1086
JAMESF. CORWIN
Vol. 62
HYDROTHERMAL REACTIONS UNDER SUPERCRITICAL CONDITIOiSS. V. REACTIONS BETWEEN SILICA AND ALKALINE EARTH METAL SALTS BY JAMES F. CORWIN Contribution from the Department of Chemistry, Antioch College, Yellow Springs, Ohio Receiaed A p r i l 86, 1968
The reaction of dilute solutions of the salts of the alkali earth metals with silica has been investigated by using the devitrification of clear fused quartz as a measure of the rate of reaction, and X-ray, optical and chemical analysis as a measure of the structure of the crystallized material. By carrying out these reactions under controlled conditions using compounds of the alkaline earth metals it has been found that the crystalline structures resulting from the reactions of these solutions vary in a systematic way from or-cristobalite and several other modifications of the silica structure to silicate minerals. The rate of reaction and the formation of crystalline structures are correlated with the initial p H of the solution and the relative dimensions of the ions involved. On the basis of these correlations, the natural occurrence of or-quartz and talc is inferred since the concentration of these dilute solutions is quite similar to that of ground water.
Introduction The investigation of the reactions of the alkaline earth metal oxides with silica glass2 and a more complete evaluation of calcium and strontium hydroxide reactions3p4has shown that the presence of these materials in high temperature water contributes t o the formation of p- and finally a-cristobalite from fused silica in those cases where the silicates formed are soluble enough t o maintain, by hydrolysis, the necessary p H for this formation.5 These results failed t o explain why &-quartz which is the result of the reactions of the alkali metal compounds with silica under the same conditions was not formed in spite of the fact that a-quartz is found in nature surrounded by alkaline earth metal minerals. Another point left unexplained is the formation of p-cristobalite as an initial, st)able, crystalline form when this form has been gerierally characterized as an unstable intermediate. I n order to t r y t o find an answer to the first of these questions, a series of experiments involving other calcium compounds usually found in nature was planned, and to answer the second, a series using the fluorides of the alkaline earth metals which are in general less soluble than the oxides or hydroxides, so that the reactions would be slower and some insight into the nature of the initial phases of the reactions could be obtained. Experimental The equipment and methods used for reactions and for the analysis of reaction products were identical with those already d e s ~ r i b e d . ~ , ~ All chemicals used conformed to C .P.A.C .S.Standards of Purity.
Results and Discussion Table I contains the results obtained when the alkaline earth fluorides were used. BeF2which is very soluble in mater at room tem(1) This research was supported in part by the United States .4ir Force through t h e Air Force Office of Scientific Research of the Air Research a n d Development Command. under contract No. AF 18(600)1490. hdditional support was received from t h e U. 9. Army Signal Corps (Contract No. D A 36-039 SC-64605) through its Signal Corps Engineering Laboratories a t Fort Monmouth, New Jersey. Reproduction in whole or in part is permitted for a n y piirpose of t h e United States Government. (2) J. F. Corwin, R . G. Yalman. J . W. Edwards and G. E. Owen, Paper No. 1, THIR J O U R X A L61, , 939 (1957). (3) J. F. Corwin, R. G. Yalman, J. W. Edwards and E . R. Shau., Paper No. 11, ibid., 61, 941 (1957). (4) J. F. Corwin, R. G . Yalman. J. W. Edwards and E. R. Shaw, Paper No. I V , ibid., 61, 1137 (1957). ( 5 ) R . G. Yalman a n d J. F. Coririn, Paper No. 111, ibid., 61, 1432 (1957).
