of clay minerals more characteristic of an earlier stage of weathering.
Hardy et al. 46 and Alexander et alY have shown that gibbsite developed
in lateritic soils may be resilicated with the formation of kaolinite.
Regeneration of the clay minerals can be a result of man-made activi-
ties. Volk 48 and Wood 49 have shown independently that illite is formed
46 Hardy, F., and G. Rodrigues, Soil Genesis from Andesite in Grenada, British
West Indies, Soil Sci., 48, 361-383 (1939).

47 Alexander, L. T., S. B. Hendricks, and G. T. Faust, Occurrence of Gibbsite in
Some Soil-Forming Materials, Soil Sci. Soc. Am. Proc., 6, 52-57 (1941).
48 Volk, G., Nature of Potash Fixation in Soils, Soil Sci., 45, 263-276 (1938).
49 Wood, L. K., The Chemical Effects of Soluble Potassium Salts on Some Illinois
Soils, Ph.D. thesis, University of Illinois (1941).

344 Clay Mineralogy
in soils under certain conditions as a consequence of the addition of potash
feJ,tilizers. Further, the formation of illite may take place quickly (in a
matter of a year or so), illustrating the rapidity with which clay materials
can adjust themselves to environmental changes.
Occurrence of Halloysite and Allophane in Soils. Several investigators
have recorded the presence of halloysite in weathering products. Thus
Bayley 50 describes halloysite in some deep kaolins of North Carolina and
ascribes its presence to weathering processes. Bates 1 has confirmed the
identification of halloysite in these same kaolins. Alexander et al. 52 and
Allen 53 have described weathering products which contain halloysite.
Although the presence of halloysite in weathering products is well estab-
lished, it is a rare component of such materials, and peculiar conditions
must be required for its formation. Alexander et al. state that plagioclase
feldspar in a neutral or slightly acid environment and in the presence of
water can weather to the hydrous form of halloysite.
Some few soil materials have been found by the author which contain
considerable quantities of poorly crystalline or amorphous components.
No detailed study has been made of these materials, and nothing is known
of the conditions which determine their origin.
Nature of the Alteration Process. Correns and Engelhardt 54 have
concluded that weathering proceeds by the breakdown of the parent min-
eral to ionic solutions, and that secondary minerals are reaction products
of such ionic solutions. Niggli 55 states that, in the alteration of feldspar,
under certain conditions, the breakdown reaches a colloidal state before
the formation of the secondary products. Demolon and Batisse 56 have
concluded that in the weathering of a granite there is a spontaneous
change to the clay minerals.
It is likely that the transition from a primary mineral or minerals to a
clay-mineral weathering product does not always take place in exactly
the same way. It can probably take place by means of the formation of
ionic solutions with the later precipitation of the clay minerals, by the
50 Bayley, W. S., Kaolin in North Carolina, with a Brief Note on Hydromica, Ecan.
Geal., 16, 236-246 (1920).

51 Bates, T. F., personal communication.
52 Alexander, L. T., G. F. Faust, S. B. Hendricks, H. Insley, and H. F. McMurdie,
(1943).
53 Allen, V. T., Formation of Bauxite from Basaltic Rocks of Oregon, Ecan. Geal.,
43, 619-626 (1948).
54 Correns, C. W., and W. Engelhardt, Neue Untersuchungen iiber die Verwitter-
ungen des Kalifeldspates, Chem. Erde, 12, 1-12 (1938).

55 Niggli, P., Zusammensetzung und Klassification der Lochersteine. Vortag
Gehalten am Erdbauku,.rs, Eidgenoss. Tech. Hach. Zurich, 6, (1939).
b6 Demolon, A., and E. Batisse, First Changes of the Spontaneous Change of a
Granite and the Origin of Clayey Colloids, Campt. rend. 223, 115-118 (1946).
formation of colloidal mixtures with later crystallization of the clay min-
erals, or by a direct structural change from the parent mineral to the clay
mineraL The nature of the transition would depend on the parent min-
eral, the resulting clay mineral, and the environmental conditions. Thua
kaolinite formed by the silication of gibbsite would perhaps be the reac-
tion product of ionic or colloidal solutions and a solid, whereas the transi-
tion of muscovite and biotite to illite would almost certainly take place
by a slight structural rearrangement without any breakdown to ionic or
colloidal components.
Synthesis of Clay il!inerals

Bowen, N. L., and O. F. Tuttle, The System MgOSi0 2 H 20, Bull. Geol. Soc. Am., 60,
439-460 (1949).
Dittler, E., Zur Frage der Kaolinbildung, Z. anorg. u. allgem. Chem., 24, 33-40 (1933).
Gruner, J. W., Hydrothermal Alteration of Montmorillonite to Feldspar at Temper-
atures from 245C to 300C, Am. Mineral., 21, 511-515 (1936).
Longuet, J., Synthese de silicates de nickel, magnesien et cobalt prcsentant des struc-
tures du type kaolinit-antigorite, Compt. rend., 226, 869-872 (1947).
Morey, G. W., and E. Ingerson, The Pneumatolytic and Hydrothermal Alteration and
Synthesis of Silicates, Econ. Geol., 32, 607-760 (1937).
Permjakov, V. M., The Hydrothermal Synthesis of Kaolin, abstracted Chem. Zentr.,
p. 1553 (1938).
Schwarz, R., and G. Trageser, Ueber die kunstliche Umwandlung von Felspat in
Kaolin, Z. anorg. u. allgem. Chem., 216, 190-200 (1933).
Yusopova, S. M., The Characteristics of Montmorillonite, Compt. rend. acado sci.
URSS, 61, 631-634 (1946).
Hydrothermal Alteration
Alderman, A. 1., Clay derived from Sillimanite by Hydrothermal Alteration, Mineralog.
Mag., 29,271-279 (1950).
Anderson, C. A., Alteration and Metallization in the Bagdad Porphyry Copper
Deposit, Arizona, Econ. Geol., 46, 609-628 (1950).
Burbank, 'V. S., Problems of Wall Rock Alteration in Shallow Volcanic Environments,
Quart. Colorado School Mines, 46,286-326, January, 1950.
Butler, B. S., Influence of Replaced Rocks on Replacement Minerals, Econ. Geol., 27,
1-27 (1932).
Butler, B. S., and J. Vanderwilt, The Climax Molybdenum Deposit, Colorado, U.S.
Geol. Survey Bull. 846C (1933).
Chuckrov, F. V., and N. A. Kozlova, Halloysites from the Deposits Berkara and
Karagaily, Compt. rend. acado sci. URSS, 27, 1011-1012 (1940).
Da Silva, D. J., Montmorillonite dans des pegmatites granitiques et Ie probleme de
la montmorillonitisation, Portugal Servo Fom. Min., 4, 48-52 (1948) .
Jung, H., Montmorillonit von Dolmai bei Meiningen, Chern. Erde, 11,217-222 (1937).
Kerr, P. F., Alteration Studies, Am. Mineral., 32, 158-162 (1947).
Lindgren, W., Copper Deposits of the Clifton-Morenci District, Arizona, U.S. Geol.
Survey Profess. Paper 43 (1905).
Lovering, T. S., The Origin of the Tungsten Ores of Boulder, Colorado, Econ. Geol.,
36, 229-279 (1941).
346 Clay Mineralogy
Lovering, T. S., Geochemistry of Argillic and Related Types of Alteration, Quart.
Colorado School Mines, 45, 231-260, January, 1950.
Pri"er, R. T., and W. F. Cole, The Alteration Products of Olivine and Leucite in
Leucite Lamprophyres from the West Kimberly Area, \Vestern Australia, Am.
Mineral., 27,373-384 (1942).
Ross, C. S., and W. F. Forbag, Anauxite, a Mineral Species, Based on Material from
Bilin, Czechoslovakia, Am. Mineral., 13, 153-155 (1928).
Sales, R. H., and Meyers, C. F., Interpretation of Wall Rock Alteration at Butte,
Montana, Quart. Colorado School Mines, 45,261-274, January, 1950.
Schwartz, G. M., Hydrothermal Alteration of Igneous Rocks, Bull. Geol. Soc. Am.,
50, 181-237 (1939).
Tomlinson, W. H., and A. J. Meier, On the Origin of Montmorillonite, Am. Mineral.,
22, 1124-1126 (1937).
Soils and Weathering
Barshad, I., A Pedologic Study of California Prairie Soils, Soil Sci., 61, 423-442 (1946).
Brown, A. L., and A. C. Caldwell, Clay Mineral Content of Some Colloidal Material
Extracted from a Solodi Soil Profile, Soil Sci. Soc. Am. Proc., 11, 213-215
(1916).
Clark, G. L., F. F. Riecken, and D. H. Reynolds, X-Ray Diffraction Studies of Two-
micron Fractions of Some Genetic Soil Profiles, Z. Krist., 96, 273-286 (1937).
Coleman, R., and M. L. Jackson, Mineral Composition of the Clay Fraction of Several
Coastal Plain Soils of Southeastern United States, Soil Sci. Soc. Am. Proc., 14,
381-391 (1945).
De Leenheer, L., Les Propri6tcs sorptives des sols et leur interpretation mineralogique,
avec application aux sols argileux des polders marin, Am. Soc. Geol. Belg. Bull., 57,
299-320 (1948).
Edelman, C. H., Relation entre les proprietes et la structure de quelque mineraux
argileux, Verre silicates indu., 12,3-6 (1947).
Favejee, J. C., Quantitative X-ray Analysis of Some Dutch Soils, Medediel, Land-
bouwhoogeschool, 43 (1939).
Feustel, I. C., A. Dutilly, and M. S. Anderson, Properties of Soils from North Amer-
ican Arctic Regions, Soil Sci., 48, 183-199 (1939).
Gieseking, J. K, The Clay Minerals in Soils, Advances in Agron., 1, 159-204 (1949).
Hardon, H. J., and J. C. Favejee, Quantitative X-ray Analysis of the Clay Fraction
of the Principal Soil Types of Java, Medediel., Landbouwhoogeschool, 43 (1939).
Jacob, A., The Composition of the Clay Fraction of Soils, Bodenkunde u. Pjlanzener-
nahr., 29,219-231 (1943).
Jenny, H., "Factors of Soil Formation," McGraw-Hill, New York (1941).
Kelley, W. P., and W. H. Dore, The Clay Minerals of California Soils, Soil Sci. Soc.
Am. Proc., 2, 115-120 (1937).
Kelley, W. P., W. H. Dore, A. O. Woodford, and S. M. Brown, The Colloidal Constit-
uents of California Soils, Soil Sci., 48,201-255 (1939).
Kelley, W. P., W. H. Dore, and J. B. Page, The Colloidal Constituents of American
Alkali Soils, Soil Sci., 51, 101-124 (1941).
Larson, W. E., W. H. Allaway, and H. F. Rhoades, Characteristics of Three Soils from
the Chernozem and Chestnut Soil Regions of Nebraska, Soil Sci. Soc. Am. Proc.,

12, 420-423 (1947).
MacEwan, D. M. C., Les Mineraux argileux quelques sols ecossais, VeTTe silicates
indu., 12, 13-18 (1947).
Mattson, S., The Laws of Soil Colloidal Behavior, IX, Amphoteric Reactions and
Isoelectric Weathering, Soil Sci., 34, 209-239 (1932).
Nagelschmidt, G., The Mineralogy of Soil Colloids, Imp. Bur. Soil Sci. Tech. Com-
mun. 42 (H)44).
Nikiforoff, C. C., R. P. Humbert, and J. G. Cady, The Hardpan in Certain 86ils of
the Coastal Plain, Soil Sci., 65, 135-153 (1948).
Pearson, R. 'V., and L. E. Ensminger, Types of Clay Minerals in Alabama Soils,
Soil Sci. Soc. Am. Proc., 13, 153-156 (1948).
Peterson, J. B., Relation of Parent Material and Environment to the Clay Minerals
in Iowa Soils, Soil Sci., 61,465-475 (1946).
Sedletsky, I. D., Mineralogic Composition of Clays and Their Genesis, Soviet Geol.,
8, 82-89 (1940).
Shearer, J., X-ray Analysis of some Tasmanian Soil Colloids, J. Roy. Soc. W. Australia,
26, (1940).
Walker, G. F., The Decomposition of Biotite in Soils, Mineralog. Mag., 28, 693-703
(1949).

CHAPTER 14
Origin and Occurrence of the Clay Minerals (Cont'd)
RECENT SEDIMENTS
Only a very limited study has been made of the clay-mineral composi-
tion of sediments accumulating in various environments at the present
time. Unfortunately many of the determinations that have been made
are incomplete or are based on inadequate analytical data. Recent sedi-
ments frequently are composed of complex mixtures of relatively poorly
crystallized minerals, and their analysis is particularly difficult. Unless
the clay-mineral analyses are made by a combination of X-ray-diffraction,
differential thermal, and chemical methods and unless the procedures are
detailed, the completeness of the results is questionable. For example,
the presence of montmorillonite in moderate amounts usually cannot be
determined unless the X-ray analyses are supplemented by the glycol-
treatment procedure. Nevertheless some trustworthy information is
available on the clay-mineral composition of Recent sediments, and it is
adequate for a preliminary consideration of di[1genetic changes in clay
minerals and of the relation of clay-mineral types ,,0 particular environ-
mental conditions of accumulation.
Marine Environment. On the basis of their Challenger studies, Murray
and Renard 1 suggested that the fine-clay fraction of the sediments accu-
mulating on the present ocean floors was noncrystalline. At that time
the idea that many clays were composed of amorphous components was
widely held (see Chap. 2), largely because there were no adequate tech-
niques available to study them.
Using -modern analytical methods, Revelle 2 in 1936 and Correns 3 in
1937 first showed that the fine-clay fraction of Recent marine sediments
is substantially all crystalline. Correns 3 and his colleagues found that
mica (probably illite) was present in all the bottom samples collected by
1 Murray, J., and A. F. Renard, Reports on Deep-Sea Deposits Based on Specimens

Collected during the Voyage of H.M.S. Challenger in the Year 1872-1876, Challenger
Rcports (1891).
2 Revelle, R., Marine Bottom Samples Collected in the Pacific Ocean by the Carnegie
on its Seventh Cruise, Ph.D. thesis, University of California (1936).

