'l .1. C rt a. A.“ P» «am ,0 R 'f \Jl.‘. .I II. {1! ABSTRACT PREPARATION OF SILYLP AND SILOXANYL CHROMATES by Paul M. Dupree The purpose of this investigation was the preparation of organosilicon-chromium compounds with thermal stabilities greater than known compounds of this type. The heterogeneous reactions, in inert organic solvents, of either silver chro- mate or dichromate with the appropriate organochlorosilane yielded bis(triphenylsily1)chromate(VI), bis(diphenylfluoro— sily1)chromate(VI), cyclo-bis and tris[d1pheny1siloxanyl- chromate(VI)] and cyclo—bis and tris[3,3,3-trifluoropropyl- methylsiloxanylchromate(VI)]. Cyclo-bis[di-3,3,3-trifluoro- propyldimethyldisiloxanylchromate(VI)] was prepared by the reaction of chromyl chloride with 3,3,3-trifluoropropy1methy1— silanediol in methylene chloride. With the exception of the light orange crystalline cyclo—tris[diphenylsiloxanylchromate- (VI)], the compounds were deep orange viscous liquids at room temperature and only cyclo-bis[di-3,3,3-trifluoropropy1di- methyldisiloxanylchromate(VI)] could be made to crystallize at lower temperatures. The assignment of cyclic structures to the siloxanyl— chromates resulted from consideration of their infrared, visi- ble, ultraviolet, and proton magnetic resonance spectra as well as differential thermal analyses. The assignments should Paul M. Dupree only be considered tentative for although each measurement indicated the existence of a cyclic structure, no completely conclusive data were obtained. The compounds prepared during this investigation were photolytically unstable. The phenylsiloxychromates were rap- idly decomposed in water while the siloxanylchromates contain- ing trifluoroprOpyl linkages appeared to be very hydrolyti- cally stable. The thermal stabilities of the silyl and siloxanyl— chromates were studied by means of differential thermal ana- lysis. Highly electronegative substituents such as fluorine or polyfluoroorganic groups on silicon gave substantial in- creases in the thermal stabilities of the silyl and siloxanyl- chromates. Cyclo—bis and tris[diphenylsiloxanylchromate(VI)] and. cyclo—bis and tris[3,3,3-trifluor0propylmethylsiloxanylchromate- (Vl)]are believed to be the first cyclic compounds with alter- nating oxygen-metal-oxygen-silicon linkages. Bis(diphenyl— fluoro)chromate(VI) is the only known transition metal-silicon compound containing a silicon-fluorine linkage and the silox- anylchromates with 3,3,3-trifluor0propy1 groups are also the first known siloxychromates containing fluorocarbon substi- tuents. The infrared absorptions of the siloxychromates at 11-12 H and 20.6-21.4 u were tentatively assigned to the asymmetric Paul M. Dupree and symmetric stretching vibrations of the chromium-oxygen- silicon linkage. Crystallographic interatomic Spacings for cyclo—tris[diphenylsiloxanylchromate(VI)] were calculated from x-ray diffraction data. PREPARATION OF SILYL- AND SILOXANYL CHROMATES b Y A\ \V C” .W Paul M?rDupree A THESIS Submitted to Michigan State University In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1965 ACKNOWLEDGEMENTS The author wishes to extend his sincere appreciation to Dr. Robert N. Hammer for his assistance and counsel during this investigation and to the United States Air Force for financial aid. The author is also deeply grateful for the understanding and encouragement shown by his wife, Joan, throughout this investigation. *#************** ii TABLE OF CONTENTS INTRODUCTION. . . . . . . . . . . . . . . . . . . . HISTORY . . . . . . . . . . . . . . . Chemistry of Chromium(VI). . . . Siloxychromates . . . Inorganic Chromium(VI) Chemistry. Chromium(VI) Organic Compounds. Chemistry of Fluorine. . . . . . . . . . Physical Properties . . . . . . . . Thermal Stability . Hydrolytic Stability. Inorganic Pelymer Chemistry. EXPERIMENTAL. . . . . . . . Preparation of Reactants . . . . . . . Preparation of Silver Chromate. . . . Preparation of Silver Dichromate. . . . Characterization of Triphenylchlorosilane Preparation of Diphenyldifluorosilane . . . Characterization of Diphenyldichlorosilane. Preparation of Diphenylchlorofluorosilane Page 30 . 30 . 30 . 31 . 32 32 . 34 . 34 Characterization of 3, 3,3-trif1uoropropylmethyl- dichlorosilane . . . Preparation of 3, 3, 3-trifonropropylmethyl- silanediol . . . . . . . . . Preparation of Silydchromates. . . . Reactions of Triphenylchlorosilane with Silver Chromate with Silver Dichromate . with Potassium Chromate. . with Potassium Dichromate. . . with Sodium Dichromate . . . . . Reactions of Diphenylchlorofluorosilane with Silver Chromate . with Silver Dichromate . . Reactions of Diphenyldifluorosilane with Silver Chromate . . . . . . . . with Silver Dichromate . . . . . . Reactions of Diphenyldichlorosilane with Silver Chromate . . . with Silver Dichromate . . Reactions of 3, 3, 3- trifluoropropylmethyldi— chlorosilane with Silver Chromate . . . . . . . . with Silver Dichromate iii 38 . 38 4O . 40 . 42 43 43 43 43 46 . 46 . 47 47 49 . 51 52 TABLE OF CONTENTS - Continued Page Reaction of 3,3,3-trif1uoropropy1methy1- silanediol with Chromyl chloride , , , , 53 Characterization of Reaction Products. . . . . . . . 55 Reactions of Triphenylchlorosilane with Silver Chromate . . . . . . . . . . 55 with Silver Dichromate . . . . 56 Reactions of Dipheny1chlorofluorosilane with Silver Chromate and Silver Dichromate. . 58 Reactions of Diphenyldichlorosilane with Silver Chromate . . . . . . . . . . 64 with Silver Dichromate . . . . . . 70 Reactions of 3, 3, 3- trifluoropropylmethyl- dichlorosilane with Silver Chromate . . . . . . . . . . 76 with Silver Dichromate . . . 85 Reactions of 3, 3 ,3-trifluoropropylmethylsilane- diol with Chromyl Chloride . . . . . . . 86 Instrumental Measurements. . . . . . . . 95 Ultraviolet and Visible spectra. . . . . 95 Infrared Spectra . . . . . . . . 96 Proton Magnetic Resonance Spectra. . . . 97 Xeray Diffraction Data . . . . . . . . 97 Vapor Phase Chromatography . . . . . . .100 Analytical Methods . . . . . . . . . . . . . .100 Chromium Analysis. . . . . . . . . . . .100 Silicon Analysis . . . . . . . . . . . .101 Fluorine Analysis. . . . . . . . . . . .102 Molecular Weight . . . . . . . . . . . .102 DISCUSSION. . . . . . . . . . . . . . . . . . . . . . . .139 Hydrolytic Stability . . . . . . . . . . . . . . . .139 Thermal Stability. . . . . . . . . . . . . . . . . .140 Reaction Mechanisms. . . . . . . . . . . . . . . . .148 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . .167 RECOMMENDATIONS FOR FUTURE WORK . . . . . . . . . . . . .170 .APPENDIX. . . . . . . . . . . .. . . . . . . . . . . . . .173 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . .180 iv TABLE I. II. III. Iv. VI. VII. "VIII. IX. XI. XII. XIII. XIV. XVII. XVIII. XIX. LIST OF TABLES Energies of Related Chemical Bonds. . . Bond Lengths and Bond Energy Terms in Halo- methanes. . . . . . . . . . . . . . . . Elemental Analysis of Diphenyldifluorosilane. Elemental Analysis of Diphenylchlorofluorosilane. Elemental Analyses of Phenyl—Siloxychromates. Infrared Spectra of Phenyl-Siloxychromates Infrared Spectra of Phenyl-siloxychromates (15 - 29p) o o o o o o o o o o o o o o o o The Chromium—Oxygen—Silicon Infrared Absorption Frequencies of Several Organosiloxy Chromates . . Interatomic Spacings. for Cyclo-bis[diphenylsilox- anylchromate(VI)]. . . . . . . Elemental Analyses of 3, 3,3-trif1uoropropy1— methylsiloxanylchromates. . . . . . . Infrared Spectra of 3, 3, 3-trif1uoropropy1methy1- siloxanylchromates and Related Compounds. . Proton Resonance Spectra. Infrared Spectra of Phenyl—Halosilane Reactants (15 - 29“). ..... . . . . . Interatomic Spacings of Silver Chromate Interatomic Spacings of Silver Dichromate . Physical and Chemical Enthalpic Reactions . Differential Thermal Analyses of Organosilyl- and Siloxanylchromates. . . . . . . . . . . Organodichlorosilane Reactions with Silver Chromate and Silver Dichromate. Elemental Analyses and Molecular Weight of Tris(triphenylsilyl)vanadate(V) . . . . . . Interatomic Spacings of Tris(triphenylsily1)- vanadate(V) . .7. .. . ... . . . . . . V Page 16 18 34 37 57 61-2 63 66 74 77 81-4 90 99 .137 .138 .143 .149 .160 .176 .177 FIGURE 10. 11. 12. 13. 14. LIST OF FIGURES Page Proton Magnetic Resonance Spectra of Cyclo- tris[diphenylsiloxanylchromate(VI)] . . . . . . . 73 Proton Magnetic Resonance Spectra of 3,3,3-tri- fluoropropylmethyldichlorosilane . . . . . . . 91 Proton Magnetic Resonance Spectra of 3,3,3,—tri—' fluoropropylmethylsilanediol . . . . . . . . . . 92 Proton Magnetic Resonance Spectra of Cyclo-bis- [di—3, 3 ,3-trif1u6ropropyldimethyld1‘3110xany1— chromate(VI)] . . . . . . . . . . . . 93 Ultraviolet and Visible Spectra of Silver Chromate and Silver Dichromate . . . . . . . . . . . . . . 104 Ultraviolet and Visible Spectra of Bis(diphenyl- f1uorosily1)chromate(VI). . . . . . . . . . . . 105 Ultraviolet and Visible Spectra of Cyclo-bis and tris[diphenylsiloxanylchromate(VI)] . . . . . . . 106 Ultraviolet and Visible Spectra of Cyclo-bis and tris[:3, 3, 3- -triflucropropylmethylsi10xanylchromate- (VI)]. . . . . . . . . . . . . . . . 107 Ultraviolet and Visible Spectra of Cyclo-bis[di— 3, 3 ,3-trifluoropropy1dimethyldisilbxanyl-chromate- (VI)]. . . . . . . . . . . . . . 108 Infrared Spectrum (2 —15u) of Triphenylcthro- silane . . . . . . . . 109 Infrared Spectrum (15 4 29p) of Triphehylcthro- silane . . . . . . . . . . . . . . . . . . . . . 110 Infrared Spectrum (2 4 15p) of Diphenyldichloro— silane' . . . . . . . . . . . . . . . . . . . . . 111 Infrared Spectrum (15 - 29g) of Diphenyldichloro— silane . . . . . . . . . . . . . . . . . . . . . 112 Infrared Spectrum (2 - 15“) of Diphenyldifluoro- silane . . . . . . . . . . . . . . . . . . . . . 113 vi LIST OF FIGURES - Continued FIGURE 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Infrared Spectrum (15 - 29p) of Diphenyldi-' fluorosilane . . . . . . Infrared Spectrum (2 - 15p) of Diphenylchloro- fluorosilane . . . . . . . . . . . . . Infrared Spectrum (15 - 29p) of Diphenylchloro- fluorosilane . . . . . . . . . . . . . . . Infrared Spectrum (15' - 29p) of Bis(tripheny1- silyl)chromate(VI) . . . . . . . Infrared Spectrum (15 - 29p) of Cyclo+bis[tetra- phenyldisiloxanylchromate(VI)] . . . . . . . . . Infrared Spectrum (2 - 15p) of Bis(diphenyl- f1uorosily1)chromate(VI)] . . . . . Infrared Spectrum (15 - 29p) of Bis(diphenyl-' f1uorosily1)chromate(VI)] . . . . . . . . . . Infrared Spectrum (2 - 15p) of Cyclo-bis[di- phenylsiloxanylchromate(VI)] from Silver Chromate Infrared Spectrum (15 - 29p) of Cyclo-bis[di- phenylsiloxanylchromate(VI)] from Silver Chromate Infrared Spectrum (2 - 15p) of Cy¢lo-tris[di-' phenylsiloxanylchromate(VI)] . . . . . . . . . . Infrared Spectrum (15 - 29M} of Cycloétris[di- phenylsiloxanylchromate(VI) Infrared Spectrum (2 — 15") of Cyclo—Bis[di- phenylsiloxanylchromate(VI)] frOm'SilVer Di- Chromate 0 O I O O O O O O O O O O O O O O O O 0 Infrared Spectrum (15 - 29p) of Cyclo-bis[di- phenylsiloxanylchromate(VI)] from Silver Di- Chromte C C C O C O C O O C C O C C O O C C C . Infrared Spectrum (2 - 15p) of 3,3,3-trifluoro— propylmethyldichlorosilane . . . . . . . . . . . Infrared Spectrum (15 - 29p) of 3,3,3-trifluoro- propylmethyldichlorosilane . . . . . . . . . . . vii Page 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 LIST OF FIGURES - Continued FIGURE 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. Infrared Spectrum (2 - 15p) of 3,3,3-trifluoro— propylmethylsilanediol . . . . . . . . . . . . Infrared Spectrum (15 - 29p) of 3,3,3-trifluoro- propylmethylsilanediol . . . . . . . . . . . . Infrared Spectrum (2 - 15p) of Cyclo-bis[3 3,3- trifluoropropylmethylsiloxanylchromate(VI) . . . . Infrared Spectrum (15 - 29p) of Cyclo-bis[3,3,3- trifluoropropylmethylsiloxanylchromate(VI)]. . . Infrared Spectrum (2 - 15u) of Cyclo—tris[3,3,3— trifluoropropy1methylsiloxanylchromate(VI)]. . . Infrared Spectrum (15 - 29p) of Cyclo-trisfi3,3,3- trifluoropropylmethylsiloxanylchromate(VI) . . . Infrared Spectrum (2 -15u) of Cyclo-bis[di-3,3 3- trifluoropropy1dimethy1disiloxanylchromate(VI)j. . Infrared Spectrum (15 - 29p) of Cyclo—bis[di-3,3 - 3-trif1uoropropy1dimethy1disiloxanylchromate(VI)j. Differential Thermal Analyses of Bis(tripheny1- silyl)chromate(VI) and Bis(diphenylf1uorosily1)- Chromate(VI) o o o o o o o o o o o o o o o o o o 0 Differential Thermal Analysis of Cyclo—bis[tetra— phenyldisiloxanylchromate(VI)] . . . . . . . . . Differential Thermal Analyses of C clo-bis and tris[diphenylsiloxanylchromate(VI) . . . . . . . Differential Thermal Analyses of Cyclo—bis and tris[3,3 3-trifluoropropylmethylsiloxanylchro- mate(VI)j. . . . . Infrared Spectrum (2 — 15u) of Tris(tripheny1— 811y1)vanadate(V). . . . . . . . . . . . . . . . . Differential Thermal Analysis of Tris(triphenyl- sily1)vanadate(V). . . . . . . . . . . . . . . . viii Page 129 130 131 132 133 134 135 136 150 151 152 153 178 179 INTRODUCTION During the past fifteen years, there has been a con— siderable intensification in the study of polymerization chemistry involving inorganic atoms. This has been brought about by the need for synthetic materials which can with- stand elevated temperatures for long periods of time; this property is not generally found in organic polymers. Limited success has been obtained with polysiloxanes which usually have better stabilities than organic polymers. However, these materials are unable to withstand tempera- tures of 500° C. or better. It has been postulated that the incorporation of metal atoms into the polysiloxane linkages would increase their resistivity toward thermal decomposition. This concept is strengthened by the apparent inertness of metal silicates found in nature. When this research group was initiated in 1958, chromium was chosen as one of the metals to be studied for the following reasons: (a) It is a strong oxidizing agent and, therefore, is resistant to oxidation. (b) Compounds of chromium(VI) have a tendency to poly— merize as in chromium(VI) oxide, polyacids, poly- chromates and chromyl chloride (1, 2, 3, 4). (c) The chromyl group is difunctional. (d) Chromium(VI) displays tetrahedral coordination and has very nearly the same radius as silicon in silicates. The geometry should therefore be favorable for substi— tution. Initial studies of the chromium-oxygen-silicon system were performed in this laboratory by Rare (5) who prepared several organosilylchromates and observed their relative thermal stabilities. Electron-releasing groups attached to silicon atoms in silylchromates appear to lower stabilities. Thus bis(trimethylsilyl)chromate(VI) explodes when heated, and bis(diphenylmethylsi1y1)chromate(VI) as well as bis(diphenyljp-tolylsi1yl)chromate(VI) decompose when warmed. Bis(triphenylsi1y1)chromate(VI), however, shows considerably more thermal stability. Similar observations have been made with silylvanadates and siloxy-titanium compounds. One of the goals of this investigation was to prepare and study some silylchromates containing highly electron with— drawing substituents. Fluorine was chosen as a silicon sub- stituent because of its high electronegativity and the apparent thermal stability of silicon-fluorine bonds, as indicated both by the energy of homolytic cleavage (143 kca1./mole.) (141) and ionic bond energies (6,’7L, Specifically, the preparation of some silylchromates containing a single Si-F bond was under- taken. Hare reported the preparation of a silylchromate telomer of the alternating A-B-C variety (5). An initial study indi- cated that this alternating form is not as stable thermally as substances containing a higher silicon to chromium ratio. A second purpose of this investigation was to prepare other A-B-C telomers for further study of this hypothesis. HISTORY Chemistry of Chromium(VI). The over-all purpose of this investigation was to prepare inorganic polymers which might be thermally stable at tempera- tures above 500° C. The high thermal stability of naturally occurring metal silicates suggested that linear metal-oxygen— silicon chains might be similarly stable yet not show the rigidity of a network structure. Hexavalent chromium was a natural choice for the metal component. Compounds of chromium(VI) are strong oxidizing agents; consequently, they are resistant towards oxidation and, like most transition metals, chromium shows a definite tendency towards auto—poly- merization. The bifunctionality of the tetrahedral chromyl group would assure formation of linear molecules. Siloxychromates Schmidt and Schmidbaur (8), in 1958, were first to report the preparation of an organosilylchromate. Bis(trimethysi1yl)— chromate(VI) was synthesized by refluxing hexamethyldisiloxane containing suspended chromium(VI) oxide. Repetition of the experiment by Hare (5) and Abel (9) resulted in explosions. The problem of instability was resolved for Abel by combining the reactants at -40°(L Hare isolated the red oil from heter- ogenous reactions of chromium(VI) oxide with both hexamethyl- disiloxane and trimethylsilanol in methylene chloride. Bis(p-tolyldiphenylsilyl)chromate(VI), bis(cyclohexy1— diphenylsilyl)chromate(VI), bis(tricyclohexylsilyl)chromate- (VI) and bis(diphenylmethylsilyl)chromate(VI) were prepared and isolated by Hare. Each of the above compounds was syn- thesized from the respective silanol and chromium(VI) oxide in refluxing methylene chloride. Bis(p—tolyldiphenylsilyl) chromate(VI) was a yellow crystalline solid which melted at 93-5-990 C., bis(cyclohexyldiphenylsilyl)chromate(VI) and bis(tricyclohexylsilyl)chromate(VI) were red—orange solids with respective melting points of 84-850 c, and 125—1260 c, Bis(methyldiphenylsilyl)chromate(VI) was an unstable red liquid. Bis(triphenylsily1)chromate(VI) (m. 153-4° C.) was prepared from the above reaction by Granchelli and Walker (10) and by Hare. Hare also prepared this substance from chromyl chloride and triphenylsilanol. Hampton (11) studied the kinetics of the reaction between chromyl chloride and triphenylsilanol by following the disappearance of the sili- axrhydroxyl stretching frequency at 2.7 p. The reactants formed an equilibrium after three minutes and the equilibrium could be shifted toward the silylchromate product by addition of calcium oxide. The kinetic data indicated that the re- action was one-half order with respect to the concentration of both triphenylsilanol and chromyl chloride. The frac- tional order was attributed to the possible dimerization of each reactant before intercombination occurred. Hampton predicted and isolated the intermediate, triphenylsiloxo- chlorochromate(VI), by treating chromyl chloride with an ex- cess of triphenylsilanol. This violet crystalline solid melted at 100° C. When Hare treated diphenylsilanediol with either chromyl chloride or chromium(VI) oxide he obtained mixtures of two products, a light orange crystalline solid and a dark orange amorphous gum. His primary characterization indicated that the respective products were cyclo-bis[tetraphenyldisilox- anylchromate(VI)] and bis(hexaphenylr6—hydroxotrisiloxanyl)- chromate(VI). The reaction of tetraphenyldisiloxane-l,3-diol and chromium(VI)oxide yielded only the crystalline dimer. Hampton, subsequently, confirmed the structure of the cyclic dimer with proton magnetic resonance spectroscopy. Hare isolated a bright red viscous oil from the methylene chloride reaction of dicyclohexylsilanediol and chromium(VI) oxide. Characterization suggested that the product was hydroxo- {penta [dicyclohexylsiloxanylchromate(VI)]} dicyclohexyl- silanol. ’ Hare tentatively assigned the strong vibrational frequency occurring between 11.5 and 12.1 p to the Si-O—Cr stretch by qualitative evaluation of the infrared Spectra of several siloxy chromates and by comparing the frequencies and appear- ances of absorption bands shown by other transition metal- oxygen-silicon compounds (12, 13). Ovrutskii (14) obtained a complex containing the silicon- oxygen-chromium linkage from the reaction of silicic acid and hexa'aquo’chromium (III) nitrate. Several organosilicon—chromium coordination compounds having the following basic structure Cl O—Cr/ / / C1 \ R3SiR'C OH \ Cl / \ O —— Cr \ Cl have been prepared by Gilkey (15) from reactions of organic acids, R3SiR'COOH, with dichlorohydroxochromium(III). Inorganic Chrdmium(VI) Chemistry. The inorganic compounds of chromium(VI) includes chromium(VI) oxide, metal chromates, polychromates, peroxo- chromates, and chromyl compounds. Coordination compounds con- taining chromium(VI) are rare. ‘Chromium(VI) oxide is obtained as a precipitate by the action of sulfuric acid on concentrated solutions of either sodium or potassium dichromate. The bright red hygroscopic crystals melt at 1970 C. (16). The crystalline structure of chromium(VI) oxide was generally considered to consist of six oxygen atoms surrounding each lattice chromium in a distorted octahedron (17), however, in 1950 Bystrom and Wilhelmi (18) determined that the structure was a distorted tetrahedron with chromium-oxygen distance of 1.79-1.81 A. The tetra- hedra are formed by the sharing of oxygen atoms. The oxide is an extremely vigorous oxidizing agent, reacting explosively with many substances, particularly or- ganic materials. The oxidative characteristic is employed quite extensively in quantitative analytical chemistry. The oxide is reduced to the more stable chromium(III) oxide. When heated, chromium(VI) oxide loses oxygen progressively and de— composition occurs by the following steps (19): Cr03—> Cr308—-b Cr205—b Cr02 -—-v' Cr203 The paramagnetism of chromium(VI) oxide was attributed to incomplete pairing of the ground state electrons (20). Crystal- line chromium(VI) oxide is piezoelectric (21), suggesting that a permanent dipole results from the distorted tetrahedra. Dissolution of the highly soluble oxide occurs with the establishment of a Chromate-dichromate equilibrium. The rela— tive concentration influences the polymerization of the solution species. In very dilute solutions the Chromate [Cr04]'fi Species predominates. As the concentration of chromium(VI) increases the predominate solution species becomes first the dichromate, [Cr207]"2 then the trichromate, [Cr3olo]", and finally the tetrachromate, [Cr4013]-'. The formation of polychromate ions also occurs by increasing the hydrogen ion concentration. In a basic medium the chromate is quite possibly the only species present. Metal chromates and polychromates are generally prepared utilizing the above equilibrium. In 1931 Mellor (27) listed two hundred and fifty-seven dichromates, thirteen trichromates and five tetrachromates; many more have been prepared during the interim. Silver chromate was originally prepared by Vauquelin (21) in 1809 by precipitating the silver salt from silver nitrate and potassium chromate solutions. Silver chromate precipitates as monoclinic crystals (46) which vary in color from bright red to deep green. The color variations were reported by Bush (22) to be dependent upon the rate of precipitation. Nisikida (44) found that the red and green products were identical by x-ray powder patterns and he attributed the colors to differences in crystal grain sizes. The shape of the precipitated particle ranged from powders to large acicular or