THE COPPER SULFATE-SODIUM SILICATE REACTION 439

the same pattern was obtained that resulted when the precipitation was
carried out a t lower temperatures under the conditions for yielding calcite
in the presence of the same amount of manganese, as exhibited in pattern
e. Both of these products emitted the same low red fluorescence, despite
the difference in preparation. The formation of the “calcite” rather than
the “aragonite” structure can therefore be accepted as the determining
factor for the occurrence of fluorescence under activation by manganese.
Attempts to prepare fluorescent aragonite by precipitation failed also
when lead, copper, and samarium were used as possible activators. Fluo-
rescent aragonite does, nevertheless, occur in nature. A specimeno from
Girgenti, Sicjly, examined by the author was unexcited by 2536 A. but
under 3650 ,4. emitted a rose fluorescence.
REFERENCES
(1) AVERELL, P . R., A N D WALDEN,G. H., J R . : J. Am. Chem. SOC.69, 906 (1937).
(2) BYLER,W.H . : J. Am. Chem. SOC.60, 1247 (1938).
(3) FONDA,G. R.: J. Phys. Chem. 43, 561 (1939).
(4) MELLOR,J. W.: A Comprehensive Treatise o n Inorgantc and Theoretical Chemis-
try, Vol. 111, p. 818. Longmans, Green and Company, London (1923).
( 5 ) NICHOLS,E . L., HOWES,H. L., A N D WILBER, D . T.: Calhodo Luminescence and
the Luinznescence of Incandescent Solids, p. 23. Carnegie Institution,
Washington, (1928’1.
(6) PRINGSHEIM, P. : Flvorescenz und Phosphorescenz, p. 312. Julius Springer,
Berlin, (1928).

T H E COPPER SULFATE-SODIUM SILICATE REACTION

HAROLD V. ANDERSON AND FRAIiK P. HOCHGESANGI
Department of Chemistry, Lehigh University, Bethlehem, Pennsylvania
Received J u n e PO, 19S9

INTRODUCTION

When solutions of copper sulfate and sodium silicate are mixed a blue-
green precipitate forms. This precipitate is of varying composition (4)
and may or may not be gelatinous depending upon the c u 0 : S i O ~ratio
of the initial solutions. The precipitate or gel may be boiled, or the dried
material may be heated a t lOO”C., without turning black. Neville and
Oswald (6) have shown that a copper hydroxide suspension does turn black
upon boiling but that gelatin will stabilize and protect this precipitate.
Because of this fact Weiser (9) suggests that the precipitate obtained upon
1 William L. Heim Research Fellow, 1937-39.
440 HAROLD V. ANDERSON AND FRANK P. HOCEGESANG

mixing copper sulfate and sodium silicate solutions may be blue copper
hydroxide stabilized by hydrous silica. It is the purpose of this investiga-
tion to establish the composition of this precipitate with the aid of x-ray
powder analysis.
PREPARATION AND ANALYSIS

The materials were precipitated by mixing M/10 copper sulfate and

M/10 sodium metasilicate solutions in varying proportions. The majority
of the precipitations were carried out a t the boiling point, but a few were
made at room temperature. The principal difference detectable by x-rays
in the products obtained by precipitation at various temperatures appears
to be one of particle size only. The materials to be discussed here will be
those obtained by precipitation from boiling solutions, since the larger
particle size is more satisfactory for x-ray analysis. After mixing, the
solutions and precipitates were maintained at the boiling temperature for

TABLE 1
Results of chemical analysis

€I€? cent €I€? e a t p a colt p aMI1 par c a t

2 to 1 65.29 16.63 1.12 33.03 99.4
1 to 1 59.48 14.93 9.94 29.w 99.4
1 to 2 48.30 0.97 35.65 14.57 98.5

10 to 15 min., washed by decantation, transferred to a filter, and washed

again. They were then dried to constant weight in an oven at 65-7OoC.
The results of a chemical analysis of these materials are tabulated in table 1.
X-ray spectrograms were obtained of all the materials. The patterns
were taken using copper K, radiation (0.001 in. nickel foil filter) obtained
from a Philips Metalix tube operated a‘ 38 kilovolts and 14 milliamperes.
The exposure time was 10 hr. in a previously calibrated cylindrical camera
of 59.10 mm. radius. The samples to be analyzed were ground in an agate
mortar, pasted on a hair with a little petroleum jelly, and rotated during
exposure. Also, pinhole patterns were taken of all the materials in order
to follow variations in particle size.
DISCUSSION