perature gave quite different results from those obtained from the remainder of the alkaline earth metal fluorides. When the runs were made at the same concentration, the BeFz reaction resulted in soluble compounds and only 4.0 mg. of solid was recovered. A second pair of runs then was made using 0.1 N BeFz and solid material was formed, but crystal analysis gave new materials that could not be identified by current X-ray files and optical constants. The initial pH of all the solutions is low enough that little devitrification would be expected in the time used for the experiments; however, the hydrolysis of fluorides would make the solution alkaline enough to allow the devitrification reaction t o pro~eed.~ Since the calcium salts are more prevalent in nature than the other members of the alkaline earth metal group, a number of calcium salts were used under the same conditions. A set of experiments designed to re-evaluate the reaction of MgO with water and silica was performed and these with the data obtained when calcium salts were used are contained in Table 11. The calcium phosphates showed little devitrification of the silica rod, and the X-ray data showed no silica in the solid material recovered from the reaction vessel. Chemical analysis confirmed these results since almost all the silica was found in the solution. Analysis of the solid showed 3.0% SiOz in the solids from Ca(H2P04)2.H20, 0.2% from CaHP04.2H20 and 7.8% from Ca3(P04)2. CaS04 and CaClz gave very similar results to those obtained when Ca(OH)z is used3; however, the CaClz reaction was accompanied by a considerable reaction with the vessel walls and the solid material was colored with iron and chromium compounds. The reactions of silica with CaCOj resulted in a mixture of crystalline materials which gave X-ray patterns similar to, but definitely different from, crystalline materials obtained in the formation of syiithetic Tobermorite (XCn0.Si02.YH20 where X-1 and Y-1). This material could be of similar composition, but with variations in the amount of CaO, SiOz, H 2 0 ratio. This would account for the X-ray patterns of the crystalline material. The MgO experiments were conducted as a repetition of earlier work2 where the reaction resulted in a coating on the silica rod that showed no
! i
I 1
REACTION BETWEEN SILICAAND ALKALINE EARTH METALSALTS
Sept., 1958
1087
TABLE I Fluoride concn. 0.025 N , 10 g. of silica glass rod, 400°, 125 ml. vol., 340 atni., 48 hr. Wt. loss silica rod, Fluoride
pHfh
%
5.8
3.3
2.94
Optical and X-ray d a t a of solids a n d devitrified material
Reaction products soluble. Only 4 mg. solid recovered 5.8 2.5 5.25 40-50y0 irregular isotropic. n = 1.440-1.450 BeFzc(2) 30-40y0 irregular birefringent. n = 1.510-1.520 X-ray gives pattern unidentified by A.S.T.M. file 5.7 3.2 1.64 70-80oJ, birefringent. n = >1.400 MgFz MgFz X-ray mostly MgFz. Small amount of talc, 3Mg0.4SiO~HzO 4.5 4.0 5.26 Xi-Ray mostly CaFz, 10% p-cristobalite CaFz 6 2 3 6 14.7 X-Ray p-cristobalite and SrFz SrFz 5.8 3 .2 14.6 X-Ray p-cristobalite and BaFz BaFz More concentrated 0.10 N solution used. Final p H a t room temperature. Initial p H at room temperature.
BeFz (1)
a
p H la
TABLE I1 Salt concn. 0.025 iV, 10 g. of silica glass rod, 400°, 125 ml. vol., 340 atm., 48 hr. Salt
W t . loss silica rod, PHLa
Ca( I-12P04)~H20
3.2
CaHPO4.2HZO
6.8
CadPOdz CaSOl CaC03 CaC12 MgO
6.9 6.8 9.6 5.8 10.3
MgO 10.3 Initial p H at room temperature.