tabular (micaceous) crystals. The density of silver chromate is 5.52 g/cc and the molar volume 60.1 cc (23). The solubility of silver chromate in water is quite low; Kohlrousch (24) reported 2.56 x 10‘4 g./l. from conductivity measurements while G. S. Whitby (25), using colormetric observations, found a solubility of 2.52 x 10'4 g./l. Vauquelin (21) was also the first to prepare silver dichro- mate. He obtained the dichromate from an acidified solution of silver chromate. The usual preparation consists of precipitating 10 silver dichromate from a silver nitrate solution by the addi- tion of an acidified potassium dichromate solution. Color and crystal shape variations are obtained from different preparative techniques, however silver chromate generally occurs as violet acicular crystals with a density of 4.77 g/cc (26). Schabus (45) reported that silver dichromate was a mem- ber of the triclinic crystal system and has the axial ratios: a: b:c:- 1.5320: 1: 1.0546. Metal dichromates are generally quite water soluble and Silver dichromate (8.2 x IO’Zg/l) (28) has been found to be the least soluble of any known dichromate. The reactions of silver salts with organohalometallic compounds have been discussed in the literature. Schmidt and Schmidbaur (29) prepared tris(trimethylsilyl)arsenate(V) when they preformed the heterogenous reaction of silver orthoarsenate with trimethylchlorosilane in an organic solvent, while an analogous reaction using triphenylchlorosilane yielded tris- (triphenylsilyl)orthoarsenate(V) for Chamberland and MacDiarmid (30). The reaction of silver orthoarsenate with dimethyl- dichlorotin (31) in water yielded a white polymer with the following structure: 11 T“3 0— Sn —0 0/ I \ 0 CH3 I cu, I l —— 0 As As O —— Slm —-—' C x \ Ina / CH3 0— Sn —0 “’3 Tris(trimethylsilyl)orthophOSphate(V) was prepared from the reaction of trimethylchlorosilane and silver orthophosphate (29). The reactions of silver perchlorate with triorgano— chlorosilanes (32) gave trimethylsilylperchlorate, triethyl- silylperchlorate, tripropylsilylperohlorate, triphenylsilyl- perchlorate and trijp—tolylsilylperchlorate. Harder, gt_§1:. (33) prepared chromyl borate, Cr02(802)2, from the reaction of chromyl chloride with silver borate. Numerous chromyl compounds have been reported. A few of these are chromyl fluoride, chloride, acetate, nitrate, perchlorate, borate, bromide, sulfate, pyrosulfate, azide, thiocyanate, and cyanate. The chloride is the only chromyl compound whose chemistry has been studied to any extent. The bright red transparent liquid (m. -96.5° C., b. 115.70 C.) was first prepared by Berzelius (34) from the distillation of a chromate, sodium chloride-sulfuric acid mixture. Chromyl chloride is generally prepared by the reaction of chromium(VI) 12 oxide with hydrogen chloride containing small amounts of sulfuric acid as a dehydrating agent (35). The structure of chromyl chloride was shown to be a slightly distorted tetra- hedron by electron diffraction measurements (36). The nearly symmetric structure was also indicated by its low dipole moment (37) and low dielectric constant (38); these and other measurements substantiate the non-polar nature of chromyl chloride. Chromyl chloride is soluble in most common organic sol- vents, however it will often react violently with some of them. It is highly soluble in halogenated organic compounds and inor- ganic covalent halides. It is immiscible with water and hydro- lyzes very slowly at the liquid interface. Molecular weight studies of chromyl chloride (1, 2, 3, 4) indicate a degree of polymerization of one to three molecular units. The vapor density suggests the presence of the chromyl chloride monomer in the gas phase. When exposed to light the red liquid decomposes to chlorine and a black solid. The mecha- nism for photolytic deComposition of chromyl chloride has been reported by Schwab and Prakash (39). It is thermally stable to 180° C. Decomposition begins above this temperature and at approximately 4000 C. A magnetic oxide is formed having the composition of Cr509, Chromium(III)_oxide occurs at tempera- tures above 4000 C. (35). 13 Chromyl compounds are generally very reactive. The lability of the two substituents is used fairly extensively in preparative chemistry involving chromium(VI). Chromium (VI) Organic Compounds. The first organochromium compound prepared was bis(tri- phenylcarbinyl)chromate. Gomberg (40) isolated this material from the reaction of silver chromate with triphenylmethyl- chloride in benzene. The usual method for the preparation of chromate esters is the addition of chromium(VI) oxide to a solution of an alcohol in an inert solvent. The esters have also been prepared from reactions of the alcohol with chromyl chloride in alkylhalide solvents or with alkali metal dichro- mates in glacial acetic acid. Most chromate esters are derivatives of tertiary alcohols. The reaction of primary and secondary alcohols with chromium(VI) oxides in an inert solvent yields yellow to yellow-orange solu— tions. The color, however, remains only for a short time and as yet none of these unstable chromates have been isolated. Zeiss and Matthews (41) obtained very pure chromate esters by using a freeze-drying technique for purification. The alcoholysis of chromate esters yields the parent alcohol and a reduced chromium species, possibly HZCr03. The hydrolysis of bis(tggt-butyl)chromate (42) occurs through cleavage of the chromium-oxygen bond while the hydrolysis of bis(triphenylmethyl)— 14 chromate proceeds by breaking the carbon-oxygen linkages. Other organochromium compounds include numerous carbonyl derivatives of chromium(VI) and arylchromium chemistry (43). Although these two topics are extremely interesting, they are not closely related to this investigation and are much too extensive to be included. Chemistry of Fluorine. Physical Properties. Hare (5) noted that the thermal stabilities of organo- siloxychromates were enhanced when the organosilicon substi- tuents exhibited strong electron-withdrawing tendencies (p. 140 and 141). It was postulated, therefore, that siloxy- chromates with either silicon-fluorine or silicon-fluoroorganic linkages might show even greater resistance to thermal decom- position than that seen with previously prepared compounds of this type. The term "electronegativity" has been defined by Pauling (48) as the "power of an atom or molecule to attract electrons to it— self." Although this concept is not susceptible to direct experi- mental measurement, the three principal methods for electro- negativity assessments (Pauling, Mulliken, and Malone) suggest that fluorine is the most electronegative element (47). It is‘ also quite possible that atomic fluorine exhibits this electron- withdrawing tendency to a greater degree than any known molecular 15 species. The physical property of di- and poly-atomic mole- cules containing the halogen homologues generally follow predictable trends through iodine, bromine, and chlorine only to go completely off course upon reaching the fluorine species. A classic example of this anomalous behavior is shown by the melting points, boiling points, and heats of fusion and vapori- zation of the hydrogen halides (49). The property values decrease progressively from hydrogen iodide to hydrogen chloride, however hydrogen fluoride property values are considerably higher and with the exception of the melting point, all hydrogen fluor- ide values were greater than that of hydrogen iodide. This deviation from progressive trends is attributed to intramole- cular association of hydrogen fluoride resulting from hydrogen bonding which in turn is caused by the abnormally high electro- negativity of fluorine. Not all anomalies shown by fluorine- containing molecules can be explained by electronegativity alone. For example, fluorine, unlike the remaining members of the series, does not have g_orbitals available for bonding which results in a somewhat different chemistry. The silicon-fluorine bond has one of the highest known energies for a single bond (143 kcal./mole.), being only slightly weaker than the boron—fluorine linkage (150 kcal./mole.) (Table I). It should be noted, however, that high bond energy is not the only criterion for stability. Many other factors, such as kinetic considerations, influence molecular thermal resistance (p. 142). 16 TABLE I ENERGIES OF RELATED CHEMICAL BONDS BOND ENERGY REF. BOND ENERGY REF. (kcal./mole.) (kca1./mole.) Si-F 143 48 Si-Cl 95 50 F-F 37 49 Sl-Br 69 48 H-F 135 49 Si-H 76 49 C-F 111 52 C-C 83 49 B-F 150 50 o-C 146 49 81-0 110 50 CIC 200 49 Sl-C 64 51 CIN 213 50 The high strength of the silicon-fluorine bond has been attributed to partial double bond character. Kriegsmann (53) calculated the force Constants for ideal Si-X single bonds where X - F, Cl, Br, I, 0, S, N, C, and H. When X is a highly electronegative element, he found a strengthening of the link- age that corresponded to appreciable double bond character. Pauling (48) also reached this conclusion for silicon-fluorine and silicon-chlorine linkages by means of theoretical interper- tations of the interatomic distances in various silanes. Rochow (54) attributed the magnetic shielding of the fluorine exhibited in the F19 nuclear magnetic resonance spectra of methyl- and ethyl-fluorosilanes to both the inductive effect and a T! bonding effect. Interpertations of electric moment measurements 17 of halosilanes (55) and alkylhalosilanes (56) also indicated appreciable double bond character for the silicon-chlorine and silicon-fluorine linkages. Recently the Dow Corning Corporation has been marketing silicon polymers with considerably higher thermal stability than has been seen previously. This greater resistance to elevated temperature resulted from the replacement of the normal organic groups with fluoroorganic substituents. The bond energy term for the carbon-fluorine linkage varies from 107-116 kca1./mole. The strength of the carbon— fluorine bond increases as fluorine atoms successively dis- place hydrogen atoms attached to the same carbon. The energy term for carbon tetrafluoride is one-quarter of the heat of atomization of this compound or 116 kca1./mole. and the CFg, -CF2, and monofluorinated structures have respective carbon— fluorine energy terms of 114, 109, and 107 kcal. A change in bond strength does not accompany the successive increase in the number of chloro or bromo substituents on a particular carbon atom. This hypothesis is clearly illustrated by a comparison of the bond energy term for the carbon-halogen linkages in the fluoro, chloro, and bromo halomethane families (Table II). 18 TABLE II BOND LENGTHS AND BOND ENERGY TERMS IN HALOMETHANES (57) X - F X - C1 X - Br Compound B(C-F) r(C-F) B(C-Cl) r(C-Cl) B(C-Br) r(C—Br) (kca1.) (X) (kca1.) (X) (kcal.) (X) cx4 116 1.317 73.2 1.766 66.6 1.942 CX3H 114.6 1.332 78.3 1.767 66.1 1.930 cxznz 109.6 1.353 77.9 1.772 65.6 cxas . 107 1.335 73.0 1.73 66.6 1.939 B a bond energy, r - bond lengths The strengthening of the carbon-fluorine bond is accompanied by a corresponding decrease in bond length. The energy terms and bond lengths for the chloro and bromo methanes remain relatively constant. Pauling (48) explained that the increased strength and contraction in length of the carbon-fluorine linkage can be accounted for by the contribution of resonance structures of the type: F' F' 3* . \. _ F—C=X F—C+ --x F—CX / /' ./ F F F The degree of influence that each of the above resonance contri- butors has upon an individual molecule depends upon the identity of X. The number of fluorines in each molecule determines the number and importance of each resonance structure. Meyers and 19 Gutowsky (58) found that the F19 chemical shift in nuclear magnetic resonance Spectra of the fluoromethane family de- creased progressively with decreasing carbon-fluorine bond lengths. They postulated that if this bond shortening re- sulted from an increased ionic character of the bond, the trend of the chemical shift values would be in the opposite direction. However, if the contraction in bond length was related to a greater degree of double-bond character the observed chemical shift trend would be expected. Lagowski (59) has shown the electronegativity of the trifluoromethyl group to be approximately 3.3 on the Pauling scale. The value was obtained fnam a number of calculations involving the use of bond dissociation energies and infrared absorption frequencies. According to Lagowski the electro- negativity of the trifluoromethyl group is equal to or greater than chlorine and considerably larger than the alkyl groups which are of the order of 2.0. Cullen, gt_gl., (60) sub- stantiated Lagowski's electronegativity value from ultra- violet measurements of trifluoromethyl arsines; however, at a later date (61), he determined the ionization potential of a series of perfluoroalkylarsines by electron impact mass Spectrometry and suggested that the electronegativity of the CF3 group was slightly less than that of the chloro group. The approximate covalent diameter of the perfluoromethyl group is 3.3 angstrom units, which is slightly larger than the methyl 20 group at 2.8 X. It has been suggested (62) that the larger size of the fluorine atom compared with that of hydrogen, provides much greater protection of the carbon skeleton in. a fluorocarbon in contrast to a hydrocarbon. The greatest physical significance of the highly electronegative fluorine atoms in organic fluorine molecules is their electron-with- drawing or "inductive" effect which leads to a decrease in electron density at adjacent reaction centers. Direct evi— dence of this effect is the repeatedly encountered shifts to shorter wavelengths by other infrared absorbing linkages with- in a fluorine-containing molecule. Another clear example of the inductive effect exerted by organic fluorine groups is illustrated by trifluoroacetic acid which has a considerably greater acidic nature than acetic acid (63). Thermal Stability. The increased thermal stability obtained by the inclusion of fluorine in molecules has been effectively demonstrated by the Dow Corning Corporation with their high heat resistant fluororganic—silicones (64). Haszeldine, gt_gl., (65) and Pierce (64) found that the thermal and hydrolytic stability of polyfluoroalkyl silicones and polysiloxanes depends markedly on the position of fluorine relative to silicon. Both groups concluded that molecules in which fluorine is attached to the gamma carbon atom exhibit the greatest thermal and hydrolytic stabilities. Thermal stability increases, resulting from the 21 replacement of fluorine for lower electronegative substituents, have been observed with many varied molecular types. Emeléus, Haszeldine, and Brandt (66) found that perfluoroalkyl sulfides exhibited much greater resistances to heat and chemical attack than did their corresponding alkyl analogues. The boron— phosphorus linkages in certain compounds have shown considerable increases in bond strength from the replacement of fluorine for alkyl or hydrogen substituents attached to the phosphorus atom of the molecule (67). Tetrakis(trifluoromethyl)diphosphine, diarsine, and distibine (68, 69, 70, 71, 72) compounds of the type (CF3)2M-M(CF3)2 have been found to be more resistant to thermal cleavage of the metal-metal bond than their correspond- ing methyl analogues. Burg suggested that this greater stability resulted from the electronegative trifluoromethyl groups stabil— izing the metal-metal bonds by strengthening the T1 bonds which involve the metal lone-pair electrons of the 3d orbitals of the neighboring metal atom. Inorganic polymers with phosphorus- phosphorus and phosphorus-boron back-bone linkages with thermal stabilities approaching 5000 C. have been prepared (73). Perfluoroalkyl-transition metal derivatives such as RfFe(CO)4I (Rf = CF3, CZFS: C3F7) and C3F7Re(C0)5 are much more stable to thermal decomposition than their alkyl analogues (74, 76). The thermal decomposition of RfFe(CO)4I involves cleavage of the Fe-I and apparently not the Rf-Fe bonds. McClellan (75) prepared and studied eleven 22 manganese and three cobalt carbonyls containing fluorocarbon linkages and found them all to show greater thermal stabili- ties than their hydrocarbon analogues. It is interesting to note that the authors listed in the preceeding section invari- ably explained that the thermal stability increases shown by fluorine-containing molecules resulted from the highly electron— withdrawing property of the fluorine atom. Hydrolytic Stability. The relative hydrolytic stability, like thermal stability, of fluorine-containing compounds as compared to their hydrogen or haloranalogues can generally be explained in terms of electro— negativities. Alkyl-, trialky1-, and triaryl-silyl fluorides have been found to be considerably more resistant to hydrolysis than the corresponding chlorides and bromides (77, 78, 79). The hydroly- sis of triphenylsilyl fluorides in aqueous acetone was retarded by pgmethyl substituents and the reactions of these fluorides were more than a million times faster in alkaline solutions than in water. A nucleophilic bimolecular substitution has been proposed as the mechanism for hydrolysis (77). The silicon bond in alkyl fluorosilicates (80) is considerably less reactive than the silicon-chlorine bond in analogous compounds; however, the hydrogen fluoride that is formed acts as a powerful catalyst for the further hydrolysis or alcololysis of the compounds. 23 A study of the nuclear magnetic resonance spectra of the tetrafluoroborate and trifluoromethylfluoroborate ions in aqueous solutions (81) indicated that the degree of hydro— lysis of these substances was quite small. The hydrolytic stability of fluoroalkylsilanes and polysiloxanes has been discussed on p. 22. Trifluoromethyl-trihalogermanes, CF366X3, were found to be fairly stable in cold water but evolved trifluoromethane in hot water or alkaline solutions (82). The reduction in donor character (decreased bascity) of perfluoroalkyl amines results in a much more hydrolyti- cally stable species than the corresponding alkyl tertiary amines (83). The perfluoroalkyl derivatives of phosphorus, arsenic, and antimony are fairly resistive to aqueous hydro- lysis. However, in the presence of alkali these materials de- compose quickly. Haszeldine and West (84) found that the ease of attack by hydroxyl ions was facilitated by the replacement of trifluoromethyl for methyl substituents in trimethyl com- pounds of group V elements. They established the following order of increasing ease of hydrolysis: “(053’s < CF3M(CH3)2 < (CF3)2M053 < (CF3)3M The ease of hydrolysis also increased as the atomic radius of the group V atoms increased, i.e.. P As Sb. Kolditz and Range (85) prepared a linear arsenate having an unknown molecular weight based upon the following structure: 24 They found that complete hydrolysis could only be obtained by refluxing the polymer in an alkaline solution for several hours. They attributed the stability to a fluorine screening effect which limited the availability of the arsenic—oxygen- arsenic linkage to hydroxyl attack. The perfluoroalkyl derivatives of sulfur such as CF3SF5, CF2(SF5)2, CF2(SF3)2, (CF3)2SF4, and (CF3)2S02 have been found to be completely unreactive in alkaline solu- tions at 1000 C,(86). The limited number of perfluoro deriva- tives of iron, manganese, cobalt, and rhenium that have been prepared have shown a much greater resistance towards alkaline hydrolysis than either their hydrogen analogues or the perfluoro— alkyl derivatives of the main group metals (74, 75, 76). Inorganic Polymer Chemistry. It is generally assumed that inorganic polymer chemistry is a relatively new field, 143,, that no more than two decades have seen investigations concerning macromolecules composed ‘essentially of skeletons other than carbon. In a sense, this assumption is both true and false. The investigation of inor- ganic macromolecules is in actuality far from being a new field. 25 The structures of silicates and borates as well as those of the polymeric forms of phosphorus and transition metal oxy— acids have been known for some time. The structures of rhombic sulfur and metallic selenium are considered polymeric by some authors. In fact, under a strict definition of the term, inorganic polymers would necessarily include all crystal- line lattices in which the bonding is primarily covalent. Transition metal compounds are often polymeric in the solid state; for example, vanadium pentoxide tetrahedra form large sheets through shared oxygen atoms. The rather complex poly— phosphomolybdates and tungstates have confused and delighted structural chemists for decades. Polymeric forms of phospho- nitrilic chloride were discovered in 1834 by Liebig and Wahler. In 1897 Stokes reported the preparation of a so—called "inor- ganic rubber" from thermal treatment of phosphonitrilic chloride. During the past two decades, the space development program has prompted the chemist to initiate investigation of inorganic polymeric systems. This came about primarily because polymers with carbon-carbon or carbon-oxygen skeletal systems had been unable to withstand the temperature gradient required for practical application in space. The inorganic polymeric forms which have been isolated and investigated in the past also were unsuited for these applications primarily because of their somewhat rigid mechanical properties. The ngw_inorganic polymer investigations have consisted primarily of searching 26 for polymeric materials which combine the temperature resist- ant properties of inorganic material with the flexibility and toughness of plastics or the viscosity of oils. Many different combinations of elements have been used in an attempt to prepare these utopian materials; however, complete success has continued to elude the inorganic polymer chemist. The synthesis of materials which decompose at fairly high temperatures (300—400°(L) has often been negated by their inability to withstand the attack of chemical reagents. A perusal of the many review articles concerning inorganic polymers would lead one to believe that the most difficult problem facing the inorganic polymer chemist is that of classi- fication of the many polymeric systems developed during the past few years. These systems include materials such as phosphoni- trilic halides and related phosphorus-nitrogen macromolecules; the phosphorus-phosphorus, boron-boron, phosphorus-boron sys- tems studied extensively by Burg; silicon-oxygen-metal polymers which include the vast number of silicones developed by numerous private and commercial laboratories, coordination polymers, and many other systems. The listing or discussion of these numerous inorganic polymer systems is completely out of the scope of this dissertation; therefore, an attempt will be made to limit the coverage to polymerization reactions which are directly related ‘to this investigation. 