3 Correns, C. W., et al., Die Sedimente des aquatorialen atlantischen Ozeans,
Deutsche atlantische Exped. Meteor 1925-1927, Wiss. Ergeb., 3, Teil 3 (1937).

348


the Meteor in the South Atlantic. In addition to the mica, the clay
fraction in some of the samples was composed predominantly of kaolinite, J
in others of halloysite, and in a few of montmorillonite. A map was pre-
sented showing the distribution of the predominant clay minerals, omit-
ting from consideration the mica. Montmorillonite was found to be
limited to relatively small regions where there were neighboring areas of
basic volcanic activity. Correns himself pointed out that the differentia-
tion of halloysite and kaolinite was not certain, and the presence of halloy-
site in Recent marine sediments has not yet been definitely established.
It seems probable from his data that the sediments in the South Atlantic
which he studied are chiefly mixtures of the illite and kaolinite types of
clay minerals. Montmorillonite appears to be limited to certain
areas, but this may be questioned; i.e., montmorillonite may be
more widely distributed because of the great difficulty of detecting small
amounts of this mineral by the techniques available when Correns did
his work.
Some of the Meteor samples were cores, and Correns reported no varia-
tion in the clay-mineral composition of any of his cores from the top down-
ward. He found no evidence of diagenetic changes in the clay-mineral
components of his samples.
Leinz 4 in a further study of the Meteor samples reported that all sam-
ples of Blue Clay showed mica and that most of them showed also a
kaolinite type of mineral. Montmorillonite was found in some of the
Blue Clay samples. All samples of globigerina ooze also contained mica,
but only a few showed a kaolinite type of mineral reported as halloysite.
Montmorillonite was also found in some of them.
Dietz" determined the clay-mineral composition of 39 samples collected
from various oceans of the world. Most of his samples were collected by
the Challenger. Dietz found illite in all his samples, and it was the pre-
dominant clay mineral in 23 of them. Illite was particularly abundant
in the deep-sea clays. Kaolinite, although generally less abundant than
illite, was present in all the samples and predominant in 9 of them, most
of which were relatively close to the shore. Small amounts of montmoril
lonite were found in the near-shore samples, but this mineral was not
detected in most deep-sea sediments. Again it should be stated thai.
adequate methods for detecting small amounts of montmorillonite in mix-
tures of poorly crystallized material had not been developed when Dietz
.did his work.
Dietz" presented strong evidence for the development of illite on the
Leinz, V., Die Mineralfazies der Sediments des guinea Beckens, Deutsche atlanti-
sche Expedition Meleor 1925-1927, Wiss. Ergeb., 3, Teil 3,245-262 (1937).
& Dietz, R. S., Clay Minerals in Recent Marine Sediments, Ph.D. thesis, University
of Illinois (1941).
350 Clay Mineralogy
sea floor in sediments accumulating at the present time. It was his con-
ecept that the illite develops largely by the alteration of montmorillonite
from the source material being carried to the sea.
Revelle'ss study of the Carnegie samples showed the finest clay fraction
to be crystalline, but unfortunately his analytical data did not permit a
positive identification of the predominant clay minerals. In a few sam-
ples kaolinite or halloysite was identified definitely, but most of the sam-
ples were said to be composed largely of a "beidellite-like" clay mineral.
From Revelle's data it appears certain that most of his samples were mix-
tures of clay minerals, and that in many, if not most of them, illite was
the predominant clay mineral.
Bramlette and Bradley 7 made no serious attempt to determine the clay-
mineral components in their study of the samples collected by the Lord
Kelvin in the North Atlantic. They pointed out that the clay fraction
possessed optical properties similar to those of beidellite or hydrous mica,
and from their data it appears probable that their clay samples were a
mixture of clay minerals and that illite was an important component in
many, if not all, of them.
Keunen 8 obtained X-ray-diffraction data for the clay fraction of a few
of the samples collected by the Snellius in the Netherlands Indies area.
He found muscovite (probably illite), kaolinite, and montmorillonite in
each of the samples investigated. The relative amounts of the clay min-
erals varied in the different samples, but too few samples were studied to
permit any correlation of clay-mineral composition with type of sediment
or oceanographic conditions.
Grim, Dietz, and Bradley 9 studied a series of samples collected by the
E. W. Scripps from the Gulf of California and the Pacific Ocean off the
California coast. They found that all the samples contained illite, mont-
morillonite, and kaolinite and that generally illite was the most abundant
and kaolinite the least abundant clay-mineral type. A chloritic clay
mineral could be identified definitely in some samples. The clay minerals
were very complex mixtures, including mixed crystallizations as well as
mechanical mixtures of discrete phases. In general the crystallinity was
lower, the individual size smaller, and the intergrowth more intimate in
6 Revelle, R., Scientific Results of Cruise VII of the Carnegie during 1928~1929,
Carnegie Inst. Wash. Pub. 556 (1944).
7 Bramlctte, M. N., and W. H. Bradley, Geology and Biology of North Atlantic
Deep Sea Cores between Newfoundland and Ireland, U.S. Geol. Survey Profess. Papm
196A (1940).
8 Keunen, P. H., "The Snellius Expedition in the Eastern Part of the Netherland
East Indies," vol. V, part 3, Brill, Leidcn (1943).

9 Grim, R. K, R. S. Dietz, and W. F. Bradley, Clay Mineral Composition of Some
Sediments from the Pacific Ocean off the California Coast and the Gulf of California,
Bull. Geol. Soc. Am., 60, 1785~1808 (1949).
"".
- Origin and Occurrence of the Clay Minerals 351
these Recent sediments than in ancient sediments which these authors
have studied.
Grim and his colleagues conclude from their data that kaolinite is vefy
slowly being lost in these marine sediments, probably by alteration to
illite or chloritic mica. They postulate that potassium is being taken up
by the accumulating sediments mostly by" degraded illite," which is illite
which reaches the environment of deposition with a relatively low potas-
sium content as a result of weathering processes, and that magnesium is
also being taken up by the accumulating sediments, probably in the for-
mation of illite or chloritic micas. The widespread occurrence of mont-
morillonite indicates that this clay mineral is not lost completely or
quickly in diagenetic processes.
Diagenetic Processes. The marine environment is alkaline; there is no
leaching; and the water contains a good deal of dissolved calcium. These
environmental conditions favor the formation of a montmorillonite, illite,
or chlorite clay mineral rather than kaolinite. Millot lO has recently pre-
sented evidence from a study of the composition of many sediments to
show that the presence of Ca++ tends to block the formation of kaolinite.
A study of weathering processes indicates that kaolinite does not form
from calcareous parent material until all the carbonate has been removed.
Correns l l pointed out that in environments in which the pH is less
than 5, silica is insoluble and R 2 0 3 is relatively soluble. At pH values
from 5 to 9, the R 2 0 3 is very slightly soluble and the silica becomes more
and more soluble; and at pH values above 9, silica and alumina are both
soluble. Therefore, at low pH values, there would be little silica in solu-
tion, and kaolinite is likely to be the stable form, whereas at higher pH
values with much silica in solution, the clay minerals with higher propor-
tions of silica to R 2 0 3 are likely to be the stable form.
MilloVo has shown that silica sols are relatively stable and that a con-
siderable cation content is necessary for flocculation, whereas alumina sols
are not very stable, being flocculated by a relatively low cation content.
Therefor~, according to Millot, in an environment of low cation content,
there would be a relatively large amount of available alumina with a con-
sequent clay mineral of low silica-to-alumina ratio, e.g., kaolinite. In an
environment of high cation concentration, the available silica would be
relatively high, because of the flocculation of the alumina, with a conse-
quent clay mineral of high silica-to-alumina ratio, e.g., montmorillonite.
The land-derived muds carried into the sea from many areas probably
consist to a considerable extent of degraded illite, degraded chloritic mica,
10 Millot, G., Relations entre la constitution et la genese des roches sediment aires
argileuses, "Geol. Appliq. et Prosp. Min.," vol. II, Nancy, France (1942).
11 Correns, C. W., T. F. Barth, and P. Eskola, "Die Enstehung der Gesteine,"
Julius Springer, Berlin (1939).
352 Clay Mineralogy
and kaolinite. The degraded minerals represent material that has been
partially leached of its constituent alkalies and alkaline earths, but not
sufficiently to transform it into new minerals. Such degraded minerals
would have deficiencies of alkalies and alkaline earths and would quickly
adsorb available potassium and magnesium with an attendant increase
in the perfection of the mica structures. It seems likely that such recon-
stitution of the micas is a most important diagenetic change and one
which would take place quickly.
It seems certain that chloritic mica and illite also tend to form during
marine diagenesis from other minerals. Thus illite probably forms from
montmorillonite; chloritic material appears to develop from kaolinite and
possibly also from montmorillonite; and glauconite forms from a variety
of raw materials. It seems that these changes, which may represent
major alterations in atomic structures, would take place relatively
slowly. Thus, the diagenetic alteration of kaolinite may be so slow
that it frequently persists in varying amounts into ancient marine
sediments.
Marine conditions vary from oxidizing to reducing, depending on the
topography of the ocean floor and hence on water movement, on the rate
of sedimentation, and perhaps on other factors. It is not known exactly
how the variation from oxidizing to reducing conditions would influence
diagenetic processes. Perhaps oxidizing conditions, by tending to fix the
R 2 0 S in a relatively inert form, would retard diagenetic changes.
Variations in the concentration of potassium and magnesium would
also affect diagenesis. Deficiencies of the alkalies und alkaline earths
would retard the formation of the micas. Diagenetic changes would be
expected to be most intense near shore, where the sediments first encoun-
ter the marine environment.
The nature of the marine environment and diagenesis are such that
variations in the character of the source material might be reflected in the
sedimentary accumulation. 12 For example, a kaolinitic marine sediment
means a kaolinitic source area, since kaolinite is not formed in the sea.
Also it means relatively rapid accumulation with the persistence of
kaolinite through an environment fundamentally unfavorable for it.
A completely micaceous marine sediment means a source area produc-
ing little kaolinite, or possibly extremely slow accumulation. It seems
likely that source material containing a considerable amount of kaolinite
would give rise to a marine sediment in which at least some kaolinite _
persisted.
Glauconite. Glauconite is a rather unique illite type (see page 68) of
clay mineral that is formed during marine diagenesis. Many investiga-
12 Grim, R. E., The Depositional Environment of Red and Green Shales. J. Sedi-
ment. Petrol., 21, 226-232 (1951) .
,
."'1
tors have recorded the frequent association of glauconite with organic
residues, and it has been generally concluded that the presence of organic
material is necessary for the formation of the mineral.
Galliher13 has shown that some glauconite has formed as an alteration
product of biotite, with the alteration taking. place under reducing condi-
tions and slow deposition. Unless deposition is slow so that alteration
takes place before burial, the biotite mica persists. Hendricks and ROSS14
have reviewed the literature which shows that minerals other than biotite
may alter to glauconite in a marine environment.
Hadding, 15 who made an extremely complete study of glauconite, showed
that glauconite occurs in flakes and as pigmentary material, in addition
to the well-known large rounded pellets. According to Hadding, glau-
conite forms in the shallow sea in agitated waters which are not highly
oxygenated. It forms during times of decreased deposition of detritus
and often during negative sedimentation. Revelle 2 has also emphasized
that glauconite is currently forming off coasts lacking important rivers,
where deposition is slow.
Hendricks and Ross,14 after a careful study of the variation in the
composition of glauconites, concluded that the mineral is formed in a
reducing environment maintained by bacterial action. The environment
probably remains unchanged for long intervals of time. The very uni-
form content of magnesium and the rather constant ratio of Fe H to Fe++
indicate that a critical content of magnesium and a particular oxidation-
reduction potential might be required for the mineral to form. The ratio
of Na+ to K+ in interlayer positions in the structure of glauconite is rather
distinctive, suggesting that a certain concentration of theses ions may be
required for its formation.
In summary, glauconite forms during marine diagenesis, at times of
slow or negative sedimentation, and in relatively shallow water. In
addition, peculiar reducing conditions and particular concentrations of
magnesium and alkalies seem to be essential for its formation.
Nonmarine Conditions. Lacustrine. The paucity of clay-mineral
data for Recent nonmarine sediments is so great that few conclusions can
be reached. Cuthbertl 6 has shown that illite is the dominant clay min-
eral in sediments accumulating in Lake Erie.
It would seem that there would be few diagenetic changes in fresh-
13 Galliher, E. W., Glauconite Genesis, Bull. Geol. Soc. Am., 46, 1351-1366 (1936).
14 Hendricks, S. B., and C. S. Ross, Chemical Composition and Genesis of Glauconite
and Celadonite, Am. Mineral., 26, 683-708 (1941).
15 Hadding, A., The Pre-Quaternary Rocks of Sweden, IV, Glauconite and Glau-
conite Rocks, Medd. Lunds. Geol.-Mineral. Inst., 61, (1932).
16 Cuthbert, F. L., Clay Minerals in Lake Erie Sediments, Am. Mineral., 29, 378-388
(1944).
354 Clay Mineralogy
water lake sediments and that the sediments would reflect the source
maJerial supplied to the lake. Lacustrine environments in which there
is any appreciable movement of water through the sediments, with some
consequent removal of alkalies and alkaline earths, would be expected to
show the development of kaolinite as a diagenetic mineral. As a general
thing, lakes in which there is no accumulation of lime would probably
show a tendency toward the development of kaolinite. In lakes where
lime is accumulating, the lime would tend to block the formation of
kaolinite, and any diagnetic changes would probably be very slight.
In salt lakes, the montmorillonite and illite clay mineral would form at
the expense of whatever material was supplied. Investigations by Long-
chambon,17 Millot,IO and unpublished work by the author have shown
that the sepiolite-attapulgite minerals are frequently found in Recent
sediments accumulating in dry desert basins. Probably a particular com-
position, chiefly in regard to the abundance of magnesium, leads to the
formation of these minerals.
Ril'crs. Holmes and Hearn 18 have investigated the clay-mineral com
position of Recent alluvium deposited by the Mississippi River and its
tributaries, with the results shown in Table 46.
TABLE 46. l\fINERALOUICAL COMPOSITION O~' ALLUVIAL MATERIAL DEPOSITED BY

