862
WALTER [CONTRIBUTION FROM THE
c
SCHUMB AND
DEPARTMENT OF
WALTER J. BERNARD
VOl. 77
CHEMISTRY, MASSACHUSETTS INSTITUTE OF TECHNOLOGY]
The Thiochlorides of Silicon1 BY WALTERC. SCHUMB AND WALTER J. BERNARD~ RECEIVED SEPTEMBER 7, 1954 An investigation has been made to determine the existence of a homologous series of chlorosilthianes, as an analog of the homologous series of chlorosiloxanes. It was found that the thermal decomposition of trichlorosilanthiol results in a small yield of hexachlorodisifthiane, the first member of the expected series, and octachlorocyclotetrasilthiane. No evidence was found for higher members of t,he series. Hexachlorodisilthiane was also synthesized by the reaction between silver sulfide and iodotrichlorosilane a t 250 .
Within recent years it has been shown that the alysis of the gaseous products of decomposition partial hydrolysis of silicon tetrachloride3 leads to a showed that both H2S and HCl were formed. The homologous series of oxychlorides of the general source of HCI was presumed to be the intramolecuformula S i n + ~ 0 , C k + 4 and in a like manner lar decomposition of CI3SiSH, giving rise to (Sihexachl~rodisilane~ yields a series represented by sc12)4which was also observed as a product of the the general formula S ~ Z ~ + Z O ~ C Schumb ~ ~ + ~ reaction. . and Towle have also prepared the first member of a The small yields of both Si2SCl6 and (SiSC12)4, series of chlorosilazanes by the ammonolysis of and the fact that their boiling points lie close tosilicon tetrachloride.6 From these considerations gether (approximately 75 and 80°, respectively, a t it was believed that a series of chlorosilthianes 17 mm.) prevented complete separation by fracwould also be capable of formation by the reaction tional distillation. Positive identification of Si2between hydrogen sulfide and silicon tetrachloride. SCla as a product of the reaction was obtained by The work of previous investigators indicated that comparison of its Raman spectrum with the specthe only volatile products of this reaction were the tra of pure Si2SC16and a 40% solution of (SiSCIz), in colorless liquid, C13SiSH,6 and the white crystalline benzene. The two strongest lines of (SiSC12)4were solid, SiSC12.' observed in the spectrum of the mixture and all the In the present work i t was found that HzS and rest were attributed to SizSCl6, as shown in Table I. S i c 4 give no evidence of reaction below 500' but TABLE I above that temperature a slow reaction takes place. THERAMAN SPECTRA OF SizSCla AXD (SiSCl.), The initial product is trichlorosilanthiol, C4SiSH, ClaSiSH decomp. (SiSClr)r in and further reaction leads to a small yield of SiSC12. SinSCls product benzene The latter compound, reported by Qtienne to be A6 AS AS (cm.-I) Intensity (cm. -1) Intensity (cm.-l) Intensity dimeric,8 in our measurements was found to be tet118 5 116 5 rameric. Calcd. for Si4S4C&: mol. wt., 524. 130 5 134 5 135 2 Found: mol. wt., 536, 497. The tetrameric formula 191 5 193 4 175 2 indicated a further analogy with the chemistry of 228 4 222 4 215 6 silicon with oxygen, corresponding to the well 266 2 262 1.5 known cyclic tetramers, (Si0Cl2)4 and (Si0Br~)d. 277 0 280 6 The systematic name for this substance is octa397 10 396 10 chlorocyclotetrasilthiane. 43.5 1 5 433 10 The major products of reaction are silicon disul546 0 543 0 fide and a polymerized solid of uncertain composi592 21%' 582 2w tion. The higher the temperature a t which the 63 1 1 630 1 reaction is carried out, the smaller is the yield of Cl,SiSH, due to the increased formation of nonThere was no evidence of any higher members of volatile products, the optimum temperature for reaction being 650'. Members of the homologous the expected series of silthianes in the products of series of chlorosilthianes were not observed by decomposition of C13SiSH. Si2SC16and (SiSCl2)r the continuous reaction between H2S and SiC14; were actually minor constituents, the major prodbut the first member of the series, hexachlorodisil- ucts being Sic14 and a non-volatile solid correspondthiane, Si2SCle,was identified in the thermal decom- ing closely to the composition SiZS3Cl2. The action of C12 on silicon disulfide was also position products of ClSiSH in the absence of H2S. This previously unreported compound would be studied as a possible source of chlorosilthianes. The expected to result from the condensation of two major products a t 700' were Sic14 and (SiSC12)4, as molecules of C13SiSH with elimination of H S . An- well as free sulfur. The yield of (SiSC12)4 was about lo?