27 As was stated previously, chromium(VI) was chosen as a constituent in the silicon-oxygen-metal system because of its polymeric tendencies and the difunctionality of the tetrahedral chromyl group. The polymeric tendencies of chromium(VI) are evident in the polyanionic aggregates which exist in either acid solution of chromium(VI) or highly con- centrated aqueous solutions (87). Large linear molecules such as the metal tetrachromates can be precipitated from these solutions. The crystalline form of chromium(VI) oxide consists of Cr06 octahedra which make up~a continuous network stretching throughout the entire crystal. Cryoscopic and ebullioscopic measurements of the molecular weight of chromyl chloride in the liquid state have suggested a degree of polymerization equal to 1.5-3 molecular units (1, 2, 3, 4). During the thermal decomposition of chromyl chloride an elimination of chlorine occurs at approximately 200°“C. Analysis of the residue indicates the existence of the trimer and tetramer of (CrO)nCL2 (88). Numerous coordination polymers of chromium have been pre- pared. Block, gt_31,, have reported the preparation of a mono- meric chromium(II) compound (89) with polymeric properties, dimers of chromium(III) coordination compounds (90), and chro- mium(III) coordination polymers with molecular weights from 400 to 1400 (91). Schmitz-DuMont (92) has prepared a variety of chromium(III) amide and alkoxo chromium(III) amide polymeric 28 coordination compounds. A chromium(III) coordination polymer marketed under the trade name of "Valan" has been used pri- marily as an adhesive between organic resins and glass (87). The only known chromium(VI) coordination polymer was prepared by Podall and Iapalucci (93) from photochemical oxidative decarbonylation of chromium hexacarbonyl by diphenylhypophos- phorus acid, (PhZPOZH). The resulting solid was stable up to 360° c insoluble in water or hydrochloric acid, and had the ’ 9 following structure: CO Cr / 0P CO CO 8 Silicon polymers containing silicon-fluorine linkages have been prepared in several laboratories. J. Goubeau and H. Grosse-Ruyken (94) reported the preparation of the hydro- lytically unstable hexafluorodisiloxane, octafluorotrisiloxane, and a difluoropolysiloxane having an unknown molecular weight. Schmeisser (95) and Pease (96) simultaneously reported the preparation of Polydifluorosilanes, (SiF2)x, with as many as ten molecular units. In 1949 a British patent was granted to the Dow Chemical Company for phenylfluoropolysiloxanes of un- known molecular weights (97). Numerous polysiloxanes with fluoroorganic silicon linkages have been prepared in the 29 laboratories of McBee, Haszeldine, and the Dow Corning Corporation. Silicon-oxygen-X polymers and telomers have been prepared using many different elements. Silicon, titanium, vanadium, aluminum, tin, lead, germanium, arsenic, antimony, boron, and phosphorus as the X atom have accounted for most but not all of the work accomplished in this field. A complete discussion of the Silicon-oxygen—X systems is outside the scope of this thesis, primarily because of the vast number of publications relating to compounds of this type. EXPERIMENTAL Preparation of Reactants. Preparation of Silver Chromate. Silver chromate was prepared by the method of Vauquelin (21). A 1.2 molar solu- tion of potassium chromate (2.4 moles) containing a catalytic amount of aqueous ammonia was added to a 1.2 molar silver nitrate solution (2.4 moles). The silver nitrate solution was stirred vigorously throughout the addition and the result- ant dark reddish-brown precipitate was extensively washed with warm water. When this preparation was carried out with more concentrated solutions of silver nitrate and dilute potassium chromate solution or when both solutions were concentrated, the resulting silver chromate was found to be badly contaminated with the potassium salt. The product was dried at 110° C. for twenty-four hours. When the pure silver chromate was heated on an aluminum melting block it Showed a gradual color change between 300 and 400° C., indicating the possibility of decomposition at this point. A differential thermal analysis of this material, how- ever, showed that little or no physical change took place over this temperature range. The thermogram of silver chromate gave a pronounced exothermic peak extending from 480 to 580° C. with a maximum at 535° C. It is believed that this is the actual temperature range for the thermal decomposition of silver chro- mate. 30 31 The molecular Spectrum (Figure 5, p.2HL4) from 200 to 500 millimicrons gave absorption peaks at 503 and 274 mp with respective molar extinction coefficients of 4.27 x 103 and 3.14 x 103. The x-ray powder pattern of silver chromate is tabulated on p. 137 . The chromium analysis of the compound was 15.62 percent. (Theory: 15.68% Cr). Preparation of Silver Dichromate. Silver dichromate (80% yield) was precipitated from an acidic one molar potassium dichromate solution by the addition of a one molar solution of silver nitrate (98). The resulting deep violet precipitate was purified by recrystallization. A concentrated solution of silver dichromate was prepared by adding an excess (approx. 100 g.) of the product to 400 m1. of water and 50 m1. of concentrated nitric acid. The mixture was stirred at its boiling point for thirty minutes, filtered hot to remove the excess reaction product and stored in a refrigerator for six hours. Approximately twenty grams of silver dichromate crystal- lized as long violet needles. After filtration, the pure material was dried at 110° C. The melting range of the Silver dichromate was found to be from 358 to 362° C. when observed visually. The thermogram also Showed this melting point as a very intense and Sharp endothermic peak at 330° C. A broad exothermic peak between 480 and 580° C. and a maximum at 540° C. corresponded to the decomposition of Silver dichromate. 32 The visible and ultraviolet spectra (Figure 5, p. 104) showed strong absorption at 258, 344, and 470 mu, with molar extinction coefficients of 1.75 x 103, 1.98 x 103, and 3.36 x 103 respectively. The infrared Spectrum of silver dichromate in a potassium bromide matrix gave a strong ab- sorption at 11.3 p which correSponded Closely to the MEO-Cr absorptions found in several dichromates by Miller and Wilkins (99). The x—ray powder pattern is tabulated on p. 138. The chromium analyses were 24.09 and 24.01 percent (theory: 24.11% Cr). Characterization of Triphenylchlorosilane. Commercial grade triphenylchlorosilane was obtained from K and K labora- tories and purified by recrystallizing the white solid from methylene chloride. The melting point of 96 to 97° C. agreed well with the literature value of 97° C. (6). Silicon analyses of this material gave experimental values of 9.61 and 9.58 per- cent (theory: 9.54% Si). The infrared spectrum of triphenyl- chlorosilane in the sodium chloride and cesium iodide regions are Shown in Figures 10 and 11 (p. 109 and 110). Triphenlehlorosilane hydrolyzed very easily and was, therefore, purified prior to each reaction. . Preparation of Diphenyldifluorosilane. Diphenyldifluoro- silane was prepared by the method of Sommer et al. (133). Aqueous hydrogen fluoride (49%) (40 g., 1.0 mole) was treated 33 with twenty grams (0.074 mole) of diphenyldiethoxysilane in a polyethylene beaker. This immiscible mixture was stirred magnetically for four hours during which the mixture was cooled in an ice bath. Diphenyldiethoxysilane had been puri- fied previously by vacuum distillation. The two resulting layers were separated with a separatory funnel. The organic layer (diphenyldifluorosilane) was then added to an equal volume of methylene chloride containing five grams of an- hydrous magnesium sulfate which removed the water trapped in the organic layer. After standing for at least one day, the magnesium sulfate was removed by filtration and the methylene chloride by evaporation at reduced pressure (water aspirator). The dry diphenyldifluorosilane was distilled at several reduced pressures with boiling points of 151° C. at 43 mm. Hg, 75° at 0.4 mm. Hg, and 68° at 0.05 mm. Hg. Emeléus and Wilkins (100) cited a boiling point of 156 to 1600 C. at 50 mm. of Hg. Yields of 65 to 75 percent were obtained. Diphenyldifluoro— silane has a density of 1.15129 g./cc. at 250° C. and a re- fractive index of 1.5269 at 20° C. The infrared spectrum in the sodium chloride and cesium iodide regions are shown on Figures 14 and 15 (p.113 and 114). The elemental analysis of diphenyldifluorosilane is listed on the next page. 34 TABLE III ELEMENTAL ANALYSIS OF DIPHENYLDIFLUOROSILANE Element Theory (%) Experimental (%) C 65.42 64.35, 65.61 H 4.58 4.55, 4.74 Si 12.75 12.37 F 17.25 17.45 Characterization of Diphenyldichlorosilane. Crude diphenyldichlorosilane was obtained from General Electric Company and purified by vacuum distillation. The boiling point was 110° C. at 1.5 mm. of Hg and 304° C. at 1 atm. The melting point was -22° C. and the refractive index 1.5792 at 20° C. The infrared Spectrum of diphenyldichlorosilane is shown in Figures 12 and 13 (p.111 and 112). The chlorine analysis was 28.00 percent (theory: 27.93% Cl). Preparation of Diphenylchlorofluorosilane. Pletcher and Nutting (101) prepared diphenylchlorofluorosilane by the fluorination of diphenyldichlorosilane with lead tetrafluoride. Booth gt_al. (102-107) using the Swarts reaction (antimony trifluoride) prepared dimethylfluorochlorosilane and several trihaloalkylsilanes with various degrees of fluorination. 35 Antimony trifluoride (5.95 g., 0.033 mole) was added quickly to a three necked flask containing 25 g. (0.10 mole) of liquid diphenyldichlorosilane. The reaction was cooled in an ice bath while being stirred magnetically. After three hours the mixture was distilled at reduced pressure (1.0 mm.). The first fraction to be removed was antimony trifluoride which collected as white crystals on the cold finger. Several frac- tions of the remaining liquid were obtained; however, analysis of these fractions by vapor phase chromatography showed only two of the expected three peaks. The retention time of these components corresponded exactly with diphenyldifluorosilane and diphenyldichlorosilane. Distillation of a second reaction mixture was performed at 1 atm., after which the vapor phase chromatograph gave an entirely new peak equidistant between the two previously shown. It was therefore assumed that diphenylchlorofluorosilane was produced by elemental exchange between diphenyldifluorosilane and diphenyldichlorosilane at the high temperatures (245-3050 C.) required for distillation. Equal molar quantities of diphenyldichlorosilane and diphenyl- difluorosilane were refluxed at temperatures between 275 and 285° C. for ten hours. The vapor phase chromatograph showed three distinct equidistant peaks. The above reaction was continued and samples were analyzed (V. P. C.) at approximately ten-hour intervals. An equilibrium was apparently established after one hundred and fifty hours. The reaction rate did not 36 appear to increase when catalytic amounts of silicon tetra- chloride were used. The resulting diphenyldihalosilanes were separated by fractional distillation using a vacuum Jacketed column containing glass helices. This method was somewhat cumbersome because several distillations were necessary before pure diphenylchlorofluorosilane could be obtained. When molecular distillation was attempted at pressures of 10'3 mm., the separation was even poorer than with the less refined vacuum distillation. Yields of twenty to twenty-five percent were obtained. Diphenylchlorofluorosilane has a boiling point of 274° C. at 1 atm. and 720 C. at 1.0 mm. of Hg. The three homologues in this series (PhZSiClz, b. 304° C.; PhSiClF, b. 274° C. and PhZSiFZ, b. 247° C.) follow the Swarts rule (108) which states that the boiling point lowering is constant as successive fluorine atoms replace other halogen atoms in a given poly— halogen. The clear liquid, diphenylchlorofluorosilane, was viscous at -780 C. and froze to a clear glass-like substance when it was immersed in liquid nitrogen (-196° C.); however, upon warming, no definite melting point could be distinguished. The density at 25.0° C. was 1.1860 gu/cc. which is in good agreement with Pletcher (101) who obtained a value of 1.181 gu/cc. at 25.5° C. The refractive index was 1.5539 at 20.0° C. 37 The infrared spectrum of diphenylchlorofluorosilane (Figures 16 and 17, p. 115 and 116) exhibits the normal ab- sorptions characteristic of silicon-phenyl linkages. The strong carbon-silicon stretch at 8.89 p was resolved into two peaks similar to diphenyldichlorosilane. This splitting is not observed in triphenylmonohalosilanes or phenyltri- halosilanes (115). The two out-of-plane hydrogen deforma- tion peaks which normally occur between 13.5 and 14 u were masked by solvent (CC14) absorption. The silicon-fluorine asymmetric stretching mode (123) at 11.42 p was extremely strong and sharp. The silicon-chlorine asymmetric stretching mode (113) was similar in appearance and absorbed in the cesium iodide region at 17.6 p. This absorption masked the silicon-fluorine asymmetric deformation mode which appears at 17.6 p in the spectrum of diphenyldifluorosilane. The elemental analysis is tabulated below. TABLE IV ELEMENTAL ANALYSIS OF DIPHENYICHLOROFLUOROSILANE Element Theory_(%) Experimental (%) C 60.78 60.77 H 4.26 4.38 Si 11.86 11.70 F 8.03 7.85 CI 14.98 14.79 .u. PA..— .\~ ‘hr ‘1 ‘1‘ 38 Characterization of 3,3,3-trifluoropropylmethyldichloro- silane. The 3,3,3-trifluoropropylmethyldichlorosilane was received from the Dow Corning Corporation. The elemental analyses gave 33.40 percent C1 and 27.18 percent F (theory: 33.40% Cl and 26.87% F), and the clear liquid was used as a reactant without further purification. The infrared spectra are shown in Figures 28 and 29 (p. 127 and 128) and are tabu- lated on p. 81—4. There was no absorption in the ultraviolet region. The nuclear magnetic resonance spectrum is shown in Figure 2 (p. 91) Table XII (p. 90) and discussed on p. 87. This material was first prepared by McBee gt_gl. (109) from the reaction of 3,3,3—trif1uoropropene with methyldi- chlorosilane using tert—butyl peroxide as a catalyst. A sealed tube was charged with the reactants and the catalyst under a nitrogen atmOSphere and then heated at 125 to 130° C. for sixty hours. The major product, 3,3,3-trifluoropropyl- methyldichlorosilane, was purified by distillation. Its boil- ing point at 751 mm. of Hg was 112.8 to-ll3.7° C. Preparation of 3,3,3-trifluoropropy1methylsilanediol. The method of Holbrook and Brown (110) for the hydrolysis of 3,3,3-trif1uoropropylmethyldichlorosilane was used to prepare 3,3,3—trifluorOpropylmethylsilanediol. Over a period of three hours 0.1 mole (21.1 g.) of 3,3,3-trifluoropropylmethy1— dichlorosilane in 340 m1. of anhydrous ether was added to 250 ml. of water containing twenty-four grams of twenty-eight 39 percent ammonium hydroxide. This mixture was stirred magnet- ically and cooled with an ice bath. When the addition was complete the water layer was removed by means of a separatory funnel. The ether layer was washed with equal volumes of water until the wash water was no longer alkaline; normally three washings were sufficient. Purification by recrystallization from the reaction solvent (ether) as proposed by Holbrook and Brown (110) was found to be unsatisfactory. The reaction mixture was cooled in a Dry Ice bath. After a short time, crystals of water appeared, which were removed by filtration. Repeated crystal— lization removed most of the water trapped in the ether layer by the vigorous stirring. The ether was evaporated under a stream of air, leaving a white solid. This solid was easily sublimed at room temperature with pressures of approximately 1.0 mm. Hg. Yields of the pure material ranged from eighty- five to ninety—five percent. The melting range was 82-83° c. which agreed with the 82-85° C. range found by Holbrook and Brown (110). The silicon analyses gave 16.35 and 16.22 percent (theory: 16.12% Si). The infrared spectrum is shown in Figures 30 and 31 (p. 129 and 130), and the proton magnetic resonance spectrum is recorded in Figure 3 (p. 92). 40 Preparation of Silylchromates. Generally silyl- and siloxanylchromates have been pre— pared by the reactions of organosilanols or disiloxanes with chromyl chloride or chromium(VI) oxide. The preparation of organosilicon compounds having both a fluoride and a hydroxyl group attached to silicon is quite improbable; therefore it was necessary to investigate other synthetic routes for the prepara- tion of silylchromates containing a Silicon-fluoride bond. Numerous organosilicon compounds having both chloride and fluoride groups bonded to Silicon have been prepared. Investi- gations of these substances have shown that the reactivity of the chloride substituent is far greater than that of the fluoride. Consequently, the initial experiments in this study were performed to determine whether organochlorosilanes react with metal chro- mates to yield silylchromates and if this were the case whether only the chloride substituent on an organochlorofluorosilane would be replaced. I. Reactions of Triphenylchlorosilane. A. With Silver Chromate. Yields of up to ninety percent of pure bis(triphenylsilyl)chromate(VI) were obtained by the reaction of silver chromate with triphenylchlorosilane. A heterogenous mixture of 3.0 x 10'2 mole (10 g.) of silver chromate in 150 m1. of freshly distilled methylene chloride was magnetically stirred in a 300 ml. reaction flask. As a 41 means of eliminating hydrolysis of both the reactant and the product, the vessel was flushed with a continuous stream of dry nitrogen. The reaction flask was also covered with aluminum foil to reduce photolytic decomposition of the product. The triphenylchlorosilane (1.70 x 10‘2 mole or 5.00 g.) in fifty milliliters of methylene chloride was added dropwise to the silver chromate slurry over a period of thirty minutes. The reaction continued for twelve hours at room temperature. The excess silver chromate was removed by filtration in a dry box after which the solution was condensed to approximately thirty milliliters by means of a rotary evaporator (Rinco). Bright orange crystals of bis(triphenylsilyl)chromate(VI) precipitated from the concentrated reaction mixture at Dry Ice temperatures. The material was recrystallized from fresh methylene chloride and gave a melting point of l55-l56° C. which corresponds well with the value of 153.5—4° C. found by Hare (5). The unreacted chromate was subsequently dissolved in six hundred milliliters of boiling 0.6 M nitric acid leaving a light tan solid residue which was identified as silver chloride by comparing its x-ray powder pattern with a known pattern. The calculated yield for silver chloride from this reaction was 1.70 x 10'2 mole (2.43 g.); however, because of the crudeness in the separation technique the 1.47 x 10"2 mole (2,11 g,) ob- tained was believed to be a satisfactory indication of the completeness of the reaction. 42 B. With Silver Dichromate. A ten-fold excess of silver dichromate (3.4 x 10'2 mole or 14.7 g.) was treated with 6.8 x 10'3 mole (2.0 g.) of triphenylchlorosilane using a procedure similar to the above silver chromate reaction in all but one respect i;g., several preparations were carried out at the reflux temperature of methylene chloride (34° C.) and several at room temperature. The reaction times were varied from twelve to eighty hours and the best yields (I>90%) were obtained at reflux temperatures and with Shorter times. The large orange twinning crystals which were precipi— tated from the concentrated reaction solution had a melting point of 153—6° C. The calculated amount of silver chloride for the reaction was 6.8 x 10’3 mole (0.97 g.) and 5.9 x 10"3 mole (0.85 g.) was found. This product was shown to be bis(triphenylsilyl)chromate(VI) by its x—ray powder pattern (a), infrared spectrum, and elemental analysis. It was sur— prising, as well as disappointing, when the product of this reaction proved to be bis(triphenylsilyl)chromate(VI) instead of the dichromate. A mechanism for this reaction has been proposed and is discussed on p.161 . (a) The x-ray powder pattern for this material did not correspond to either the pattern or the table of interatomic spacings shown by Hare (5). After careful inspection it was found that Hare had interchanged the pattern and tabulated values of bis(triphenylsilyl)chromate(VI) with those of bis(p-tolyldiphenylsilyl)chromate(VI). 43 C. With Potassium Chromate. A four—fold excess (3.1 x 10"2 mole or 6.0 g.) of potassium chromate was treated with 1.67 x 10"2 mole (5 g.) of triphenylchloro- silane in a manner similar to that of silver dichromate. The mixture was refluxed for two days and at unevenly spaced intervals 10 ml. samples of the reaction mixture were re- moved and checked for a color change. The solution remained clear and colorless for the forty-eight hour reaction period. The mixture was filtered and condensed, and the infrared spectrum of the white solid residue corresponded to that of the starting material, triphenylchlorosilane. D. With Potassium Dichromate. Similar amounts (grams) of potassium dichromate and triphenylchlorosilane were re- fluxed and stirred in methylene chloride. A reaction between these two substances was not evident after twenty-one hours. E. With Sodium Dichromate. A gram of triphenylchloro- silane in five milliliters of methylene chloride was mixed with approximately five grams of sodium dichromate in a six- inch test tube. The mixture was shaken for ten minutes, allowed to stand overnight, and then filtered. The colorless filtrate indicated that a reaction had not occurred. II. Reactions of Diphenylchlorofluorosilane. A. With Silver Chromate. The procedure described below for the reaction of silver chromate with diphenylchlorofluoro- silane was the general method used for most of the syntheses 44 in this study and will apply to most of the succeeding reactions. A three-fold excess of silver chromate (3.0 x 10‘2 mole or 10 g.) which had been stored at 1100 C., was added to 250 ml. of methylene chloride in a 500 ml. three—necked round bottom flask. All solvents used during this study were freshly distilled from drying agents, generally phos- phorus pentoxide, and then stored over molecular sieves for no longer than three days. A reflux condenser and a dropping funnel were connected to the reaction vessel which was covered with aluminum foil. A continuous stream of oil-pumped nitrogen was passed through a sulfuric acid drying tower and then through the reaction system. The reacting compound was stirred by means of an internal magnet and heated by a mantle to maintain reflux conditions. A solution of 9.7 x 10"3 mole (2.3 g.) of diphenyl— chlorofluorosilane in fifty milliliters of methylene chloride was added to the silver chromate slurry dropwise over a period of thirty minutes. The reaction continued for a total of six and one-half hours. When the reaction was complete, a filtering tube was con- nected to both the reaction vessel and a 500 ml. one-necked round bottom flask. The one inch (0. D.) tube was curved so that the standard taper male ends were separated by ninety degrees. A glass frit at the center of the bend allowed the separation of excess silver chromate from the solution in a ii! " CC. 45 relatively inert atmOSphere. A decanting action.was used in order to maintain a constant pressure throughout the system. The 500 ml. flask containing the product and solvent was transferred quickly to a rotary evaporator where the solvent was removed at pressures of approximately 1 mm. of Hg. One preparation was attempted using diethyl ether as the solvent; after fifteen hours no color change was observed. Bis(diphenylfluorosilyl)chromate(VI) (80 to 90% yields) was a dark orange viscous liquid which failed to crystallize with numerous crystallizing techniques. The amount of silver chloride separated from the reaction residue was 1.01 x 10‘2 mole (1.45 g.) (theory: 0.97 x lo-2 moie or 1.40 g.). At —78° C. the liquid froze to a red—orange glass which melted between -45 and -25° C. Attempts to purify the product by molecular distillation at 10"3 mm. of Hg and 80° C. and by column chromatography were unsuccessful. The product was purified to some extent by dissolving the viscous liquid in fresh methylene chloride, filtering, and evaporating the sol— vent. Trace amounts of solvent were removed at room tempera- ture'and a pressure of 10"3 mm. of Hg. Because of this crude purification technique, fresh product was prepared before each physical measurement to insure that decomposition would be at a minimum. Analytical data of bis(diphenylfluorosilyl)chromate- (VI) are given in Table V (p. 57) and the spectrum is shown in Figures 20 and 21 (p. 120 and 121). up ~a 54 46 B. With Silver Dichromate. A solution of 4.85 x 10"3 mole (1.15 g.) of diphenylchlorofluorosilane in methylene chloride was added toa three-fold excess of silver dichromate (1.15 X 10-2 mole or 5.00 g.). A reaction time of four hours for this synthesis was established by following the growth of the chromyl absorption in the visible spectrum. The reaction did not take place when pentane was used as the solvent. Characterization of the product revealed that diphenylchloro- fluorosilane and silver dichromate also yielded bis(diphenyl— fluorosilyl)chromate(VI). Crude yields of greater than ninety percent were obtained. Analytical data are shown in Table V (p. 57). III. Reactions of Diphenyldifluorosilane. A. With Silver Chromate. Pure diphenyldifluorosilane (0.105 mole or 23.0 g.) was added dropwise to twenty—five grams of silver chromate in a one hundred milliliter flask. The heterogenous mixture was stirred at room temperature for fifteen hours after which time no color change was observed. The mixture was then heated to 1000 C. for another two hours and an infrared spectrum of the liquid phase was identical to that of diphenyldifluorosilane. Three hundred milliliters of methylene chloride were added to this mixture. The components were refluxed for fifteen hours and again the lack of color and the infrared spectrum revealed that a reaction had not occurred. 