THE l\1ISSISSIPPI RIVER AND ITS TRIBUTARIES
(From Holmes and Hearn 18)
Kaolinite Montmoril- Illite

lonite
Eastern tributaries: Ohio, Cumherland, Tennessee,

Duck, and Clinch Rivers. . . . . . .. . .... I 10-20 10-15 65-75
Mississippi River. .... . . .. . ............ i 5-1.5 25-45 40-60
\Vestern tributaries: Milk,. Yellowstone, ::vlissouri, I I
Platte, and Arkansas Ibvers ................. . 10-20 20-45 40-60
I
Values given are percentages.
The clay minerals in fluviatile sediments should reflect the composition

of the soils from which they are derived. In the data of Holmes and
Hearn, there is a relatively larger amount of montmorillonite in the west-
ern tributaries of the Mississippi River, which derive their material from
dry areas which favor the formation of montmorillonitic soils.
17 De Lapparent, J., Structure et origine des terres naturelles sus.ceptibles d'etre

utilisees pour la decoloration des huiles minerale, Compt. rend. congr. rnondial pet role
Congr., Paris (1937).
18 Holmes, R. S., and W. E. Hearn, Chemical and Physical Properties of Some of
the Important Alluvial Soils of the Mississippi Drainage Basin, U.S. Dept. Agr. 'Pech.
Bull. 833 (1942).
In rivers themselves one might expect a certain amount of leaching
action, and if this is the case, kaolinite would tend to develop. Otherwise
any diagenetic changes would probably be insignificant. "
Glacial Deposits. A considerable study of the composition of the
glacial deposits in North America has been made by students of soils.
In unweathered till and loess, the dominant clay mineral is illite, fre-
quently with considerable amounts of chloritic mica. Kaolinite and
montmorillonite are occasionally present in small amounts.
The composition of varved glacial lake sediments is not well known,
but Cooling 1S" has shown that, in at least some varved sediments in
Canada, illite is the dominant clay mineral, with montmorillonite also
present in the dark layers but not in the light layers.
ANCIENT SEDIMENTS
The clay-mineral composition of sediments of widely varying modes of

origin, types, and geologic age has been determined. Unfortunately
many of these analyses are incomplete in that only the most abundant
clay minerals were determined. However, sufficient trustworthy analyt-
ical data are available for a first approximation to the relation of clay-
mineral composition and sediments of various modes of origin and of
various lithologic types.
Clay-mineral Composition in Relation to Mode of Origin. Millot lO has
recently presented thorough clay-mineral analyses of a large series of sed-
iments from eastern France and the adjacent sections of Germany and
Switzerland whose mode of origin is well established on paleontological
and stratigraphic evidence. In sediments of marine origin, Millot found
illite invariably present, making up from 50 to 100 per cent of the total
clay-mineral-content. The illite was particularly abundant in calcareous
marine sediments, in which it made up from 70 to 100 per cent of the total
clay-mineral content. Kaolinite was often present and made up from
oto 50 per cent of the clay-mineral content. Millot emphasizes the point
that calcareous sediments are likely to have little or no kaolinite and that
calcium tends to inhibit the formation of this clay mineral. Pyritiferous
marine sediments were particularly likely to contain appreciable kao-
linite. Chloritic mica and vermiculite were frequently present as minor
consti tuen ts.
In 26 sediments of lagoonal origin, illite was found to be the dominant
clay mineral in 19, montmorillonite in 4, and attapulgite-sepiolite in 3.
Chloritic mica and vermiculite were frequent minor clay-mineral compo-
nents. Kaolinite was completely absent in 19 samples and in no case
18a Cooling, L. F., Personal communication.

356 Clay Mineralogy
totaled more than 30 per cent of the clay-mineral content. It is probable
,that the sediments containing attapulgite-sepiolite represent lacustrine
rather than lagoonal conditions.
In sediments of lacustrine origin where the lake is believed to have con-
tained "aggressive" water, i.e., where there was actual leaching of alkalies
and alkaline earths in the accumulating sediments because of active move-
ment of water through them, or because of relatively low pH of the water,
the dominant clay mineral is kaolinite. In such sediments the total con-
tent of illite, chloritic mica, and montmorillonite ranged from 0 to 30 per
cent.
In sediments of lacustrine origin of the" nonaggressive" type, where
salts and carbonate are likely to accumulate because of slight movement
of water through the lake and/or relatively alkaline water, the dominant
clay-mineral components are illite, montmorillonite, and sepiolite-atta-
pulgite. The latter type of clay mineral is particularly prevalent in sed-
iments that have accumulated in dry desert lakes.
Clay Minerals in Relation to Geologic Age. Analyses of many ancient
sediments in the author's laboratory from the United States and elsewhere
have shown that montmorillonite is generally absent in sediments older
than the Mesozoic. Except for a few beds of probable bentonitic origin,
montmorillonite has not been reported in the older sediments. This
point is of particular importance, for montmorillonite is abundant in
many Mesozoic and Cenozoic sediments, in Recent marine sediments,
and in present-day weathering products.
It seems likely that metamorphic processes would cause the alteration
of montmorillonite to a mica type of mineral, and, therefore, the mineral
would tend to disappear in more ancient sediments. When the structure
of montmorillonite is completely collapsed so that all the water is removed
from between the unit layers, rehydration of the mineral is difficult, and,
the mineral has many of the attributes of mica. Further in the course of
geologic time, montmorillonite would probably slowly pick up potassium
and magnesium from ground waters, which might in turn obtain such
constituents from the disintegration of ferromagnesian minerals and
feldspars. The adsorption of potassium and magnesium would favor a
shift to a mica type of mineral. Thus in the course of time, because of
dehydration due to compaction, montmorillonite would tend to be altered
in the direction of the micas.
There is, however, another aspect to this problem of the absence of
montmorillonite in ancient sediments. It appears that certain times and
certain places in the geologic past have been either particularly favorable
or particularly unfavorable for the formation of montmorillonite. Thus,
the Cretaceous formations of the Great Plains and the Gulf Coast of the
United States contain large amounts of bentonite and montmorillonitic
,', ,
shales. Volcanic activity producing ash which altered to montmorillonite
is largely responsible for the flood of montmorillonite in these sedime'1ts.
MilloVo studied many Cretaceous sediments from France and found little
montmorillonite in them. Thus the abundance of montmorillonite in
some ancient sediments varies geographically because of peculiar, rel-
atively local conditions.
Montmorillonite is developing in many soils widely scattered over the
land today, and hence the argillaceous material carried to areas of
accumulation at the present time generally has a fairly large content of
montmorillonite. Recent sediments, therefore, have a fairly high mont-
morillonite content. This probably was not always the case in the
geologic past. An interval of geologic time without much volcanic
activity, without large areas of relatively low rainfall, and/or with large
areas of lateritic weathering would produce little montmorillonite. Thus,
because of variations of weathering conditions and/or volcanic activity,
either on a world-wide scale or in restricted areas, montmorillonite would
be expected to be absent or abundant in sediments of a particular geologic
interval or of a particular geographic area.
There is a suggestion from available data that kaolinite also is less
abundant in very ancient sediments than in those deposited after, say,
the Devonian. Kaolinite may therefore be lost in ancient sediments,
because of metamorphic processes complemented by adsorption of alkalies
and alkaline earths, with the formation of micas. This would be expected
to be a much slower process than the alteration of montmorillonite to
the micas.
Since kaolinite is formed during weathering under certain conditions
and not under others, it would be expected that ancient sediments of a
particular age and/or from a particular area might have either a flood of
kaolinite or none of it.
Since montmorillonite and kaolinite tend to disappear in sediments of
increasing age, it follows that very ancient sediments must be composed
largely of the illite and chlorite types of clay minerals. Attapulgite-
sepiolite alsQ appears restricted to relatively young sediments. No
information is available on the distribution of vermiculite in sediments.
Halloysite is generally absent in sedimentary rocks, except possibly in
association with some coals. It has been suggested that the halloysite
minerals would change under metamorphic processes to kaolinite, but
there are no specific data on this point.
Clay Minerals in Relation to Lithology. Shales. The dominant clay
mineral in many, probably most, shales that have been studied is illite.
Chloritic mica is frequently present, and montmorillonite is a common
component in many shales of Mesozoic or younger age. Kaolinite is a
common component of some shales, but usually in minor amounts. In
358 Clay Mineralogy
shales of relatively high kaolinite content, the lamination is frequently
not well developed.
Slates. The mica clay minerals are the dominant components in the
slates that have been studied. Fairbairn 19 reports kaolinite in some
slates and finds that kaolinitic slates have relatively low schistosity.
It appears that the illite and chloritic clay minerals in the slates have a
higher crystallinity than these same clay minerals in clays. This means
that the clay minerals are composed of layer units more regularly oriented
and with fewer structural defects. The increased crystal perfection
developed in the metamorphism of illite yields a material approaching
muscovite or biotite, depending on the original composition and structure
of the illite.
No montmorillonite, attapulgite-sepiolite, or halloysite clay minerals
have been reported in slates. Vermiculite may occur in some slaty
materials, but it is usually difficult to establish that it is not an alteration
product.
Clays. Material which can be classed as clay has been described con-
taining each of the clay minerals in substantially pure form. Except for
mixtures of kaolinite with sepiolite-palygorskite, any combination of the
various clay minerals has been described in such sediments.
Underclay. Many beds of coal, particularly of Carboniferous age, the
world over are immediately underlain by a nonbedded carbonaceous clay
characterized by many small slickensided fracture surfaces. Such
material, frequently called underclay, contains either kaolinite or illite,
or a mixture of these clay minerals. Occasionally the underclay is
calcareous, and then illite is usually the only clay-mineral component. 20
Brindley and Robinson 21 have shown that the English under clays con-
tain kaolinite of the poorly crystalline variety. The kaolinite in the
plastic under clays in Illinois is also the poorly crystalline variety. In
several instances very small amounts of dickite 22 have been identified in
underclays. This dickite could not have had the usual hydrothermal
origin for the mineral, and its genesis is obscure.
Flint Clay. Flint clay is a dense, hard, nonslakable, massive, non-
plastic clay with flint-like characteristics. Flint clay is frequently
associated with underclays, but other occurrences, as, e.g., in association
with the diaspore clays of Missouri (see page 365), are also known. The
19 Fairbairn, H. W., X-Ray Petrology of Some Fine-grained Foliated Rocks, Am.
Mineral., 28, 246-256 (1943).

20 Grim, R. E., and V. T. Allen, Petrology of the Pennsylvanian Underclay of Illi-
nois, Bull. Geol. Soc. Am., 49, 1485-1513 (1938).

21 Brindley, G. W., and K. Robinson, X-Ray Study of Some Kaolinitic Fireclays,
Trans. Ceram. Soc., (Eng!.) 46, 49-62 (1947).

22 Honess, A. P., and F. J. Williams, Dickite from Pennsylvania, Am. Mineral., 20,
462-466 (1935).
"
,Origin and Occurrence of the Clay Minerals 359
flint clays that have been studied in detail are composed of well-crystal-
lized kaolinite in extremely small particle sizes. Sometimes excess ~lica
or alumina is present and may serve to cement the kaolinite particles
together. In many cases, however, the precise factors which give flint
clays their distinctive physical properties have not been established.
Limestones and Dolomites. An analysis of 35 Paleozoic limestones from
Illinois 23 showed illite in all of them. Kaolinite in small amounts was
present in about half of them. No montmorillonite was determined in
these limestones, but the analyses were made before the development of
the glycol method for the, determination of montmorillonite, and small
amounts of this clay mineral would probably have been missed.
The determination of the clay minerals in limestones and dolomites is
particularly difficult since it is frequently necessary to dissolve away the
carbonate in order to concentrate the clay minerals to get adequate
analytical data, and some of the clay minerals themselves are fairly
soluble in acid (see pages 296-299). i
Coal. Hicks and Nagelschmidt2 4 showed that a two-layer mineral was

