THE AMERICAN MINERALOGIST, VOL.

52, JANUARY-FEBRUARY, 1967

CLAY MINERAL SYNTHESIS. II.1 A RANDOMLY

INTERSTRATIFIED ALUMINIAN
MONTMORILLONOID'
W. T. GnaNeursr AND S. S. Por-r-acr<,
M ellon I nstitute, Pi.ttsburgh, P a.

Assrnacr
A randomly interstratified layer lattice silicate (a heteropolytypic montmorillonoid)
was synthesized from a starting slurry of diatomite and bayerite in an aqueous solution of
NaOH. The synthesis was carried out for varying reaction times at 300oC and the cor-
responding 1uro.
Samples quenched at various reaction times were subjected to X-ray difiraction, DTA,
surface area and cation exchange capacity determinations, and infrared absorption studies
with particular emphasis on the 400-900 cm-r region.
The synthesis process is a crystallization from solution, and is made up of several steps:
solution of the solid reactants, nucleation of the layer lattice, and growth of these nuclei to
the final crystalline product. The growth step is sufficiently rapid so that the solution
and/or nucleation steps become rate controlling.

INrnooucrroN
Synthesis of the three-layer clay minerals, such as montmorillonite,
dates from the classicwork of Noll (1930). More recently, the subject
has been pursued vigorously by Roy (Iiyama and Roy, 1963). For some
years, the Research Project at Mellon Institute sponsoredby Baroid
Division National Lead Companlr has been studying various facets of
clay mineral synthesis,with primary emphasison the variation of the
physical propertiesof the synthetic clay as a function of the nature and
extent of isomorphoussubstitution. In general,in the work discussedin
this paper,reactionconditionsweremore moderateand vesselsizeslarger
than those describedby the other workers in this field.
One phaseof our study has been concernedwith the composition

n]- . *M*
[1Alu;""."iAl"si6_,)1"1"oo26(oH)
In this paper, we describea "big-bomb" synthesisof a heteropolytypic
(Iiyama and Roy, 1963)montmorillonoid (or smectite) in which, on the
average,*c21. Reaction times and temperaturesare explored and the
various products characterized.

I For the first paper in this series, which was concerned with the synthesis of hectorite,
see Granquist and Pollack (1960).
2 The word
"montmorillonoid" is used in the sensein which it was originally proposed
by MacEwan (1951); i.e., meaning a mineral of the montmoriilonite group. British workers
have more recently adopted the name "smectite" for this same group.

212
 INTERSTRATIFIED ALUMINIAN MONTMORILLONOID 2T3

ExpBnrlrnNrer
The feed slurry for each synthesis was prepared by addition, with
stirring, of 20 g of acid-washeddiatomaceousearth to one Iiter of distilled
water, solution of AlCIa.6}{1o- (48.2 g) in the resulting mixture, and
precipitation of AI(OH)3 by a 75 ml addition of aqueous ammonia (28/6
NHa). This final mixture was filtered and washed by redispersion and
filtration through three cycles. The final filter cake was redispersedin
distilled water, sufficient NaOH added to establish a NaOH/AlzOr mole
ratio of 0.4, and the volume adjusted to one liter with additional distilled
water.
This slurry was introduced into a Type 347 stainlesssteel Aminco su-
perpressurebomb, with an inside diameter oI 2 9/16 in. and an inside
depth of 21 in., equipped with a standard Aminco closure. The head con-
tained three ports; one leading to a rupture disk assembly, a vent, and a
pressuregauge: one equipped with a probe thermocouple assemblyl and
one containing a long tube running from near the bottom of the vessel
through the head and terminating in a needle-valve.Heating and stirring
were furnished by a standard Aminco heating jacket mounted on a
rocker assembly. The jacket temperature was controlled by an off-on
temperature control device, and the reading from the thermocouple
probe was recorded.Variation from the desiredcrystallization tempera-
ture was approximately *2'C.
The bomb was vented, at the boiling point of the contents, without
rocking until the air had been displaced from the vessel. The vent was
then closed,rocking started, and the temperatureallowedto climb to the
control point. At the end of the scheduledreaction time, the rocker was
stopped with the head of the bomb at the maximum angle above the
horizontal, a water-cooledcondensingdevice was connectedto the needle-
valve, and the contents of the bomb discharged through the condenser.
Flow was controlled to maintain total discharge times of the order of
15 min.
The products were washed with distilled water by filtration and re-
dispersion, and oven-dried at 105"C. Examination of the products in-
cluded: X-ray diffraction studies with both film and diffractometer
techniques;DTA of samplesequilibrated with 50 percent RH; infrared
absorption spectroscopy, with KBr pellets, over the range 400 to 4000
cm-1 (a Beckman IR-9 grating instrument was used in this work); sur-
face area by Nz adsorption (standard Brunauer-Emmett-Teller [BET]
method; Emmett, 1942) ol samples which had been outgassedat 500"C
and at least 10-a mm Hg; and cation exchangecapacity (CEC) by a
suitable variation of the ammonium acetate technique (Lepper, 1945).
2t4 W. r. GRANQUTST
AND S. S. POLLACK