47 B. With Silver Dichromate. Similar preparations were attempted using silver dichromate. In both instances, with and without a solvent, the reaction failed to take place. The product resulting from the reaction of diphenylchlorofluoro— silane and the unreactivity of diphenyldifluorosilane shows clearly that the silicon-fluorine bond in organochlorofluoro- silanes is quite stable under the conditions employed during this study. IV. Reactions of Diphenyldichlorosilane. A. With Silver Chromate. The preparation was performed using the general method described on p.443 . A solution of 3.96 x 10"3 mole (1.00 g.) of diphenyldichlorosilane in methy- lene chloride was added to a slurry of a ten-fold excess of silver chromate (3.98 x 10"2 mole or 13.2 g.). The mixture was stirred at room temperature for two hours followed by isolation of the product in the manner described previously. The product was a deep orange gum which was soluble in most organic solvents. The silver chloride obtained from the reaction was one hundred three percent of the calculated amount (7.92 x 10'3 mole) which indicated that both chloride groups of the diphenyldichlorosilane had been removed during the reaction. Attempts to crystallize this oil were unsuccessful. The 011 would not distil at 10'3 mm. of Hg and 50° C. and column chromatography proved unsatisfactory. The methods for preparing 48 this material for physical measurements varied with the measurement being made. For qualitative elemental analysis trace quantities of solvent were removed from the gum at pressures of 10'3 mm. of Hg and 50° C. and then the material was stored for at least two days in a vacuum oven at 60° C. This treatment resulted in a brittle black glass which was shown to be amorphous by an x—ray powder pattern. The spectra were analyzed immediately after the reaction had been completed and the product separated. This procedure was carried out in order to minimize the amount of decomposed product present. Specific syntheses were made before each molecular weight determination and cryoscopic measurements immediately followed each synthesis. After the reaction solvent had been removed by evaporation, the product was washed with approximately fifty milliliters of the very pure cryoscopic solvent (benzene) and filtered. The bulk of this benzene was removed by rotary evaporation and trace amounts at 50° C. and 10"3 mm. of Hg. Analytical data are given in Table v (p. 57). This characteri- zation indicated that this compound was the cyclic dimer of 0 Ph I Cr —— 0 - Si -— 0 . The compound will therefore II | X 0 Ph be designated as cyclo-bis[diphenylsilexanylchromate(VI)]. 49 Other synthetic techniques were employed in an attempt to prepare Species larger than the dimer. Silver chromate. (13.2 g.) was stirred vigorously in one liter of methylene chloride under anhydrous conditions. Diphenyldichlorosilane (1.00 g.) in two hundred milliliters of solvent was added to the chromate slurry by means of a Hershberg dropping funnel (111). The reaction was stopped at the end of this addition period (forty hours). A mixture of twenty-five grams of silver chromate, 1.2 grams of diphenyldichlorosilane and ten milliliters of methylene chloride were sealed in a thick walled glass tube and heated to 100° C. for seventy hours. In both cases the isolated product was the red-orange viscous liquid observed previously. B. With Silver Dichromate. Two products resulted from the reaction of a large excess (4.9 x 10'2 mole or 21.0 g.) of silver dichromate with 3.96 x 10-3 mole (1,00 3,) of diphenyldichlorosilane. The reactants, in methylene chloride, were stirred at room temperature for twenty hours and separa— tion followed the general procedure. Bright orange crystals formed as the solvent was removed. After the solution had been reduced to a minimal amount (10 ml.) the crystals were removed by filtration, washed with very dry, cold carbon tetrachloride and dried in an Abderhalden apparatus (112). .0.— RN v.7.“ DIG 50 A deep red—orange viscous liquid was obtained by removal of the solvent from the filtrate fraction. Attempts to purify the crystalline fraction by recrystallization were unsuc— cessful for regardless of the purity of the solvent or the concentration of the solution, recrystallization did not occur and the crystals were not reformed with evaporation of the solvent. In all attempts the residue was a red-orange oil. For molecular weight determinations the crystals were washed with cold benzene instead of carbon tetrachloride. Yields of from ten to twenty percent of the crystalline organo- siloxanylchromate were obtained while separation of silver chloride gave 8.16 x 10’3 mole (1.17 g.) (theory: 7.92 x 10"3 mole or 1.14 g.). Melting point determinations of the orange crystals showed a color change between 93 and 97° C., indicating decomposition. The thermal properties of this material are dis- cussed in greater detail in relation to the analysis of its thermogram (p. 147). Purification treatments of the viscous oil were similar to those used for cyclo—bis[diphenylsiloxanyl— chromate(VI)]. The liquid product freezes to a glass at -78° C. and melts over a wide range between —50 and 0° C. Elemental analyses established the basic structure of both 0 Ph products as gr -—— 0 -—- Si ——- 0 . The molecular II I 0 Ph IM -4 2". L4) I." 51 weight and the nuclear magnetic spectrum of the crystalline material indicated that it was the cyclic trimer or cyclo— tris[diphenylsiloxanylchromate(VI)]. The analytical data of both reaction products are listed in Table V. V. Reaction 0f 3,3,3-trifluoropropylmethyldichlorosilane. A. With Silver Chromate. The reaction of 1.19 x 10"2 mole (2.51 g.) of 3,3,3-trifluoropropylmethyldichlorosilane and a two-fold excess of silver chromate (2.4 x 10'3 mole or 8.0 g.) in methylene chloride occurred over a period of fifteen hours. The reaction mixture was transferred to a dry box for filtration, followed by removal of the solvent by means of rotary evaporation. Yields of from ninety to ninety-two percent of a very dark orange gum were obtained. The silver chloride recovered from the reaction (2.39 x 10"2 mole or 3.43 g.) agreed quite well with the theoretical value of 2.38 x 10'2 mole (3.42 8.). Purification of the crude product was accomplished by molecular distillation at 10'3 mm. of Hg at 50° C. Because of the photolytic instability of the viscous material, separate distillations were performed before each measurement. The complete distillation apparatus was sealed off and transferred to a dry box for the preparation of solu- tions for spectrophotometric examination. For molecular weight determinations and elemental analyses the distillation appara- tus was dismantled in the vicinity of an analytical balance and an, h‘q, v 52 a swab of the viscous material was removed from the collection tube with a nickel spatula. The Spatula and liquid were weighed and immersed in the appropriate solvent as quickly as possible. Pure microanalytical samples were prepared by dis— tilling into a narrow collection tube which then was sealed at the reduced pressure. The orange oil, like other viscous materials prepared during this study, froze at Dry Ice tempera- ture and softened over a very wide temperature range. The analytical and molecular weight data (Table X) suggests a CHZCHZCF3 0 cyclic dimer of the basic - -Si -—— 0 ——— gr -—— 0 I ll X CH3 0 . unit or cyclo-bis[3,3,3-trifluoropropylmethylsiloxanyl- chromate(VI)]. B. With Silver Dichromate. A ten—fold excess of silver . -2 - dichromate (4.73 x 10 mole or 20.4 g.) and 4.74 x 10 3 mole (1.0 g.) of 3,3,3-trifluoropropy1methyldichlorosilane were stirred for twelve hours in methylene chloride. A distillable red-orange gum was produced in yields of ninety percent. Separation and purification techniques were similar to those used in the previous reaction. Ninety-eight percent of the calculated amount of silver chloride was obtained from the excess silver dichromate. The analytical date (Table)() 53 indicated a basic unit similar to the previous product, 149. CH2CH2CF3 fl -1- Si -—— 0 -—— Cr -—— 0 , and the molecular I N x CH3 0 _ structure is believed to be the cyclic trimer or cyclo- tris[3,3,3-trifluoropropylmethylsiloxanylchromate(VI)]. An unusual phenomenon was observed during the purifica- tion of this material. On several occasions the black tar-like residue left in the distillation apparatus ignited explosively when it was scraped with a nickel Spatula. The residue from this violent reaction was a light green powder indicating the presence of a chromium(III) compound. Similar explosions were not observed with residues from the silver chromate 3,3,3-trifluoropropylmethyldichlorosilane reaction. The or— ganic substituents in the residue are possibly oxidized by chromium(VI) in a manner similar to the explosive decomposition of bis(trimethylsilyl)chromate(VI); however, neither the reaction nor its products were investigated further. VI. Reaction of 3,3,3-trifluoropropy1methy1silanediol with Chromyl Chloride. A solution of 5.08 x 10’3 mole (0 79 g.) of chromyl chloride in five hundred milliliters of methylene chloride was stirred magnetically in a three-necked one liter round bottom flask while a stream of dry nitrogen was passed over the solution and through a reflux condenser connected to 54 the reaction flask. A slurry of 9.34 x 10'3 mole (1.74 g.) of 3,3,3-trifluoropropylmethylsilanediol in one hundred milliliters of methylene chloride was added to this solution over a period of thirty minutes. The reaction was shown to be complete in one hour by following the disappearance of the "free" hydroxyl stretching frequency at 2.75 p. Excess chromyl chloride was destroyed by addition of sodium bicar— bonate to the reaction mixture until there was no longer any effervescence. The reaction mixture was filtered and the filtrate condensed with a vacuum rotary evaporator. The residue from the condensation was a bright orange liquid which yielded orange crystals from a concentrated methylene chloride solution at —78° C. The solvent was removed by decantation and the crystals melted slightly above Dry Ice temperatures. The extremely low melting point made the possibility of filtration difficult; therefore, purification was accomplished by crystallization at -78°(L; deCantation of the excess solvent followed by removal of trace amounts of the solvents at 50° C. and 10"3 mm. of Hg, The analytical data, molecular weight, and nuclear magnetic resonance spec— trum indicated that this substance was cyclo—bis [di-3,3,3-tri- fluoropropyldimethyldisiloxanylchromate(VI)]. 55 Characterization of Reaction Products. I. Reactions of Triphenylchlorosilane. A. With Silver Chromate. The reaction of silver chromate with triphenylchlorosilane gave high yields of bis(triphenylsilyl)chromate(VI). This compound has been prepared previously by other synthetic means (5, 10). The chromium analyses (Table V) agreed well with the theoretical values. Both the x-ray powder photograph and the vibrational spectrum in the sodium chloride region were identical to the results reported by Hare (5). The spectrum between 15-29 m (Table VII, p. 63 ) showed absorptions at 16.5, 19.5, and 22.4 p which Smith (113) has assigned to in-the-plane ring bending, out-of-plane ring bending, and the phenyl—silicon antisymmetric stretching mode. The remaining peak at 21.2 u will be tentatively assigned to the Cr—0-Si symmetric stretch— ing mode, and the absorption found between 11 and 12 p to the Cr-O-Si asymmetric stretching mode (a). The absorption between 20.6 and 21.2 p was seen only with compounds containing the Cr-O-Si linkage. The ratio of the intensities of the out-of- plane ring bending to the Si-O-Cr absorption for bis(tri- phenylsilyl)chromate(VI) and cyclo—bis[tetraphenyldisiloxanyl- chromate(VI)] were 17 and 5 respectively, which compares (a). These Cr—O-Si vibration assignments were tentatively made with particular reference to their relative intensities, order of occurence, and their Similarity to Si—O-Si absorp- tion frequencies (113). 56 favorably to calculated values of 3 and l. The cesium iodide spectrum of cyclo-bis[tetraphenyldisiloxanyl- chromate(VI)] is shown in Figure 19 (p. 118). B. With Silver Dichromate. It was predicted that the reaction of silver dichromate and triphenylchlorosilane would yield bis(triphenylsilyl)dichromate(VI). However, extensive evaluations established conclusively that the orange crystalline product was bis(triphenylsilyl)chromate(VI). The analytical data shown in Table V agree well with the chromate structure. The sodium chloride infrared spectrum and the x-ray powder photograph are identical to the results found by Hare (5) for bis(triphenylsilyl)chromate(VI). Exami- nation of the spectrum of silver dichromate between 200 and 560 mp (Figure 5, p. 104) shows a peak with its maximum at 343 mp which has no counterpart in the Spectrum of silver chromate. This absorption results from the Cr-O-Cr linkage (134) and was not observed with the silver dichromate tri- phenylchlorosilane reaction product. It has also been shown by Stammreich (114) that the Cr-O-Cr linkage has a weak sym- metric vibration at 17.7 p and a strong asymmetric vibration at 13.1 p. These absorptions were not observed in the spectrum of the reaction product. Although attempts to prepare silyldichromates failed during this study, it is felt from the results of this investi— gation that their preparation is possible. This topic is dis- cussed in more detail in the section concerning recommendations for future work. Molecular calc. found calc. found calc. found calc. found calc. found calc. found Weight calc. found Compound (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) ELEMENTAL ANALYSES OF PHENYL—SILOXYCHROMATES from 8.19 8.24 8.23 bis(triphenyl- silyl)chromate— (VI) from A82Cr04 Ag20r207 AgZCr04 Ag20r207 8.85 9.08 8.89 68.11 67.10 67.02 5.60 5.36 57 TABLE V bis(diphenyl- fluorosilyl)- chromate(VI) from 10. 9 10. 10. 10. 11. 10. 55. 51. 52. 52. .46 53 3 4. 03 .92 00 13 83 01 91 58 96 13 85 .89 22 4.38 4.28 4.40 7.33 6.54 8.28 12.34 14.34 from 10.03 10.09 9.84 9.87 cyclo-bis di- phenylsiloxanyl- chromate(VI) from Ag2CrO4 Ag2Cr207 17. 17. 17. 17. .39 17 9. 9. 9. 48. 46. 48. 45 77 46 55 38 54 40 3O 91 1 .38 37 4:33 597 541 570 574 from 17.45 17.23 17.27 48.30 47.93 597 463 453 470 cyclo-tris— [diphenyl- siloxanyl- chromate(VI)] 17.45 17.32 17.34 17.30 .38 .26 9 9.47 9 9.23 48.30 47.54 47.31 3:29 895 902 858 835 865 58 II. Reactions of Diphenlehlorofluorosilane with Silver Chromate and Silver Dichromate. The deep orange liquid re— sulting from the reactions of diphenylchlorofluorosilane with silver chromate and silver dichromate had a silicon-chromium ratio of 2.02. The elemental analyses (Table V) of the re- maining structural elements agreed fairly well with the calculated values for bis(diphenylfluorosilyl)chromate(VI). The infrared spectrum (Figures 20 and 21, p.119 and 120) (Tables VI and VII) showed the normal silicon-phenyl absorp— tions. The presence of chromium in the molecule is confirmed by the unresolved Cr=0 stretching absorption at 10.15 p and the two Si-O-Cr vibrations at 11.17 p (5) and 20.6 p. The shoulder at 11.40 p corresponds to the Silicon-fluorine asymmetric stretching frequency (115). The weaker silicon- fluorine asymmetric deformation mode-(115) appears at 17.4 p in the cesium iodide region. The spectrum of bis(diphenyl- fluorosily1)chromate (VI) from 240 to 600 mp (Figure 6, p. 105) shows one chromyl absorption at 475 mp with a molar extinction coefficient of 1.93 x 103. The intense phenyl absorption at 260 mp masked the second chromyl absorption. This second chromyl absorption which occurred between 240 and 300 mp did not appear in any of the siloxychromates containing phenyl groups. The presence of the ultraviolet chromyl absorption could possibly account for poor resolution of phenyl peaks. The deep orange viscous liquid was quite sensitive to light and 59 decomposed in direct sunlight. The material was soluble in aliphatic and aromatic hydrocarbons, alkylhalides, ether, acetone, dioxane, tetrahydrofuran, carbon disulfide, and pyridine; it was insoluble in water, formic acid, and mineral acids. The oil was soluble but decomposed in simple alcohols and acetic acid. When placed in water the liquid formed small droplets at the bottom of the container. These droplets changed to a light orange solid in approximately twelve hours after the solution had been made slightly acidic. This solid was identified as cyclo—bis [tetraphenyl- disiloxanylchromate(VI)] by means of its infrared spectrum, x—ray pattern, melting point, and chromium analysis. It is believed that hydrolysis of the fluorine-containing silyl- chromate occurred in the following manner: 60 Ph Ph Ph |‘¢;\ + l H+ | Ph — Si ‘ Ph — Si—0( Ph—Si—OH I I H I 0 If + 0 + 0 l H30 | -2H 0 - Cr—o -—> o — Cr—o ———> 0==Cr—0 I -2HF I I 0 0 + 0 I; \ + I H I Ph — Si\ 7 Ph -— Si—Cfil Ph—Si—OH Ph Ph Ph 0 ll Cr / \ Ph 0 H 0 Ph 2\Si’ 0 \Si/ 2 I l 0 0 SI l i 0 81 Ph/ \0 II ’0/ \th \‘Cr II 0 This reaction scheme resembles the mechanism proposed for the formation of cyclo-bis[tetraphenyldisiloxanylchromate(VI)] from chromyl chloride and diphenylsilanediol (5). 61 cm Amv 3> H «we see a.So ma.w Hm.m om.b No.5 mN.m om.m ansv am Ame _Au>vopmsoaso IazsmonHmstona IapumfiaquHomo opoa ofia Imwuopoassso gmlwm neuouum oaupoaasns amIoIam msaaammssv soeospn ouam NNHAHmIAAV noamHam Glam en me.m en «w.m Hus mas was has mm.m en oo.m ASen mm.m om.a a sm.m A3m mm.m . a mm s A Vn ve.m sea as.» a oo.m a no.» sea Has a an om.m A vs mm.m has mm.m sees mm.m Hus was Ase H. has a ”ass 4So 1.So eSo reeoaaom masseuse consume Iona sosm _AH>vopesouno AH>vopssoamo Iaassxoaama>soss IHmHHmouoan Iapamasuososo -Hseoseapvnas .56: A: cause muessoesoaxoqsmiqszmme so H> amdaht codenamom Imp mean amuse noumuum «ulna opoa was .Imoaospm mean 010 secs usesououum ouuuosshm mic woos msasowomum eupmaonm MID soaospm moi-Hm 200%“: usosswfimmd V h'.. lime-AI.» .l m! 1' A I K'.‘ I It! In a .c I '0!- i a Ice.IIIlII.|InI in!!! IllvUI-DII0\IIJIIII.I 62 m> mm m> H N man>vopmsonso IHhsmonwmahsmsn IfionHvaoHOho Ase one o .e .moe adv .masouw Hmmsm sass seen mamsflm a on maso>oa esp .SOOHHam seam on» so anemone one masomw assess 039 some mxmoa 03» one“ mo>~omou scpmo soHpmsomnm Hmsosalsoofiaem : m.m was ADV mv.va hm.ma om.m~ nH.HH bN.OH HH.es Ase no one 4So m ov.vH m> mo.HH m mN.oH m $H.OH esoo ECNHUNM< scam m N¢.¢H m> wo.HH am GN.OH m- ms.os H. has” vsoo schemes sous HAH>voumsosso lazsmNoHfimahsosm IsoHnahIoHomo m oe.ea mm ov.~a Asses 8H.HH s nH.oH HM any. vsoo as>voamaosso IaazafimososHm Isasosoeovmas A.usoov H> sofiu ImEHOHSc mafia ossan Hus aeoInso .Hseoss sows . .Imasomop mews ocean Hus Imoueso .Hseoss seas Imsnowom usfin ocean HNH Include .Hhsmnm sonouum was oasaoaasna MIam o nepomum A cm oesuoaamns amnouso mononum NH owuuossmmm oumo souonvm NH ofispoaammm onso 4¢G0>Hom .mom uneaswumm< Smash. .nopasonmIsm “mamsmIQm “pmomnIn ”ssfiposIs mwsoupmlm ”msosum zuo>Im> mammals Amy m.ma m.mH m.HH IHH n.0H H.os any name unsound TOGO Ihomom bH ma ma wu ma NH HH Sana 63 3 m.HN B msHN. m m.om E b.om m> ®.mH m @.mH .somasosmlmm memosnln masfiposls mwsoapmlm mwsomum mso>Im> a.m Hm Assam e.e~ E.m.om 3 m.om Ame .m mama m> m.mH a v.8s a m.ms 5 e.ms a e.ms A33 s.ca Hg Any. A Any H Amy. H Amv a no «mo vsoo eSo someones 1.9834 son“. scum HAH>V Ioussosso AH>V Iaasexoafim Ioumsossoamsm AH>vopmaouso Iazeosmepa IonHmHmsons lamammonous Imfiauloaomo IapHmfinIoHomo IHmsosaapvan ooOEMwSH IsouonsmpOumnme 3 H.Nm a e.mm was-rasnaaee saIam woos Amv wswsouompm Own» a w.om a m.sm are unseen amuouso mswpson Amv . - mean mamas n m.ms n m.ms mas mousse ssIam Amy newsman can» 5.8 m.os mas Inseam emIoIam soap Issuomom can» Iossmmm hIfim msfipson . mean weeds 5 s.cs .Aeva m.es was Immense snIam H Ase In .ASV «no «mo encasom HAH>voumsonao Ismemaosauae AH>Voaa Iamsosmasuoufi IsonaoAHmHmm ImfinIoHomo Iamsonswnpvmfin .mcm psossmumm< A: mNInHv mHBNOJHm|JMZHEQ ho ¢fl80flhm Gmm¢mh2H Hg amass? nhm .9 .moV Any ”xsoBIB new ¢.NN mm mm h.cN HN n.5H Io.bH om ma N.