the dominant clay mineral in English coals and that this mineral appeared
to be halloysite in the anthracites and kaolinite in the bituminous coals.
This relationship is difficult to understand, since it is contrary to the
usual concept of the transition of halloysite to kaolinite by metamorphic
action. Some illite is also present in the English coals. The clod and
roof shale of the coals is largely composed of illite containing some sodium
in interlayer positions. Nagelschmidt found that the sodium content of
the illite decreased from about 1.8 to 0.6 per cent with a decrease in the
rank of the coal.
Endell and Endell 25 have shown that illite is the dominant clay mineral
in the German coals. Their analytical data indicate that there is no sub-
stantial change in the clay-mineral component with the rank of these coals.
In Illinois coals, Ball 26 found that kaolinite was the dominant clay
mineral in the coal, both in fractures and cavities and in the coal itself.
A relatively small amount of illite is usually present also in these coals.
Salines. Unpublished work by the author and studies by MillotI and
De LapparentI 7 show that the sepiolite-palygorskite-attapulgite minerals
are frequent in the argillaceous sediments associated with salines.
23 Grim, R. E., J. E. Lamar, and W. F. Bradley, The Clay Minerals in Illinois Lime-
stones and Dolomites, J. Geol., 45, 829-843 (1937) .
24 Hicks, D., and G. Nagelschmidt, The Chemical and X-Ray Diffraction Analysis
of the Roof and Clod of Some South Wales Coal Seams and of the Mineral Matter in
Coal, Med. Research Council (Brit.) Special Rept. 244 (1943).
25 Endell, K., and J. Endell, Ueber die Bestimmung der R6ntgenfeinstruktur
mineralischer Bestandteile von Kohlen und Aschen sowie ihre technische Bedeutung,
Feuerungstech., 31, 137-143 (1943).
26 Ball, C., Kaolinite in Illinois Coal, Eeon. Geol., 29, 767-776 (1934).
360 Clay Mineralogy
Montmorillonite and illite clay minerals are also common. Frequently
the ...clay minerals in such sediments are poorly crystallized, and their
precise identification is difficult.
Kuhn27 has studied the composition of the Zechstein salt in Germany
and has found that illite is the dominant clay-mineral component. He
believes the mineral to be of diagenetic origin. He reports also that the
crystallinity of the illite is quite low in these sediIUents.
Till. Any of the clay minerals may be found in glacial till, depending
on the character of the source material, since there is likely to be sub-
stantially no alteration of the clay minerals during transportation and
deposition. Older tills might be expected to have a relatively higher
mica content, because later metamorphic changes probably would be
in the direction of the formation of well-crystallized micas from any
degraded mica or montmorillonite. It seems likely that kaolinite would
tend to persist unless the tills were very old or had been metamorphosed
substantiaIIy.
The Pleistocene till of North America has been extensively studied,
and illite is the dominant clay mineral in it. Chlorite, montmorillonite,
and kaolinite are frequently present, but of subordinate importance. .
Gumbotil, a weathering product of till found at the surface of some
pre-Wisconsin tills in North America, contains illite, with considerable
amounts of montmorillonite and occasionally more kaolinite than the
normal till. It also has a higher clay-mineral content than the unweath-
ered till as a consequence of the concentration ot colloidal material by
downward-moving waters.
Fuller's Earth. This name is used for any natural earthy material
which will decolorize mineral or vegetable oils to a sufficient extent to
be of economic importance. The name as generally used at the present
time has no genetic or mineralogical significance. An exception is the
use of the term in England as a stratigraphic term for certain beds long
used as commercial bleaching clays. Fuller's earths are known which
are composed of attapulgite,28 montmorillonite,29 and kaolinite. 30 Some
fine silts of relatively low clay content have high decolorizing ability.
On the other hand there are many silts and clays composed of mont-
morillonite and kaolinite and possibly also of attapulgite which are not
fuller's earths.
21 Kuhn, R., Ueber die Mineralgehalt der Salztone, Schriften Mineral. Petrog. Inst.
Kiel, 6 (1938).
28 De Lapparent, J., Les Argiles des terres it foulon, CQngr. intern. mines met. et
geol. appl., 1,381-387 (1935).
29 Grim, R. Eo, Petrography of the Fuller's Earth Deposit, Olmsted, Illinois, with a
Brief Study of Some Non Illinois Earths, Econ. Geol., 28, 344-363 (1933).
30 Piersol, R. J., J. E. Lamar, and W. H. Voskuil, Anna "1:{aolin" as a New Decolor-
izing Agent for Edible Oils, Illinois State Geol. Survey Rept. Invest. 27 (1933).
...
+ "
' ....
Miscellaneous Clay Materials. Under this heading will be considered
the origin and occurrence of clay minerals in argillaceous materials of
unique origin or whose origin is not well established.
Bentonite. The term bentonite was first applied by Knight 31 to a
particular, highly colloidal, plastic clay found near Fort Benton in the
Cretaceous beds of Wyoming; it has the unique characteristic of swelling
to several times its original volume when placed in water, and it forms
thixotropic gels with water even when the amount of bentonite in such
gels is relatively small. It was subsequently shown by Hewett3 2 that
this clay was formed by the alteration of volcanic ash in situ. Later,
Ross and Shannon 33 studied a number of clays which had been formed by
the alteration of ash and redefined the term bentonite to limit it to clays
produced by the alteration of volcanic ash in situ. They pointed out
that such clays are largely composed of montmorillonite clay minerals
and that they are generally highly colloidal and plastic. The term as
now used by many mineralogists and geologists has no reference to the
physical properties of the clay.
Until about 1930 bentonite was known in few places outside of the
original Wyoming area, and until about 1940 the term bentonite, espe- .
cially in industrial usage, was synonymous with the Wyoming material.
Even today (1951) the term bentonite to most people, other than some
mineralogists and geologists, carries with it the connotation of a highly
plastic, colloidal, swelling clay. Also, today the term is used frequently
for clays with these properties without reference to a particular mode of
onglll. Thus, montmorillonite clay from the Island of Ponza, Italy, is
generally called bentonite and sold commercially as such, even though
it is apparently of hydrothermal origin.
Because of the high commercial value of bentonite, a search for it has
been made in all quarters of the globe since about 1930, and now clays
composed largely of montmorillonite and formed by the alteration of
volcanic ash in situ have been found in almost all countries and in rocks
of a wide variety of ages. Bentonites are particularly widely distributed
in the Tertiary, as, e.g., in California and the Gulf Coast states of the
United States, in Italy, and in New Zealand. Bentonites are also very
widely distributed in the Upper Cretaceous, as, e.g., in the Gulf Coast and
High Plain states of the United States, in England, in Germany, and in
Canada. Beds older than the Cretaceous appear to contain little
bentonite. Some so-called metabentonite has been found in the Paleozoic
rocks (e.g., the Ordovician of Illinois and of the Appalachian states). In
31 Knight, W. C., Bentonite, Eng. Mining J., 66, 491 (1898).

32 Hewett, D. F., The Origin of Bentonite, J. Wash. Acad. Sci., 7, 196-198 (1917).
33 Ross, C. S., and. E. V. Shannon, Minerals of Bentonite and Related Clays and
Their Physical Properties, J. Am. Ceram. Soc., 9,77-96 (1926).
t.
362 Clay Mineralogy
such metabentonite the montmorillonite usually has been almost com-
pletf'-iy collapsed by compaction and metamorphism, so that the mineral
swells very little and the clay does not have the usual high colloidal
properties of bentonite.
Usually the volcanic-ash parent material of bentonites can be detected
by the persistence of shared struetures of the ash as pseudomorphs in the
day. Also the presence of a suite of nonclay minerals, such as feldspars,
biotite, hornblende, etc., characteristic of igneous material, may provide
evidence for the origin from ash.
Beds of bentonite vary in thickness from a fraction of an inch up to
many feet. Rarely are beds thieker than about 10 ft, but a few are
known which reach a thickness of 50 ft. The clay called bentonite which
has a hydrothermal origin is found in irregular masses without any
bedding. Bentonite varies in color from white to gray, yellow, green,
blue, and black but is perhaps most often yellow or yellowish green.
Frequently a yellow color develops on the weathered outcrop, and the
weathered material may have physical properties quite different from
these of the fresh material. Also the weathering outcrop frequently
develops a characteristic cellular structure, because of the swelling and
shrinking on wetting and drying, which permits it to be recognized easily
in the field. The bentonite frequently has a characteristic waxy appear-
ance when freshly dug. On exposure to the air, it often develops a
characteristic" jigsaw puzzle" set of fractures accompanying the shrink-
ing on drying.
Bentonites interbedded with shales, limestones, and sands are known.
Perhaps most frequently it is found in a section of marine beds. Often
the bentonite is immediately underlain by a hard siliceous bed in an
association that suggests that there has been an enrichment of silica in
the underlying bed accompanying the formation of the bentonite.
As stated above, the dominant clay-mineral component of bentonites
is montmorillonite. Other clay minerals, particularly illite and kaolinite,
are present in many bentonites, sometimes in amounts up to about 50
per cent of the total clay-mineral content. N on clay minerals charac-
teristic of igneous material as well as of detrital origin are also present in
varying amounts. Some bentonites are substantially pure montmoril-
lonite, but rarely is there less than about 10 per cent of nonclay minerals.
Cristobalite 34 is frequently present in varying amounts up to about
30 per cent of the total rock. This mineral is often intimately mixed
with montmorillonite and is very difficult to identify, so that it has
certainly been missed in many clay-mineral analyses of bentonites. It
undoubtedly has formed from the ash at the same time as the mont-
34 Gruner, J. W., Abundance and Significance of Cristobalite in Bentonite and
Fuller's Earth, Econ. Geol., 35, 867-875 (1940).
"".
morillonite. In a few cases zeolites 35 are found in bentonites intimately
mixed with the montmorillonite.
The composition of the montmorillonite itself varies greatly in different
bentonites, as has been shown in detail by Ross and Hendricks. 36 The
variation may be within the montmorillonite lattice itself or in the nature
of the exchangeable cations. There is considerable variation in the
relative abundance of magnesium and iron and also apparently in the
R 203-to-silica ratio. That is to say, there is considerable variation in
the population of the silica layer. With regard to exchangeable cations,
most bentonites that have been described carry Ca++ as the most abun-
dant ion. Only a few are known which carry N a+ as the dominant ion,
and of these the Wyoming bentonite is the main example. Also only a
few bentonites have been reported that carry H+ or K+ as the dominant
ions. Mg++ is frequently present as an exchangeable ion in relatively
small amounts, and this is particularly the case when Ca++ is the domi-
nant ion. I
The geologic column contains many beds of volcanic ash which have
not altered to bentonite. In order for bentonite to form, it is probably
necessary for the ash to fall in water. The kind of water, i.e., fresh or
saline, undoubtedly is important in determining whether bentonite forms
at all and, if it does, the precise character of the resulting montmorillonite.
Since much bentonite is associated with marine formations, it seems
certain that the alteration can take place in sea water. Whether or not
the alteration can also take place in even more saline waters or in fresh
water is not known definitely. The composition of the ash is also
an important factor. Apparently the ash must have a moderate con-
tent of MgO, since ash devoid of magnesia does not seem to alter to
montmorillonite.
The evidence strongly indicates that the alteration of the ash to
montmorillonite takes place soon after accumulation or possibly almost
contemporaneously with accumulation. It does not seem, at least in
most cases, to be a later process and certainly is not a weathering process.
The Cretaceous bentonites in Mississippi frequently grade upward with-
out a break into glauconitic sands. The basal parts of these sands con-
tain rounded masses of bentonite, indicating that the bentonite was
already formed when the sand began to accumulate.
The process of formation of the montmorillonite is essentially a
devitrification of the natural glass of the ash and the crystallization of
the montmorillonite. The ash probably in most cases contains an excess
'5 Kerr, P. F., and E. N. Cameron, Fuller's Earth of Bentonitic Origin from Tehach-
api, California, Am. Mineral., 21, 230-237 (1936).
36 Ross, C. S., and S. B. Hendricks, ;\1inerals of the Montmorillonite Group, U.S.

364 Clay Mineralogy
of silica and alkalies. The silica in some cases undoubtedly remains in
the ylay and gives rise to the cristobalite. Also the excess silica is prob-
ably the source of the free silica in the hard siliceous beds immediately
underlying many bentonites. The excess alkalies and alkaline earths
are probably carried away in solution in the water in which the ash fell.
The population of the exchangeable-cation positions would be a conse-
quence of the composition of the original water and the alkalies and
alkaline earths from the ash which dissolved in it.
Bentonites possess important and unique properties which give them
great commercial value in decolorizing oils, in the manufacture of
catalysts, in bonding molding sands, in the preparation of oil-well drilling
muds, and in many other relatively minor uses. The properties of
individual bentonites vary widely, and not all of them are suitable for
commercial use. Also many bentonites have properties fitting them for
one particular use, whereas they cannot be used at all for other purposes.
Thus, the Wyoming type of bentonite is excellent for drilling muds and
foundry use but is of no value for oil decolorizing or the making of
catalysts. Similarly, some bentonites from Mississippi are good for
decolorizing clays and for foundry use but are of no value for drilling
muds. The cause of the variation in properties fundamentally resides
in the composition of the montmorillonite itself and in the character of
the exchangeable ions. Much remains to be learned of the factors con-
trolling the properties of bentonites, and the subject cannot be pursued
at any length here. However, some examples of pertinent correlations
between properties and composition can be given. The very high swell-
ing and highly colloidal bentonites carry Na+ as the principal exchange-
able cation. The Wyoming bentonite is an example of such a bentonite.
That the presence of N a + is not the sole cause of the properties of the
Wyoming clay is shown by the fact that other clays treated so that they
will carry Na+ as the exchangeable cation do not fully develop the prop-
erties of the Wyoming material. Another cause for the relatively unique
properties of the Wyoming bentonite must reside in some feature within
the montmorillonite lattice composing it.
Bentonites with montmorillonite low in iron are required for catalyst
manufacture. Decolorizing clays are usually those carrying Ca++
and/ or Mg++ as exchangeable cations. Again, the montmorillonite
itself must be important, for good decolorizing clays cannot always be
prepared by a simple ion-exchange reaction. Montmorillonite clays t!
carrying either Na+ or Ca++ are suitable for foundry molding-sand use,
but they have quite different properties in such sands. Thus a sodium
montmorillonite provides relatively high dry strength and lower green
strength, whereas calcium montmorillonite gives high green strength
and lower dry strength.
I , ._
Highly Aluminous and Ferruginous Clays. It is not proposed in this
work to discuss in detail the highly aluminous or ferruginous materW;tl,
such as bauxites and laterites, since adequate volumes 37- 39 on these
subjects are already available. The usual clay mineral associated with
deposits of the oxides or hydroxides of aluminum and iron is kaolinite.
Halloysite has been reported 4o in some high-alumina clays, but the
halloysite may have been confused with poorly crystalline kaolinite.
There appears to be kaolinite of two modes of origin in such materials:
one which has formed prior to the development of the hydrates and is the
parent material from which much of the aluminum hydrate has developed,
and another 41 which has formed after the development of the aluminum
hydrates by the reaction of silica with such hydrates.
A particularly distinctive type of high-alumina clay is the diaspore
and burley clay of Missouri, in which kaolinite is associated with diaspore.
The origin of these clays is not entirely clear, but in general terms they
seem, as suggested by MacQueen 42 and Allen,4o to have formed by some
kind of leaching action associated with the development of the sinkholes
in which they are found. The genetic relation of the kaolinite to the
diaspore is not clear.
Indiana-type H alloy site Deposits. The deposits of halloysite near
Bedford in Indiana, first studied by Logan 43 and later in great detail by
Callaghan,44 typify an occurrence of halloysite known in several other
places. The halloysite occurs as lenses and cavity fillings at the contact
of the Mansfield sandstone of Pennsylvanian age with the underlying
Chester clastics. According to Callaghan, the halloysite was precipi-
tated from solution from ground water which derived its alumina and
silica from the residual soils on the Chester beds. The formation of the
halloysite was post-Mansfield. The waters were believed to be acid in
nature because of a sulfate content which was derived from the oxidation
of pyrite in the Mansfield sandstone.
The halloysite is relatively pure in some places but is often mixed with
a good deal of alunite. Also the clay sometimes contains P 20 5 (1O%)
37 Fox, C. S., "Bauxite," London (1927).
38 Harder, E. C., Stratigraphy and Origin of Bauxite, Bull. Geol. Soc. Am., 60,887-
908 (1949).
39 Harrassowitz, H., Laterite, Fortschr. Geol. Pal. Min., 14,253-566 (1926).
40 Allen, V. T., Mineral Composition and Origin of Missouri Flint and Diaspore
. Clays, Missouri Geol. Survey Ann. Rept. 58 (1935) .