&, Fractional distillation of the product showed (1) Based on part of a thesis presented by W. J. Bernard to the Dethat a small amount of Si2SCls was also formed. partment of Chemistry in partial fulfillment of the requirements for This is presumably the liquid which Besson reported the degree of Doctor of Philosophy. (2) National Science Foundation Fellow, 1953-1954. as a product of the reaction, but did not identify.g (3) W. C. Schumb and A . J. Stevens, THIS JOUKNAL, 72, 3178 It was found that below 700' the reaction resulted (19.50). almost completely in SiCl?, and again above 750'. (4) W. C. Schumb and R. A. Lefever, ibid., 76, 2091 (1954). The failure t o obtain increased yields of chlorosil(5) W. C. Schumb and L. H. Towle, ibid., 75, 6085 (1933). (0) C. Friedel and A. Ladenburg, Ann., 145, 179 (1868); A x % .chim. thianes a t higher temperatures was found to be due phys., 27, 416 (1872). to the instability of such compounds. A run a t 800' (7) M . Blix and W. Wirbelauer, Bcr., 36, 4220 (1903). 18) Y.ktienne. Bull. soc. chim. France, 791 (1953).
(9) A . Resson, C o m p l . rend., 113, 1040 (1891).
Feb. 20, 1955
THETHIOCHLORIDES OF SILICON
863
Eabornlo was able t o prepare hexamethyl- and hexaethyldisilthiane in good yield by refluxing trimethyl- and triethyliodosilane with ApS, but in this case simple refluxing of Sic131 with AgzS failed t o bring about reaction and it was necessary to use a higher temperature. The reaction was carried out in a bomb constructed of heavy-walled 25 mm. 0.d. Pyrex tubing. The bomb was charged with 42 g. of Sic131 and the theoretical amount of AgzS needed for c o a plete reaction, sealed, and heated for 17 hours a t 230-250 The liquid product of reaction was decanted and the yellow solid in the bomb was extracted three times with benzene, the washings being added to the liquid product. The benzene was removed by distillation a t atmospheric pressure and the remaining liquid was distilled a t 50 mm. ShSCls boiled a t 100-100.5', 5 ml. being collected. A slight disExperimental coloration of the product, due to free Iz, was eliminated by Procedure.-In the thiohydrolysis experiments Sic& was shaking with Hg and redistilling under vacuum a t room temrecycled by use of an all-glass apparatus constructed of Pyrex perature. The resulting clear, colorless liquid ;hen cooled and Vycor tubing. This piece of equipment was similar to -78' formed a white solid melting a t -45 The ret o that described by Blk7 except that standard-taper ground covered yield was 7 g., or 30%, based on the amount of Sijoints were used t o connect the flask and the water conden- ClaI used. Anal. Calcd. for Si2SCle: Si, 18.64; S, 10.66; ser to the system, and the return line was joined by a side- C1, 70.70; mol. wt., 301. Found: Si, 19.22, 18.76; S, arm ring seal to the entrance line. The section of the ap- 10.82, 10.81; C1, 69.78, 70.00; mol. wt., 289, 306. paratus made of Vycor was used as the reaction tube. The Sic181 was prepared by the high temperature reaction flask containing S i c 4 was equipped with a thermometer between H I and SiC1,.l1 Hydrogen was passed over the well and connected to the cyclic apparatus by means of a surface of boiling SiCL and then over iodine heated just be24/40 standard-taper joint. The flask contents were elec- low its melting point. The mixture of gases was then trically heated and kept constantly boiling during the course passed through a Vycor tube heated to 700" and the prodof a run. Hydrogen sulfide, which was bubbled into the ucts were collected in a flask cooled to -78". Free 11 was flask, was previously dried by passing through Drierite and eliminated from the products by distilling over mercury. PnOs. The volatility of Sic4 (vapor pressure = 76 mm. The fraction boiling between 110-117' was collected in glass a t 0') did not permit satisfactory condensation by the water receivers which were immediately sealed. condenser and to prevent undue loss a trap cooled to -78' Si& was prepared by the metathetical reaction between was used beyond the condenser. HpS was also condensed aluminum sulfide and silica,lZand the monosulfide was preto a certain extent by this trap and before returning Sic& pared by the reaction between Si& and Si.18 Chlorination to the flask it was necessary t o allow the trap to warm to of these compounds was carried out using helium as a diluent. room temperature. A P2Os drying tube was connected to A diluting gas was found to be particularly necessary in the the exit of the trap t o prevent diffusion of atmospheric mois- case of the chlorination of SiS, since that compound reacts ture into the system. Before each run the entire apparatus with great ease with oxidizing agents. was thoroughly flushed with an inert gas. Analytical.