®H ma .Asv new» Insomn< mono Imomom Mama 64 III. Reactions of Diphenyldichlorosilane. A. With Silver Chromate. The reaction of silver chromate with diphenyldichlorosilane yielded a deep orange viscous liquid. The chemical analyses (Table V) gave a silicon- chromium ratio of 1.00 and the carbon hydrogen analyses were in good agreement with the calculated percentages for Ph 3 Si-——- 0 -—- Cr -—- 0 I H x Ph 0 '— Cryoscopic molecular weight determination of 574, 541, and 570 corresponded fairly well to the calculated value for the dimer, 597. The infrared spectrum (Figures 22 and 23, p. 121 and 122) (Tables VI and VII) exhibited the normal silicon-phenyl absorption frequencies in both the sodium chloride and the cesium iodide regions. The chromyl absorp- tions were present at 10.12 and 10.26 p. A broad and strong Cr-O-Si asymmetric stretching vibration appeared at 11.08 p and the weaker symmetric mode at 20.9 p. The absence of a Si-O-Si absorption at either 9.2-9.8 p (116) or 17.0-17.5 p (113) tends to substantiate the A-B-C structure type. Termination of the dimeric unit could be accomplished by cyclization or end-blocking. In characterizing a linear siloxyanylchromate as being end-blocked with hydroxyl groups. Rare (5) rested the full burden of proof for the existence 65 of these two terminating ligands upon the free hydroxyl stretching frequency at 2.7 p. Examination of the spectrum of bis(hexaphenyl-S-hydroxotrisiloxyanyl)chromate(VI) re- veals a minute shoulder at approximately 2.7 p. This absorp- tion could be traced to background noise, the presence of small amounts of starting material (diphenylsilanediol) or to a solvent containing water. The fact that the diphenyldichlorosilane-silver chromate reaction product was a liquid at room temperature and becomes a glass—like material at -78° C. is highly indicative that the structure is not cyclic. A cyclic (therefore symmetrical) molecule having such a high molecular weight would be ex- pected to be a solid under normal conditions. This conjecture, however, is strictly intuitive and is far from being con- clusive. The Spectrum of the dimer gave an extremely weak shoulder at 2.95 p which in all honesty cannot be assigned to the Si-OH stretching mode for the reasons stated pre- viously. The viscous liquid hydrolyzed to cyclo-bis[tetra— phenyldisiloxanylchromate(VI)] in a slightly acidic solu- tion. Both proposed structures could lead to this product. The hydrolytic cleavage of the cyclic compound would yield the hydroxyl-terminated form. Hydrolysis of this linear structure would give chromic acid and bis(diphenylhydroxo- silyl)chromate(VI) which in turn dehydrates to cyclo-bis- [tetraphenyldiSiloxanylchromate(VI)] . Examination of the bond angles and lengths show that the cyclic structure is 66 feasible (5) and that little or no intramolecular stress would be developed. Hare (5) produced several trialkylsilylchromates and assigned the Cr-O-Si stretching vibration to the region be- tween ll and 12 p. It can readily be seen from the tabula- tion below that all but one of these linear silylchromates absorb above 11.4 p. TABLE VIII THE CHROMIUM-OXYGEN-SILICON INFRARED ABSORPTION FREQUENCIES OF SEVERAL ORGANOSILOXY CHROMATES Compound Ref. Si-O-Cr Stretch p l (Me3Si)2Cr04 5 12.1 vs 2 (Ph3Si)2Cr04 5 11.4 vs 3 (p—tol*Ph2Si)2Cr04 5 11.5 vs 4 (thfh'tSi) 2Cr04 5 11 . 4 vs 5 (Ch3Si)2CrO4 5 11.6 vs 6 (PheSiaOzOH)2Cr04 5 11.5 s broad 7 (thFSi)2Cr04 11.2 0 vs 8 [(Ph4S120)Cr04]2 5 11.0 vs 9 (PhZSiCr04)3 11.1 vs 10 (thSiCrO4)2 11.1 vs rp-tol = p-tolyl; ch = cyclohexyl 67 The Cr-O—Si absorption of bis(diphenylfluorosilyl)- chromate(VI) at wavelengths lower than the other linear sil- oxychromates is easily understood for it is well known that fluorine increases the stretching frequencies for adjacent groups (113). The Cr-O-Si stretching vibration of the known cyclic siloxanylchromates (Nos. 8 and 9) appear at lower wavelengths (11.0 to 11.1 p). Several extensive and conclu- sive infrared studies dealing with the Si-O-Si asymmetric stretching frequencies of linear and cyclic siloxanes(ll7, 118, 119) have shown that the cyclic siloxanes absorb at lower wavelengths than linear siloxanes. It is therefore possible to draw a parallel between these results and the Cr-O-Si vibrations shown above. This would suggest that the dimer in question could well be cyclic for the Cr-0-Si asym- metric stretching vibration occurs at 11.1 p. It should also be noted at this point that the above argument would favor Hare's characterization of bis(hexaphenyl-G-hydroxotrisilox- anyl)chromate'(5VI)i as being end-blocked for this substance shows an asymmetric Cr—O-Si absorption at 11.5 p. The Si-O-Si asymmetric stretching mode of open-chain siloxanes containing more than one Si-O-Si group resolved into separate peaks for each Si-O-Si linkage present in the molecule while cyclic siloxanes yield only one absorption maximum regardless of the number of siloxy-linkages present. This phenomenon has been observed with linear organosiloxanes with as many as five k}? 68 linkages (117, 119). The Si-O-Si symmetric vibration at higher wavelengths also shows this characteristic (113). Hare (5) obtained a bright red viscous oil when he treated dicyclohexylsilanediol with chromium(VI) oxide which he characterized as hydroxo{penta[dicyclohexylsiloxanyl— chromate(VI)flIdicyclohexylsilanol, or Ch 0 Ch 0 Ch 0 Ch 0 Ch I Ch I II I I l I I I I I l HO—fii-0-Cr-0—Si-0-€r—0-Si-0-Cr—0-Si—0-Cr-0-Si—0-Cr-0-fii-0H. l I I I I I I I Ch 0 Ch 0 Ch 0 Ch 0 Ch 0 Ch The infrared spectrum between 11 and 12 p gave two strong absorption maxima, two definite shoulders, and with a great deal of imagination the fifth Cr-O-Si asymmetric stretching mode can be discerned. This evidence although far from being conclusive indicates very strongly that the Cr-O-Si stretching frequencies exhibits the same resolution phenomenon shown by linear siloxanes. Both Cr-O—Si stretching frequencies of the diphenylsiloxanylchromate(VI) dimer were unresolved, lending further evidence to the possibility of a cyclic structure. The proton magnetic resonance Spectrum of this oil was determined in carbon tetrachloride using a varian A-60 proton magnetic resonance spectrometer. With a linear hydroxyl end- Vblocked material three absorption peaks would be expected, one rather intense phenyl hydrogen peak and two weaker hydroxyl 69 hydrogen absorptions. The hydroxyl absorptions would be split, Ph 0 l I for in H0 -n— Si -—- 0 ——— Cr -sr- 0H one hydrogen I II 2 Ph 0 would be influenced by silicon and the other by chromium. The Si-OH hydrogen generally absorbs in the vicinity of 80 c.p.s. (11). The liquid dimer showed only one intense peak at 460 c.p.s. which corresponds well with the spectrum of cyclo-tris- Idiphenylsiloxanylchromate(VI)] (Figure l, p. 73 ) and the phenyl hydrogen absorption at 460 c.p.s. shown by cyclo-bis- [tetraphenyldisiloxanylchromate(VI)] (11). The possibility of a linear structure with a.terminating linkage other than hydroxyl groups was not investigated. It is fairly obvious from the bias emanating from the above dis- cussion that the author favors the cyclic structure; however, the assignment of a cyclic structure to the diphenylsiloxanyl- chromate(VI) dimer is made with considerable reservation. This deep orange liquid becomes exceedingly more viscous and eventually turned to a black brittle solid on standing in a vacuum oven at 60° C. The x-ray powder pattern of the brit— tle glass-like material had two wide bands at 6.7 and 19.3 2 indicating a very low order of crystallinity. The liquid was soluble in numerous aliphatic and aromatic hydrocarbons, alkyl halides, and carbon disulfide; it decomposed in simple 70 alcohols. The material, like most siloxychromates, decomposed in the presence of light. It also decomposed slowly in water, and as mentioned previously, it changed fairly rapidly to cyclo-bisItetraphenyldisiloxanylchromate(VI)] in the presence of a weak acid. The spectrum between 240 and 600 mp of bis[diphenylsilox— anylchromate(VI)] (Figure 7, p. 106) exhibited two absorption maxima, the chromyl peaks at 475 mp and the unresolved phenyl absorption at 246 mp. The molar extinction coefficient for the chromyl absorption was 8.1 x 103. B. With Silver Dichromate. The bright orange crystal— line solid resulting from the reaction of diphenyldichloro— silane with silver dichromate had a silicon-chromium ratio of 1.00, and the carbon-hydrogen analyses (Table V) also agreed with the calculated values for a basic diphenylsiloxanyl- chromate(VI) structure. The infrared spectrum from 2-29 p (Figures 24 and 25, p. 123 and 124) (Tables VI and VII) showed the normal vibrations associated with silicon-phenyl linkages. The chromyl absorptions were resolved at 10.11 and 10.27 p, and the Cr-O-Si asymmetric stretching mode appeared at 11.13 p. The Cr-O-Si symmetric stretching mode at 20.9 p was considerably stronger than similar absorptions seen with other siloxanyl and silylchromates. The two Si-O-Si stretch- ing vibrations which normally appear at 9.3-9.8 p and l7.0-l7.5 p were not present. The analytical and infrared data 71 therefore suggest that the compound is constructed from the Ph 0 l I basic . ;Sfi -—— () -—— Cr -—— O Tunit. Molecular H x Ph 0 '— weight determination (Table V) indicates that the crystals are a trimer of this unit. The proton magnetic resonance spectrum of these crystals (Figure l, p. 73 ) was obtained in a mixed carbon tetrachloride-carbon tetrabromide solvent, a rather- intense absorption maximum at 463 c.p.s. was the only peak observed. This absorption was quite similar in appearance to the 460 c.p.s. phenyl-hydrogen peak of cyclo-bis[tetra- phenyldisiloxanylchromate(VI)] (11). The infrared spectrum also indicated that the compound was cyclic. An Si-OH absorption at 2.7 p was very definitely not present. The Cr—O—Si asymmetric stretching frequency appears at wavelengths lower than those normally shown by linear siloxychromates and both the asymmetric and symmetric Cr-O-Si absorptions were obviously singular in nature. The x-ray powder diffraction pattern of cyclo-tris- [diphenylsiloxanylchromate(VI)] (Table XI) gave rather large values, indicating a large unit cell. An attempt at indexing these interatomic Spacings was unsuccessful; however, it was 72 shown that the crystal was not cubic and a comparison of the spacings with the Hull-Davey charts (120) indicated that it might possibly be hexagonal. The visible and ultraviolet spectrum between 240 and 600 mp (Figure 7, p. 106) showed a chromyl absorption at 490 mp with a molar extinction coefficient of 2.4 x 103. The phenyl absorption occurred at 268 mp. The crystals were soluble in most organic solvents; however, like cyclo-bis[tetraphenyldisiloxanylchromate(VIH the degree of solubility was much less than that of liquid siloxychromates. Once dissolved, cyclo-bis [diphenylsilox- anylchromate(VI)] could not be made to recrystallize. Re- evaporation of the solvent left a red-orange oil. Hydrolysis in an acidic medium resulted first in an orange gum and then cyclo-bis[tetraphenyldisiloxanylchromate(VI)]. It is believed that the crystals were initially changed to cyclo-bis[diphenyl- siloxanylchromate(VI)] and then by hydrolysis and subsequent dehydration to the light orange crystals of cycle-bis[tetra- phenyldisiloxanylchromate(VI)]. The material was light sensitive but to a much less degree than any siloxychromates prepared to date. The thermal proper- ties of cyclo-tris[diphenylsiloxanylchromate(VI)] are covered in the DisCussion'section. The elemental analyses of the liquid product resulting from the reaction of diphenyldichlorosilane with silver Intensity 73 (MW I 500 Figure l. I 450% 400 cycles per second oton magnetic resonance spectrum of cyclo-tris- dinhenvlsiloxanvlchromate(VI)I- 74 TABLE IX INTERATOMIC SPACINGS FOR CYCLO-BIS[DIPHENYLSILOHANYLCHROMATE(VI)I Line H @QQGUIIh-WN #4 H IA 6: P‘IO ha hi hi h‘ h‘ h'IH Id .9 In (n -a a: clue to N H .(Cr K“ radiation, 2.? 2.2909 A) .9Z(Degrees) 6.97 7.35 7.95 8.85 9.53 10.55 10.78 12.35 13.63 14.23 14.90 16.45 17.05 19.28 20.00 20.80 23.10 23.73 24.43 dhk1.(z) 9.42 .95 .28 .45 .92 .25 .13 .35 .86 .66 .45 .05 .91 upsehumczchooo 3.47 3.35 3.23 2.92 2.85 2.77 Relative Intensity 10 QNI-‘nhIhUIOINN H 03¢ <1 (1 ha I4 Id .5 In .s 75 dichromate were in fair agreement with calculated values for 0 Ph II I the -w- fir -—— 0 -—— ii -—- 0 -v- structure (Table V). x 0 Ph — The infrared spectrum between 2 and 29 p (Figures 26 and 27, p. 1.25 and 126) (Tables VI and VII) was identical to that of cycle-bis[diphenylsiloxanylchromate(VI) ]. There was no visible evidence of an Si-OH absorption in the range of 2.7 p, and the Cr-O-Si (11.05 p) absorption appeared at a lower wavelength than linear siloxychromates. The red-orange liquid was soluble in most organics but decomposed to cyclo-bis[tetraphenyldisiloxanylchromate(VI)] in water. The spectrum between 240 and 600 mp of the liquid sub- stance showed absorptions at 475 and 265 mp with molar ex- tinction coefficients quite similar to cyclo-bis[diphenyl- siloxanylchromate(VI)]. So far all the evidence given has been identical to that shown by cyclo-bis[diphenylsiloxanylchromate(VI)]; however, the values obtained during molecular weight determinations this material (Table V) were all somewhat lower than the cal- culated molecular weight for the dimer, 597. It is the belief of this investigator that the liquid was cyclo-bis[diphenylsiloxanylchromate(VI)] and that the low 76 molecular weight values resulted from small amounts of an impurity not removed by the somewhat crude purification techniques. This impurity was quite possibly a by-product which formed during the silver dichromate diphenyldi— chlorosilane reaction but did not appear during the silver chromate synthesis. The mechanisms proposed for these re- actions (cf. p. 161 and 162) indicate that the situation could exist. IV. Reactions of 3,3,3-trifluoropropylmethyldichlorosilane. A. With Silver Chromate. The reaction of 3,3,3—tri- fluoropropy1methyldichlorosilane with silver chromate re- sulted in a distillable deep orange viscous liquid. Ele- mental analyses (Table X) showed a silicon-chromium ratio of 1.00. The carbon—hydrogen determinations were in fair agreement with the calculated values for 3,3,3—trifluoro- propylmethylsiloxanylchromate(VI). The infrared spectrum (Figures 32 and 33, p. 131 and 132) (Table XI) gave the normal vibrational modes associated with the 3,3,3—trif1uoropropyl and methyl-silicon linkages. The chromyl stretching absorptions were very well resolved and occurred at their predicted frequency. The Cr-O—Si asymmetric stretching mode appeared at a slightly lower wavelength (10.9 n) than those shown by phenylsiloxychromates. This shift could be attributed to the fluorines; however, their influence on the 81-O-Cr stretch should be very slight 77 Hmb cam now mmm wvw HNm «mm own man mnm ovm venom .aB .HOS. Nm.nN cm.NN mb.om an.NN $5 $3 ossom .oamo mason mm.m N¢.m hm.N ©N.N om.N @b.N wN.MN mo.mN om.mN mv.wH mv.wH mb.wH mN.mH mu.mH Q; 33 bosom .oHao mv.m~ mm.MH Av.md wv.m~ ch.OH hw.OH mm.o~ 0m.cH vo.- mm.cH awv bosom Nw.MH mm.ou wa.c~ A 6Q \/ .UHNU Hm Nm.NH mm.Nfi mn.NA vv.NH Hm.NA mH.ON HM.ON OM.ON w¢.ON hc.cN H~.ON NN.ON cm.ON cwv Afiv punch .OHsU H0 _An>vmuslounu uaaaaxcasmseasapwsnu luhnonaouosdmwuu um.m.muac_manno~osu ~AH>vousaoanu Iazsaonmemnvoa Iahnouaouoaamwnu um.m.m.mauanososm ~AH>voumlounoamsmonwm Iahaumlahnounouosau usuanm.m.m_msauo~omu vmsonfloo mmeesommoqwzaxoqquwmemsqamommomODAmHmaum.m.m mo wmm>q 3 om.b m Hm.b a mm.b a om.b mHH InvisOOfiHHm mm.» m newumsnom lop oauaoaahm a mo.s 3» HH.> a so.» 3 so.» a so.» was unpaa mmonam as.» s .I Aamnoun a hm.o 3> Na.o a vm.m s mm.o a mm.® mad usvlsooHHflm om.w o woos mew Aav Inoaonam o a» a sv.m a> m¢.m a> me.m ms -.»:maaam o m¢.m n opoa usfiAOHonum ofiuwma 3 me.m 3 me.m a m¢.m was [asmfiaaa o Hv.m v woos unfinopouam .. , oanpoa 9.3 bw.m 3.3 mN.m mad Lahmdpsm mic nm.m m a A vs mm.m Ame n .a mo.m Ans mm.m « Amy nououum an .s mb.N mad molam :oouu: N>.N H H Any H A15 _H Ans .H Aav H Aav Aav V . . eSo no a see «we a eBo «we a eSo «we a eSo .Taao>som H5H>V Ioumaouno . nasaaxosam _AH>V _AH>V Iapamnpoa-lowmaounoaznm nauseounonsm lupahnoun IonHmamnuoa IonHmahnuoa Hpr sofiu .Iouosamuwup Iamnounomosam Iazaonnouosaw losaawmamnwoa osmflfimonoHnoflp Imuomn< um.m.mufic_ uaaaum.m.m. ufluuum.m.m_ usamonaouossm uaanamaasmoum mono manloHomo ImfiHHIOHomo Inflatedomo. Iwuplm.m.m Ioaosamanulm.m.m .mom asoasmMmm< uneven Mama mozsoasoo onseqmm oz< mmez2 < MFF‘ I h 7: I.>hlvi\oII-\I .I alf\filllf.\fllll\sl Ivhl I III! ‘1 82 Aamnonm m mu.m s m>.m a mb.m m mw.m s vu.m mHH IM¢|moo«me om.m ma nonoapm 0 up m> He.m m> mm.m m> mm.m m ww.m m> hm.m and seesaw .ho mm.m ma maa nououpm am mm.m wHH canvas w.m .n uma.m and nasma amnousm um.e vs Amv .I Aahaoua :m NN.m am NN.m am mN.m am mN.m mHH usvlsoouawm om.m ma and nononpm mud can we m> mw.m m> mw.m m> mm.w m ww.w m> mw.w mad Ishmfiasm ho mu.w NH mma nououum mug. causes m> ma.m m> mm.m .m> um.m m om.m m> mm.m was .aamspaa «no Hm.m as new» any Imauomo camp m> om.u m> um.> m> om.b m nm.> m mm.b mHH Ioaamm nonfim om.b oH mNH mononum a no.» a me.s m no.5 a ve.u a oo.e ems ofiupmaaam who em.s m H. Age. a any H. .Asvv a Aav s Ase has vsoo «we a esoo «we a eSo «no a esoo «we a vsoo +¢=m>som .AH>V nauseouno IamsmonHm —AH>V HAH>V Iapaznpoa nauseousoazsm nauseounonnm Iapamnona Ionfimamnpoa IonfimHmnuoa HOac sewn tonedunanu Iamnonnouosam Iaznonnonosam nonmawmahnpos osmafimOHOasofip Ianomnm Im.m.mlwpa lunalm.m.m_ napalm.m.mg IHmQOAnouosHm Iazauoaamnoua ooso ImHanHomo InfianoHoao unannoHomo Iwnutm.m.m nonosamfluulm.m.m .mom uneaswamm< luomom mama mozsonsoo oma-LL£$T - >1-Av-.-Av-a:-.-1- - lion. ti- u. an nuts I nullulu--. 83 Hum .mv Hne scan a om.mH m mH.mH n .m> om.mH emH uasnomme who om.mH mm .3. mo.ma pommflmmmss mo.m~ wN o.NH m> hb.NH mHH clam In.HH MN 3 mv.NH am mv.NH a NV.NA maa nouonum clam mN.N~ am now mason a mm.HH a em.HH a we.HH n .a om.HH a mm.HH mHH. u-mme mouHm s.HH Hm .I Aahnoun m> mH.HH m 0N.HH an wH.HH a .s 5H.HH m> bH.HH mad Isvlsoouafim N.HH on nonohum ADV OHpflum—B HoNH m om.oH an oa.oH m> om.oH m .aama Hmneuuo uo.HH 6H newsman came a mm.oH an om.cH .m cm.oH «H aoaasm ammo m.cH mH noaouam Oahu 5 wo.o~ m> NH.OH m ¢H.ou NH Ioaahms Oluo H.9H ha H H 3.... H.. 3” H. 3. H. c: H. 3 3 «me «me a H.Hou «we a eHoe «we a H.Hoo «we a HVHoe +5ao>Hom .HH>V nauseouno uHaaaonHm _HH>V _HH>V lupfimnpoa Ioaaaounoamsm Impmaounoamsm Iapamnoun IKOHHmamnuoa Ixoawmsmnuoa Heap new» IouoaHMHuu Iamaounouosam Iamnounouosfiw Iosmawmamnuoa osaHHmOHOHaoHp Inhomn< Im.m.mnwpa Ianulm.m.ma Ianulm.m.m_ Ithouqouosam lamauoaaaaona mono ImHn|o~o>o ImHupIoHozo ImHDIOHomo [Huplm.m.m «chosamflhpIMHm.m .mom unmasmwmm< Imogen lawn mnzsomsoo cabana: nz< musesommoq>z II:I.H-..¢. .>II"UQI.QI..‘V§I|~.I~ .‘Qa.« vs...‘ ‘\V§.i\. '..In § 1... 5‘ J! i. I\\h .n‘z I 84 mmon s n.wm N.mm 3 ®.wm mHH.>..xoou OH soHu m.mm a m.mm mmH uaaHonc mmo wanmu m nououum cameos a v.Hm m m.Hm m o.H~ Hoe uaHm Hmuouuo m nououam cameos b.HN m c.HN mHH lama HouHm ue.mH s soHu a o.om a H.om a H.om m.ou a H.cm mmH uaaHoHoo who m.mH e newsman owuu h.wH m m.wH mHH :oaamma HonHm uw.eH .m. soHa m H.wH m H.mH m H.wH H4: 92 -2533 «so «.3 v moHa rmanowmp 0 ma m.hH m m.ba mNH uoaammm no use: m nopouam cameos n.bH a e.mH mHH name HmnouHm uo.sH m H . uses“ a m.mH a mamH a. mHmH a a.nH a a “H. mmH ..uaapHmu «mu mth H H HaHi H HaHH 8. Has. H Has H H15 Has v . .. . - H00 vaoo - .vfioo zommu Nmo. 4¢mo>Hom .HH>V Impasouao IamsmonHw _AH>V ”An>u Iapamnpoa Ioumaounoammm Imumaouaoammm Iapamaonm IonHmaznuoa IonHmHmnaoa aofip neg» .Ionosaanp Iahaougouosam lamaouaouosam nonmafimamnpoa osmHHmouoHnoHp Innomn< Im.m.m|HoH Iwuplm.m.mH Iwupum.m.ma lamaounouoaam Iamnuoaaanonn mono umHnloHomo ImHauloHomo ImHnnoHomo Iauulm.m.m neuosamfinulm.m.m asosswflmm< onemm mnHooH saHmmo mazsomzoo cayenne 92¢ mmez51 s1\ CH3 I l CH3 0 o CH3\| | / CH3 /s1 s1\ CF3CH2CH2 \ o o o / CHZCHZCF3 \cu: r 90 TABLE XII PROTON RESONANCE SPECTRA 3,3,3-trif1uoropropy1- 3,3,3-trif1uoropro- cyclo-bis[di-3,3,3-tri- methyldichlorosilane pylmethylsilanediol fluoropropyldimethyl— disiloxanylchromate(VIH Chemical Shift Chemical Shift Chemical Shift (c.p.s.) I (c.p.s.) I (c.p.s.) I CH3 17 w 24 vw 21 sh 22 vs 27 vs 25 vs 25 w 14-32 at one— half height BCHZ 43 m 56 m 48 m 61 m 50 m 64 m 61 (max.) b 52 m 57 m 72 m 47—74 at one- half height 61 s «CH2 130 w 97 m 133 w 100 w 104 w 137 vw 107 m 139 vw 109 w 141 w 134 b 110 w 143 w 113 w 145 vw 118-148 at one- half height 114 w 147 w 116 w 151 vw 119 w 154 vw 124 w 159 w 162 w 175 vw 176 vw 177 179 w 181 w 91 .ecmHHmOHoHnoHUHmapwEH>Q0HQ0HOSHHHHpum.m.m we Esupoon mosmsomou ofiposwms cowosm psooom Hon moaozo on OOH and com 7 .N mHSMfim H fi Kitsueiul 92 .Hofloocmaema>auosazaosmoaosawfipptm.mHm mo asspoonm mosmsowoa oHuocwme :OaOHm ocooom you meaomo .m esswfim om OOH omH com — 2DRJfl 4S\K)\\\J<<(FAHA)<(\#(&75/ «K47bx2/}€¢<2€ Kitaueiul 93 ”fit-I." -0 r -v.”o"' .HAH>voamaonnoH>smonHmeamagma IHoasaoumonousHsplm.m.mnfipHmHn|oHomo mo sapwooam mosmsomou ofiuoswms :OHOHA .v muswwm csoomm Hon moao>o OCH 4 and CON ‘3 I 1 l i 5-- --_._-}__ ..— o “’ ~— ””‘b-‘ovmww—H m.m. -____ ”g Airsueiul 94 The infrared spectrum of this compound also favored a cyclic structure. The Cr-O-Si asymmetric stretching frequency absorbed at low wavelengths and was unresolved. The symmetric stretching vibration mode was also unresolved. Both Si-O-Si stretching modes were found at wavelengths considerably shorter than those exhibited by linear silox- anes. The symmetric peak was definitely singular in nature, but the area around the asymmetric band was much too congested for any determination of peak multiplicity. Consideration of a cyclic molecule in terms of the bond angles and lengths did not indicate any apparent intramole- cular strain. The spectrum of cyclo-bis[di-3,3,3-trif1uoropropyldi- methyldisiloxanylchromate(VI)] between 220 and 600 mp (Figure 9, p. 108) showed chromyl absorptions at 470 and 240 mp. The respective molar extinction coefficients were 7.0 x 2 10 and 2.9 x 103, This light orange liquid was slightly soluble in water and was apparently stable for short periods of time. It also was photochemically sensitive and would decompose to a brown powder in diffuse light. The solubilities of this substance in organic solutions were similar to that of the cycle-bis and cyclo-tris[3,3,3-trifluoropropylmethylsiloxanylchromate(VI)]. 