41 Goldman, M. I., and J. I. Tracey, Relations of Bauxite and Kaolinite in the
Arkansas Bauxite Deposits, Econ. Geol., 41, 567-576 (1946).
42 McQueen, H. S., Geologic Relations of the Diaspore and Flint Fire Clays of
Missouri, J. Am. Ceram. Soc., 12,687-697 (1929).

43 Logan, W. N., "Kaolin of Indiana," Division of Geology, Indiana (1919).
44 Callaghan, E., Endellite Deposits in Gardner Mine Ridge, Lawrence County,
Indiana, Indiana Geol. Survey Bull. 1 (1948).

366 Clay Mineralogy -
in a mineral structurally similar to alunite. Some of the halloysite
co~tains allophane as an additional component.
Supergene Kaolinite Deposits. Occurrences of supergene kaolinite are
frequently found associated with metalliferous sulfide ore bodies. Such
clays are developed during the downward movement of acid water pro-
duced by the oxidation of the sulfides. The conditions of origin of such
clays would seem to be quite similar to those of the Indiana-type halloy-
site noted above. It is not known why halloysite is produced in one case
and kaolinite in the other.
Igneous Clay Minerals. A few occurrences of clay minerals in igneous
rocks have been described which suggest that some clay minerals can
occur as primary igneous components. Thus Rosetti 45 has described an
association of anauxite and cristobalite in the spherulites of a rhyolite
from Sardinia which seems to have this origin. Prider and Cole 46 have
described some clay minerals in a lamproite from West Kimberly,
Australia, which they believe are o~ late magmatic origin.
It is manifestly difficult to distinguish with certainty between clay
minerals of hydrothermal origin and those of possible igneous origin.
There seems to be no reason why the clay minerals could not be formed
directly in some igneous rocks.
Recent Sediments
Andre, K., Die Diagenese der Sedimente, Geol. Rundschau, 2, 61-74, 117-130 (1911).
Conway, E. J., Exchange of K, Na, and H Ions between the Cell and Its Environ-
ment, Irish J. Med. Sci. (1947).
Hummel, K., Green Earths of South Tyrol and Halmyrolytic Iron Silicates, Chem.
Erde, 6,468-551 (1931).
Lundquist, G., Depositional Environment of Lake Sediments, Sveriges Geol. Undersokn.
Arsbok, 36 (1942).
Norin, R., Mineralogical Composition of Some Clays from the Rhaetic and Liassic
Beds of Northwest Scania, Sweden, Geol. Foren. i Stockholm Forh., 71, 215-257
(1949).
"Recent Marine Sediments," Trask, ed., American Association of Petroleum Geolists
Tulsa, Oklahoma (1939). Has a series of papers by many authors on all aspects
of Recent sediments.
Renngarten N. V., Boehmite from Northern Caucasus, Doklady Akad. Nauk SSSR,
69, 425-427 (1949).
Ancient Sediments
Albareda-Herrera, J. M., and R. Munoz, Fenomenos de Ordenacion y reoanisotropia
de Arcillas, An. Inst. esp. Edafol. Ecol. Fisiol. veget. 6, 475-541 (1947).
45 Rosetti, V., Anauxite Rhyolite of Salamura, SardinifL, Periodico mineral. (Rome),
20, 61-92 (1951).

46 Prider, R. T., and W. F. Cole, The Alteration Products of Olivine and Leucite in
the Leuci.te Lamll1'oi.tes f1'om the We~t Ki.mberly Area, t\.ustralh, Am. MineraL, a't.
373-384 (1942).
Allen, V. T., Mineral Composition and Origin of the Missouri Flint and Diaspore
. Clays, Missouri Geol. Survey Ann. Rept. 59 (1937).
Bates, T. F., Investigation of the Micaceous Mineral in Slate, Am. Mineral., 32, 62fi-
636 (1947).
Bossazza, V. L., "The Petrography and Petrology of South African Clay," privately
printed, Johannesburg (1948).
Burst, J. F., The Clay Mineralogy of Two Typical Missouri Fireclays, Ph.D. thesis,
University of Missouri (1949).
Carobbi, G., Sediments of the Passo delle Rodici, Atti. soc. toscana sci. nat. (Pisa) M em.,
49, 189-215 (1941).
Correns, C. W., Die Tone, Geol. Rundschau, 29, 201-219 (1938).
Davidson, C. F., The Archean Rocks of the Rodil District, South Harris, Outer
Hebrides, Trans. Roy. Soc. Edinburgh, 41, 71-112 (1942).
De Lapparent, J., Les Argiles d'EI Golea (Sahara), Bull. service carte geol. Algerie, ser.
5, no. 3 (1937).
Dosogne, C., Contribution 11 l'etude des argiles, belges, Soc. geol. Belg. Bull., 54, 18-28,
31-38, 89-105 (1945).
Endell, J., Composition and Shapes of Minerals in English Ball Clay, Ber. deut. keram.
Ges., 27, 13-16 (1950).
Endell, K., U. Hofmann, and E. Maegdefrau, Ueber die Natur des Tonanteils in
Grim, R. E., The Constitution of Various Ceramic Clays, J. Am. Ceram. Soc., 19,307-
315 (1936).
Gude, A. J., Clay Minerals of the Laramie Formation, Golden, Colorado, Bull. Am.
Assoc. Petroleum Geol., 34, 1699-1710 (1950).
Hebert, R., Contributions to the Study of the Physicochemical Structure of Clays,
Ann. mines, 139, 3-54 (1950).
Keller, W. D., and P. T. Chuen, The Petrology of a Specimen of the Perry Farm
Shale, J. Sediment. Petrol., 20, 123-132 (1950).
Noll, W., Ueber die geochemische Rolle_der Sorption, Chem. Erde, 6, 552-577 (1931).
Riviere, A., Contribution 11 1'6tude des sediments argileux, Bull. soc. geol. France, 16,
43-55 (1946).
Ross, C. S., Minerals and Mineral Relationships of the Clay Minerals, J. Am. Ceram.
Soc., 28, 173-183 (1945).
Sedletsky, 1. D., Mineralogical Composition of Clays and Their Genesis, Soviet Geol.,
8, 82-89 (1940).
Ventriglia, U., Mineralogical Constitution and Minerogenic Classification of Clays,
Periodico mineral. (Rome), 17, 189-217 (1948).
Wiklander, L., Differential Analyses of Some Quaternary Clays of Sweden, Geol. Foren.
i Stockholm Forh., 72, 119-132 (1950).
Zwetsch, A., The Kaolins of the Western Zones of Germany, Tonind.-Ztg., 74,166-171,
196-198 (1950).
Bentonites
Bechtner, P., Bentonite, "Industrial Minerals and Rocks," Chap. V, American
Institute of Mining and Metallurgical Engineers, New York (1949).
Bhola, K. L., Bentonite in India, Quart. J. Geol. Mining Met. Soc. India, 19, 55-77
(1947).
Bordas, A. F., Argentenian Bentonites, Rev. minera geol. y mineral. Soc. argentina
mineria y geol., 14, 3-60 (1943).
Conley, J. E., Bentonite, Its Properties, Mining, Preparation, Use, U.S. Bur. Mines
Tech. Paper 609 (1940).
368 Clay Mineralogy
De Lapparent, J., Deposits of the Smectic Clays of North Africa, Campt. rend., 221,
335-337 (1945).
Theribere, M., "Le Bentonite," Dunod, Paris (1943).
Endell, K., Bentonite: Properties and Uses, Tonind.-Ztg., 77, 122-126 (1950).
Fisk, H. G., Bentonite, Natl. Resources Research Inst. Univ. Wyoming Bull. 2 (1946).
Hagner, A. F., Adsorptive Clays of the Texas Gulf Coast, Am. Mineral., 24, 67-108
(1939).
Kerr, P., Bentonite from Ventura, California, Econ. Geol., 26, 153-168 (1931).
Lee, D. E., Bentonite and Bleaching Clays in Japan, Natl. Resources Section Supreme
Command Allied Powers Rept. 139 (1951).
Rich, A. D., Bleaching Clay, "Industrial Minerals and Rocks," Chap. VI, American
Institute of Mining and Metallurgical Engineers, New York (1949).
Rios, E., "Bentonitas Espanolas," Madrid (1949).
Rios, E. G., and L. Hernando, The Bentonite Deposits of Spanish Morocco, Ann. inst.
espan. edafol. ecol. fisiol. veget., 6, 53-77 (1947).
Ross, C. S., Altered Paleozoic Volcanic Materials and Their Recognition, Bull. Am.
Assoc. Petroleum Geol., 12, 143-164 (1928).
Savelli, F., "La Bentonite," Ulrico Hoepti, Milan (1943).
Sazonova, N. K., Bentonite Clays in U.S.S.R., Trudy Lab. Keram., 7, 217-244 (1939).
Schroter, G. A., and 1. Campbell, Geologic Features of Some Bleaching Clay, Am.
Inst. Mining Met. Engrs. Tech. Pub. 1139, pp. 1-31 (1940).
Spence, H. S., Bentonite, Can. Dept. Mines Resources Mines Branch Bull. 626 (1924).
Wherry, E. T., Bentonite as a One-Dimensional Colloid, Am. Mineral., 10,120-123
(1925).
Worcester, W. G., Saskatchewan Bentonites, Trans. Can. Inst. Mining Met., 40, 438-
451 (1937).
APPENDIX
Chemical Analyses
A vast number of chemical analyses of clay materials have been published. In

many cases no attempt was made to determine the mineral composition of the material
analyzed, or the material is clearly a mixture of minerals. There are, however, many
analyses of materials whose clay-mineral composition is well established and which
seem to be mono mineral. In the tables to follow an attempt is made to present a
sampling of these analyses. The objective has been to show the spread of the compo-
sition of the various clay minerals so that their range in composition is illustr[tted.
In the case of the vermiculite and chlorite minerals, there are substantially no data
a.vailable for clay-mineral specimens, and the data given are for larger specimens.
370 Clay Mineralogy
TABLE A. CHEMICAL ANALYSES OF KAOLINITE MINERALS
t 1 2 3 4 5 6 7
Si0 2 46.90 44.81 45.20 46.77 44.59 54.32 48.80

AbO. 37.40 37.82 37.02 37.79 36.83 29.96 35.18
Fe 2O. 0.65 0.92 0.27 0.45 1.14 2.00 l.24
FeO ..... . ....... 0.06 0.11
MgO 0.27 0.35 0.47 0.24 0.39 0.14
CaO 0.29 0.43 0.52 0.13 1.02 0.32 0.22
K 20 0.84 ...... . 0.49 l.49 0.32 ...... 0.40
Na.O 0.44 ....... 0.36 0.05 0.13 0.37 0.25
TiO. 0.18 0.37 1.26 ..... . 2.17 ...... 0.61
H 2O- ..... . l.10 1.55 0.61 ....... 0.84 l.16
H.O+ 12.95 ]4.27 13.27 12.18 13.63 11.80 12.81
Total .... 99.92 100.07 100.47 99.82 100.22 99.75 100.67
Kaolinite Anauxite
1. Zettlitz, Czechoslovakia 6. Bilin, Czechoslovakia
2. Mexia, Texas 7. lone, California
3. Macon, Georgia
4. St. Austell, England
5. Anna, Illinois
Analyses 1, 2, 6, and 7 from C. S. Ross and P. F. Kerr, U.S. Oeol. Survey Profess.
Paper 165E (1931); 3 and 4 from P. F. Kerr et ai., Rept. 7, American Petroleum
Institute Project 49 (1950); 5 from R. E. Grim, Econ. Geol., 29, 659-670 (1934).
TABLE B. CHEMICAL ANALYSES OF NACIUTE, DICKITE, AND HALLOYSITE
1 2 3 4 5 6 7
SiO. 44.75 46.86 40.80 44.75 43.79 44.46 43.98

Al 2O, 39.48 37.12 35.70 36.94 36.10 36.58 38.46
Fe 2O, 0.53 1.43 ...... 0.31 0.11 0.36
FeO ...... 0.06 0.04 . ..... 0.01 0.07 0.03
MgO 0.19 0.09 Tr ...... 0.18 0.18 Tr
CaO 0.13 0.22 3.96 0.11 0.36 0.19 0.32
K.O . . .... 0.60 0.20 ...... 0.51 0.48
Na 20 ..... . 0.07 0.20 }0.60 ...... 0.01 0.14
TiO. ...... 0.51 0.02 . ..... 0.01 0.15 0.01
H.O- 0.61 0.29 1.95 2.53 5.92 4.05 2.58
H 2O+ 14.40 12.77 17.68 14.89 13.53 13.38 14.59
Total. ........ 100.09 100.02 100.55 100.13 100.01 99.94 100.59
N acrite H alloysite
1. Brand, Saxony 4. Liege, Belgium
Dickite 5. Cadouin, France
2. Anglesey, vYales 6. Bedford, Indiana
3. Ouray, Colorado 7. Eureka, Utah
Analyses 1 from C. S. Ross and P. F. Kerr, U.S. Geol. Survey Profess. Paper 165E
(1931); 2, 3, 5, 6, and 7 from P. F. Kerr et ai., Rept. 7, American Petroleum Institute
Project 49 (1950); 4 from C. S. Ross and P. F. Kerr, U.S. Geol. Survey Profess. Paper
185G (1934).
Appendix 371
TABLE C. CHEMICAL ANALYSES OF MONTMORILLONITE MINERALS
1 2 3 4 5 6 7 ..
--
---<~.---
Si0 2 52.09 50.30 50.20 51.14 55.44 57.55 49.91