As will be evident, all of thesetechniqueswere not applied to all product

samples.
In the X-ray studies, powder diffraction patterns (North American
Philips camera,diameter 114.6mm) w€re first obtained. Diffractometer
traces were then made for both random and oriented specimens.Nickel
filtered copper radiation was used in both diffractometer and camera
work.
Oriented specimenswere preparedby suspendingthe various products
in water and then letting the solids settle on gla,ssslides. No dispersing
agent was added in this step. A diffractometer trace was made of each
of two oriented slides prepared for each clay sample. One slide was then
treated with a 10 percent glycerol in ethanol solution and again stuCied
with the diffractometer.The secondslide was heated at 440oCand then
500"C (two hours at temperature) and diffractometer traces were ob-
tained correspondingto each temperature.
The intensity of the 06 reflection was measured as follows. Finely
ground sampleswere packed in aluminum holders 20X10 mm in area
and 1 mm thick. All intensities were measuredunder the same instru-
mental setting on the Geigercounter; that is, a scalef actor of 2, multiplier
setting of 1, and time constant of 1. The goniometerscannedat a rate
of ]o per minute. The resulting peaks on the chart were symmetrical
except for the weak peak for the 12-hour 300oCsample.The integrated
intensitl' was obta,ined by multiplying the width at half-maximum by
the peak height.

RBsur,rs qno DrscussroN

The results of this study appea,rin Tables I and 2 a,nd Figures 1
through 8. The time-dependentfeatures of particular interest are the
change in the integrated intensit.vof the 06 reflection (see Fig. 1); the
behavior of the 001 reflection (before and after glycerol soivation, Figs.
2 and 3);the variation in surfacearea and CEC; the shifts seenin the
400-900 cm-l region of the infrared; and the changesin the differential
thermal analysis patterns.
In the 300oCproducts,exceptthe 72-hoursample,the montmorillonite
is randomly interstratified with mica-like la,yers.In the 72-how sample,
however, the product is a mixture of montmorillonite a,nd regularly
interstratifiedmontmorillonite-mica.This interpretation is ba"sedmainly
on the behavior of the oriented specimensfollowing gl1'ceroltrea,tment.
All the products except the 72-hour sample show an irrationa,lbasa,lse-
quenceand an 001 reflectionwith a shoulderon the high angle side. The
72-hoursampleha.san 001sequencewhich is closeto ra.tionaland in a,ddi-
 I N TERST RAT I FI ED A LU M I N I AN MON TMORI LLONOI D

Teelr 1. Suuu,qny ol Sonr, Trun-DrpetnnNl Pnopnn:rrrs or trre

SyNrnrsrs Prooucrs

Time, Bulk density Relative2 Surface area3 CEC4

Temp.,'C
hours g/ccr Intensity (06) m2/g meql100 g

0 o-22 0 oo. /
24 279 .56 0.68 128
35 279 .86 0 .8 4 120
46 279 .88 o.97 96.4
JY 279 .94 1.0 92.3

0 .22 0 o o ./
12 300 .47 040 133 7t -64
24 300 .44 0 .8 4 80 90.46
36 300 .93 I.IJ 83 llo.67
48 300 .99 1.08 77 9 1. 2 8
72 300 .93 1.13 62 8 1. 9 2

t 30/60 mesh particles, packed to minimum volume.