The wet precipitates varied from those which were easily filtered to those
which came down in a voluminous, gelatinous mass, while the color varied
from a pale greenish blue to deep blue. These properties were found to be
 THE COPPER SULFATE-SODIUM SILICATE REACTION 441

in direct order with increasing silica content and with decreasing particle
size, as shown by the pinhole patterns. With the exception of those
precipitates which contain a relatively large amount of silica, Le., those
made using more than one volume of M/10 sodium silicate to one of M/10
copper sulfate, the materials are definitely crystalline. Also, all the
materials prepared exhibit the same basic diffraction pattern. Even in
the dense background scattering which appears on the patterns of the
essentially non-crystalline materials, one can easily find weak lines cor-
responding to the strongest lines of the crystalline materials.
Diffraction patterns of a great variety of known materials were obtained
in an effort to establish the identity of this unknown material which is
common to all the precipitates. That the material is not blue copper
hydroxide is quite evident by reference to figure 1. This figure shows
typical patterns of the unknown material with comparison patterns of
blue copper hydroxide (stabilized by the addition of gelatin), the rare
copper silicate mineral dioptase, and the more common mineral chrysocolla.
Mineralogical literature (2) records several analyses of dioptase which show
conclusively that it is CuSiO3.H20. Chrysocolla is thought to be the
dihydrate, but it occurs in nature intimately mixed with impurities so that
its composition is not certain.2 I t is improbable that anhydrous copper
silicate can exist as such Upon heating either dioptase or chrysocolla to
a temperature a t which water is given off (ca. 500°C.) the samples de-
crepitate, decomposing completely into cupric oxide and the a-quartz
modification of silicon dioxide.
Posnjak and Tunell ( 7 ) investigated with considerable thoroughness the
system cupric oxide-sulfur trioxide-water. Among their results they
recorded powder x-ray diffraction data for several of their basic copper
sulfates, the identities of which are conclusively established. Table 2
records their data for several of these compounds and compares them with
those obtained for the precipitated ‘‘copper silicates.” Both the relative
intensities of the lines and the interplanar spacings agree very well.3 The
unknown crystalline material present in all the prepared precipitates is
thus shown to be a mixture of the basic sulfates 4Cu0.S03.3Hz0 and
3CuO.SO3.2HzO.
Chemical analysis of the prepared materials supports the fact that they
consist of a basic copper sulfate plus some other amorphous material,
presumably hydrous silica. Attempts to calculate the original analysis
Chemical analyses made on the mineral samples from which the x-ray patterns
were obtained (figure 1) agreed well with the analysis recorded by Doelter (2).
The sllght discrepancies are accounted for by the fact that Posnjak and Tunell
obtained their data using molybdenum radiation and General Electric cassettes,
whereas the data recorded in column 1 of table 2 were recorded in a cylindrical
camera using copper radiation.
442 ELAIW1.D V . ANDERSON AND FRANK P. HOOHGESANG

FICA. I, I~iiiinrtionpatterns taken i n s i rylindriciil puwdcr cainei.~,tising copper

iirdiat,ion. it, t y p i r d pal,tcin obtained wlmn eqonl, or excess, volumex of c o w e r
rwihtc were used i n precipitation; 11, typii,al p i ~ t , t f i i i iobtained from rnatsriale u i
higher silica c m t m t ; c, b l t cuprir
~ hydroxide; d, dioptuar; e, clirysocolla.

to mole ratios of CuO, SiO,, SO3,and HIO which were very nearly whole
numbcrs ended in failure. However, if tlic silica i s subtract,ed from the
aiid t,hc resulting composition recalculated to 100 per cent,
then tahlc 3 s h o w that the mole rabios for aomc of the compounds are
 Diffraction pattern data
UNKNOWN YATTERIAL 4Cu0~501~3HD 1 ~CUO.SOI.ZH~O'
~-
1e.t. Spacings lent. Spacinga {eat. 1 Spacings
A. 2. d.
4 6.5 7 6.20
2 5.9 5 5.83
8 5.36 7 5.25 4 5.35
1 4.99 7 4.80
1 4.33
10 3.90 9 3.83
0.5 3.52 7 3.58
5 3.20 4 3.14 1 3.31
3 2.92 2 2.88 2 3.07
3 2.79 3 2.99
5 2.68 7 2.65 2 2.73