95 Instrumental Measurements. A. Ultraviolet and Visible Spectra. The ultraviolet and visible spectra of the siloxanylchromates prepared during this investigation were measured in methylene chloride with a Beck- man DK-2 recording spectrophotometer. The spectrum of silver chromate, determined in aqueous ammonia, showed peaks occurr- ing at 274 and 503 mp while water solutions of silver dichro- mate absorbed at 258, 344, and 470 mp. The absorptions be- tween 250-280 mp and 470-500 mp corresponded to transitions characteristic of the chromyl linkage while the peak in the vicinity of 350 mp results from the dichromate bridge. Each siloxanylchromate exhibited one absorption maximum in the ultraviolet region occurring between 240-2701m1and one plateau between 450 and 5OOIM11n the visible region. The absorption peak in the ultraviolet region shown by phenylsiloxanylchromates resulted from the phenyl groups and it masked the chromyl absorption which also occurred at these wavelengths. The ultraviolet chromyl absorption in 3,3,3-tri- fluoropropylmethylsiloxychromates was observed because the organic substituents did not absorb in this region. The ab- sence of phenyl resolution in phenylsiloxanylchromates has been attributed to interaction of the chromium with the phenyl groups (5). This decrease in fine structure could also result from "double-absorption" within a specific range. The chromyl 96 absorptions in the visible range exhibited shifts which were related to the number of atoms within the ring. The cyclic dimer of diphenylsiloxanylchromate(VI) showed a chromyl transition plateau at 460 mp while the cyclic trimer absorbed at 480 mu. Both chromyl absorptions of cyclo—bis and cyclo- tris[3,3,3—trif1uoropropy1methylsiloxanylchromate(VI)] showed corresponding shifts to lower wavelengths (498 to 480 mp and 250 to 243 mu). Cyclo—bis[di-3,3,3-trif1uoropropyldimethyl- disiloxanylchromate(VI)] extended this shift in the visible region even further (465 mp). These shifts suggest that interactions between the chromyl group and some other substi- tuents in the molecule did exist because as the ring size be— came smaller, thereby increasing the proximity of intramolecu- lar substituents to the chromyl linkages, the wavelengths of the absorption became longer. This chromyl interaction most likely takes place with the phenyl (128) and the trifluoropro- pyl groups. These spectral shifts mentioned above are also indica- tive of the cyclic structure of the compounds prepared during this investigation. B. Infrared Spectra. The infrared spectra of the come pounds prepared and used during this investigation were recorded in carbon disulfide, carbon tetrachloride solutions, and in po- tassium bromide pellets. The spectra in the sodium chloride region were obtained with a double-beam Perkins-Elmer Model 21 97 spectrophotometer using both sodium chloride and potassium bromide cavity cells. The spectra in the cesium iodide region were recorded with a double-beam Beckman IR-7 spec— trophotometer and potassium bromide cells. Most of the fine points concerning the infrared spectra of the materials prepared during this study have been dis- cussed at some length in the preceeding section. Because of the qualitative nature of the Cr-O-Si vibra- tional assignments, these assignments should be considered tentative. Rare (5) assigned the 11 to 12 p absorption to the Cr-O-Si stretching vibration using a process of elimina- tion and comparisons of the wavelengths to literature values for transition metal-oxygen-silicon vibrational assignments. His characterization of the vibrational mode was substantiated during this study by similar means. The absorption appearing between 20.6 and 21.4 p was also assigned to Cr-0-Si vibra— tions by means of a process of elimination and by comparisons of the peak intensities shown by bis(triphenylsilyl)chromate(VI) and cyclo-bis[tetraphenyldisiloxanylchromate(VI)] . The struc- ture of these compounds had been previously established (5, 11). The assignment of the absorption modes as stretching vibrations, and more specifically as asymmetric and symmetric stretching vibrations, was performed by comparing the absorption posi- tions and appearances with the previously assigned SiaO-Si asymmetric and symmetric stretching frequencies (113, 115). 98 The cesium iodide spectra of the phenylchlorosilane reactants are recorded in Table XIII (p. 99). The silicon-fluorine asymmetric deformation mode which should be present in the spectrum of diphenylchlorofluoro- silane was masked by the stronger silicon-chlorine asymmetric stretching mode. The shifts to shorter wavelengths by the phenyl-silicon out-of-plane ring bending modes are caused by the presence of fluorine in the molecule (115). C. Proton Magnetic Resonance Spectra. All measurements were performed with a varian A-60 spectrometer employing a sweep width of 500 c.p.s. The solvents used were carbon tetra— chloride, diethylether, methylene chloride, and carbon tetra- chloride-carbon tetrabromide solvent mixture. The standard was tetramethylsilane. D. X-Ray Diffraction Data. The interatomic spacings for silver chromate and silver dichromate were calculated from x-ray diffraction patterns obtained with a Siemens Kristalloflex-4 diffractometer using copper, K, radiation. The atomic spacings for cyclo-tris[diphenylsiloxanylchro— mate(VI)] were obtained from chromium K, powder diffraction film using a North American Philips powder diffraction unit with a Debye-Scherrer camera. 99 noemuem oHnu m.ma 3 «.mm 3 m.om a e.mm 5 «.mm mHH neaasmHuee Hmugm u¢.mu s soHumEHOHop s o.om mHH oHepoaasm stem e ms“ . Ibsen mafia ocean b.cN m3 o.om m> v.mH m b.om m3 mw.ma mHH IHOIHSO Hmlsn Im.mH n nepoupm m> m.mH mHH oHeeeaaHm HouHm m.wH v soflumaHOHop m e.sH mHH oHapmaaame muHm m noponpm o.mH m> o.eH m> m.eH m> H.mH mHH oHnuoassme HouHm um.sH m assume: mafia mamas N.@H 3 ~.ma 3 «.mH a «.mH 3 c.0H mad losulsfi “mush no.mH H H Any H Hay H Hie H sz Any new» Ianomn< esmafimonosam osmafimouosam esaHHmoaoHno osmHHm come .02 IonoHnoahsonan Iapamsozafio twoamsonawp IoaoHconsonnHuu .mem uneasmfimm< uneven Mama 3 3.33 mazfiofim szHisHLSzmHE .3 53.8% HHBHEEH HHHN mam<8 100 E. Vapor Phase Chromatography. Vapor phase chromato— grams were obtained with a Beckman GC-Z recording chromato- graph and a twenty percent silica column. The flash plate was maintained at maximum temperature (2200 C.) A carrier gas of helium was passed through the system at positive pressures and the analyses were performed using samples of approximately two microliters of the non-diluted liquid silane. Analytical Methods. The following methods were found to be satisfactory for silicon and chromium analyses. Carbon, hydrogen, and chlorine analyses were carried out by Spang Microanalytical Laboratory, Ann Arbor, Michigan and by Alfred Bernhardt Microanalytical Laboratory, Mulheim, Germany. Chromium Analysis: The decomposition of thirty to fifty milligram samples took place in platinum crucibles with a flux of approximately ten grams of sodium bicarbonate. The crucibles were heated with a steam bath for several hours. After cooling, the solid residue was dissolved in two hundred milliliters of water and carefully neutralized with sulfuric acid. A slight excess of acid was added and the solution was diluted to four hundred milliliters. After the addition of an electrolyte (ten grams of concentrated potassium chloride), 101 the chromate was titrated with a ferrous sulfate solution which had been previously standardized against potassium dichromate. The titration was followed amperometrically using a rotating platinum electrode and a calomel cathode. A recording ammeter plotted changes in the diffusion current with increased ferrous sulfate concentration and the equiva— lence point was obtained by extrapolation of the post- equilibrium curve to the pre-equilibrium curve. The fluorine substituent in bis(diphenylfluorosilyl)- chromate(VI) appeared to interfere with the determination. Therefore, ten grams of calcium nitrate were added and the neutralized solution was left standing overnight. After filtration, the above procedure was followed. Silicon Analysis: Oxidation of fifty to one hundred milligrams of the unknown sample was accomplished by the addition of ten milliliters of concentrated sulfuric acid to the material in a one hundred and fifty milliliter beaker. The contents were heated to relatively high temperatures for several hours followed by the careful addition of an excess of ammonium peroxydisulfate (50 g.). The solution was heated for approximately three more hours, cooled, and diluted to four hundred milliliters. The insoluble silica was removed by filtration through ashless paper and ignited in platinum crucibles. The silicon dioxide content was determined by weight difference. 102 It was predicted that the fluorine substituents present in several of the siloxanylchromates would interfere in the above analytical procedure. However, for reasons unknown to this investigator the suspected interference did not occur. Fluorine Analysis: Several classical analytical pro- cedures for the determination of fluorine were investigated, however none were found to be satisfactory. Fluorine analyses of diphenyldifluorosilane, diphenylchlorofluorosilane, and 3,3,3-trifluoropropylmethyldichlorosilane were performed by the Spang Microchemical Laboratory. Alfred Bernhardt de— termined the fluorine content of cyclo-bis and cyclo-tris— [3,3,3-trifluoropropy1methylsiloxanylchromate(VIXland bis- (diphenylfluorosilyl)chromate(VI). The fluorine and oxygen content of bis(diphenylfluorosilyl)chromate(VI) were also determined with activation analyses by the Dow Chemical Company. The activation analyses followed procedures des- cribed by Anders (129, 130). Molecular Weight: Cryoscopic molecular weight deter- minations were performed with benzene solutions. Current variations resulting from temperature changes in the area of a thermister were fed through a Wheatstone bridge whose corresponding out-put was received by a Sargent Model XXI Polarograph and the cooling curves of the benzene solutions were automatically plotted on this current-time recorder. 103 Variations in temperature could be measured to 0.0020 C. The freezing point of solutions were obtained by extra- polation of the freezing curve to the cooling curve. The procedure consisted of determining the freezing point of the pure benzene, addition of weighed pellets of the unknown to the solution, dissolution, and then multiple determinations of the resulting lower freezing points. Weighed pellets of hexachlorobenzene were then added to the solution and five freezing curves were recorded. Averages of the above measure- ments were made and the freezing point depressions were expressed in terms of chart divisions, A.D (mm), which corre— lated directly to AlT. The molecular weight of the unknown samples were calculated from the following relationship: LXTé W1 M = M ———-—— 1 2 - ATl W2 M—molecular weight of unknown. MZ-molecular weight of hexachlorobenzene. lel-freezing point depression of the unknown solution. ZSTz-freezing point depression of the hexachlorobenzene solution. Wl-sample weight of the unknown. W2-sample weight of hexachlorobenzene. The above procedure is described in greater detail by Skelcey (131). 104 4. 0L Molar extinction coefficient (KID-O) AgZCr207 ‘ I Ag2Cr04 / I l / 1 ‘~ // J— ‘_—- =_ _ __L _ l . 200 300 400 555 wavelength in millimicrons Figure 5. Ultraviolet and visible spectra of silver chromate and silver dichromate. 24 105 1.0- .75 .50. .25 I J l I 240 300 340 400 500 600 wavelength in millimicrons Figure 6. Ultraviolet and visible spectrum of bis(diphenylfluoro— silyl)chromate(VI). ') Molar extinction coefficient (Xiu 2. 106 ---cyclo-bis[diphenyl- siloxanylchromate(VI)] -——cyclo-tris[dipheny1— siloxanylchromate(VI)] Figure 7. l l 340 ‘6‘4 0» ’55?) wavelength in millimicrons Ultraviolet and visible spectra of cyclo—bis[diphenyl— siloxanylchromate(VI)] and cyclo-tris[diphenylsiloxanyl- chromate(VI)]. ' L 300 600 L‘pllv \lluL-v \oVUJ—A‘la CA k-AIIU DLUAI JVIU-LuL 107 10.0» cyclo-bis[3,3,3—trif1uoropropyl- methylsiloxanylchromate(VI)] - - - cyclo-tris[3,3,3-trif1uoropropyl- methylsiloxanylchromate(VI)] - - _, ,_. " ’ I ‘\ \ \ 2.5- \\\ \\ l 1 . 2‘ o 300 3'40 400 160 60 wavelength in millimicrons Figure 8. Ultraviolet and visible spectra of cyclo-bis[3,3,3-tri- fluoropropylmethylsiloxanylchromate(VI)] and cyclo-tris [3,3,3-trifluoropropylmethylsiloxanylchromate(VI)]. Molar extinction coefficient (X10 2.00 1.00 108 4 1 I - l 220 300 340 420 500 600 wavelength in millimicrons Figure 9. Ultraviolet and visible spectrum of cyclo-bis[di-3,3,3- trifluoropropyldimethy1disiloxanylchromate(VI)]. 109 vH .sofimeu oofiHoHao ssfinom map a“ osmaemoaoHsoaxsenaflnp mo sapwooam poamamsu .OH madman mcoaofls Ce cumseao>m3 ma NH HH OH m w b o m v m - d d‘ d - u d .- I ‘ .1 5\ 2i notsstmsuaxl iueoaed OOH nueuuvlviq-ssmwa-usfi.svlv J-ueuuv..~..-A~ 110 100 G O 'H 0) m H E m E: m a. E-t 4.: I: Q) 0 H :2 o l l I L L 1 1 I 1111_1 15 16 17 18 19 20 21 22 23 24 25 26 27 28 wavelength in microns Figure 11. Infrared spectrum of triphenylchlorosilane in the cesium iodide region. 111 .sofimea swayedno Ezeoom on» as osmaamoaoHuofioamsonafio mo Esnpooam consumsH .NH onswfih msoaofia sfi sumswao>m3 vH ma NH HA OH m w h m m w m A _ _ _ _ _ _ _ _ H _ notsstmsusam ineoaad OOH 10 Percent Transmission 112 1 Figure 13. 17 1 19 20 1 2 5282728 wavelength in microns Infrared spectrum of diphenyldichlorosilane in the cesium iodide region. 113 .sofiweu wownofino Esfioom one sH osmHHmouosHmfipamsosawc mo Esauoonm ponsuwsm .¢H onawwm msouoaa a“ sawsoao>m3 ea ma NH HH ,0H m m b m n v m _ _ J 4 H _ 4 _ H _ _ I o a... D _ .1 _ m J O 9 u fi/ _ 1. :12. -i..- .. m u S I T? 8 8 I. _ O u .. x .” \Ls/c l.\\\\\ /\> M OOH 190 114 Percent Transmission I... Er is Figure 15. ...--/ N‘ “. l7 1 ’ wavelength in microns Infrared spectrum of diphenyldifluorosilane in the cesium iodide region. 8.53 I 115 .sofimen_oanoHno,Bsaoom,onp se mamaHmoaosagoaoHnonsonQHp mo asauooam conmnmsH .ma madman moonofia mu newseao>mB «A ma Na , .3 6H m w b w n v M _ e u _ _ _ _ _ — _ D noun-sues], guessed OOH 116 100 Percent Transmission I 14 1 1 I I 4 I l I I l I 15 16 17 18 19 20 21 22 23 24 25 26 27 2 wavelength in microns Figure 17. Infrared spectrum of diphenylchlorofluorosilane in the cesium iodide region. 117 100 A ‘v’ / X i / '1 i G O I .H I m 1 m i T l s . m I = I 2 [-I a . = 1 m I o a “ i 3: i i l O I I I J J I I J l I I I I 15 16 17 18 19 20 21 22 23 24 25 26 27 2 wavelength in microns Figure 18. Infrared spectrum of bis(triphenylsily1)chromate(VI) in the cesium iodide region. 118 100 Percent Transmission 1 J- 1 ‘15 15 17 18 i9 ’“20W21 22 :24' W5 6 7 wavelength in microns Figure 19. Infrared spectrum of cyclo—bis[tetraphenyldisilox- anylchromate(VI)] in the cesium iodide region. 119 .sonon opHaoHno ssHoom one sH AH>voamaoasoAHmHHmonosHmstonQHpvan Ho Esmaooam poaeamsH .om omsmHm msoaoHa sH newseHo>a3 «H 2 NH H H S m w b mun v m N u d a u d u 4 a J.‘ u *l] c m o e u 1 In J \ .. u s I I m 1.. o u OOH. 120 100 s ‘l O v-I m m fl 1 s w I s m a I— .p a m o 1.. 3’. 0 1 1 I 1 L 1 1L 1 n 15 l6 17 18 19 20 21 22 23 24 25 26 27 28 wavelength in microns Figure 21. Infrared Spectrum of bis(diphenylfluorosilyl)- chromate(VI) in the cesium iodide region. ‘\ b 121 .sOHwou oOHHoHso EsHpom onu.sH voaumw< seam HHH>vepmaonsonssnoHHmHmseaaHpHmHnloHomo mo asapooam penmmmsH .NN eastm msomoHl sH nuwson>m3 eH mH NH HH oH m w h min v m N 4 d - H u u q - q u - O notsstmsusam guessed OOH 122 100 Percent Transmission LIE—‘13 17 18 125 WWW wavelength in microns Figure 23. Infrared Spectrum of cycle-bis[diphenylsiloxanyl— chromate(VI)] from Ag2Cr04 in the cesium iodide region- 123 .sonon ooHaoHso Echom one SH HAH>voumaonnoH>smx0HHmHmsonQHpHMHHuuoHomo mo asapooam nonmamsH .eN onstm mSOHOHE SH numsoHe>m3 . VH mH NH HH OH O m b mum v m N H H J H H H H H 1 H H - O _T «me + HVHoe L notsstmsusai guessed OOH 124 100 Percent Transmission - 0 1 A V ‘24 '5 3 wavelength in microns Figure 25. Infrared spectrum of cyclo-tris[diphenylsiloxanyl- chromate(VI)] in the cesium iodide region. 125 .sonoH mpHHoHno asHoom can sH >0NHONm< Sosa HAH>VopmsoasonsmonHmHmsonaHpHmHnuoHoao mo asauooam poumnmsH .ON ousth mmoHoHS SH sumsoHo>m3 Illhw nu Nu Hm JH n? m » Lowm w m so J o e u 1 L J m I t. s s T? 0 u OOH 126 100 Percent Transmission U . 3 . I X wavelength in microns Figure 27. Infrared Spectrum of cycle-bis[diphenylsiloxanyl- chromate(VI)] from Ag20r207 in cesium iodide region. .SOHon ooHHoHSo achom 0S3 SH oSmHHmOHoHSOHpHmSpwEHonnaoaousHHulm.m.m Ho asnpooam oonHmSH mSonoHS SH SuwSoHo>m3 vH NH NH HH 0H m . w b O m .wm eHsmHm H. m H H H _ _ — _ — H H NW0 i T VHOO 127 . ._. notsstmsuszm iueozed OOH 100 Percent Transmission 128 tr—Tr Figure 29. \ / ”*— \ J 17 1819262122—21525-28-29-30- wavelength in micron Infrared spectrum of 3,3,3-trifluoropropylmethy1di- chlorosilane in the cesium iodide region. 129 .SonoH eoHHoHSo ESHoom oSH SH HoHnoSmHHmHmSpeeHmuonaouosHmHaulm.m.m mo muwoeam SoHSHHSH .om mSoaoHa SH SuwsoHo>m3 «H mH NH HH 0H m w h min v m H H H H H d H H H H 1 5 «me + H.Hoo m . i . _ . i i i _ I i 1 .1 1 .. ... \ .._ i . i . m. 1 _ w i . ‘ M i i 1 .1 . i _ . 1 KO .\ /\H \( uotsstmsusal iueoaad OH 130 100 _ / \‘\\ / l/i’ .. F \ . 1 I, P\\ /\J /’ ,/ 1 z’ \,. \ /" 8 \ ,1" -H ,' g / S 3 // a V / H a \ / 1. a' G o \/ O H 8i I I 0 W I I I I I I I 15 18 19 20 21 22 24-26 28‘30“ Figure 31. wavelength in microns Infrared spectrum of 3,3,3-trifluoropropylmethylsil- anediol in the cesium iodide region. 131 .Sonoa ooHooH ESHmoo 0:» SH HAH>vonEonsoH>SmonHmHNSHoEHmaoaaoaosHHHHuIm.m.mHmHnIoHo>o mo asupooam pmHmHHSH .Nm oastm msonoHs SH SuwseHo>m3 VH OH NH HH OH O m h min e m N H H H H H H J H1 H H 4 l O "P «we - + eHoo + /\/\ / (. uotsstmsucam ineoaad OOH 132 100 . ion Percen OL I I I L_ J I I I I g E H 1.8. 19‘ 2O 21 22 23 24 25 I 15 16 17 wavelength in microns Figure 33. Infrared spectrum of cyclo—bis[3,3,3-trif1uoropro- pylmethylsiloxanylchromate(VI)] in the cesium 133 .Sonon opHuoHSo SSHpom 9S» SH HAH>VopmaOHSOHmSmmoHHmHSSHmaHmuonmouosHmHauIm.NHNHmHnquHozo Ho asnvooqm peasamSH .vm onsth mSonoHa SH SpwsoHo>a3 vH OH NH HH OH O m b mum e m N . H H {H + H H H H 1— H A] O m mmo + HVHoe w! notsstmsusxm iueoaed OOH 134 100 1“! I I I G O 'H g m D a £1 .9 O O) O H 81 0 I I L I L L I I I L I I 4 15 16 17 18 19 20 21 22 -23 24 25 26 27 28 wavelength in microns Figure 35. Infrared spectrum of cyclo-tris[3,3,3-trifluoropro- pylmethylsiloxanylchromate(VI)] in the cesium iodide 135 .Sonon opHnoHSo SSHoom on» SH HAH>VopmSonS0 IHNSsonHmHOHNSuochHmooaaoaoSHHHnHIm.NHNIHSHmHSIoHoho Ho asuuoonm SSHSSHSH mSonoHa SH SHMSoHo>a3 ¢H OH NH HH OH O m h min N H; 1 .Om mhflth m OOH. notsstmsusam guessed 136 100 a O -H m 2 a \ 2 g U ['1 +3 I: 0 U H 8’. 0L _l l J I I I l I J I 15 16 17 18 19 20 21 22 24-26 28 30 wavelength in microns Figure 37. Infrared spectrum of cyclo-bis[di-3,3,3-trifluoro- propyldimethyldisiloxanylchromate(VI)] in the cesium Line @QQQUIIBWNH I-‘I-‘I-‘I-‘l-‘I-‘I-‘I-‘H QQGOQWNHO 137 TABLE XIV INTERATOMIC SPICINGS 0F SILVER CHROIATE (Cu 11.. radiation, a - 1.54183) 9 (Degrees) 5.40 10.75 15.44 15.69 16.05 17.72 19.50 20.98 21.49 22.01 22.56 24.34 25.05 25.89 27.78 28.39 30.83 31.40 6 hkl 7.53 4.14 2.90 2.85 2.79 2.53 2.31 2.15 2.11 2.06 2.01 1.87 1.82 1.77 1.65 1.62 1.50 1.48 (3) Relative Intensity 100 30 16 16 21 11 15 20 12 10 138 TABLE XV INTERATOMIC SPACINGS 0F SILVER DICHROMATE (Cu Kg, radiation, A - 1.54182) Line 9 (Degrees) dhkl (8) Relative Intensity l 5.53 7.00 20 2 8.72 5.09 5 3 10.80 4.11 5 4 15.42 2.90 95 5 15.55 2.88 92 6 16.00 2.80 ' 100 7 16.69 2.69 15 8 19.47 2.31 19 9 20.87 2.16 49 10 21.44 2.11 15 11 21.96 2.06 24 12 22.54 2.01 28 13 25.87 1.77 13 14 28.26 1.63 20 15 28.86 1.60 30 16 31.28 1.48 14 17 31.63 1.47 9 DISCUSSION Hydrolytic 4 Stability. Silyl and siloxanylchromates prepared prior to this study have shown very little stability in the presence of water. Cyclo-bis[tetraphenyldisiloxanylchromate(VIM (5) decomposes very slowly in water but very quickly in wet organic solvents such as acetone or ether. Its Slow re— action with pure water apparently results from the non- wetability of the crystalline solid. The hydrolytic sta- bilities Shown by siloxanylchromates containing the tri— fluoropropyl groups were therefore somewhat surprising. McBee (109) found that 3,3,3-trifluoropropylmethyldichloro— silane and similar silanes are not only stable in the pres- ence of water but do not hydrolyze with prolonged refluxing in the presence of forty percent potassium hydroxide. Clark, gt_al. (136) found diperfluoropropyldialkylsilanes to be impervious to water but unstable to fifty percent po- tassium hydroxide solutions. Haszeldine (65) has shown that the thermal and hydrolytic stabilities of polyfluoroalkyl- siloxanes are dependent upon the position of the fluorines relative to the Silicon. He found the following stability relationships: 8 fluorine» B > a: . The hydrolysis of the Silyl and siloxanylchromates quite possibly results from a nucleophilic attack on the 139 140 silicon which has been given partial positive character by the electron—withdrawing organic substituents. The phenyl radicals in phenylsiloxanylchromates and the fluorine sub- stituent of bis(diphenylfluorosilyl)chromate(VI) would therefore create favorable conditions for hydrolysis; how- ever the 3,3,3-trifluoropropyl substituent would also enhance hydrolysis for it has been shown to have definite electronegative character (137). The fluoroorganic and methyl groups must in some manner inhibit the attack of the nucleophilic agent. Steric hinderence is not a likely explanation, for the phenyl substituents would afford more protection than the 3,3,3-trifluoropropyl and methyl groups. The silicon-oxygen bond could possibly be shielded from the nucleophilic agent by the strong polar environment established around the silicon by the 3,3,3-trifluoropropyl and methyl groups (85). Thermal Stability. The decomposition temperatures of several organosilyl— chromates prepared by Hare (5) indicated that the nature of the organic groups has a direct influence on the thermal Preperties of the compound. Molecules containing organo- silicon substituents with electron—withdrawing character are considerably more thermally stable than those with electron- releasing groups. Bis(trimethylsilyl)chromate(VI) is 141 explosive at room temperature; bis(diphenylmethylsilyl- chromate(VI) and bis(diphenylep-tolylsilyl)chromate(VI) de- compose when warmed, while bis(triphenylsilyl)chromate(VI) is unaffected by temperatures below 300° C. (Figure 39, p. 151). Similar trends have been observed with organosiloxy- vanadates (138, 139) and organosiloxytitanates. Burg (140) reported that the thermal stability of resins with phOSphorus- phosphorus bonding were greatly enhanced by the addition of the highly electronegative perfluoromethyl substituents. Andrianov (143, 159) found that the "oxidative destruction" of polyorganosiloxanes was dependent upon the nature of the organic radical. He established the following order of diminishing thermal stabilities: C6H5 > CH2 — CH2 > CH3 > C2115. Because of the apparent connection between greater thermal stability and highly electronegative substituents, it was felt that organosilyl and siloxy chromates containing fluorine and fluoroorganic radicals would show even better resistance towards thermal oxidation than had been seen previously. Fluorine was selected for several reasons. It is the most electronegative element (47) and is also believed to exert a greater electron-withdrawing tendency than any POIy-elemental radical. The silicon-fluorine ionic bond energy of 143 kcal/mole (141) and the energy of homolytic 142 cleavage (135 kcal/mole) (6, 7) is the highest of any known silicon linkage. The silicon-oxygen bond which shows such phenomenal thermal properties in naturally occurring sili— cates has a considerably lower bond energy (89 kcal/mole). It should be noted that bond energies, although important, are not the only, or even the primary, influence upon thermal stabilities. Although the energy of the silicon-oxygen bond is more than twice as large as the silicon-carbon bond, linear polysiloxanes yield cyclic siloxanes upon heating, leaving the silicon-carbon bond intact. Explanations for this occurrence usually attribute it to reaction mechanisms in which bond rearrangement occurs through a transition state with a low energy barrier relative to that required for direct thermal rupture (51, 73). Fluorine also increases the thermal stability of the carbon-carbon bond. The decomposition temp- erature for tetrafluoroethylene polymers (510° C.) is one hundred degrees higher than for polyethylene (159, 142). Kriegsmann (53) examined the force constants for Si-X bonds (X - F, Cl, Br, I, O, S, N, C, and H) and found that a strengthening of the Si—X bond and neighboring bonds in the silicon occurred when X was a strongly electronegative element. He stated that the electronegative tendency imparted partial double bond character to the Si-X linkages. The thermal properties of various silyl and siloxanyl- chromates were studied by differential thermal analysis (DTA). 