COMMUNICATIONS TO THE EDITOR
380
Vol. 66
THE STABILITY OF SILICA
listed in most of the recent thermodynamic compilations could not be employed because it was derived Sir: using the hcat of formation of silica that is in quesChipman' recently reviewed five lines of evidence tion. that indicate that the presently acceptcd heat of The heats for rcactions V, VI, and VI1 were formation of silica (-209.9 kcal./molc)2 calciilstcd tnkcn in each case from the most recent thcrmodyfrom the heat of combustion of silicon in an oxygen namic compilations in which they appeared. bomb may bc as much as 5 kcal./mole too positive. The derived value of -215.8 kcal./mole for Still more recently, Good3 measured tJhc heat of AHf0298(c,quartz) indicates that quartz is about 6 formation of aqueous fluosilicic acid and calculated kcal. more stable I han previously thought, in good a value of -217.5 0.5 kcal./mole for AHfoZmagreement with the values given by Chipman and (c, quartz) which is 7.6 kcal./molc more negative Good. than the litcmture value. Somc recent effusion COMPANY OF AMERICA measurements that the authors have made of the ALUMINUM ALCOA RESEARCH LABORATORIES C. N. COCHRAN reactions between gallium and quartz and gallium PHYSICAL CHEMISTRY DIVISION L. M. FOSTER PENNSYLVANIA and magnesium oxide in an alumina cell4provide an NEWKENSINGTON, RECEIVED NOVEMBER 22, 1061 independent basis for calculating this value. The heats listed for reactions I and 11 in the table were obtained by third law treatment of the T H E HEAT O F FORMATION OF SILICA' effusion results. These two heats are cornbincd with other thcrmochcmical data in the table to Sir: The free energy of formation of silica is a subject drrive the heat of formation for crystalline quartz. (The heat for reaction I has been decreased by 0.3 of current interest. Chipman* has discussed evikcal. to correct for the fact that vitreous quartz was dence that the free energy of formation of silica (in employed in the effusion experiments whrreas any of its various forms) is about 5 kcal. mole-' thermodynamic functions for crystalline quartz more negative than the presently accepted value. were used in deriving the former heat for reaction Recent work in this Laboratory has shown that the I from the effusion data. Values from JAXAF thermochemical discrepancies noted by Chipman Thermochemical Tables6 were employed in making are the result of an m o r in the reported heat of formation d a h 3 The earlier heat of formation this correction.) value was based on measurements of the heat of TABLEI combustion of silicon in an oxygen bomb. The Reaction AHOm value reported here was obtained by a different I 2Ga(l) + Si02(vitrcous))r thermochemical process, which avoided uncertain4-170.8 f 0 . 7 SiO(v) + Gs/rO(v) ties that are inherent in t he earlier method. In the present experiments, the heat of formation I1 2Ga(l) + MgO(c) )r 1-159.3 It 0 . 7 of aqueous ffuosilicic acid was determined in a Mdv) Gs20(v) rotating-bomb calorimeter. Mixtures of silicon I11 SiOz(d, quartz) SiOr and vinylidene fluoride polymer were burned i n $0.546 0.10 (vitreous) oxygen in the presence of aqueous HF, the product IV I/k3i(c) + l/zSiOn(c,triof combustion being fluosilicic acid in excess Hl? $82.98 f 0.23'J dymite) SiO(v) solution. Experimental procedures werc similar to V 1/,SiOl(c, quartz) Jr 1/2SiOsthose already dc~cribcd.~A sample of high purity (c, tridymite) $0.258 f 0.19 silicon (99.96% Si, 0.040/, SiOz; 50 to 75 micron 0.V VI Mg(c) f 1/202(v) z2MgO(c) -1143.7 particle size) was obtained through the courtesy of $35.3S6 f O . Z g Dr. J. E. Kunzler, Bell Telephone Laboratories, VI1 Mg(c) nfdv) Murray Hill, N. J. The combustion samples mere VI11 = I - I1 + I11 - IV - v prepared by mixing the silicon and powdered vinyliVI + VII l/zSiOl(c, quarts) I_ dene fluoride polymer in scaled polyester bags, +107.9 f 1 . 1 1/2Si(c) 1 / ~ 0 2 ( v ) which then were rolled and pelleted. The samples Si(c) + Oz(v) SiO?(c,quartz) -215.8 f 2 . 2 burned completely. and all of the silicon was conThe heat of formation of silicon monoxide from verted to aqueous fluosilicic acid. The result for the heat of formation of fluosilicic tridymite and silicon in reaction IV was calculated from data given by Brewer and Edwardss in their acid and those for the heat of solution of silica in analysis of the effusion data of Schafer and H o r n l ~ . ~aqueous HF reported by Kings permit calculation The heat of formation of silicon monoxide that is of the heat of formation of silica. The thermochemical equations are (1) J. Chipman, J . Am. C k m . Soc., 83, 1782 (1981).