1
;
1
j
I
2.60
2.52
2.45
1
10
1
2.58
2.49
2.44
4
9
1
2.67
2.55
2.48
2 2.38 1 2.36 2 2.42
2 2.28 2 2.26 1 2.37
3 2.19 3 2.17 1 2.31
1 2.14 1 2.12 1 2.25
2 2.08 2 2.06 10 2.12
2 2.02 1 2.00 1 2.06
2 1.962 2 1.95 1 2.01
1 1.88 1 1.94
2 1.818 1 1.81 1 1.83
6 1.744 5 1.73 1 1.81
1 1.712 1 1.70 5 1.63
2 1.672 1 1.66 1 1.56
1 1.633 2 1.62 1 1.55
2
3
2
2
2
2
1.596

1.534
1.500
1'561
1.463
1.432
[
:
1
~
1
2
2

1
1
1 1.59

1.56
1.53
1.50
1.46
1.42
7

21
2
1 ;;
I 1.31
1.28

:;;
3 1.409 2 1.39
1 1.369
1
3
~

i
1.336
1.311
1
2
1.33
1.31 11
1 ~

2
1
0.5
1.288
1.143
1.095
1 1.28
1 ~ ;:;:
0.5 1.083
0.5 1.073
1 1.058 ~

0.5 0.999
1 0.978
0.5 0.906
I ~

* Columns 2 and 3 reproduce the data recorded by Posnjak and Tunell (7).
444 HAROLD V. ANDERSON AND FRANK P. HOCHGESANG

very nearly whole numbers. The values used here as equal to the water
content were obtained by subtracting the per cent of sulfur trioxide from
the weight loss on ignition. It is highly probable that some water would
be held rather tenaciously by hydrous silica, resisting the original drying
a t 70°C., thereby making the values recorded in table 3 too high. Sam-
ples made from solutions containing more silica than lCuO:lSiO2 ap-
parently do not contain enough sulfate sulfur to make it possible for them
to combine all the cupric oxide as basic sulfate.

S p a c i n g s in A .
FIG.lb. A, copper sulfate-sodium silicate reaction; B, broohantite, 4CuO~SOs
3H10 (by Posnjak and Tunell).

TABLE 3
Mole ratios
Analysis, as in table 1, recalculated t o 100 per cent after subtracting the Si02 from
the total as found
MOLE wnos
VOLUMES O F MI10 CUSoi TO MI10 NalSiOa
USED IN PREPARA’IlON
CUO
- ~~

2 to 1 4.00 1.01 4.44

1to 1 4.00 1 .OO 4.47
1to 2 4.00 0.08 4.97
P & T calculated*. . . . . . . . . . . . . . 4 1 3
P & T found.. . . . . . . . . . . . . . . . . . 4.00 1.01 3.15

* P & T are the values of Posnjak and Tunell for 4 Cuo*S01.3H20.

Attempts to recalculate the original analysis subtracting sulfur trioxide
(instead of silicon dioxide as in table 3) showed a constantly varying ratio
of CuO:Si02 which approximated whole numbers in only one case. The
sample made from one part of sulfate plus two parts of silicate showed a
1:l:l: ratio of Cu0:SiO2:H20, but the few lines obtained on the x-ray
pattern agreed with those of the basic copper sulfate and not the mono-
hydrate of copper silicate.
While igniting the materials in a porcelain crucible over a Bunsen burner
a yellow “center” was found to develop in many cases, especially in those
 THE COPPER hULFATE-80DfLlM SILICATE REACTION 445