2 Relative to sample at 59 hours, 279'C.
3 Samples outgassed at 500'C and ca. 10-a mm Hg.
a Dry clay.

tion a peak at 9.2 h. This peak is probably the third order of a 1:1
regularly-interstratifi ed structure.
Application of MacEwan's (1961) technique (tor l0/12.4 A and 10/
17.5 A random interstratification) to the observed doorfor the oriented
specimensof samples near equilibrium before and after treatment with
glycerol,leads to a p(10 A)=0.+. Thus, it is concludedthat the samples
contain approximately 40 percent mica-like layers and 60 percent mont-

e
t
o
N

E
@ o

Frc. 1. Variation of intensity of 06 reflection with reaction time,

216 W. T. GRANQUIST AND S. S. POLLACK

Tlgr,n 2. X-R.rv Drlrnacrrorq Dlte lon 36-Hout ar.ro72-Houn

Pnooucrs SvNtnnsrzrr .q'r300'C

Treatment of Difi. Half- Peak height

Time, in
oriented scale width
hours deg 20 chart units
slide settings

11.8 r.7 62
7.2
5.79 B 7
4.45 hh 2
4.05 crist. 8
.t- lo B t2

glycerol t 7. 3 2.5 26
solvated 9.3-rt shoulder B 16
25"C 5. 9 8 VB J

4.61 B 9

4.05 crist. S 1t
3.49 t.4 20

103 1.5 56
7.2 kaol B a

5. 0 0 I.J 28
4.46 hk 0.5 5
4.05 crist. 0.3 15
333 1.6 34

500"c 9.9 1. 3 0 60
2 hrs 4.90 1.0 38
4.46 hk 7
4.07 crist. 0.3 20
3.22 1. 3 0 70

72 r1.8 1.95 43
7.2 kaol. S IJ

5.86 B 6
4.05 crist. MS 2
j
.1..)/ kaol. S
3.12 B 10

glycerol 4n a
2.0 49
solvated 9.2 1.6 tl
25'C 7.2 S 18
5.90 B 5
4.52 MS 10
q
4.05 5

S:Sharp. B:Broad. MS:Medium Sharp.

 I NT ERSTRATI FI ED A LU M I N IAN MON TMORI LINNOI D 217

Trsre 2-(Continued.)

Treatment of Difi. Half- Peak height

oriented scale Comments width ln
slide settings deg 20 chart units

3. 5 6 kaol.
+mont.
2.96 mont. B 6

4q"c 32-0.6-2 10.2 L.4 47

2 hrs 7.r kaol. 0.6 18
4.95 t.4 t6
4.48 hk 0.8 2
4.05 crist. 0.4 ,
5.JO kaol. 1.0 t.)

3.30 r.9 22

500"c 9.9 0.8

2 hrs. 4.84 o.7 18
4.48 hh 0.2 J

4.W crist. 0.15 8

3.20 0.8 31

morillonite-like layers randomly interstratified. Therefore, in the unit

cell formula presented in the introduction to this paper, r, although
over-all approximately equal to unity, must assumetwo ranges of value:
0 ( r ( 1 montmorillonite-like,
llsl2micalike.

In a random stacking sequenceof platelets A and B, when the numbers

o f A a n d B a r e a p p r o x i m a t e l y e q u a l , s e q u e n c e sAoAf A . . . ( o r B B B . . . )
must occur in the over-all stacking. W eak hkl lines appear in these prod-
ucts, presumably becauseof mica-like sequencesoccurring for the above
reason. These hkl lines are best seenin the Debye-Scherrerpatterns, but
also appear in the diffractometer traces (Fig. 4).
The sharpest clay peak in the difiractometer traces of unoriented
specimensof these products is the 11.02 peak, and it is also the 6.rst to
develop during the crystallization. Figure 5, which shows the region from
19 to 25"20,demonstratesthe growth of the 11.02 peak and the slow
decreaseof the 101 peak of accessorycristobalite as a function of reaction
time. In general, the crystalline forms of SiOz are not suitable as source
material for the crystallization; these results show, however, a slow loss
of cristobalite to the reaction system.
The integrated intensity of the 06 reflection was considered to be a
218 W. T. GRANQLIIST AND S. S. POLLACK

semiquantitative measure of the amount of montmorillonoid produced

in the synthesis. However, since th^e06 reflectionsof the montmorillonoid
(-1.50 A) and kaolinite (-I.49 A) superimpose,a small part of the 06
intensity of the 24- and 72-hoursamplesis due to kaolinite. This crystal-
lization of kaolinite introduces another complication by somewhat de-
pleting the system of alumina, causing a slight increasein the SiOs/AlzOr
for the 3-layer material. A further complication arisesfrom the fact that

iB i6 14 I 6 4
ltrro,,l,o,

Frc. 2. Behavior of 001 as a function of reaction time, oriented slides, 300oC series.