Al 20 3 18.98 15.96 16.19 19.76 20.14 19.93 17.20
Fe 203 0.06 0.86 4.13 0.83 3.67 6.35 2.17
FeO ...... . ...... . ....... . ..... 0.30 0.95 0.26
MgO 3.80 6.53 4.12 3.22 2.49 3.92 3.45
CaO 3.28 1.24 2.18 1.62 0.50 1.94 2.31
K.O ....... 0.45 0.16 0.11 0.60 0.59 0.28
Na.O ....... 1.19 0.17 0.04 2.75 0.33 0.14
Ti0 2 ...... . ....... 0.20 . ..... 0.10 0.32 0.24
H 2O- 14.75 15.58 14.81 ....... 15.77
7.46 }23.61 7.99 }14.70 8.53
H 2O+ 7.57 7.70
Total ... 100.42 100.14 100.50 99.52 100.69 100.41 99.43
Montmorillonite
L Tatatilla, Mexico 5. Upton, vVyoming
2. Otay, California 6. Pontotoc, Mississippi
3. Polkville, Mississippi 7. Chambers, Arizona
4. Montmorillon, France
Analyses 1 to 5 from C. S. Ross and S. B. Hendricks, U.S. Geol. Survey Profess.
Paper 205B (1945); 6 from R. E. Grim and R. A. Rowland, Am. Mineral., 27,746-761
(1941); 7 from P. F. Kerr et al., Rept. 7, American Petroleum Institute Project 49
(1950).
TABLE C. CHEMICAL ANAI,YSES OF MONTMORILLONITE MINERALS. (Continued)
1 2 3 4 5 6 7
SiO. 40.54 41.38 46.06 44.0 53.95 50.8 44.0

Al 20 a 5.19 9.84 12.22 3.6 0.14 9.4 10.60
Fe20a 31.24 27.47 18.54 29.0 0.03 2.0 Tr
FeO 0.39 Tr? 0.28
MgO 0.06 Tr? 1.62 2.1 25.89 26.5 24.30
CaO 1. 92 Tr? 1.66 .... 0.16 0.7 2.00
K 20 0.24 Tr? .... . .... 0.23
Na 20 0.14 Tr? .... . .... 3.04
TiO. ...... . ..... .. 0.84
H 2O-
H 2O+
14.75
6.00
12.10
9.25 } 17.26 } 18.7
9.29
5.61
} 10.5
12.60
6.20
Li 20 ....... ...... . .... . .... 1.22
Total .... 100.47 100.04 98.48 97.4 99.56 99.9 99.70
Nontronite Hectorite
L Manito, Washington 5. Hector, California
2. Sandy Ridge, North Carolina Saponite
3. Spokane, Washington 6. Svardsjo, Sweden
4. Nontron, France 7. San Bernadino County, California
Analyses 1 and 5 from P. F. Kerr et al., Rept. 7, American Petroleum Institute
Project 49, 1950; 2 to 4, 6, and 7 from C. S. Ross and S. B. Hendricks, U.S. Geol.
Survey Profess. Paper 205B (1945).
372 Clay Mineralogy
TABLE D. CHEMICAL ANALYSIS OF ILLITE MINERALS
1 2 3 4 5 6 7 8 9 10 11
-_. - - --- - - --- --
- - - - - - ---
Si0 2 52.23 51.22 40.87 38.18 49.26 47.55 51.65 53.12 48.66 49.4 51.95
Al 2O. 25.85 25.91 20.45 15.88 28.97 32.45 21.67 27.36 8.46 10.2 17.81
Fe 20a 4.04 4.59 12.81 18.88 2.27 0.76 6.20 2.64 18.80 18.0 6.17
FeO ..... . 1. 70 ..... 2.99 0.57 1.85 1.24 1.06 3.98 3.1 3.87
MgO 2.69 2.84 6.86 4.92 1.32 1.70 4.48 2.62 3.56 3.5 4.76
CaO 0.60 0.16 0.89 1.22 0.67 0.06 0.00 0.53 0.62 0.6 0.53
K 20 6.56 6.09 3.25 3.66 7.47 6.22 6.08 3.51 8.31 5.1 4.04
Na 20 0.33 0.17 0.70 0.46 0.13 1.05 0.31 0.49 0.00 1.4 1.06
Ti0 2 0.37 0.53 2.13 4.79 0.05 0.64
H 2O- ..... . ...... . .... 2.38 3.22
H 2O+ 7.88 7.14 11.84 6.70 6.03 7.73 6.44 8.60 6.56 8.3 10.06
TotaL ... 100.55 100.35 99.80 100.06 99.96 100.01 98.07 99.93 98.95 99.6 100.25
1. Alexander County, Illinois 6. South Wales

2. Fithian, Illinois 7. Geoschwitz, Germany
3. Northeast Scotland, calculated com- 8. Capalbio, Italy, "Illidromica," calcu-
position of trioctahedral mica, minUS lated composition minus impurities
impurities 9. St. Joseph Lead Co. mines, glauconite
4. Scotland, weathered biotite 10. Norwalk, Wisconsin, glauconite
5. Ballater, Scotland 11. South Tyrol, Germany, glauconite
Analyses 1 and 2 from R. E. Grim and R. A. Rowland, Am. Mineral . 27, 746-761
(1941); 3 from G. F. Walker, Mineralog. Mag., 29, 72-84 (1950); 4 from G. F. Walker,
Mineralog. Mag., 28, 693-703 (1949); 5 from R. C. Mackenzie et al., Mineralog. Mag.,
28, 704-713 (1949); 6 from G. Nagelschmidt and D. Hicks, Mineralog. Mag., 26,297-
303 (1943); 7 from E. Maegdefrau and U. Hofmann, Z. Krist., 98, 31-59 (1937); 8
from C. Andreatta, Periodico mineral. (Rome) 18, 11-31 (1949); 9 to 11 from S. B.
Hendricks and C. S. Ross, Am. Mineral., 26, 683-708 (1941).
TABLE E. CHEMICAL ANALYSES OF CHLORITE MINERALS
1 2 3 4 5 6 7
Si0 2 31.44 27.78 29.87 26.68 23.69 20.82 21.71

Al 20 a 17.62 24.30 14.48 25.20 21.26 17.64 21.35
Fe 2O. ..... . 1.43 5.52 . ..... . ...... 8.70 0.82
FeO Tr 0.35 1.95 8.70 26.52 37.96 43.01
MgO 37.64 32.71 33.06 26.96 17.60 4.15 2.33
CaO Tr Tr ....... 0.28 3.32
H 2O- ...... 0.06
H 2O+ 13.19 13.01 13.60 11.70 7.63 10.30 10.10
Cr 2O. ..... . ...... 1.56
Total. .. '" 99.89 99.64 100.04 99.52 100.02 99.57 99.23
1. Leuchtenbergite, Phillipsburg, Montana 4. Chlorite, Ducktown, Tennessee

2. Sheridanite, Miles City, Montana 5. Proehlorite, Trumbull, Connecticut
3. Chlorite, Brinton Quarry, West Ches- 6. Thuringite, Schmiedefeld, Thuringia
ter, Pennsylvania 7. Bavalite, Bas VaUon, Brittany
Analyses 1 to 5 hom R. C. McMurchy, Z. Krist., 88,420-432 (1934); 6 and 7 from
W. von Englehardt, Z. Krist., 104, 142-159 (1942).
Appendix 373
TABLE F. CHEMICAL ANALYSES OF VERMICULITE
1 2 3 4
'"' ,
SiO. 36.12 35.92 36.54 34.04
AI.0 3 13.90 10.68 16.96 15.37
Fe 203 4.24 10.94 2.78 8.01
FeO 0.68 0.82 0.95
MgO 24.84 22.00 19.78 22.58
CaO 0.18 0.44 0.06
Ti0 2 0.24
H 2O-
H 2O+ } 18.94 19.84 20.40 19.93
NiO .... . ...... 2.32
Total. .... 99.14 100.64 99.79 99.93
1. Bare Hills, Maryland 3. Webster, North Carolina

2. Pilot, Maryland 4. Kenya
Analyses 1 to 3 from J. W. Gruner, Am. Mineral., 19,557-575 (1934); 4 from G. F.
Walker and A. Milne, Trans. 4th Intern. Congr. Soil Sci., 2, 62-67 (1950).
TABLE G. CHEMICAL ANALYSES OF SEPIOLITE, PALYGORSKITE, AND ATTAPULGITE
1 2 3 4 5 6 7
SiO. 52.50 54.83 55.34 51.17 54.71 46.67 55.03