precipitates of high sulfate content. This same yellow material was found
when copper sulfate pentahydrate was heated a t MO-65O0C. for a short
time. Binder (1) reported finding such a compound and ascribed to it the
formula S03.2Cu0. Comparison of the spacings of this material with
those given by Hanawalt, Rinn, and Frevel (3) for anhydrous cupric
sulfate finds very little agreement, showing that the yellow compound
is apparently not anhydrous cupric sulfate. All the prepared materials
completely break down into cupric oxide and a-quartz upon heating to
constant weight over a Bunsen burner.
Basic copper sulfates readily form upon the addition of alkali to dilute
solutions of copper sulfate (8). Thus the condition of alkalinity brought
about by hydrolysis of a sodium silicate solution favors the formation of a
basic sulfate. Also, hydrous silica is not necessary to stabilize or protect
the complex sulfate, for equal volumes of M/10 copper sulfate and M/10
sodium oxide (M/5 sodium hydroxide) solutions upon mixing yield a
precipitate which is identical in every respect with the “copper silicate”
precipitates.
The silicate ion in aqueous solution is protected by a hydrated shell (8)
which would tend to make it inactive. Also, upon the least provocation,
such as slight acidity, hydrous silica is thrown out of solution. Several
stoichiometrically possible equations may be written for the copper
sulfate-sodium silicate reaction which will account for the products. If
we write for the reaction the following equation:

4CuS04 + 2NazSi03+ 6 H z 0 --j 4Cu0.S03.3Hz0 + 2HzSi03 +

2NazS04 + HzSOc

then silicic acid itself may be one of the products. Also, acid is generated
by the reaction, thereby supplying the condition necessary to throw down
silica gel from any excess sodium silicate which may be present. If
hydrous silica does accompany the basic copper sulfate precipitate, no
matter what the process by which it is formed, it would account for the
excessive background scattering present on the diffraction patterns.
SUMMARY

1. M/10 copper sulfate and sodium silicate solutions mixed a t the boiling
point yield characteristic precipitates. These precipitates vary from
light bluish green to deep blue in color and from crystalline to colloidal
in size as the concentration varies from excess sulfate to excess silicate,
respectively.
2. These prepared materials were shown by means of x-ray analysis,
supported by chemical analysis, to consist of basic copper sulfate,
4CuO. SOa.3&0, or a mixture of this compound with 3CuO so3 2Hz0.a
446 IRVIN BACHMAN

Also, all these materials were found to occlude varying amounts of an

amorphous material, essentially silica.
3. Hydrous silica is not necessary for the stabilization of the complex
compounds.
4. Reasons are discussed to explain the formation of such a compound
rather than the copper silicate to be expected in such a reaction.

The authors are indebted to Dr. Harvey A. Seville, Head of the De-
partment of Chemistry and Chemical Engineering of Lehigh University,
for his valuable criticism of their paper.
REFERENCES
(1) BINDER:Ann. chim. 6,337-409 (1936).
(2) DOELTER:Handbuch der Mineralchemie, Vol. 11, pp. 775-9. Theodor Steinkopff,
Dresden and Leipeig (1914).
(3) HANAWALT, RINN,A N D FREVEL: Ind. Eng. Chem., Anal. Ed. 10, 457 (1938).
(4) JORDIS AND HENNIS:J. prakt. Chem. [21 77, 238-61 (1908).
( 5 ) MELLOR:Treatise o n Inorganic and Theoretical Chemistry, Vol. 111, p. 261.
Longmans, Green and Co., London (1923).
(6) NEVILLEAND OSWALD:J. Phys. Chem. 36, 60-72 (1931).
(7) POSNJAK A N D TUNELL:Am. J. Sci. [51 18, 1-34 (1929).
(8) STERICKER:Chem. Met. Eng. 26, 61 (1921).
(9) WEISER:Inorganic Colloid Chemistry, Vol. 111, p. 369. John Wiley and Sons,
Inc., New York (1938).

CONVERGENCE OF TIE LINES I N TERNARY LIQUID SYSTEMS

IRVIN BACHMAN
U.S. Industrial Chemicals, Inc., Fairfield, Maryland
Received J u n e 20, 1959

Tie lines connecting compositions of conjugate solutions on ternary

liquid diagrams appear to converge to a point on the extension of the base
of the triangle. Tarasenkov and Paul’sen (2) have attempted to locate
this point mathematically by writing an equation for tie lines, but these
authors failed to take into account the fact that compositions ona ternary
chart must be translated to the rectangular system of coordinates before
substituting such compositions in the equation. These authors, more-
over, concluded that tie lines converge to a common point.
The mathematical method of the present article corrects the error made
by Tarasenkov and Paul’sen and presents data to show that tie lines for
less concentrated solutions have no common focal point.