these products are not homoionic, but rather contain a mixture of Na+
and NIL+ in the exchangesites. The ratio Na+/NHa+ undoubtedly varies
from platelet to platelet, but we suspect that the NHr+ is tlie predomi-
nant cation for the mica-like component and Na+ for the montmorillonite'
It is concluded that the product crystallizing initially was montmo-
rillonitelike platelets randomly interstratified with a somewhat lesser
amount of mica-Iike layers, an arrangement that remained essentially
constant until the equilibrium amount of crystalline material was ob-
tained (note behavior of the 06 reflection). Beyond this point, the total
amount of crystalline material remained constant, but there was an in-
creasing tendency for the interstratified material to separateinto a mix-
 I N T ERSTRAT I FI ED A LU M I N I A N MONTLTORI LLONOI D 2t9

o2e(CuKd)

Frc. 3. Rehavior of 001 upon glycerol solvation as a function of reaction time,

oriented slides, 300oC series'

ture oI two separate phases,montmorillonite plus regularly interstrati-

fi ed montmorillonite-mica.
It is of interest to compare the variation of the CEC (Table 1) with
these structural changes.The increaseof CEC to a maximum in the 36-
hour sample parallels the increase in the relative intensity of the 06 re-
flection. That is, in a mixture which was diatomite and bayerite initially,

4o4A
lu r,oz
| +r+A
il
M hkl
2 a89A
o
: srA
l"oor
M 33106 "3.'sa
nr,
r+sgA M3l;15:24 l1sA
I 168oA
I

Prcdwt oftd 59hr d 279'c

43 41 39 37 35 33 31 29 27 25 17 15 13 ll
63 61 59 57 55 53 5l

Frc. 4. X-ray difiractometer trace of an unoriented specimen of the

59 hour. 279oC samPle.
220 W. T. GRANQUISTAND S, S. POLLACK

the cEC of the product (clayfdiatomitegbayerite) increased as the

amount of clay increased,as would be expected. rrowever, the decrease
in cEC beyond 35 hours, corresponding to an increasing separation of
the interstratifi.ed product into a mixture of two phases,is more dificult

Monfmorillonoid
ll iOZ

C r y s l o l l i z o l i oTni m s
=24 ht

Time
=36 hr.

llizolion Time
=48 hr.

2790C
Series
tttll
23 22 20 t9
ozle

Frc. 5. Region 19 ro 23" 20for unorientedspecimensat severalreaction times.

to explain. In this region, the relative intensity of the 06 reflection, and,

thus, the total amount of clay is constant. Since the cEC obtained by the
NHeAc technique doesn't necessarily reflect the total charge deficiency
of the lattice because some of the charge-balancing cations may be
"bound" by the clay, this decreasing ,,apparent CEC,' probably results
from an increase in the number of bound cations, perhaps in the regularly
interstratified part.
 INTERSTRATIFIED ALUMINIAN MONTMORILLONOID 22I

The surface area undergoes an increase during the initial stages of

hydrothermal treatment, probably due to the joint effects of breakdown
oi the diatoms and appearanceof montmorillonoid nuclei of small size
and correspondingl.vhigh area. After this early increase,however, surface
area decreasesas a function of synthesis time, and this decreaseparallels
the increase in the amount of crystalline product present. Continuing
d.ecrease of area in the region of constant 06 intensity is due to increasing
crystallite size. In this d.iscussionof surface area, it is not implied that

ir /

li' t = 5 9h r s .

t = 4 6h r s .
i =3 5 h r s .
t = Oh r s .
(Control)

/f\
I
\
I

To strong peok ol
47Ocm-'
900 800 700 600 500 400
I
Wovenumberin cm

Frc. 6. Infrared spectra in the region 400-900 cm-l for varying

reaction times at 279oC.
 222 w. r. GRANQUIST AND S. S. POLLACK

t = 4 8h r s .
, I =36 hrs.