*
+
*
+
(2) J. P. Coughlin. U. 9.Bureau of Mines Bull. 542 (1954): 0.L. Humphrey and E. G . King, J . Am. C k m . Soc.. 74, 2041 (1952). (3) W. D. Good, presented st the International Calorimetry Conference, Ottawa, Ontario, Canada, August 14-18, 1861. (4) Accepted for publication in J . h7erlrochem. SOC. ( 5 ) Joint Army-Navy-Air Force Interim Thermochemical Tablea. Thermal Laboratory, Dow Chemical Co.. hiidlaad, Michigan. (8) L. Brewer and R. K. Edwards, J . Phys. Ckrn., 68, 351 (1954). (7) I€ Schtlfer and R. Htirnle, 2. anorg. u. allgem. Chem., 263, 281
(1950). (8) J. P. Coughiin, U. 8. Bureau of Mines Bull. 542 (1954). (B) Estimated by the suthora
+
+
Si(c) O&) 47HF.l72IJ~O(liq.)= HnSiF~.41HF~17411~O(liq,) AH = -250.3 f 0.3 (I)
(1) Presented at the International Calorimetry Conference, Ottawa, Ontario, Canada, August 14-18, 1981. (2) J. Chipman, J . Am. Chem. Soc.. 88, 1782 (1981). (3) J. P. Coughlin, U. S. Bur. Minee Bull. 542 (1954); G. L. IIurnphrey and E. G. Icing, J . Am. Chem. Soc., 74, 2041 (1952). (4) W.D. Good, D. R. Douslin, D. W. Scott, A. George, J. L. Lacina, J. P. Dawson and G. Waddington, J . Phyr. C k m . , 63, 1133 (1959); W. D. Good, D. W. Scott and G. Waddington, ibid., 60, 1080 (1960). (5) E. 0.King, J . Am. Chem. SOC.,73, 868 (1851).
Feb., 1962
COMMUNICATIONS TO THE EDITOR
381
+
dentally err in the same sense and amount, Chipman's observation implies either that the standard heat of formation of quartz at 25' is about 5 kcal. -+ mole-' more negative than the currently accepted value4 of -209.9 f 1.0 kcal. mole-' or that the accepted entropies of crystalline silicon or quartz a t 25" are in error by absurdly large amounts. In a + paper which apparently has gone unnoticed in this country, a Russian thermochemist, Golutvin,6 has The heats of reaction are expressed in kilocalories presented evidence that Humphrey and King4 a t 25". Equation I gives the result of the measure- erred in their analytical determination of the ments of this laboratory. Equation I1 gives the quantity of unburned silicon remaining after oxyresult of the heat of solution measurements of gen-bomb combustions, and that the true value of King,6 extrapolated to 25" and corrected for the the standard heat of formation of silica lies between change in the atomic weight of silicon from 28.06 to -215 and -219 kcal. mole-l. To resolve this 28.09. Equation I11 represents a dilution reaction problem we sought the simplest combination of studied in this Laboratory by a somewhat crude reactions that would yield an unequivocal value method already described6 The heat of eq. I11 for the heat of formation of silica. The method we was found to be very small and probably is ther- chose is based on fluorine-bomb calorimetrya; it mally insignificant, but this result should be veri- consists of combining the thermochemical equations fied by more sensitive dilution calorimetry. The Si(4 2Fdg) = SiF4g); m l (1) heat of dilution for eq. IV was computed from values in Circular 500.' Addition of eq. I, 11, I11 and and IV results in eq. V, the equation for the formaSiOz(cu,c) -I- 2Fz(d - SiFdg) 02; AHz (2) tion of quartz from the elements and the heat of formation. t o give The previously accepted value for the heat of Si(c) Oz(g) = SiOz(or,c); 1.0 kcal., about 8 formation of auartz3 is -209.9 AHf = AH1 AH2 (3) kcal. less negative than the present value. Thus, the free energy of formation of silica at 25" is For the measurement of AH1 and AHz weighed about, 8 kca.1. mole-1 more negative than the earlier amounts of crushed crystals of the high