pHtb
%
Optical a n d X-ray d a t a devitrified material
50-60% crystalline. n = 1.630-1.640 X-Ray complicated, but CazPz07-I was identified 5.6 1.45 n = 1.64-1.65. X-Ray hydroxy apatite structure, c a d Po& or C a d O H )z(PO&I 5.9 1.45 Same as CaHPO4.2Hz0 5.2 3.19 X-Ray 0-cristobalite. Anhydrous Cas04 6.3 4.60 Not identifiable by A.S.T.M. file 5.2 9.26 X-Ray p-cristobalite 5.8 4.56 X-Ray very poorly crystallized. 3Mg04SiOz.HzO (talc) amorphous material also 8.8 No silica used X-Ray-small crystalline size Mg(OH)z (Brucite) Final pH a t room temperature. 2.5
1.27
crystalline character. I n this work the rod was scraped vigorously in order t o obtain all of the devitrified material. This technique resulted in the identification of some crystalline material. Conclusions Due t o solubility of the reactant and of the products when BeFz was used, the results definitely separate this reaction from that of the other alkaline earth metal compounds. Unlike the others, the hydrolysis of the salt cannot be used t o explain the reaction since Be02'did not show any devitrification reaction under the same conditions. Before conclusions can be reached concerning the reactions of beryllium salts further investigation is indicated. The remainder of the fluoride work shows that the alkaline earth metal fluorides react by two steps, first the hydrolysis of the fluoride to form HF and the corresponding metal hydroxide and the OHion from the hydroxide causes the devitrification to proceed. This is the same reaction postulated for NaF.S Further proof that the reaction is the same is shown by the similarity in solid reaction products to those r e p ~ r t e d . ~ -The ~ difference in reaction is primarily one *of rate since the fluorides are so insoluble even in hot water the reactions are slowed down by the reduction in concentration of the OHion. The faster reactions of the hydroxides2 in some cases caused coatings over the surface, and a subsequent reduction in devitrification which did not coincide with the basicity of the hydroxide. When the fluorides are used no such interference
wit'h tJhe reaction is encountered, and the amount of devit'rification coincides with the increase in basic quality of the alkaline earth metals. The formation of talc (Table I) in the MgFz run is similar to the other alkaline earth metals since talc is an alt,ernate layer structure of p-crist,obalite and brucite, Mg(OH)2.6 MgO alone under the same conditions of temperature and pressure forms brucite (Table 11). Talc is also formed when MgO is reacting with Si02, but much amorphous material also is formed due to the speed of reaction. The formation of p-cristobalite is expected due to the low pH of the solutions in all of the alkaline earth metal reactions, but in the case of Ca, Sr and Ba salts the difference in structural dimensions of the hydroxides and the packing in their crystal forms, they will not fit between the layer structure which is the proposed form of the P-~ristobdite.~ Thus the two insoluble materials, P-cristobalite and the fluorides, are found separate in the solid phase. Although crystalline a-quartz is found in nature a,long with alkaline earth metal minerals, none has been found in any of our experiments. If the quartz were formed at the same time as the other minerals from ground water reactions, it was probably due to the presence of soluble alkali metal salts that have since been leached from t8he mass, and not to reactions of any of the alkaline earth metal salts. Even when reaction times were ( 0 ) L. Pauling, Plot. N u t . Acad. Sci., 14, GO3 (1928).
(7) F. Laves a n d W. Kieuwendkamp, Z . Kristallogruphie, 90, 273. 270, 377 (1935).
1088
VERNONW. ARNOLDAND E. ROGERWASHBURN
extended t o thirty days under supercritical conditions, no a-quartz was formed. Acknowledgments.-The author wishes to acknowledge the help of Mr. Joseph Strauch who per-
Vol. 62
formed the chemical analysis and optical work, and of Dr. Warren 0. Groves and Mr. Ralph Ferguson of the Monsanto Chemical Company, Dayton, Ohio, for the X-ray patterns and their analysis.
TERNARY SYSTEM ISOAMYL ALCOHOL-ISOPROPYL ALCOHOLWATER AT 10, 85 AND 40' BY VERNONW. ARNOLDASD E. ROGERWASHBURN Department of Chemistru and Chemical Engineering, University of Nebraska, Lincoln, Nebraska Received April 88, 1968
Isopropyl alcohol is completely miscible with water and with isoamyl alcohol at ordinary temperatures while the latter two liquids have limited miscibility in each other. The ternary solubility curves and the distribution of isopropyl alcohol between the conjugate liquids have been determiued a t 10, 25 and 40".