,/ : !) r '
H
i) I
rl lO
t-
t@
t=24hrs.
I

t =l 2 h r s .

t=Ohrs.
(Conirol)

3OO"CSeries

900 Boo 700 600 soo qoo

Wovenumberin cm-l
Fro. 7. Infrared spectrain the region 400-900cm-r for varying
reaction times at 300oC.

the BET area, obtained by Nz adsorption,of samplesheated in vacuum

to 500oc is indicative of the true area of the crystallites in the reaction
mixture. Rather, BET area is reported as an interesting time-dependent
parameter.
The apparent density of 30/60 mesh particles of the various products
also follows the increase in 06 intensity, and, thus, the increase in the
concentration of the montmorillonoid phase. This density increase oc-
curs becausethe geometry of the platelet stacks is such that more mass
can be accommodated within a given volume. From the data in Table 1
 I NT ERSTRATI FI ED A LT]M I N I A N MONTMORI LLONOI D 223

it would seemthat the apparent density of the product is a simple method

of following the course of the crystallization.
The infrared absorption spectra, in the region 400-900 cm-1, of many
of the quenched samples appear in Figures 6 and 7. Stubidan and Roy
(1959, 1960) have discussedassignmentsof the various infrared bands
of layer silicates. On the basis of their work the band at approximately
480 cm-l is attributed to Si-O, while the band at 540 cm-1 is due to some
Si-O-Alvr mode. They do not discuss the fairly strong band at 780-800
cm-r which we observe for diatomite and for the 0-hour control sample'
but it seemsreasonableto attribute this also to an Si-O mode, possibly
arising from the accessorycristobalite.
Because of its relationship to the formation of the layer lattice, the
band of most interest is that for Si-O-Alvr. That is, this band is apparent
only as a trace (if at all) in the control sample, but as crystallization pro-
ceeds,it increasesin intensity almost in direct proportion to the increase
of the 06 reflection intensity. A more quantitative approach is not possible
at this stage because of uncertainty in the base line position, but the
change in intensity of the 540 cm-1 Si-O-AlYr band relative to the 480
cm-1 Si-O band certainly supports this statement' Note that the band
at co. 800 cm-i, present in the diatomite and in the control sample'
diminishes as the crystallization proceeds and the amount of diatomite
decreases.
The DTA patterns of the 279"C series of products are presented in
Figure 8. The free-water endotherm becomes well-defined alter 24 ht'
synthesistime; the double-peakfeature of this endotherm is undoubtedly
related to the presence of Na+ and NH++ on the exchange sites' The
dehydroxylation endotherm increases in intensity and shifts towards
higher temperatures as crystallization proceeds. The apparent double
dehydroxylation endotherms for the 46- and 59-hour samplesresult from
superposition of an exotherm on an endotherm. In many instances, in
our work involving synthetic three-layer clay minerals, a strong sym-
metrical exotherm has been seenin this region. The processinvolved may
be combustion of NH3, but tests to establish this point have been incon-
clusive. The high temperature exotherms increasein intensity with syn-
thesis time and also shift to higher temperatures, indicative of increasing
stability of the structure. The high temperature processinvolved here is
the crystallization of cristobalite (major) and mullite and r-AlzOa (both
minor).

some speculotionconcerningmechanism.The comments which follow can-

not be completely supported on the basis of the preceding information;
however, neither are these comments at odds with the data'
 224 W. T. GRANQUIST AND S. S. POLLACK

fi
/\
t\
J\

tl
100 2o0 300 400 500 600 900 rooo

Frc. 8. DTA patterns for the various products synthesized at 279"C.

The processinvolves the solution of the solid reactants (alumina and

silica), nucleation of the layer lattice, and growth of these nuclei to the
final crystalline product. Thus, the reaction is a transfer of materiar from
solid reactant to solution to solid product in some variation of the foilow-
ing scheme:

[SiOr] ""ria, mo4bous+ [A1(OII)B] solid,sibbsite

or bayerite

[Al4AlSi?Oro(OH)n]-. M* (nucleus)

J (s'o*tt)
[A]4AlSi?Or0(OH)rl-. M+ (final crystallite)

It is suggestedthat the growth step is sufficiently rapid so that the solu-

tion and/or nucleationprocesses becomerate-controlling.The nucreation
step probably follows the route suggested by Cailltsre, H6nin and
 I NT ERSTRAT I FI ED A LU M I N I AN MON TMORI LI'ONOI D 225

Esquevin (1953). These authors have stated that the formation of clay
minerals under ordinary conditions of temperature and pressure (and
surely also under the synthesis conditions used in this work) appears to
be determined by the existenceof a brucite-type hydroxide layer which
induces the SiOatetrahedra to develop a layer lattice. If they are correct
in this supposition, then the mechanism would involve the nucleation of
Al(OH)s followed by immediate condensationwith Si-OH to produce the
desired structure, as illustrated.