purity matevalue, a in agreement with Chipman's conclusion.2 rials, silicon or alpha quartz, were mixed with a Full details of this investligationwill be given in a small known amount of powdered silicon and placed paper in preparation. The results presented here on a nickel support dish. On exposure to four are confirmed by those in the accompanying letters, atmospheres of fluorine in a two-chamber nickel which were determined by entirely different ther- bomb and tank system (described elsewhere'), the powdered silicon ignited spontaneously and fired mochemical methods. the entire sample. Because the products of comCONTRIBUTION NO. 112 FROM THE bustion were solely the gases indicated in (1) and THERMODYNAMICS LABORATORY BARTLESVILLE PETROLEUM RESEARCH CENTERW. D. GOOD (a), the extent of the reactions was determined by BUREAU OF MINES,U. S.DEPARTMENT OB weighing the residues. AH1 and AHz, at 25", INTERIOR, BARTLESVILLE, OKLAHOMA were found to be -386.02 i 0.248 and -168.27 RECEIVED DECEMBER 2, 1961 k 0.24 kcal. mole-l, respectively. Hence, €or a-quartz AHf'298.15 = -217.75 k 0.34 kcal. mole-'. (6) W. D. Good, D. W. Scott, J. L. Lacina and J. P. MoCullough, The agreement between this determination and J . Phya. Chem., 6 8 , 1139 (1969). (7) F. D. Rowaini, D. D. Wagman, W. H. Evans, 8. bcvine and I. that of Goodg (preceding communication) is exJaffe, "Selected Values of Chemical Thermodynarnio Properties," cellent; the conclusions of Chipman3 and Golutvin6 Natl. Bur. Standards Circular 500 (1952). are substantiated. Full details of this investigation and of its extension to other forms of silica will be presented THE HEAT OF FORMATION OF SZLICA AND later. SILICON TETRAFLUORIDE'P' UNIVERSITY OF WISCONSIN S. S.WISE
HzSiF~.759HB1*3378HzO(liq.) = SiOe(c, quartz)
AH = 32.3 f 0.1 (11) 718HFm3204HpO(liq.) = HzSiFa.759HF.3378H20(liq.) AH = 0.0 t 0.3 (111) ?65HP~3376Ha0(liq.)= 71SHF~3204H~O(Sq.) AH = 0.5 k 0.2 (IV) 47HF.l72€IzO(liq.) Si(c) Oz( g) = SiOn(c, quartz) AH = -217.5 i 0.5 (V) 765HF-3376H*O(liq.) HgSiFn.41HFs 174H20(1;s.)
+
+
+
+
Sir:
Chipmana has presented evidence, based on a number of observed equilibria, that the standard free energy of formation of silica is about 5 kcal. molee1 more negative than the currently accepted value. Unless all the observed equilibria coinci(1) This work wa8 performed under the auspices of the U. S. Atomic Energy Commission. (2) Abstracted from a thesis submitted by S. S. Wise to the faculty of the University of Wisconsin in partial fulfillment of the requirements for the Ph.D. degree. (3) J. Chipman, J . A m . Chena. Soe., 88, 1762 (1981). (4) J. P. Coughlin, U. S. Bur. Mines Bull. 642 (1954),based on the work of C. L. Humphrey and E. G. King, J. Am. Chern. SOC.,74,2041 (1952).
-
J. L. MARQRAVE MADISON, WISCONSIN ARGONNE NATIONAL LABORATORY H. M. FEDER ARCONNE, ILLINOIS W. N. HUBBARD RECEIVED DECEMBER 2, 1961
(5) Yu. M.Golutvin, Zhur. Fka. R h i n . , 80, 2251 (1958). (6) Previous papers dealing with thia subject are (a) E. Greenberg, J. L. Settle, H. M. Feder and W. N. Hubbard, J . Phys. Chena., 66, 1168 (1961): (b) J. L. Settle, IH. M. Feder and W. N. Hubbard, ibid., 66, 1337 (1961); (c) S. 8. Wise, J. L. Margrave, H. M. Feder and W. N. Hubbard, ibid., 66, 2157 (1961). (7) R. L. Nuttall, S. S. Wise and W. N. Hubbard, Reu. Sci. Inatr., in press. (8) This value for the heat of formation of SiFd(g) differs considerably from literature values; the latter, however, either are baaed on AH/ (SiOS or depend on unreliable experimental methods. (9) W. D. Good, J , Phya. Cbem., 66, 380 (1962).