-All samples for analysis and molecular C1,SiSH was prepared in desired quantities by use of this weights were prepared in a dry, inert atmosphere. Liquid equipment. In a typical run 65 g. of Sic14 yielded,.upon samples were pipetted into weighed thin-walled glass bulbs, distillation, 26 g. of ClsSiSH in 14.5 hours, corresponding t o which were immediately sealed with an oxygen torch after a yield of 40%. HzS was passed in a t the rate of 4 bubbles being removed from the dry-box. per second and the temperature of the reaction tube was Sulfur was determined by decomposing the sample under 650'. As well as being easily affected by moisture, ClsSiSH an excess of standard acidified 1 2 solution, the excess Ip being is extremely sensitive to oxygen, which causes sulfur to pre- titrated with standard sodium thiosulfate solution. Samples cipitate. For this reason vessels and other apparatus were for chlorine and silicon determinations were decomposed thoroughly flushed with carbon dioxide before introducing under 200 ml. of distilled water so that the resulting acid the liquid. ClrSiSH may be distilled a t atmospheric pres- solution was about 0.01N H S , which would interfere sure; its boiling point is 95.0' a t 768 mm. Anal. Calcd. with the subsequent titration, was removed by boiling the for ClsSiSH: Si, 16.77; S, 19.14; C1, 63.49. Found: Si, solution cautiously for a few minutes. This procedure does 16.75, 16.25; S,18.38, 19.13; C1, 63.80, 63.78. not result in the loss of any HCl.14 Silicon was determined The thermal decomposition of C1,SiSH was carried out by in the resulting solution as silica by the usual procedure. passing the vapors of the thiol in a stream of helium through Molecular weights were determined by the Beckmann a tube heated to 650" and condensing the products a t -78". freezing-point method in benzene. Repeated fractionation of the products resulted in a colorless Acknowledgment.-The authors wish to acliquid boiling at 98' a t 47 mm. Anal. Calcd. for Si2SCle: Si, 18.64; S, 10.66; C1, 70.70; mol. wt., 301. Calcd. for knowledge the help of Professor Richard C. Lord ( S i s C l ~ ) ~Si, : 21.43; S, 24.47; C1, 54.10; mol. wt., 524. and hlr. George 'Cl'ilmot in connection with the Found: Si, 19.68, 19.97, 20.15; S, 11.48, 11.29; C1, 68.89, Rarnan spectra. 68.90; mol. wt., 313, 327, 334. On the basis of the analysis for C1, it was calculated that the liquid contained 90% CAMBRIDGE, MASS. Si2SCle. (SiSC12)d was not obtained as a pure substance from this distillation but was wet with SinSCle. It could be (IO) C. Eaborn, J . Chcn. Soc., 3077 (1950). (11) P.Hautefeuille, Bull. SOC. chim., (21 '7, 198 (1867). separated from the liquid by filtration and pressing dry on (12) E.Tiede and M. Thimann, Bcr., 69, 1703 (1926). the filter paper. Further purification could be achieved by (13) W. C. Schumb and W. J. Bernard, THIS JOURNAL, 1 7 , 904 recrystallizing from benzene in a stream of dry nitrogen. Anal. Found: Si, 20.60, 21.85; S, 23.54, 23.67, 23.87; (1955). (14) I. M. Kolthoff and E. B. Sandell, "Textbook of Quantitative C1, 53.93, 54.10. Inorganic Analysis," The Macmillan Co., New York, N . Y.,1947, p. The preparation of Si2SCl, was carried out by means of the reaction between iodotrichlorosilane and silver sulfide. 645.
resulted in the deposition of Si& in a zone of the reaction tube beyond the point at which the disulfide had been placed. Since SiSz has a negligible vapor pressure a t this temperature, the formation of the disulfide must have been due to the decomposition of (SiSCl& and SizSCb, the other product being SiCh,. The action of chlorine on silicon monosulfide was studied a t lower temperatures. At 170' the products of reaction were identified as SC12, SZC12, Sic&and SizCle.
.
.