Introduction The solubility curves were determined by an extension of Alexejeff's method' as used for ternary systems by Jones and Grigsby.2 This method, in common use for binary solubilities, has the advantage that repeated observations may be made on the same sample. The temperatures, at which cloudiness indicated the appearance of a second liquid phase, were measured with a thermistor. The proportions of water in the less dense, or water poor, phases were determined by titration with Karl Fischer reagent as was done for some solutions by Purnell and Bowden. Experimental
pany, fitted with a ground glass connector was incorporated in one of the necks of the cell and was used as a resistance thermometer to determine the temperature of the contents of the cell. Measurements had been made of the resistances of the thermistor a t temperatures determined with a platinum resistance thermometer the constants of which had been determined by the Bureau of Standards. An equation of the type
was found to be suitable. The constant C, 323.5, was found by a selected point method. The constants A and B were then determined by the method of least squares to be 2091.2 and -2.7168, respectively. The final equation was used in the form
-
2091*2 323.5 log R 2.7168 The thermistor had a resistance of 3577 ohms a t 10" and 1086.5 ohms a t 40'. The cell containing the heterogeneous mixture of isoamyl alcohol and water was heated in a variable temperaIsopropyl alcohol was then ture water-bath, to about 45'. added until homogeneity resulted, and the amount of alcohol added was determined by weight. The mixture was fihaken mechanically in the bath while the temperature was lowered slowly until a uniform cloudiness throughout the mixture indicated the appearance of a second liquid phase. The temperature then 'was raised slowly until the cloudiness disappeared. This was repeated until temperatures corresponding to the appearance and disappearance of the secqnd liquid phase showed satisfactory agreement. The amount of isopropyl alcohol in the mixture then was increased by a small addition. The described procedure was repeated t o determine the new temperatures corresponding to the appearance and disappearance of the second liquid phase TABLE I for the new concentrations. Five or moresuch pairs of conB.p. (cor.), OC. d'," n%= centration-temperature measurements were determined Isoamyl alcohol A 131.0 0.8051 1.4048 over the temperature range from 10 to 40" for each of fourIsoamyl alcohol B ... .8073 1.4052 teen different weight ratios of isoamyl alcohol to water. .4 curve, mean solution temperature u s . weight per cent. Isopropyl alcohol 82.2 .7809 1,3748 of isopropyl alcohol, was then plotted for each of the fourSolubility Curves.-The compositions of ternary solutions teen ratios. In determining the mean solution temperasaturated with respect to isoamyl alcohol or water or both ture, the temperature corresponding to the appearance of B of these materials were determined in the following manner. second liquid phaw was weighted twice as heavily as that Isoamyl alcohol and water, in a known ratio by weight, corresponding to its disappearance. The concentrations were placed in a specially prepared saturation cell. This of is0 ropy1 alcohol corresponding to 10, 25 and 40" were cell was made from a 50-ml. volumetric flask by the addition read g o m these curves. The concentrations of the other of B second glass stoppered neck to the bulb of the flask. two components were calculated from their weight ratios A thermistor, Type 14B from the Western Electric Com- and the concentration of isopropyl alcohol. These results are recorded in Table 11. The solubility of isoamyl alcohol in water was determined (1) M. W.Alexejeff, Bull. SOC. china.. 38, 145 (1882). by Alexejeff's method' using sealed tubes. The solubility of (2) H.E. Jones and W. E. Grigaby, Ind. EnQ. Chem., 44, 378 (1952). water in isoamyl alcohol was determined by analyzing satu(3) J. H. Purnell and S. T. Bowden. J . Chem. Soc., 539 (1954). rated solutions at each of the temperatures with Kar (4) J. Tirnmermans and E. Hennaot-Roland, Anal. soc. espan. ]Ea. Quim. 8 1 , 400 (laan). Fisc:hrr rengcnt. Materials.-The isoamyl alcohol, 3-methyl-1-butanol, used in most of this work was prepared from isobutyl bromide by a met,hod similar to that described by Timmermans and Hennaut-Roland .4 Isobutylmagnesium bromide was allowed to react with paraformaldehyde. The resulting compound was hydrolyzed and the alcohol which was formed was purified until the constants recorded in Table I for sample A were obtained. Some distribution measurements were made a t 40' with isoamyl alcohol, 3-methyl-1-butanol, obtained from the Fisher Scientific Company. This material had the constants listed for sample B in Table I. The optical rotation indicated the presence of some of the optically active isomer 2-methyl-1-butanol. The isopropyl alcohol was obtained from the Eastman Kodak Company. After intensive drying with active calcium oxide, the alcohol was distilled yielding material having the constants recorded in Table I.
p =
+