143 The thermal effect for various physical and chemical type reactions are listed in Table XVI. TABLE XVI PHYSICAL AND CHEMICAL ENTHALPIC REACTIONS (145) Reaction Type Thermal Effect* Physical Absorption Adsorption Crystalline transition Desorption Fusion Second-order transition Sublimation Vaporization IIIH'++ Chemical Chemisorption Decomposition Dehydration Desolvation Oxidation in gaseous atmosphere Oxidative degradation Redox reaction Reduction in gaseous atmosphere Solid-state reaction I+I+I+H-H-l I l+ * + Exothermic reaction, - Endothermic reaction The analyses of the liquid organosiloxychromates were accomplished by mixing the liquid with alumina to form a thick paste, thereby preventing sample leakages or absorption into the ceramic holder. This technique also decreased the intensity of the thermogramic bands to convenient size for the recorder 144 being used. Crystalline materials were ground to fine pow- ders before analysis to increase the definition of the result- ing thermal peaks (144). All analyses were performed at a temperature elevation rate of 10° C./min. Thermograms obtained from analyses of organosiloxy- chromates exhibited broad and intense endothermic peaks with maxima lying between 300 and 500° C. Smaller endothermic bands were always found at 532 i 12° C. The crystalline com- pounds showed sharp endotherms which occurred at temperatures corresponding to their respective melting points. The liquid siloxychromates containing methyl and 3,3,3—trifluoropropyl radicals exhibited a third broad endotherm at 245 1 10° C. The intense bands occurring between 300 and 500° C. were assigned to the decomposition temperatures of the various organosiloxychromates and the following reaction was proposed: I l I I I I -— s1 -—— o 'f Cr -— o 4— Si ——£L_a. Cro3 + ——s1 -— o —— s1-— : : | I This reaction would, of course, include ring breaking as part of the decomposition. The temperatures at which the bond breaking occurred varied with the different organosiloxy- chromates (Figures 38 - 41). The strongly electronegative effect of the organic radicals gave partial positive character to the silicon (137) which in turn possibly shifted the charge 145 from the oxygen toward the reaction site. + 0 + \‘ 6 .7 H t s I, 3:81 1- O -—- fir -—- 0 ==*'Si :: This charge transfer 0 creates partial double bonding which strengthens the sili- con-oxygen link. In phenyl-containing siloxychromates the Thelectrons of the benzene ring may partially overlap vacant g orbitals of silicon to strengthen the silicon-oxygen bond even further by partial conjugation with the phenyl rings (53, 55, 146, 149, 150, 151, 152, 153, 154, 155). The sample holder contained a green powdered residue after each organosiloxychromate analysis. This high melting solid was believed to be chromium(III) oxide (m. 19900 C.). The endotherm which occurred consistantly at 532 i 12° C. was therefore attributed to the thermal reduction of chromium- (VI) oxide. ZS Hare (5) in observing the melting points of crystalline organosiloxychromates reported that bis(triphenylsilyl)- chromate(VI) melted and immediately decomposed at 154° C. and that cyclo-bis[tetraphenyldisiloxanylchromate(VIM showed a sharp decomposition point at 169° C. The sharp endothermic DTA bands associated with these compounds and with cyclo-tris- Idiphenylsiloxanylchromate(VI)] indicated that the specific 146 temperatures produced fusion and not decomposition. The solid compounds were dissolved in methylene chloride after fusion and cooling, yielding orange solutions with no apparent insoluble decomposition products. It is possible that the black substance observed by Hare after melting the organosiloxychromates was actually deep orange in color. Some of the liquid organosiloxychromates appeared to be black because of their extremely intense orange color. The endotherms appearing at 245 1 10° C. with 3,3,3-tri- fluoropropylmethylsiloxanylchromates were believed to result from cleavage of the methyl-silicon linkage which is known to be quite unstable (159). The relative thermal stabilities of several organo— siloxychromates were compared by observing the temperatures at which bond breaking maxima occurred (Table XVII). The replacement of a phenyl radical in bis(triphenyl- silyl)chromate(VI) (Figure 38, p. 150) by the highly electro- negative fluorine, yielding bis(diphenylfluorosilyl)chromate(VI), increased the thermal stability of the species by more than 60° C., thereby substantiating the original hypothesis. The resistance to thermal degradation of diphenylsiloxy- chromates appeared to be slightly better than that of 3,3,3- trifluoropropylmethylsiloxychromates; thus cyclo-bis[di-3,3,3- trifluoropropyldimethyldisiloxanylchromate(V1)] decomposed at 25° C. below cyclo-bis[tetraphenyldisiloxanylchromate(VI)]. 147 Cyclo-bis[3,3,3-trifluoropropylmethylsiloxanylchromate(VI)] and cyclo-bis[diphenylsiloxanylchromate(VI)] degraded at approximately the same temperature (375° C.). The greater stability shown by the phenyl—containing siloxanylchromates indicated that the combined electron-withdrawal of the two phenyl radicals was slightly better than the 3,3,3-tri— fluoropropyl-methyl combination; however, the phenylsilox— anylchromates would have some added stability from the partial double bond conjugation with the aromatic rings. It is be- lieved that replacement of the methyl group by a more electro— negative substituent such as fluorine, phenyl, or even another 3,3,3—trifluoropropyl group would result in compounds with even greater thermal resistances. The degree of electron-withdrawal was apparently not the only factor influencing thermal stabilities for if this were the case bis(triphenylsilyl)chromate (d. 390° c.) with three phenyl radicals for each Si-O-Cr would show greater stability than cyclo-bis[tetraphenyldisiloxanylchromate(VI)] (d. 4250 C.) with two phenyl groups per Si-O-Cr linkages. Probably the cyclic structure accounts for the added stability (73). Cyclo-bis (d. 375° C.) and cyclo—tris[diphenylsiloxanyl- chromate(VI)] (d. 322° C.) with one phenyl group per Cr-O-Si linkage degraded at even lower temperatures. The difference between decomposition temperatures of these two compounds indicated that the dimeric structure was the more stable form. 148 The thermal decomposition points of the above compounds further suggests a cyclic structure for the liquid dimer. Assuming that both compounds were linear, their decomposi- tions should occur at approximately the same temperature and if the dimer were linear the decomposition point would have been lower than that of the crystalline trimer. Reactions Mechanisms. The heterogeneous reactions of organohalosilanes with silver chromate and dichromate did not yield the expected products; however, the data accumulated during the course of this investigation suggested routes which could possibly explain the formation of the compounds which were obtained. The proposed mechanisms for the reaction of organohalo— silanes with silver chromate and dichromate require as an initial step the dissolution of the salt in methylene chloride followed by the immediate establishment of a chro— mate-dichromate equilibrium. Reactions did not occur when either pentane or diethyl-ether was used as the solvent and proceeded at a much greater rate in methylene chloride than in carbon tetrachloride. If crystal dissolutions were con- sidered as part of the over-all reaction then the necessity for a polar nonreactive solvent becomes apparent. The possibility of a reaction at the crystal surface may be ruled out if the silver chloride by-product is .5 once vvm .m ovum oov .nm once Nvm .s ovum mam H .m.a no.2 HHe eHemHHV HAH>voumsonno namsaxoafim -HeHHeHea IHpamnoun nono=HMHHp um.m.muHe. umHnnoHozo Q n n .mpfimsmpsfi u H muoowmm Hmauone n . .gswam H mm mesons n n “Hooasoam u :m ”Amos u 3 ”ssficms u a m .5 once own .5 once man .a once man a n n .m once mum .m once Nam .m a .s ooze mmm am .m> ouso moH H .m.a no.2 H .m.e o..m H ovum mum .m.a 00.: n .a ocso mmm . a .m coco va an .m comm mmH H .fl.9 Uo.S HHe ensmHHv Hoe enemHHV Hoe eeewHHv Hem oeemHHv .HH>V loamaonno HAH>V IamsmonHm Imumaouno IH>QOHQOH05HH IamsmonAm nHHHum.m.m. uHHeeeeHeL ImHQIOHomo ImHHuloHomo mmezV Impaaosno Iamsmonfim nHHeeheHeH ImHnloHomo .AH>V loumaonao nahsmonHm nHeHHeene Imuwmpa ImflnnoHomo n .3 am once omm n .m ocsm mmv H .m.a no.2 Hem enemHHv An>voamaosno :HHHHHmouoaHH uHaeeaeHevan ms Iqwqmmoz<0mo hO mmm>q<2¢vqX mqm<8 .9 ”.0 e asafixms u .00. sum mso> u m> ”msosum u a sofluomcm .s ooze 0mm sawsoun a msfixmmn .m some 0mm use a .3 once com saosxe owa>aoH mmoH usuo a once mud wsHuHm H .H.a .oo.: Hem eusmHHv Au>vmpa neonaoAHHHHm lamsonnfisuvmfin mxmm 150 .Au>vomeOHnoAahaamonosamamsmssfipvmfln use Au>voumaosnoAamafimamsmnnfinpvmwn mo mommamsm awakens Hmflusmnmmawn .oo .ossumhmnsos cam mmm . one . pww An>vmpmsosnoAamawmososawamsmnnwpvmun III: AH>VepaaoHeeHHHHHmHHeeeeHHHVmHn .wm enstm Inmxeqaoxa ° temaeqiopua 151 .HAH>voumsoun0H>smonHmeHmsonmmeopHmeuoHozo mo mHmHHmsm Hashesp Hmflpsmummmfln .oo amusvmuonsma. 2 com eee eon . nae . mom . emu eeH .mm enstH Iu'JGQ1OPUH in lo Ismaeqioxa 152 .HAH>vmumaonaonsaonHmHhumanfioamHHHIOHOAO one HAH>vopmsonnOHmsmonHmHmsanHchHnloHomo mo momzamsm amsnmnp Hmfiasmhmwnfin .ov onsmfim . .0. .onsumnmnsme HAu>vopaaonnonsmon«mausmaQHoHmHnlouomo-uu es H aAH>vmamsonconsaxoa«mamsonnHUHmHHpIOHomo.II. \ , ---—‘.-_> -u- - H... 4...... .. - .. I I I I x O I '\ C I3'19Q1OPUS O IBmJeqioxa 153 .HAu>vmamaonnonsmonHme IamnamsHoHmoounososHmHsulm.m.mlwpamfinloHomo one HAH>VmumEOHnoH>seonHm nahnumsHonunonosHmHspnm.m.mHmHnloHomo mo mmmmaasm Hashes» HmHusmnmmHHn .Uo .ouspaumqaoa oou cam can u _ . _ cow com CON H . J1 - q _HH>veHeaoHnoHHeeerHmHeHahueaHeHanoneouoeHHHHHum.m.muHe_mHnuoHoHo -qu _AH>vopmaonnoammmonHmflznpmaaknonnonosamfinpum.m.mHmHn|oH0ho .He ensmHH temaaqiopua temxeqioxa 154 adsorbed by the crystal in such a way as to block the surface rapidly. The absence of color in the silver chro- mate or silver dichromate methylene chloride slurries indicated that the salt was only very slightly soluble. The definite and sharp melting point of silver dichromate indicated that it was more ionic in nature than the chro— mate, and their respective solubilities in water (AgZCrO4, 7.5 x 10'7 mole/l. ; AgZCr207, 1.9 x 10'4 mole/l. tended to substantiate this hypothesis. It was, therefore, assumed that the salts were also slightly soluble in the polar methylene chloride and that the dichromate was some- what more soluble than the chromate. The chromate-dichromate equilibrium in water has been discussed on p. 8. The position of the equilibrium is affected by the presence of an acid or a base and by the concentration of the dissolved species. In very dilute solution this equilibrium is far to the right r- q-_ r- '1__ o o o u I H 1120+ o—clr—o—cr—o ;-: 2 o—Er—o + 2H+ o o o L JL L .I and as the solution becomes more concentrated, the solute species begins to polymerize yielding first the dichromate, then the trichromate and finally the tetrachromate ion [Cr4013]-- at very high concentrations (27, 147, 148). It would be expected that the dissolution of silver chromate in 155 methylene chloride would produce very little of the dichromate species for it would be unlikely that the highly purified sol- vent would contain any Species which would shift the equili- brium towards the dichromate. The dissolution of silver dichromate in methylene chloride could possibly yield the above equilibrium; however, a complete reaction would require that water be present in amounts equivalent to or greater than the molar quantity of the organohalosilane reactant. The rather elaborate percautions taken to exclude water from the solvents, reactants, and apparatus suggest that the chro— mate species was not formed by the action of water. It is possible that the following equilibrium existed in methylene chloride. 3 I 8 —fir-O-fir—-O '—;.—-‘ O—fir—O + Cr03 Chromium(VI) oxide, like the silver salts, is not very soluble in methylene chloride (5). Once dissolution has occurred and the equilibrium favor— ing the chromate Species established, reactions of organo- chlorosilanes with both the silver chromate and dichromate would follow the same path. Triphenylchlorosilane and diphenyl- chlorofluorosilane would be expected to react in much the same manner since the fluorine-silicon linkage of diphenylchloro- fluorosilane was not cleaved with the reaction conditions 156 employed. Bimolecular displacement (SnZ) of chlorine from the organochlorosilane by the nucleophilic chromate ion appears to be the most likely process. The mechanism is illustrated below: Reaction Scheme A R R o R o \/+ _ ll -AgCl I ll _ C1 ------ 1‘ ------- o—lcr—o ——, R—|Si-—O-—Cr—0 l R' o R' o —AgCl RZR'SiCl R o R I II I R—Si—O— Cr— 0—31 —R I I l R! 0‘ R! R—phenyl R'-phenyl or fluoro The formation of the very insoluble silver chloride could be considered as the driving force of the reaction. The silicon—oxygen bonding would be stabilized by the partial double bond character resulting from d fT- p'rr interaction of a vacant silicon g. orbital and a filled .2 orbital on the oxygen. The conjugation of the phenyl ring and silicon- oxygen double bonds would strengthen the bonding even 157 further. The above interactions would tend to lower the energy required for bond formation, resulting in faster rates with phenyl-containing halosilanes. The mechanism for the reactions of organosilanes con- taining two labile chlorines would in all liklihood follow a similar procedure, i;g. consecutive bimolecular chain building steps with ring closure as the termination process. The reactions could take place in the following manner: Reaction Scheme B 3 RI 0 R' o \/.+ _ Il _ -AgC1 I ll _ C1 ----- S'i ----- 0—|C|r—o -—-—> R—Si—O— fir— 0 Cl 0 Cl 0 -Ag::////RR'SiC12 R' 0 R I I l R—sli—o—fir—o—si—R C1 0 C1 Path I path II -AgCl Cro4“ -AgCl Cr04‘_ Path I o H /"Ir\ Cl, 0 O o \\\ 4F \ /Rt .R' ‘8; Si\\ ,/’ ‘~ _ // R b o o \n/ Cr 3 —AgCl 0 g r ./g\. R! \ RI >31/ )1\ R ‘x R o 0 o \\u // Path II o R' \ / Cgc—- O‘-§1 0// 0 R \\ R'\ / \ Si 0=Cr=0 R/"\c1 l o- -2AgCl CrO4-- RR'SiClZ RI QC§C—-O--Si/ x/’ o R \\ o 0 RL\ ./’ ‘\ /,Si 0=Cr=0 R \x o o \ //O _ R'G+/ ¢Cr-—0———JSi----C1 o R -AgC1 o R' ‘br-—-O-—:Si /’ “b R ‘\\ //o 0 RI \\Si 0-Cr=0 R” \\ o 158 159 The structures of the compounds which resulted from the reactions of organodichlorosilanes with silver chromate and dichromate are not fully explained by the preceeding mechanism. The apparent inconsistencies can be explained by consideration of the chromate ion concentration and the relative ease of bond formation which in turn was influenced by the organo-silicon radicals. The reaction of diphenyldichlorosilane with small amounts of silver dichromate (two-fold excess) yielded only the cyclic dimer of diphenylsiloxanylchromate(VI) while similar amounts of silver dichromate and 3,3,3-trifluoropropylmethyl- dichlorosilane produced the cyclic trimer of 3,3,3-trifluoro- propylmethylsiloxanylchromate(VI). The relative solubilities of silver chromate and dichromate would suggest that the dichromate should yield greater concentrations of the nucleophilic agent Cr04" than an equal amount of chromate. The high electronegativity of the phenyl groups along with the double bond conjugation would tend to enhance bond forma— tion in phenylsiloxychromates. The termination or ring closure in reaction Scheme B would be dependent upon the con- centration of Cr04" and the ease of bond formation. Small amounts of Cr04 would tend to lead the reaction mechanism along Path I while larger concentrations would favor continua- tion of the chain building steps (Path 11), while the stronger electrophilic nature of the phenyl-containing silicons would encourage ring closure at the dimer stage (Path I). The 160 formation of the various telomers and the probable mechanistic influences are listed in Table XVIII. TABLE XVIII ORGANODICHLOROSILANE REACTIONS WITH SILVER CHROMATE AND SILVER DICHROMATE Reaction Cr04" Mechanistic Path Actual conc. favored by Structure CrO -- Organic con . Radical Influence A82Cr04 + ph281C12 LC path I path I Dimer AgZCr207 + PhZSiCl2 MC Path I Path I Dimer or II Ag2Cr207 I ph281c12 HC path 11 path I Trimer(0.20) Dimer (0.80) A820r04 +(CF30H20H2)(CH3)SiC12 LC path I path II Dimer AgZCr207+ (CF3CH2CH2) (CH3)31C12 MC opitgll Path II Trimer LC - Low concentration (two-fold excess of Angr04); MC = Medium concentration (two-fold excess of Ag20r207); HC = High concentra- tion (ten-fold excess of AgZCr207). 161 The equilibrium proposed for silver dichromate 0 0 0 II II II 2Ag+ + 0-—fir——0——fir-—0 -1:7—* Ag+_+ 0-—Cr-—0 + Cro3 + Ag+ 0 OJ .oI h. yields two species, one silver ion and chromium(VI) oxide, which were not consumed during the reactions. It was fairly unlikely that the unreacted species were completely removed by the crude purification techniques required for the liquid product resulting from the reaction of diphenyldichloro- silane with silver dichromate. Minimal amounts of the low molecular weight impurities could account for the large error observed in the molecular weight values of cyclo-bis[diphenyl- siloxanylchromate(VI)] (p. 75 and Table V). The absence of disiloxanyl linkages in the products ob- tained from reactions of silver chromate and dichromate with organodichlorosilanes would tend to substantiate Hare's (5) hypothesis that these linkages resulted from the tendency of the diol reactants to undergo rapid self-condensation. The synthesis of cyclo-bis[di-3,3,3-trifluoropropyldi- methyldisiloxanylchromate(VI)] from chromyl chloride and 3,3,3—trifluoropropylmethylsilanediol in methylene chloride was analogous to Hare's reaction using diphenylsilandiol. Both reactions yielded two products; however, the liquid 162 product from the 3,3,3-trifluoropropylmethylsilanediol syn- thesis was not characterized primarily because it occurred in such small amounts. Hare proposed two mechanistic routes (Reaction Scheme C) for the preparation of cyclo-bis Retra- phenyldisiloxanylchromate(VI)] which also gives an adequate explanation for the synthesis of cyclo-bisIdi-3,3,3-tri- fluoropropyldimethyldisiloxanylchromate(VI)]. Reaction Scheme C Path I Path II R 2RR'Si(OH)2 HO—Si—O‘f— ,+ 1 I \\ 6| /0 R. /‘Cr 0’ \Cl ' -H20 ‘-HC1 R R H0— 81— O—- 81— OH R o | | l RI RI HO—Si— 0 —Cr—Cl I I R' O —2HC1 CrOZClz R R O R R I I N l | HO—Si-O—Si-O-Cr—O-Si-O—Si-OH v I II I R R O R R Path I Path II -HCl RR'Si(0H)2 (-HCl) CrOZClZ R 0 R 0 g. I g .. H0-—- r 1 -——-0H ,,Cr\ I u R\ /0 II o\ /R R. I | R. ° I R'\I R' Si 0 81” R/ 0\"8"' 5:!) \R Cr\‘”’ | bimolecular H Cl H condensation 0 -HCl 0 5 r x ‘\ R ol 0 R :Si/ 0 \51/ R' I I\Rv 0 R'\I I/R' //Si\\ 0 l/Si\\ R o\I/o R Cr Hare explained the formation of the liquid reaction product as being either the hydrolysis of the cyclic dimer followed by cleavage of a chromate groups during the coupling of diphenyl- hydroxosiloxy group or the initial condensation of diphenylsil— 163 164 anediol to hexaphenyl-l,5-dihydroxotrisiloxane, which attacked the chromyl chloride molecule to form bis(hexaphenyl- G-hydroxotrisiloxanyl)chromate(VI). The minimal amounts of liquid product obtained with 3,3,3-trifluoropropylmethylsilanediol suggested that the reaction involving the formation of hexaphenyl-l,5-dihydroxo- trisilane would be favored in the synthesis of bis(hexaphenyl— B-hydroxotrisiloxanyl)chromate(VI) (5), for the phenyl radicals with their higher electronegativities and double bond conjuga- tion would tend to condense the diol to a greater degree than the 3,3,3-trifluoropropylmethyl combination. These intra— molecular interactions predict that 3,3,3-trifluoropropylmethyl- silanediol would follow Path I for the preparation of the cyclic dimer. The synthesis of diphenylchlorofluorosilane from diphenyl- dichloro- and diphenyldifluorosilane involved elemental exchange. The reaction conditions required for the exchange indicated that the reaction was one of unimolecular substitution. Reaction Scheme D illustrates the probable steps. Reaction Scheme D ph ph slow I Ph—Si —01 T Ph— Si++ c1" I C1 C1 165 Ph Ph I A I+ _ Ph—Si—F ———-—> Ph—Si+F | slow I F F $h+ fast Pb ph-——'fii-+ Cl" -—————e’. Ph-—- fii-—— c1 F F Ph Ph + _ fast I Ph—Si + F ——-—> Ph— 81— F Cl Cl When equal molar quantities of the diphenyldihalosilanes were refluxed at 150° C. the exchange did not occur; however, refluxing at much higher temperatures (275° C.) yielded di- phenylchlorofluorosilane. ‘The high temperature quite probably resulted in unimolecular heterolytic dissociation. If the mechanism had been a temperature dependent bimolecular sub— stitution, at least small amounts of the mixed halosilane would have been produced during the prolonged low temperature refluxing. The composition of the refluxing mixture (275° C.) was checked by vapor phase chromatography at approximately ten- hour intervals over a period of one hundred and fifty hours. 