\ l,/ \ l/^\1,/
Si Si Si
o Hl I
H \1,/ oo o
HO OH
o
Si
-HrO
/\1.\ t , /
\/,
H , rIo-al-oH H AI
o ,/\ // l\.
I oo ./l
Si HH oo o
,/ l\ H H
o Si Si
.t ,/ lv/ l\
b1
,/ l\
This cond.ensationprocesscreates the Si-O-AlvI bonds which give rise to
the infrared band at about 540 cm-l. The gibbsite layer would continue
to grow with accompanying condensationof the gibbsite-OH with Si-OH
to yield a crystallite of increasing size in the a-b plane. I'inally, these
platelets become organized into the randomly interstratified stacks in-
dicated by the X-ray data, probably as the result of anlenergy minimiza-
tion process.
We attempted to fit these limited data to the pseudo-firSt order ap-
proach discussedby Granquist and Pollack (1960). That is,

. I(max) - I(t)
't
r(-u")

was plotted againstl, where /(max) is the maximum intensity (at a given
temperature) of the 06 reflection, and 1(l) is the intensity at time ,' The
data could be fitted satisfactorily by a straight Iine through the origin.
The resulting rate constants (r1) at the two temperatures were then
used to compute the apparent activation energy for the crystallization
by means of the integrated Arrhenius equation. The result was 3.9
kca/mole, a value to which we attach no great significance although it
does seem reasonable.
 W. T. GRANQUIST AND S, S. POLLACK

AcrNowrBnGMEN'rs
We acknowledgewith thanks the assistanceof the following: Mr.
R. C. Boteler and Mr. G. W. Hofiman, who ably operated and main-
tained the hydrothermal synthesisapparatus; Dr. G. L. Carlson, who
determined the IR spectra; and Baroid Division Analytical Section,
which furnished the exchangecapacity data. We also enjoyed the oppor-
tunity of discussingvarious phases of this work with professor J. L.
McAtee, Jr., of Baylor University. Finally, the generousaction of Baroid
Division National Lead company in releasingthis paper for publication
is appreciated.
RrlnrnNcns
carr,rinr, S., S. rrfNrw, axo J. EsqurvrN (1953) Recherchessur la synthese des mineraux
argileux. Bull. S oc. F r anc. M iner al. 76, 300-3 14.
Euuerr, P. H. (1942) The Measurement of the surface Areas of Finely Divided or porous
solids by Low Temperature Adsorption Isotherms. In E. o. Kramer, ed. Ad.aancesin
colloiil scienre, vol. l.
Gnarqursr, w. T. aNo s. S. Porr.ecr (1960) A study of the synthesis of Hectorite. cloys
Cloy Mineral,s. Proc. Nat- Conf . Cloys Clay Minerals (1959) S, 150 169.
hveul, J. T. ,lm R. Rov (1963) Controlled synthesis of heteropolytypic (mixed-layer)
clay minerals, Cloys Cloy Minerds, Proc. Nat. Con;f. Ctays CIoy Minerals (1961) fO,
+-21.
Lorrrn, H. A., ed. (1945) ofi,eial and rentati"re Methods oJ Analysis. Association of official
Agricultural Chemists, Washington, D.C., p B-2A.
M.r,cEwe.N, D. M. C. (1951) Montmorillonite minerals. In G.W. Brindley (ed,.) X-ray,
i.d.ent'i.f,calion
and crysla.lstruetures oJclay minerals,Mineral. Soc. London.
- (1961) Montmorillonite minerals. 1z G. Brown (ed.) X-ray i.d.mffication and,crystal
stnutures oJcloy minerals, Mineral. Soc. London, p. 143 207.
Nor.r,, W. (1930) Synthese von Montmorilloniten. Chem. Erde. 1O,129 lS4.
SrurrdeN, v. enn R. Rov (1959) The assignment of the infrared absorption bands in layer
lattice silicates.Tech. Rep. No.3, A.P.I. Proj. 55, Amer. petr. Inst New york.
-- (1961) Isomorphous substitution and infrared spdctra of the laver lattice silicates.
A mer. M iner al.. 46, 32-57.

M anuscript receil)eil,A pril 21 , 1966; acceptedJor publication, June 18,1966.