AbO, 0.60 0.28 1.81 13.73 13.48 9.84 10.24
Fe.O, 2.99 0.45 0.43 1.55 2.10 0.93 3.53
FeO 0.70 ..... . ..... . 0.31 ...... 1.22
MgO 21.31 24.51 22.95 6.40 5.44 8.94 10.49
CaO 0.47 0.55 0.24 2.89 2.79 8.36
K.O ~
...... 0.03 ...... . ..... . ..... . ..... , 0.47
Na.O ...... 0.35
H 2O- 8.18 8.60 10.29 8.65 8.29 9.73
} 21. 27
H 2O+ 10.74 10.20 13.24 12.63 15.84 10.13
Total. ....... 99.84 99.92 99.57 99.58 99.80 100.09 99.62
Sepiolite Palygorskite
1. Ampandandrava, Madagascar 4. Nizhnii-Novgorod, U.S.S.R.
2. Yavapai County, Arizona 5. Taodeni, Sahara, Algeria
3. Durango, Mexico 6. Nerchinsk, Siberia, U.S.S.R.
Attapulgite
7. Attapulgus, Georgia
Analyses 1 and 5 from S. Caillere, "X-ray Identification and Structure of the Clay
, Minerals," Chaps. VIII, IX, Mineralogical Society of Great Britain Monograph
(1951); 2 from A. J. Kaufman, Am. Mineral., 28, 512-520 (1943); 3 from W. F.
Foshag, Proc. U.S. Nail. Museum, 74, 136 (1928); 4 and 6 from S. Caillere, Bull. soc.
fran,. mineral., 59,352-374 (1936); 7 from W. F. Bradley, Am. Mineral., 25, 405-410
\1940).
INDEX
A Analytical methods, optical methods,
273, 292-294
Ablykite, 42 oriented-aggregate technique, 288-289
Acid solubility, 296-300 rational, 301
effect of heating on, 297 staining tests, 274-276
nature of reaction, 298-299 surface area, 273
Adsorbed water, density, 171 vapor pressure-water content, 190
freezing point, 173 X-ray diffraction, 84-86
ions adsorbed, influences of, 174-181 organic complexes, 271-272
nature, 162-171 Anauxite, acid solubility, 297
stability, 182 chemical analyses, 370
structure, 162-171 definition, 30
thickness, 162, 174-175 dehydration curve, 191
time factor, 175 dehydration data, 212-213
Air-set strength, 175 differential thermal curves, 196
Alkali solubility, 301-303 electron micrograph, 113
nature of reaction, 302 history, 30
Allophane, definition, 29 hydrothermal deposits, 324
dehydration curves, 193, 211 igneous origin, 366
dehydration data, 211-212 opti()al properties, 278
differential thermal curves, 196, 211- stru()ture, 51
212 Anion exchange, causes, 156-158
electron micrographs, 109-110 definition, 126
history, 29 phosphate fixation, 158
hydrothermal deposits, 324 types, 157
optical properties, 279 Anorthite, 225
in soils, 344 Argillite, 2
structure, 45 Argillization, 326
Allophaneton, 13, 301 Aridic soils, 339-340
a-Alumina, from muscovite, 236 Attapulgite, alkali solubility, 303
synthesis, 317 cation-exchange capacity, 129
'Y-Alumina, from kaolinite, 214 chemical analyses, 373
from muscovite, 236 dehydration data, 243-244
synthesis, 317 differential thermal curves, 199, 243-
Alumina octahedrons, 43 244
Amines, definition, 252 electron micrograph, 122
exchange reaction, 252-257 fusion point, 244
organophilic complexes, 266 heat of wetting, 184
Ammonium fixation, 153 high-temperature phases, 244
Amorphous material in clays, 19 history, 38
Amphibole structure, 44 infrared spectra, 307, 310
Analcime, 318 optical properties, 279, 286
Analytical methods, dehydration, 191 organic complexes, 264
differential thermal, 194-210,247,272 sediments, 355-360
exchange capacity, 155, 273 shape and size of particles, 122
exchangeable cations, 155 staining test, 275
high-temperature phases, 211 structure, 78
infrared spectra, 304-305 X-ray diffraction, 103
375
376 Clay Mineralogy
B Cation exchange, in relation, to tempera-
ture, 139-141
BauxPte, 365 replaceability of exchangeable cations,
Bayerite, 317 144-152
Beidellite, 39, 60, 246, 324 effect on, of concentration, 145
Bentonite, clay-mineral composition, 362 of heating, 148-149
composition, 247 ion hydration, 147
cristobalite in, 5, 362 nature, of clay minerals, 149-152
definition, 361 of ions, 146-148
distribution, 361-362 population of exchange positions, 146
geologic age, 356-357, 361-362 theory, 154
history, 361 (See also Exchangeable cations)
ion exchange, 128 Cation-exchange capacity, clay minerals,
origin, 303 129
properties, 361, 364 methods of determining, 155, 273
rehydration, 229-230 organic ions, 253
Binding forces, 9 in relation to heat of wetting, 186
Biological decomposition, organic com- Cations, exchangeable (see Exchangeable
plexes, 264-265 cations)
protein, 265 fixation, 152-154
Biotite, acid solubility, 298 hydration, 148, 174-181
dehydration, 233-234 replaceability (see Cation exchange)
differential thermal curves, 197, 235 size, 148
high-temperature phases, 236 Celadonite, 41
rehydration, 235 Chemical analyses, anauxite, 370
structure, 65 attapulgite, 373
X-ray diffraction, 94 chlorite, 372
Birefringence, clay-mineral data, 278-286 dickite, 370
in electrical field, 290-291 glauconite, 372
form, 291-292 halloysite, 370
immersion-media influence, 286-288 hectorite, 371
Boehmite, 317 illite, 372
Brammallite, 121 kaolinite, 370
Bravaisite, 36 montmorillonite, 371
differential thermal curve, 198, 246 nacrite, 370
Brucite structure, 43, 69 nontronite, 372
Burley clay, 365 palygorskite, 373
saponite, 371
C sepiolite, 373
vermiculite, 373
Catalysis, 250 Chernozem soil, clay-mineral composi-
enzymatic reaction, 265 tion, 340
Cation exchange, capacity of clay miner- definition, 335, 340
als, 129 Chlorite, acid solubility, 299
causes, 132-134 ancient sediments, 355-360
clogging, 144 cation-exchange capacity, 129
definition, 126 chemical analyses, 372
effect of grinding on, 138, 139 definition, 37
environment of reaction, 141, 142 dehydration curves, 192-238
history, 128 dehydration data, 238-240
hydrogen clays, 142-144 density, 314
ionization of cations, 134---136 differential thermal curves, 199, 239-
in non clay minerals, 131, 132 240
position of cations, 134-136 high-temperature phases, 240-241
rate of, 136, 137 history, 36
in relation to particle size, 137, 138 hydrothermal deposits, 324-328
to structure, 132-134 optical properties, 279, 284-285
Index 377
Chlorite, organic complexes, 263 Coal,359
recent sediments, 355-360 Cordierite, 225
rehydration, 241 Corundum, 236
structure, 60-72 Cristobalite, from attapulgite, 244
synthesis, 322 in bentonite, 5, 362
X-ray diffraction, 97-98 from halloysite, 219
Chloritization, 327 from kaolinite, 215-216
Chloropal, 42 from montmorillonite, 225
Clay, amorphous material in, 19 from sepiolite, 243
burley, 365
clay-mineral eomposition, 358 D
definition, 1, 3
flint, 1, 358 Decolorization, 250
particle-size term, Dehydration, allophane, 211-212
Clay material, concepts of composition, attapulgite, 243-244
11-25 biotite, 233
definition, 3 chlorite, 238-240
exchangeable ions and soluble salts in, 6 clay-mineral curves, 191-195
factors controlling properties of, 3 definition, 190
nonclay-mineral composition, 4, 5 halloysite, 217-219
organic material in, 6 illite, 233-235
texture, 7 kaolinite, 212-214
Clay-mineral concept, definition, 18 method of analysis, 191-194
history, 16 montmorillonite, 220-227
Clay minerals, acid solubility, 296-300 muscovite, 283
in ancient sediments, 355-366 palygorskite, 243
cation-exchange capacity, 129 phlogopite, 233
classification, 27 in relation, to differential therr
definition, 17 analyses, 202
dehydration, 191-195, 211-244 to optical data, 280-284
density, 312-314 sepiolite, 241-242
differential thermal data, 196-199, vermiculite, 231-232
211-244 Density, adsorbed water, 171
electrodialyses of, 143, 144 clay-mineral data, 312-314
heat of wetting, 183-188 method, 312
high-temperature phases, 211-244 Desert soils, clay-mineral compositi,
hydrothermal origin, 323-330 340
infrared spectra, 304-310 definition, 335-339
nomenclature, 29-38 Diaspore, 317, 365
organic reactions, 250-276 Dickite, chemical analyses, 370
organophilic gel complexes, 265-269 dehydration data, 213
origin (see Origin of clay minerals) electron micrograph, 114
potash-bearing, 35 history, 30
in recent sediments, 348-355 hydrothermal deposits, 324
in relation to geologic age, 356-357 infrared spectra, 306, 308
replaceability of adsorbed cations, 149- optical properties, 279
152 shape and size of particles, 108
in soils, 340-341 staining tests, 275
staining tests, 274-276 structure, 50
structure (see Structure of clay miner- X-ray diffraction, 90
als) Differential thermal analysis, accural
surface area, 311-312 210,247
(see Synthesis of clay minerals) allophane, 211-212
titration curves, 129-131 attapulgite, 243-244
(See also specific clay minerals) chlorite, 239-240
" Clayite," 14 clay-mineral data, 196-199
Clinochlore, 240 definition, 194-195
378 Clay Mineralogy
Differential thermal analysis, difference Exchangeable cations, ionization, 134-
scales, 205 136
eq,!ipment variables, 200-205 methods of determining, 155
experimental variables, 205-210 organic, 252-257
furnace atmosphere, 208-210 replaceability (see Cation exchange)
halloysite, 218-219 Expanding lattice, 56
heating rate, 207-208
history, 196-200 F
illite, 235-236
kaolinite, 218-219 Faratsihite, 42
mixed-layer structures, 246-247 'Y-Fe203, 236
mixtures of clay minerals, 245-247 Feldspar, 4, 225
montmorillonite, 222-227 synthesis alteration, 320
organic-complex technique, 272 Fixation of cations, 152-154
organic complexes, 256 Flint clay, 1, 358
palygorskite, 243-244 Forest soil, 335
particle size, 206 Form birefringence, 291-292
in relation to dehydration curves, 202 Formula, lattice substitution and X-ray-
sepiolite, 241-242 diffraction intensity, 97
theory, 202-203 Forsterite, 228, 240
thermocouple circuits, 201, 204 Fourier syntheses, 134
vermiculite, 231-232 Fuller's earth, 360
Dioctahedral structures, 59, 65 Fumaroles, 330
X-ray diffraction, 93-95 Fusion temperature, attapulgite, 244
Dolomite, 359 halloysite, 219-220
illite, 237
E kaolinite, 217
montmorillonite, 229
Electrodialysis, 143, 144, 299-300 vermiculite, 232
Electron micrographs, allophane, 109-110
anauxite, 113 G
attapulgite, 122
dickite, 114 Gedroizite, 41
halloysite, 115 Gels, organic-clay-mineral complexes,
hectorite, 119 265-269
illite, 120-121 swelling behavior, 266-268
kaolinite, 111-112 viscosity, 257
mixed-layer minerals, 124 Gibbsite, 317
montmorillonite, 117-120 Glacial deposits, 355, 360
nacrite, 114 Glauconite, chemical analyses, 372
nontronite, 118 dehydration curve, 192
saponite, 118 differential thermal curve, 197
sauconite, 120 optical properties, 279
a-sepiolite, 123 origin, 352-353
t'-sepiolite, 124 structure, 68
Electron microscopy, principle, 107 X-ray diffraction, 95
replica technique, 107 Glimmerton, 36
shadowing technique, 107 Gumbotil,360
Endellite, 32
Enstatite, from attapulgite, 244 H
from chlorite, 240
from montmorillonite, 225, 228 Halloysite, acid solubility, 297
from sepiolite, 243 aneient sediments, 365
from vermiculite, 232 cation-exchange capacity, 129
Exchangeable anions, 156-159 chemical analyses, 370
Exchangeable cations, 128-155 dehydra,tion curves, 191, 193,218
hydration, 174-181 dehydration data, 217-219