166 A maximum yield of twenty percent diphenylchlorofluorosilane was reached at this point. Calculations involving molar quantities of the ionic species available for recombination predicted a yield of fifty percent for diphenylchlorofluoro- silane and twenty five percent for each of the diphenyldi- halosilanes. The diphenylfluorosilyl cation, because of the electronegativity of the fluorine radical, should possess greater electrophilic character than the diphenylchlorosilyl cation and should preferentially attract the more electro- negative fluorides. Diphenyldifluorosilane should, therefore, be the most prevalent recombination product; however, because of the relative number of ions available, the final mixture contained equal molar quantities of diphenyldifluoro- and diphenylchlorosilane. SUMMARY AND CONCLUSIONS The previously known crystalline bis(triphenylsilyl)- chromate(VI) was prepared by the metathetical reaction of silver chromate and triphenylchlorosilane. A similar reac— tion using diphenylchlorofluorosilane yielded a deep orange liquid which was characterized as bis(diphenylfluorosilyl)— chromate(VI) . The replacement of the phenyl group with the more electronegative fluorine increased the thermal stability of the Silicon-oxygen-chromium linkage. The reaction of diphenyldichlorosilane with silver chromate or silver dichromate gave mixtures of a cyclic dimer and trimer with alternating oxygen-chromium?oxygen-Silicon linkages, cyclo—bis and tris[diphenylsiloxanylchromate(VI)]. The amounts of each cyclic Species produced were dependent upon both the nature and amount of the Silver salt reactant. The assignment of the cyclic structure to these molecules resulted from interpretation of their infrared Spectra, shifts occurring with ultraviolet and visible absorption bands, pro- ton magnetic resonance Spectra, and thermal decomposition studies. Similar cyclic A-B—C type monomers with 3,3,3—trifluoro- propyl and methyl-Silicon substituents were prepared from the reaction of 3,3,3-trifluoropropylmethyldichlorosilane with silver chromate and dichromate. These compounds were the 167 168 first siloxy-chromates containing fluoroorganic groups. Struc- tural proofs were Similar to those used for the cyclic phenyl- siloxanylchromates. It is believed that these four monomers are the first cyclic compounds with alternating oxygen-silicon- oxygen-metal linkages that have been isolated and identified. The condensation of chromyl chloride with 3,3,3-trif1uoro- propylmethylsilanediol yielded cyclo—bis[di-3,3,3-trif1uorc- propyldimethyldisiloxanylchromate(VI)], a liquid at room temp- erature, which could be made to crystallize at lower tempera- tures. This compound contained silicon-oxygen—Silicon linkages which resulted from the tendency of silanols to undergo self- condensation. Attempts to prepare organosilyldichromate from the meta- thetical reaction of triorganochlorosilanes with silver di- chromate were unsuccessful; regardless of the conditions, the reaction invariably led to the chromate product. The inability to obtain the predicted dichromate was attributed to the estab- lishment, in methylene chloride, of a chromate-dichromate equi- librium where the chromate Species was highly favored because of the extremely low concentrations involved. The investigation of the thermal stabilities of the com- pounds prepared during this study was accomplished by means of differential thermal analysis. The peaks occurring in the thermograms were tentatively assigned to various thermal phe— nomena. The decomposition peaks for each of the organosiloxy- chromates were found to correlate fairly well with the relative 169 electronegativities of their respective Silicon radicals, 143., greater electronegativities or withdrawing effects resulted in higher decomposition temperatures. In addition, hydrolytic stabilities of the organosiloxychromates, which were observed qualitatively, appeared to be a function of the electronegativ- ity of the silicon substituent. Generally high electronega- tive substituents facilitate hydrolysis. This concept was somewhat belied by the apparent inert behavior of the fluoro- organosiloxanylchromates toward water and alkaline solutions. Here hydrolytic stability quite possibly resulted from screen- ing of the Silicon-oxygen-chromium linkage by both the fluorine atoms and the methyl radicals. Infrared Spectrosc0py was used extensively in the identi- fication and structural characterization of reaction products. The linkages contained in the organosiloxychromate molecules absorbed in the region between 2 and 30 u. The absorption of Si-F, Si-Rf, Si-OH, Si-O—Si, and Si-O—Cr were of particular importance. The silicon-oxygen-chromium vibrational modes absorbing at 11-12 u and 20.6-21.4 u were tentatively assigned to the reSpective asymmetric and symmetric stretching wave- lengths. 170 RECOMMENDATIONS FOR FUTURE WORK 1. The preparation of organosiloxydichromates should be possible; however, this would entail the selection of a sol- vent which would favor the dichromate Species in the chromate- dichromate equilibrium. The solvent should also be inert to- wards the Silane reactant and organosiloxydichromate product. 2. The reaction mechanism discussed on p. 159 suggests that slower addition of diphenyldichlorosilane to the hetero- genous methylene chloride solution of a large excess of Sil- ver dichromate would possibly give greater yields of the crystalline cyclo-tris[diphenylsiloxanylchromate(VI)]. A more thorough study of the temperature-dependence of all the reactions should be made. 3. The thermal behavior of the compounds prepared during this investigation indicated that the replacement of the methyl group in cyclo—bis[di-3,3,3—trifluoropropyldimethyldisiloxanyl- chromate(VIfl with a highly electronegative substituent such as fluorine, trifluoromethyl, phenyl, or another 3,3,3—tri- fluoroprOpyl group would yield a monomer with an outstanding resistance to thermal decomposition. The addition of fluorine or trifluoromethyl groups, however, might tend to make the resulting compounds less hydrolytically stable. 4. A quantitative study of the hydrolysis, both aqueous and alkaline, of the organosilyl- and siloxanylchromate come pounds should yield Significant information for future prepa- rations. 171 5. The further polymerization of the cyclic siloxanyl- chromates should be attempted. This might possibly be at- tained by the use of a Lewis acid catalyst or possibly solid state irradiation of the crystalline cyclo—tris[diphenyl- siloxanylchromate(VI)]. 6. Sulfur tetrafluoride reacts with metal-oxygen link- ages to form difluoro—metal units in the following manner (64): MO or RMO + SF4 -—————e> MFZ or RMF2 Under the appropriate conditions the reaction of sulfur tetra- fluoride with the chromium-oxygen linkage in silyl- and siloxy- anylchromates might yield substances having very interesting physical properties. 7. Structural isomers of the cyclic siloxanylchromates containing 3,3,3—trifluoropropyl and methyl linkages can be drawn. Cyclo-bis and tris[3,3,3-trifluoropropylmethylsilox- anylchromate(VI)] Should have one cis and one trans isomer with the following structures: 0 H H Cr Rf /II \ Rf Rf /I|3\0 TH3 \\ Ii/ \41 I1/ s1 ('233\o o/c'm (|:H\o o/ If \\\fi’// 3 3 \\\fi‘// Cr fir 3 o 172 Six possible structural isomers can be postulated for cyclo- bis[di-3,3,3-trifluoropropyldimethyldisiloxanylchromate(VI)]. Although difficult, the resolution and investigations of these isomers Should prove to be an interesting project. APPENDIX PREPARATION OF TRIS(TRIPHENYL)VANADATE(V) Tris(triphenylsilyl)vanadate was first prepared by Granchelli and Walker (10) in 1955 by treating tris(n—butoxy)vanadate(V) with triphenylchlorosilane in refluxing xylene. They obtained a thirty percent yield of a pure white crystalline solid which melted at 228-229° C. In 1958 Orlov, 3341., (139) reported the preparation of this solid (m. 228° C.) by the re- action of tris(n-butoxy)vanadate(V) and triphenylsilanol and in 1962 Chamberlain (160) obtained ninety—five percent crude yields of the white tris(triphenylsilyl)vanadate(V) (m. 225° - 6° C.) by the reaction of a benzene solution of vanadium oxytrichloride with triphenylsilanol in the presence of ammonia. 173 EXPERIMENTAL Preparation of Vanadium Oxytrichloride. One hundred grams of vanadium pentoxide (0.58 mole) and 130 m1. of thionyl chloride (1.1 moles) were refluxed for twenty hours on a steam bath. Pure vanadium oxytrichloride (b. 127° CJ)waS obtained from the reaction mixture by distil- lation. Preparation of Sodium Triphenylsilanolate. An excess of sodium metal was refluxed in 150 m1. of toluene for one hour followed by the addition of ten grams (3.6 x 10'2 mole) of triphenylsilanol. After continued reflux- ing for another three hours, the hot mixture was decanted from the sodium metal. White crystals of sodium triphenylsilanolate, which formed during cooling, were washed with hot ethanol and toluene. Synthesis of Tris(triphenylsily1)vanadate(V). A slight excess of sodium triphenylsilanolate (4.97 g. or 1.67 x 10"2 mole) was slurried in 100 ml. of cyclohexane. The cyclohexane had been previously distilled into a two necked 300 m1. round bottom reaction flask. Additional cyclo- hexane (50 ml.) was distilled into a 100 ml. dropping funnel to which 0.87 gram (5.0 x 10"3 mole) of vanadium oxytrichloride 174 175 were added. The dropping funnel and a reflux condenser with a drying tube were connected to the reaction flask and the sodium triphenylsilanolate Slurry was stirred with an internal magnet. The vanadium oxytrichloride solution was added rapid- ly to the excess sodium triphenylsilanolate and the reaction mixture was stirred at room temperature for twenty minutes. The residue from the filtered reaction solution was subjected to x-ray diffraction and infrared spectral analy- sis and proved to be a mixture of sodium chloride and the excess sodium triphenylsilanolate. Light green crystals appeared when the filtrate was condensed to approximately twenty milliliters by rotary evaporation. Recrystallization of the product from fresh cyclohexane gave a seventy percent yield of pure white tris(triphenylsilyl)vanadate(V). The white crystals had a metallic sheen and were soluble in either hot or cold aliphatic and aromatic hydrocarbons, alcohols, ethers, alkyl halides, organic acids, ketones, an- hydrides, acetates, and all other organic solvents used. The solid was not wet by water and was either insoluble or decomr posed in mineral acids. The differential thermal analysis of a 1:1 mixture of tris(triphenylsily1)vanadate(V) andIubalumina (Figure 43, p. 179) exhibited a Sharp endothermic band at 230° C. which agreed with the value for the melting point determination (227.5° C.). The thermal decomposition has been assigned to the broad exothermic peak occurring at 400° C. 176 Elemental analyses and cryoscopic molecular weight de- terminations of the recrystallized product (Table XIX) agreed quite well with the calculated values for tris(triphenylsilyl)- vanadate (V). TABLE XIX ELEMENTAL ANALYSIS AND MOLECULAR WEIGHT OF TRIS(TRIPHENYLSILYL) VANADATE (V) Calculated (%) Experimental (%) Carbon 72.62 72.50 Hydrogen 5.08 5.28 Silicon 9.44 9.48, 9.56 Vanadium ' 5.70 5.67, 5.51 Molecular Weight 893 892, 889 The infrared Spectrum between 2 and 15 u for tris(tri- phenylsilyl)vanadate(V) (Figure 42, p.178) was determined in carbon disulfide and gave the normal absorption associated with phenyl-silicon molecules. The peak occurring at 9.93 u resulted from the stretching Vibration of the vanadium-oxygen linkage (10). The strong broad maximum at 11.07 n has been assigned to the VrO-Si asymmetric stretching vibration. This peak iS quite similar in appearance and absorption wavelength to the Ti-O—Si (13, 131, 158) and Cr-O—Si (5, 157) (p. 6 and 97 ) asymmetric stretching modes. 177 The x-ray powder diffraction pattern of the white crystal- line solid was obtained with copper K.¢ radiation using a North American Philips powder diffraction unit and a Debye-Scherrer camera. The interatomic Spacings are listed in Table XX. TABLE 10‘ INTERATOMIC SPACINGS OF TRIS(TRIPHENYLSILNIJVANADATE(V) (Cu K“ radiation, A = 1.54183) 0 Relative Line 9 (degrees) dhkl (A) Intensity 1 27.7 1.66 3 2 26.2 1.75 l 3 23.9 1.90 1 4 22.7 2.00 1 5 16.8 2.67 2 6 15.8 2.83 1 7 13.5 3.30 l 8 12.8 3.48 2 9 11.7 3.80 2 10 11.3 3.94 l 11 10.6 4.91 2 12 10.2 4.35 6 13 9.9 4.48 9 14 9.7 4.58 7 15 9.1 4.87 6 16 ‘8.2 5.41 3 17 6.9 6.42 3 18 6.2 7.14 4 19 5.6 7.90 3 20 5.0 8.84 10 21 4.9 9.30 2 22 4.3 10.09 23 3.8 11.63 10 178 .sowwmn opssoano aswpom may aw A>Vopmcmsm>Aazasmahsosnwspvmsnu mo asnpownm cosmhmsm msououa as sawsmam>mB VA MH d NH b’-~ ~- m... h“: .l. HA n .‘ OH ‘1 m d m u Jvopmomsm>Aamasmamsmnnflsavmwsu mo mammHmsm Hassmsu Hmfipsosmmmfla .Uo .mssussosams D¢fl DON O@H .me oasmHH Ismaeqiopug {emxaqioxg 10. 11. 12. 13. 14. 15. 16. 17. 18. 180 BIBLIOGRAPHY International Critical Tables, McGraw-Hill Book Co., Inc., New York, 1933. E. Moles and L. Gomez, Z. Physik. Chem. 89, 513-30 (1911). E. Moles and L. Gomez, Anales soc. espafi. fis. y quim. 12, 142-54 (1914); C. A. g, 3184 (1915). G. Oddo and E. Serra, Gazz. chim. ital. 22, jj, 318-29 (1899). C. R. Hare, (Michigan State Univ., East Lansing) Univ. Microfilms (Ann Arbor, Mich.) L. C. card No. Mic. 61-2695, l40pp; Didsertation Abstr; 22, 42243 (1961). C. Eaborn, Organosilicon Compounds, Butterworths Scientific Publications, London, 1960? R. A. Shaw, J. Chem. Soc.,lg§z, 2831. M. Schmidt and H. Schmidbaur, Angew. Chem. 19, 704 (1958). E. W. Abel, Z. Naturforsch. 156, 57 (1960). F. E. Granchelli and G. B. Walker, U. S. 2,863,891, December 9, .1958. Private Communication, J. F. Hampton. F. A. Miller, G. L. Carlson, and W. B. White, Spectrochim. Acta. 15, 709—16 (1959). V. A. Zeitler and C. A. Brown, J. Am. Chem. Soc. 12, 4618-21 (1957) . M. S. Ovrutskii and M. A. Goronovskaya, Legkaya Prom. 18, 43 (1958); C. A. 52, 169680. J. W. Gilkey, U. S. 2,894,967, July 14, 1958. M. J. Udy, Chromium, Vol. 1, Reinhold Publishing Corp., New York, 1956. ' H. Brakken, Z. Krist. 18, 484 (1931). A. Bystrom and K. A. Wilhelmi, Acta. Chem. Scand. 3, 1131-41 (1950). 19. 20. 21. 22. 23. ' 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 181 J. Kleinberg, W. J. Argensinger, Jr. and E. Griswold, Inorganic Chemistry, D. C. Heath and Co., Boston, 1960. D. S. Datar and S. S. Datar, Nature 158, 518-19 (1946). N. Vauquelin, Ann. Chem. phys. 19 [1], 70 (1809). Bush, J. Phys. Chem. 31, 931—3 (1927). L. Dundon, J. Am. Chem. Soc. 45, 2658-9 (1923). L F M F. Kohlrausch, Z. Phys. Chem. 59, 355-7 (1905). G S. Whitby, Z. Anorg. Chem. 61, 107-8 (1910). G . P. Baxter and R. H. Jess, Z. Anorg. Chem. g2, 331-5 (1909). J. W. Mellor, A Comprehensive Treatis on Inorganic and Theoretical Chemistry, Vol. XI; Longmans, Green and Co., New York, 1931. O. Mayer, Chem. Ber. 36, 1740-4 (1903). M. Schmidt and J. Schmidbaur, Angew. Chem. 11, 553 (1959). B. L. Chamberland and A. G. MacDiarmid, J. Am. Chem. Soc. .83, 549-51 (1961). B. L. Chamberland and A. G. MacDiarmid, J. Chem. Soc. 1961, 445-8. V. Wannagat, F. Brandmair, W. Liehr, and H. Niederprum, Z. Anorg.tL Allgem. Chem. 302, 185-98 (1963). S. Z. Harder, M. H. Khundhau, and K. De, J. Inorganic and Nucl. Chem. 24, 847-50 (1963). J. J Berzelius, Ann. Physik Chem. 1, 34 (1824). H Hartford and M. Darrin, Chem. Revs. 58, 1-61 (1958). x. J. palmer, J. Am. Chem. Soc. 69, 2360-9 (1938). P . Smyth, A. J. Grossman, and S. R. Ginsburg, J. Am. Chem. Soc. 63, 192-5 (1940). P. Walden, Z. Physik. Chem..4§, 180 (1903). 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 182 G. M. Schwab and S. Prakash, Z. Physik. Chem. 6, 387-93 (1956). M. Gomberg, Chem. Ber. 35, 2397 (1902). H. H. Zeiss and C. N. Mathews, J. Am. Chem. Scc._Z§, 1694-8 (1956). Anbar, I. Dostrovsky, D. Samuel, and A. D. Yoffe, Chem. Soc. 1954, 3603. 2s as H. Zeiss, 0r anometallic Compounds, A. C. S. Monograph 147, ReinhoId Pub. Corp., New York, 1960. T. Nisikgda, J. Chem. Soc. Japan_§l, 595-6 (1940); C. A. ‘34, 6504.. J. Schabus, Sitzben.Akad. Wien 5, 369 (1850). Alphabetical and Numerical Indexes of X—ray Diffraction Patterns, American Society for Testing Materials, Philadelphia, 1954. A. G. Sharpe, Quart. Rev. (London) 11, 49-60 (1957). L. Pauling, The Nature of the Chemical Bond, 2nd Ed., Cornell Univ. Press, Ithaca, New York, 1948. J. A. A. Ketelaar, Chemical Constitution, Elsevier Publ. Co., New York, 1958. G. Barth-Wehranalp, Chem. Eng. 68, 117-22 (1961). K. R. Eilor and R. I. Wagner, Chem.Eng. News 49, 138-50 (1959). D. E. Petersen and K. S. Pitzer, J. Phys.Chem..§1, 1252-3 (1957). H. Kriegsmann, Z. Anorg. u. Allgem. Chem. 299, 138-50 (1959). E. Schnell and E. G. Rochow, J. Am. Chem. Soc. 18, 4178 (1956). C. Curran, R. M. Witucki, and P. A. McCusker, J. Am. Chem. Soc. 12, 4471-4 (1950). L. O. Brockway and I. E. Coop, Trans Faraday Soc. 34, 1429-39 (1938). 57. 58. 59. 60. 61. 62. ‘63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 183 M. Stacey, J. C. Tatlow and A. G. Sharpe, Ed” Advances in Fluorine Chemistry, Vol. 2, Butterworths, Washington, 1963. L. H. Meyer and H. S. Gutowsky, J. Phys. Chem. 51, 481-6 (1953). J. J. Lagowski, Quart. Rev. (London) 13, 233-64 (1959). W. R. Cullen and R. M. Hochstrasser, J. Mol. Spectroscopy 8, 118-32 (1960). W. R. Cullen and D. C. Frost, Can. J. Chem. 49, 391-2 (1962). J. H. Simons and T. J. Brice, Fluorine Chemistry, Vol. 2, Academic Press, New York, 1954. A. L. Henne and C. J. Fox, J. Am. Chem. Soc. 13, 2323-5 (1951) . 0. R. Pierce and A. M. Lovelace, Chem. Eng. News 49, 72-80 (1962). R. N. Haszeldine, M. J. Newlands, and J. B. Plumb, Proc. Chem. Soc. 1960, 147. G. A. R. Brandt, H. J. Emeléus, and R. N. Haszeldine, J. Chem. Soc. 1252, 2549-52. F. G. A. Stone, Chem. Rev. 58, 101—29 (1958). J. W. Dale, H. J. Emeléus, R. N. Haszeldine, and J. H. Moss, J. Chem. Soc. 1957, 3708-13. F. W. Bennett, H. J. Eméleus, and R. N. Haszeldine, J. Chem. Soc. 1953, 1565-71.‘. G. R. A. Brandt, H. J. Eméleus, and R. N..Haszeldine, J. Chem. Soc. 1952, 2552-5. A. B. Burg and W. Mahler, J. Am. Chem. Soc. 89, 6161-7 (1958). W. Mahler and A. B. Burg, J. Am. Chem. Soc. 19, 251 (1957). A. B. Burg, J. Chem. Educ. 31, 482-90 (1960). T. A. Manuel, S. L. Stafford, and F. G. A. Stone, J. Am. Chem. Soc. 83, 249-50 (1961). 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 184 w. R. McClellan, J. Am. Chem. Soc. 83, 1598-1600 (1961). W. Hieber, W. Bech, and E. Lindnev, Z. Naturforsch. 16B, 229-31 (1961). C. G. Swain, R. M. Estene, and R. H. Jones, J. Am. Chem. Soc. 11, 965-71 (1949). A. D. Allen and C. F. Modena, J. Chem. Soc. 1957, 3671-78. E. A. Flood, J. Am. Chem. Soc. 25, 1735 (1933). D. F. Peppard, W. G. Brown, and W. C. Johnson, J. Am. Chem. Soc. 68, 76—7 (1946). R. D. Chambers, H. C. Clark, W. W. Reeves, and C. J. Willis, Can. J. Chem. 39, 258-9 (1961). H. C. Clark and C. J. Willis, J. Am. Chem. Soc. 84, 898-900 (1962). R. N. Haszeldine, J. Chem. Soc. 1951, 102-4. R. N. Haszeldine and B. 0. West, J. Chem. Soc. 1957, 3880-4. L. Kolditz and K. Haage, Z. Anorg. u. Allgem. Chem. 301, 36—53 (1959). A. F. Clifford, H. K. El-Shany, H. J. Eméleus, and R. N. Haszeldine, J. Chem. Soc. 1953, 2372-5. D. B. Sowerby and L. F. Audrieth, J. Chem. Educ. 31, 2-10 (1960). S. Z. Makarov and A. A.)Mkhrushev, Izvest. Akad. Nauk S. S. S. R. Otdel. Khim. Nauk 1960, 1731-8; C. A. 56, 15128c. ' ' ‘— A. J. Saraceno and B. P. Block, U. S. Dept. Com., Office Tech. Serv., AD 410,821, 144i(1963). B. P. Block, E. S. Roth, C. W. Schaumann, and L. R. Ocone, Inorg. Chem. 1, 860-3 (1962). A. J. Saraceno and B. P. Block, Inorg. Chem. 2, 864-5 (1963). O. Schmitz-DuMont, Chem. Tech. (Berlin) 19, 528 (1958). 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 185 H. E. Podall and T. L. Iapalucci, J. Polymer Sci. Bl (9), 457-9 (1963). J. Gobeau and H. Grosse-Ruyken, Z. Anorg. u. Allgem. Chem. 264, 230-48 (1951). M. Schmeisser, Angew. Chem. 66, 713 (1954). D. C. Pease, U. S. 2,840,588, June 24, 1954. Dow Chemical Co., Brit. 627,880, Aug. 16, 1949. W. Autenrieth, A. Anorg. Chem. 35, 2057-60 (1902). F. A. Miller and C. H. Wilkins, Anal. Chem. 24, 1252-94 (1952). H. J. Emeléus and C. J. Wilkins, J. Chem. Soc. 1944, 454-6. D. W. Pletcher and H. S. Nutting (to Dow Chemical Co.) U. 8. 2,436,777, Feb. 24, 1948. H. S. Booth and P. H. Cornell, J. Am. Chem. Soc. 68, 2650-2 (1946). S. Booth and H. S. Habedel, ibid, 2652-4. S Booth and W. F. Martin, ibid, 2655-7. 8. Booth and J. F. Suttle, ibid, 2658-60. S Booth and D. R. Spessard, ibid, 2660-2. S. Booth and A.A..Schwortz, ibid, 2662-5. Swarts, Bull Soc. Chim. 35, 1557 (1924). T. McBee, C. W. Roberts, and G. W. R. Puerckhauer, Am. Chem. Soc. 12, 2329-32 (1957). W. Holbrook and P. L. Brown, (to Dow Corning Corp.) 8. 2,915,544, Dec. 1, 1954. B. Wiberg, Laboratory Techniqge in Organic Chemistry, 208; McGraw—HiII Book Co. Inc., New York, 1960. as 99 as a s s s s s A. I. Vogel, ngsical Organic Chemistry, 3rd Ed., p. 139; SpOttiswoode, Ballantyne and Co., Ltd., London, 1950. 186 113. A. L. Smith, Spectrochim. Acta, lg, 849-62 (1963). 114. H. Stammreich, D. Bassi, O. Sala, and H. Siebert, Spectro- chim. Acta, 13 192-6 (1958). 115. A. L. Smith, Spectrochim. Acta, 16, 87-105, (1960). 116. L. J. Bellamy, The Infrared Spectra of Complex Molecules, John Wiley and Sons, Inc., New York, 1958} 117. N. Wright and M. J. Hunter, J. Am. Chem. Soc., 62, 803 (1947). 118. C. W. Young, P. C. Servais, C. C. Currie, and M. J. Hunter, J. Am. Chem.Soc., 12, 3758 (1948). 119. R. E. Richards and H. W. Thompson, J. Chem. Soc., 1949, 124. 120. H. P. Klug and L. E. Alexander, x-Ray Diffraction Proce- dures, John Wiley and Sons, Inc” New York, 1954. 121. L. Spialter, D. C. Priest, and C. W. Harris, J. Am. Chem. Soc. 11, 6227 (1955). 122. D. H. Whiffen, J. Chem.Soc. 1956, 1350. 123. E. A. V. Ebsworth, M. Onyszchuk, and N. Shephard, J. Chem. Soc. 1958, 1453-60. 124. R. R. Randle and D. H. Whiffen, J. Chem.Soc. 1955, 1311. 125. 0. W. Steward and O. R. Pierce, J. Org. Chem. 26, 2943 (1961). 126. J. E. Griffiths and A. B. Burg, J. Am. Chem.Soc. 84,3442-50 (1962). 127. Private Communication, L. LaPlanche. 128. F. G. A. Stone and D. Seyferth, J. Inorg. and Nuclear Chem., .1, 112-18 (1955). 129. 0. U. Anders, Anal. Chem. 32, 1368-9 (1960). 130. O. U. Anders and W. H. Beamer, Anal. Chem. 33, 226-30 (1961). 131. J. S. Skelcey, Univ. Microfilms (Ann Arbor, Mich.), Order No. 62-1680, Dissertation Abst. 23, 4117-8 (1962). 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 187 R. D. Chambers, H. C. Clark, and C. J. Willis, Can. J. Chem. 39, 131-7 (1961). L. H. Sommer, N. S. Marons, and F. C. Whitmore, J. Am. Chem. Soc. 13, 5127 (1951). U. Klaning and M. C. R. Symons, J. Chem.Soc. 1961, 3204-8. W. F. Edgell, T. R. Riethof, and C. Ward, J. Mol. Spec. 1}, 92-108 (1963). H. C. Clark, J. T. Kwon, and D. Whyman, Can. J. Chem. 41, 2628-33 (1963). 0. W. Steward and O. R. Pierce, J. Am. Chem. Soc. 83, 1916-21 (1961). M. Schmidt and H. Schmidbaur, Angew. Chem. 11, 220 (1959). N. F. Orlov, B. N. Dologov, and M. G. Voronkov, Doklady akad. Nauk S. S. S. R. 122, 246-9 (1958); C. A. _§_3_, 2076a. A. B. Burg, G. Brendel, A. P. Caron, G. L. Juvinall, W. Mahler, K. Modritzer, and P. J. Slota, WADC TR 56—82 Pt. 11 ASTIA Document No. A. D. 118301, OTS release, May, 1957. T. Moeller, Inorganic Chemistry, John Wiley and Sons, Inc., New York, 1952. L. A. Wall, SPE(Soc. plastics Engrs.) Trans., 16, [8], 1-5 (1960). K. A. Andrianov, Soc. Chem. Ind. Monograph 13, 75—88 (1961). F. H. Norton, J. Am. Ceram. Soc. 22, 54-6 (1939). L. Meiter, Ed., Handbook of Analytical Chemistry, McGraw- Hill Book Co., Inc., New York, 1963. R. L. Mital, J. PhyS. Chem. 68, 1613-4 (1964). H. Remy, Treatiseton Inorganic Chemistry, Elsevier Publ. Co., New York, 1956. W. Hfickel, Strggtural Chemistry of Inorganic Compounds, Elsevier Publ. Co., New York, 1951. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 188 D. P. Craig, A. Maccoll, R. S. Nyholm, L. E. Orgel, and L. E. Sutton, J. Chem. Soc. 1954, 332-353. E. G. Rochow and E. Schnell, J. Am. Chem. Soc. 18, 4178 (1956). L. A. Wall, R. E. Donadio, and W. J. Pummer, J. Am. Chem. Soc. 82, 4846-8 (1960). Pople, Schneider, and Bernstein, High-Resolution Nuclear Magnetic Resonance, McGraw—Hill Book Co., Inc., New York, 1959. J. D. Roberts, E. A. McElhill, and R. Armstrong, J. Am. Chem. Soc. 11, 2923-6 (1949). I. R. Beattie and T. Gilson, Nature 193, 1041-2 (1962). G. Fritz and J. Grobe, Z. Anorg. u. Allgem. Chem. 311, 325-30 (1961). M. M. Chamberlain, G. A. Jabs, and B. B. Wayland, J. Org. Chem. 21, 3321-2 (1962). J. R. Ferraro, J. Chem. Educ. 38, 201-8 (1961). C. G. Barraclough, D. C. Bradley, J. Lewis, and I. M. Thomas, J. Chem. Soc. 1961, 2701-5. K. A. Andrianov, USpekhi Khim. '26, 895-922 (1957). "IIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIES 3 1293 03071 2867