Index 379
Halloysite, density, 312 Hydromica (see Illite)
. differential thermal curves, 196, 218- Hydrothermal clay minerals, formation,
219 325-328 "
electron micrographs, 115 occurrence, 323-325
fusion tern perature, 219-220 parent materials, 328-329
high-temperature phases, 219 in relation to mineralization, 329
history, 31 zonal arrangement, 325
hydrothermal deposits, 324-328 Hydrothermal solutions, 328
infrared spectra, 305-306
nomenclature, 32 I
optical properties, 279, 281
organic complexes, 257-261 Ice structure, 168
recent sediments, 349 Igneous clay minerals, 366
rehydration, 220 Illite, acid solubility, 297
shape and size of particles, 110 ancient sediments, 355-360
soils, 344 anhydride structure, 236
staining tests, 275 cation-exchange capacity, 129
structure, 52-53 chemical analyses, 372
surface area, 311 definition, 35
thermal expansion, 219 dehydration curves, 1\)2, 195, 233-235
X-ray diffraction, 87-91 dehydration data, 233-235
Heat of wetting, causes, 186-187 density, 312
definition, 183 differential thermal curves, 197, 198
effect on, of cations, 185 235-236
of firing, 187-188 electron micrographs, 120-121
in electrolyte solutions, 187 fusion point, 237
in relation, to exchange capacity, 186 heat of wetting, 185
to particle sizc, 185 high-temperature phases, 236-237
values for clay minerals, 184-186 hydrothermal deposits, 324-328
Hectorite, chemical analyses, 371 infrared spectra, 305, 310
definition, 60 organic complexes, 263
dehydration curve, 194 recent sediments, 348-354
dehydration data, 222 rehydration, 237-238
differential thermal curve, 197 shape and size of particles, 119, 122
electron micrograph, 119 soils, 340-341
high-temperature phases, 223 staining test, 275
infrared spectra, 308 structure, 65-69
optical properties, 279 surface area, 311
staining test, 275 synthesis, 318-322
Heptaphyllite structures, 59, 65 thermal expansion, 216-237
high-temperature alteration, 223 X-ray diffraction, 93-97
High-temperature phases, analytical Immersion liquids, optical-property in-
methods, 211 fluence, 286-287
attapulgite, 244 specifications, 288
biotite, 236 Indianaitc, 19
chlorite, 240-241 Indices of refraction, clay-mineral data,
halloysite, 219 278-286
illite, 236-237 effect of heating on, 280-284
kaolinite, 213-217 immersion-media influence, 286-288
montmorillonite, 222-229 Infrared spectra, clay-mineral data, 304-
muscovite, 236 310
palygorskite, 244 method, 303-304
phlogopite, 236 Inheritance, structural, 226
sepiolite, 242-243 Interlayer water, 161
vermiculite, 232 Ion exchange, definition, 126
Hot springs, 330 importance to, agriculture, 126
Hydrogen clay, 142-144 engineering, 127
Clay Mineralogy
Ion exchange, importance to, geology, Lattice substitution and X-ray-diffrac-
127 tion intensity, 95-97
8industry, 127 Lepidolite, 95
Ion hydration, 148 Leucite, 236
Ion size, 148 Leverrierite, 39
Ionization of exchangeable cations, 134- Levisite, 49
136 Limestone, 359
Lorenz factor, 97
K
M
Kaolin, tttoiinition, 29
history, 29 Magnesium fixation, 153
Kaolinite, acid solubility, 297 Marine sediments, 348-358, 355
ancient sediments, 355-360 Meerschaum, 37
cation-exchange capacity, 129 Mellorite, 31, 49
chemical analyses, 340 Metahalloysite, 32
definition, 30 Methylation of montmorillonite 269-'270
dehydration curves, 191, 193,212 Mixed-layer minerals, classification, 28
dehydration data, 212-214 dehydration data, 246-247
density, 312 differential thermal curves 198
differential thermal curves, 196, 213- electron micrograph, 124 '
214 optical properties, 289-290
electron micrographs, 111-112 structure, 79-80
fusion temperature, 217 X-ray diffraction, ]02-104
geologic distribution, 357 Monothermite, 40
heat of wetting, 185 Montmorillonite, acid solubility, 297
high-temperature phases, 213-217 alkali solubility, 303
history, 30 ancient sediments, 355-364
hydrothermal deposits, 324-328 anhydride structure, 222-224
infrared spectra, 305-306 bentonite, 361-364
optical properties, 278-280 cation-exchange capacity, 129
organic complexes 262-263 chemical analyses, 371
classifica~ion of group members, 60
recent sediments, 348-354
dehydratlOn curves, 194, 220-221
rehydration, 217
dehydration data, 220-226
shape and size of particles, 108
effect of adsorbed cations on 174-
soils, 340-341
181,221 '
specific gravity, 216-217 diff(,I'ential thermal curves 197 222-
staining tests, 275 227 ' ,
structure, 46-51 electron micrographs, 117-120
surface area, 311 fusion temperature, 229
synthesis, 317-320, 322 geologic distribution, 356
thermal expansion, 216 heat of wetting, 185
X-ray diffraction, 87-89 h!gh temperature phases, 222-229
Kaolinization, 826 hIstory, ;~3
Kaolinton, 301 hydrothermal deposits, 324-328
kX diffraction units, 84 infrared spectra, 306-309
optical properties, 279-28:3
L organic complexes, 253-261, 269-270
polar mol~cules, 257-261
Lacustrine sediments, 353-356 recent sedIments, 348-354
Lago~nal sediments, 855 rehydration, 229-230
Latente, 365 shape and size of particles 116 119
Laterite soils, clay-mineral composition, soils, 340-341 "
341 specific gravity, 229, 313-314
definition, 335-339 stability of hydration, 182
weathering process, 342-343 staining test, 275
Index 381
Montmorillonite, stepwise hydration, Optical methods, discussion, 292
181-182 immersion media, 286-288
struGture, 55-64 limitations, 278, 2!)2-294 Il
surface area, 311 organic complexes, 273
synthesis, 318-319 oriented-aggregate technique, 288-289
thermal expansion, 229 Optical properties, allophane, 279
X-ray diffraction, 91-93 attapulgite, 279, 285
Mountain leather, 192 chlorite, 279, 284-285
Mullite, from biotite, 236 dickite, 279
from halloysite, 219 effect of heating on, 280-284
from illite, 236 in electrical field, 290-291
from kaolinite, 214-216 form birefringence, 291-292
from montmorillonite, 225 halloysite, 279, 281
Muscovite, acid solubility, 297 illite, 279, 284
dehydration, 233-234 immersion-media influence, 286-288
dehydration curves, 195 kaolinite, 278-280
differential thermal curves, 197, 235 mixed-layer structures, 289-290
high-temperature phases, 236 montmorillonite, 279-283
rehydration, 238 nacrite, 279
structure, 65 palygorskite, 279, 285
synthesis, 319-320 sepiolite, 279, 285
X-ray diffraction, 94 vermiculite, 279, 285
N Organic-clay-mineral complexes, analyti-
cal techniques, 271-276
Nacrite, chemical analyses, 370 attapulgite, 264
dehydration data, 213 chlorite, 263
electron micrograph, 114 halloysite, 257-262
history, 30 history, 250-251
hydrothermal deposits, 324 illite, 263
optical properties, 279 ionic reactions, 252-257, 261-262
shape and size of particles, 108, 110
kaolinite, 262-263
staining test, 275
methylation, 269-270
structure, 50 montmorillonite, 252-262
X-ray diffraction, 90 structural implications, 269-270
Nonclay-mineral components of clay
materials, 4 organophilic complexes, 265-269
methods of analysis, 5 polar molecules, 257-262
Nontronite, chemical analyses, 371 thermal reactions, 256
composition, 60 vermiculite, 263-264
definition, 35 Organic material, methods of analysis, 6
dehydration curve, 194 Organic molecules, adsorbed, decomposi-
dehydration data, 212-222 tion resistance, 265-266
differential thermal curve, 197 geometry, method of determining.
electron micrograph, 118 273
history, 34 hydration influence, 183, 254
hydrothermal deposits, 324 orientation, 254
infrared spectra, 308 Organophilic clay-mineral complexes,
optical properties, 279 265-269
staining test, 275 swelling behavior, 266
structure, 58 Oriented-aggregate techniques, micro-
synthesis, 319 scopic, 288-289, 292-294
X-ray diffraction, 85, 92
o Origin of clay minerals, diagenesis, 351-
Octahedral structural unit, 43 352
Octaphyllite structures, 58, 65 hydrothermal, 323-330
high-temperature alteration, 223 soils, 342-343
Olivine, 240 . synthesis, 316-323
Clay Mineralogy
P Quartz, differential thermal curve, 199
high-temperature development, 226-
Palygorskite, cation-exchange capacity, 228
129
chemical analyses, 373 R
definition, 38
dehydration curves, 192, 195, 243 Rational analysis, 301
dehydration data, 243-244 Rehydration, biotite, 238
density, 314 chlorite, 241
differential thermal curves, 199, 243- definition, 190
244 halloysite, 220
high-temperature phases, 244 illite, 237-238
history, 37 kaolinite, 217
optical properties, 279, 285 montmorillonite, 229-230
soils, 341 muscovite, 238
structure, 77 sepiolite, 244
Paramontmorillonite, 37 vermiculite, 232-233
Pedalfer, 337 Rendzina soils, clay-mineral composition.
"Pelinite," 14 341
Penninite, 98, 240 definition, 335
Periclase, 225 River deposits, 354
Permutites, 126, 139
Phlogopite, 65, 95 s
dehydration, 233
high-temperature phases, 236 Saline sediments, 359-360
Pholorite, 41 Saponite, chemical analyses, 371
Phosphate fixation, 158 composition, 60
'Planosols, clay-mineral composition, 340 definition, 34
definition, 335 electron micrograph, 118
Plasticity, definition, 1 optical properties, 279
theory, 173 structure, 58
Podsol soils, classification, 338 Sarospatite, 36
clay-mineral composition, 340 differential thermal curve, 198, 246
definition, 333, 335, 338 Sauconite, 58
weathering process, 343-343 electron micrograph, 119
Polar molecules, clay-mineral complexes, optical properties, 279
257-262 Sediments, ancient, clay-mineral com-
nature, 257 position, clays, 358
organophilic complexes, 266 coal, 359
Polarization factor, 97 dolomite, 359
Pore water, 161 flint clay, 358-359
Potash-bearing clay mineral, 35 fuller's earth, 360
Potassium fixation, 154 gumbotil, 360
Prairie soils, 335 limestone, 359
clay-mineml composition, 340, 341 salines, 359-360
definition, 335, 339 shale, 357-358
Protein, exchange reaction, 252 slate, 358
pyrophyllite, 57 till, 360
differential thermal curve, 199 underclay, 358
high-temperature phases, 228 clay minerals and geologic age, 356-
infra red spectra, 307-308 357
staining test, 275 lacustrine, 356
synthesis, 317-320 lagoonal, 355-356
marine, 355
Q recent, diagenetic processes, 351-352
Quartz, <X-, 225-228 fluviatile, 354
fJ-, 225-228 glacial, 335
Index 383
Sediments, recent, glauconite, 352-353 Soil profiles, 332-334, 337
lacustrine, 353-354 Solonchak, 335
marine, 348-351 Solonetz, clay-mineral composition, 340-;
varved,335 341
Sepiolite, ancient sediments, 355-360 definition, 335
cation-exchange capacity, 129 Solubility of clay minerals, in acids (see
chemical analyses, 373 Acid solubility)
definition, 37 in alkalies, 301-303
dehydration curve, 192, 195, 241 effect on, of heating, 297
dehydration data, 241, 242 of particle size, 301
density, 314 electrodialysis, 299-300
differential thermal curve, 199, 241-242 nature of reaction, 298-299
electron micrographs, 123, 124 neutral salts, 300
high-temperature phases, 242-243 rational analysis, 301
history, 37 Specific gravity, chlorite, 314
optical properties, 222, 285 halloysite, 312
rehydration, 244 illite, 312
shape and size of particles, 125 kaolinite, 216-217, 312
soils, 341 montmorillonite, 229, 313-314
structure, 77 palygorskite, 314
thermal expansion, 243 sepiolite, 314
X-ray diffraction, 102 vermiculite, 314
Sericite, 325 Spinel, from biotite, 236
Sericite-like minerals, 35 from chlorite, 240
Sericitization, 326 from illite, 236-237
Shale, clay-mineral composition, 357-358 from montmorillonite, 225-227
definition, 2, 3 from phlogopite, 2:16
Sheridanite, 98, 240 Staining tests, clay-mineral data, 275
Siderite, 209 description, 274-275
Silica tetrahedrons, 44 limitations, 275-276
Silicification, 327 Structural inheritance, 226
Sillimanite, 244 Structure, of adsorbed water, 162-171
Silt, 5 of clay minerals, allophane, 45
Slate, 358 chlorite, 29-72
Smectite, 34 halloysite, 52-55
Soil, aridic, 339-340 illite, 65-69
azonal,334 kaolinite, 46-51
Chernozem, 335-340 mixed-layer, 79-80
classification, 334-335
montmorillonite, 55-64
clay-mineral composition, 340-341
in relation to cation exchange, 132-
definition, 2
134
Desert, 335-340 sepiolite, palygorskite, attapulg:ite,
forest, 335, 338
77
intrazonal, 334 vermiculite, 72-77
laterite (see Laterite soils)
Planosol, 335, 340 Surface area, clay-mineral data, 311-312
Podsc,!s (see Podsol soils) method, 273, 308-311
Prairie (see Prairie soils) Synthesis of clay minerals, from feldspars,
profile development, 332-334 319-321
Rendzina, 335, 341 from oxides and hydroxides, 313-319,
Solonchak, 335 321-322
Solonetz, 335, 340-341 transformations, clay minerals, 322
tilth, 127
Soil, tundra, 335, 337 T
weathering cycle, 342-343
weathering processes, 330-332 Talc, differential thermal curve, 199
zonal,334 high-temperature phases, 223
LIBRA.RY
, A6r\cul tU"
""."~'e(\r"
J_. b - 01 r-
Andhra. Pradesh .
b. mirl1ltura\ UniverSl:!_
384 Clay Mineralogy
Tetrahedral structural units, 44
Texture, 7
binding forces, 9 Wall-rock alteration, 325-328
methods of ana lysis, 7 in relation to mineralization, 329-330
significance, 8 Water, adsorbed (see Adsorbed water)
Thermal diffusivity, 205 channel, 161
Thermal expansion, beidellite 216 dipole character, 163
halloysite, 216, 219 interlayer, 161
illite, 216, 237 pore, 161
kaolinite, 216 structure, 163-165
montmorillonite, 229 zeolitic, 242
sepiolite, 243 Weathering, factors controlling, climate,
Thin-section study, 8 331, 342-343
Till, 360 parent rock, 330-331, 342-343
Tilth of soil, 127 time, 332, 342-343
Titanium in kaolinite, 47, 49 topography, 331, 342-343
Titration curves of clay minerals, 130, 131 vegetation, 332, 342-343
significance, 129, 131 processes, 342-344
Trioctahedral structures, 58, 65 profile development, 332-334
X-ray diffraction, 93-95 in relation to ion exchange, 127, 141
Tundra soil, clay-mineral composition, Weathering cycle, 342-343
337 reversion of, 343-344
definition, 335, 337
X
U X-ray analysis, application to clay
minerals, 84-86
Underclay, 358 high-temperature technique, 211
history of application of clays, 16
V organic complexes, 271-272
X-ray-diffraction data, attapulgite, 103
Varved sediments, 335 biotite, 94
Vermiculite, ancient sediments, 355-360 chlorite, 97-98
cation-exchange capacity, 129 dickite, 90
chemical analyses, 373 halloysite, 87-91
definition, 35 illite, 93-97
dehydration curves, 192, 231 kaolinite, 87-91
dehydration data, 23] -232 mixed-layer structures, 102-104
effect of adsorbed cations on, 231- montmorillonite, 91-93
232 montmorillonite anhydrite, 223-224
density, 314 muscovite, 94
differential thermal curves, 199, 231- nacrite, 90
232 sepiolite, 102
fusion temperature, 232 vermiculite, 99-101
high-temperature phases, 232 X-ray-diffraction intensity and lattice
history, 35 substitutions, 95-97
optical properties, 279, 285
Z
organic complexes, 263-264
rehydration, 232-233 Zeolite minerals, 15
structure, 72-77 ion exchange, 126
X-ray diffraction, 99-101 Zeolitic water, 242
Volkonskoite, 58 Zinnwaldite, 95

Clay Mineralogy: Ralph Grim

Uploaded by

Copyright:

Available Formats

Clay Mineralogy: Ralph Grim

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Clay Mineralogy: Ralph Grim

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are discussed in the book?

What topics are discussed in the book?

CLAY MINERALOGY

New York Toronto London

THE MAPLE PRESS COMPANY, YORK, PA.

2. CONCEPTS OF THE COMPOSITION OF CLAY MATERIALS 11

3. CLASSIFICATION AND NOMENCLATURE OF THE CLAY MINERALS 27

4. STRUCTURE OF THE CLAY MINERALS. 43

6. SHAPE AND SIZE-ELECTRON MICROGRAPHS 106

Reactions with Montmorillonite and Halloysite 252

11. OPTICAL PROPERTIES 278

12. MISCELLANEOUS PROPERTIES 295

13. ORIGIN AND OCCURRENCE OF THE CLAY MINERALS 316

14. ORIGIN AND OCCURRENCE OF THE CLAY MIl\fERALS (CONT.) 348

FACTORS CONTROLLING THE PROPERTIES OF CLAY MATERIALS

The factors which control the properties of clay materials or the

3 Grim, R. E., P. F. Kerr, and R. H. Bray, Application of Clay Mineral Technique

Properties of Certain Bentonites, Ph.D. thesis, University of Illinois (1951).

Chemical Analysis, Am. Mineral., 30, 1-26 (1945).

leum, Bull. Am. Assoc. Petroleum Geol., 31, 1491-1499 (1947).

Mining Met. Engrs. Tech. Pub. 2400 (1948).

Concepts of the Composition of Clay Materials

Because of the importance of clay materials in ceramics and other

vol. 7, pp. 1-60, Springer, Berlin (1933).

Res. Bull. 60 (1923).

der Ackererde, Landw. Verso Sta. 35, 69-136 (1888).

Geol. pp. 205-211 (1911).

Soil Sci., 30, 459-495 (1930).

Isoelectrie Weathering, Soil Sci., 34, 209-239 (1932).

and Comments, Soil Sci., 56, 443-456 (1943).

CLAY -MINERAL CONCEPT

10, 204-211 (1887).

Zusammensetzung andert, Z. anorg. Chem., 63, 69-139 (1909).

8ubstanzen zu identifizieren, Z. Krist., 58, 108-112 (1923).

alien Kunsprodukten und dichten Gesteinen, Z. Krist., 60, 55-69 (924).

Chem. Soc., 45, 2669-2678 (1923).

ties of Beidellite, J. Wash. Acad. Sci., 15, 467-468 (1925).

Their Physical Properties, J. Am. Ceram. Soc., 9, 77-96 (1926).

105E, pp. 151-175 (1931).

Material of Bentonites, Soils, and Zeolites as Revealed by Chemical and X-ray

Petrol., 1, 55-65 (1931).

Paper 185G, pp. 135-148 (1934).

Intern. Congr. Soil Meeh., 3,8-12 (1948).

48 Ross, C. S., and S. B. Hendricks, Minerals of the Montmorillonite Group, U.S.

Halloysite, Am. Mineral., 23,275-300 (1938).

California Soils, Soil Sci., 48, 201-255 (1939).

ican Alkali Soils, Soil Sci., 51, 101-124 (1941).

of Cry~tal Lattice, Am. Mineral., 35, 275-238 (1950).

Ceram. Soc., 22, 141-151 (1939).

and Other Hydrous Materials, Am. Mineral., 27, 746-761 (1942).

U.S., 16, 123-129 (1930).

68 Gruner, J. W., The Structure of Vermiculites and Their Collapse by Dehydration,

Am. Mineral., 19,557-574 (1934).

Lattices, J. Chern. Phys., 10, 147-167 (942).

morillonit, Z. Krist., 86, 340-348 (1933).

Nachweis von Kaolinit, Halloysit, und Montmorillonit, Z. Krist., 94,337-348 (1936).

80 Correns, C. W., et al., Die Sedimente des aquatorialen atlantischen Ozeans,