. Lettvren r...:,p~. . ‘ . x151 Satin; . Z .43. 3 1.3 , z .i .z. ‘ , 4.31. 3.22321- .3: I 7.5a. (i... v .1. .mu :53, naizrfi ‘ .. . (£2. 5.“? .1311¥!‘|Pt . "J M‘TJ.‘ J-a——I '- uI- —- ~v ”'1 LIBRARW 0‘ .3 ’ {\r'lichjgm .5 Un‘w era“: _ r 'l ‘- l - f M “4. w: - "I' THESIS T;TICTIFJ,2.“.€‘I| STATE UNIVERSITY u; AG.:;.:I=:::?:-: ."af‘D ’IPP'J‘ZC) SJIEKCE DEFART TVTEI“«!T OF Ci-LEIK’EESTRY EAST LANSING, MICHIGAN ABSTRACT PREPARATION OF SILOXY-CHROMIUM COMPOUNDS by Curtis Robert Hare The purpose of this investigation was the preparation of inor- ganic heteropolymers containing silicon and hexavalent chromium. Prior to the preparation of the heteropolymers, a series of bis(triorgano- silyl)chromates were prepared to study the effects of the silicon sub- stituents on the stability of the si1icon-oxygen-chromium(VI) linkage. The preparation of these silylchromates was carried out by reaction of the corresponding silanol with chromium(VI) oxide in methylene chloride. The silylchromates prepared and characterized were bis(p-tolyldiphenyl- silyl)chromate, bis(tricyclohexylsilyl)chromate, and bis(cyclohexyldi- phenylsilyl)chromate; the probable preparation of bis(methyldiphenyl- silyl)chromate was also carried out. The compound bis(triphenylsilyl)- chromate was prepared and characterized in this investigation prior to discovery of a brief description of its synthesis in the patent litera- ture. These silylchromates decompose in light and when heated above their melting points. The most stable is bis(triphenylsilyl)chromate; aliphatic groups reduce the thermal stability. The new silanol cyclo- hexyldiphenylsilanol was prepared by the reaction of cyclohexyllithium and diphenyldichlorosilane in petroleum ether. Reaction of either chromium(VI) oxide or chromyl chloride with diphenylsilanediol gives a mixture of two siloxy-chromium compounds and siloxane impurities. The complete separation of the chromium compounds was difficult. Compound l is a yellow-brown, brittle, glassy solid which is easily melted to a viscous fluid° The structure of this com- pound is (Ph = phenyi): Ph Ph Ph fi Ph Ph Ph HO-Si-O-Si-0-Sfi-O-fr-O-Si-O-Si-O-Sfi-OH (I) I I I I Ph Ph Ph 0 Ph Ph Ph Compound l; is a yellow-orange, crystalline solid with structure: 0 g r /’H \\ Ph2 /,O O O\\ .//Ph2 1 Si I I I I (H) Si Si \ / , Ph2/ O \P. / o \ph2 Cr H 0 Studies of the reactions of either chromium(VI) oxide or chromyl chlor~ ide with the siloxane condensation products of diphenylsilanediol as well as of the reaction of compound T; with the diol enabled postulation of a mechanism for the formation of compounds l_and 1;. Dicyclohexyl- silanediol reacts with chromium(VI) oxide in methylene chloride to give a dark red, amorphous solid with the structure (Ch 2 cyclohexyl): Ch F Ch 0 Ch 0 Ch 0 Ch 0 Ch Ho-SL-o-gr-o—SL-o-gr-o-SI-o-gr-o-SL-o-ur-o-sL-o-gr-o-sg-oH (III) I II I I I I II II II Ch 0 Ch 0 Ch 0 Ch g Ch 0 Ch Compounds T and TI; are believed to be the first compounds of their kind to be isolated and identified. Compound 3; is believed to be the first reported cyclic siloxane containing chromium. Compound Tl; decomposes on standing, compound 1; has a sharp decomposition point at 169°C, and differential thermal analysis of compound l indicates decomposition at 279°C, which leads to the conclusion that increased chromium(VI)-to- silicon content in chromium(VI) heterosiloxanes brings on reduced thermal stability. Ultraviolet and visible spectra of the compounds prepared suggest an interaction involving the substituents of the silicon atom with the chromate group. The infrared absorption of the silylchromates in the ll.h-ll.5p. region was tentatively assigned to the chromium—oxygen-silicon linkage. PREPARATION OF SILOXY-CHROMIUM COMPOUNDS By Curtis Robert Hare A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1961 '6123/87 0/2/62, ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Professor Robert N. Hammer for his guidance and c00peration during this investigation. Appreciation is also extended to Professor Jack B. Kinsinger for many helpful suggestions pertaining to this work. Grateful acknowledgment is also made to my associates, John F. Benner, James S. Skelcey, and Paul M. Dupree, for enlightening discussions of the problem, and to Sister M. Clarus Strouth for assistance with the differential thermal analyses. Dr. Michael N. Hoechstetter's suggestions, aid, and contributions to the apparatus for the determination of molecular weights are also very much appreciated. And finally, the author gratefully acknowledges the COOpera- tion and assistance of his wife, who was more than the typist of this thesis. ii TABLE OF CONTENTS INTRODUCTION 0 O O O O 0 O O O O O O O O 0 O O O o O O l 0 HISTORY. O O O O O O O O O O O O O O ‘ O O O O O O 0 O O O 0 Chemistry of Hexavalent Chromium. . . . . . . . . . . Chemistry of Chromyl Chloride. . . . . . . . . . Other Chromyl Compounds. . . . . . . . . . . Chromium(VI) Oxide, Chromates and Polychromates. Chromate Esters. . . . . . . . . . . . . . . . . Toxicity of Hexavalent Chromium. . . . . . . . . Chemistry of Organosilicon Compounds. . . . . . . . . Bonding Considerations in Organosilicon Compounds and Chromates. . . . . . . . . . . . . . . . . . . Silicon. . . . . . . . . . . . . . . . . . . . . Chromium . . . . . . . . . . . . . . . . . . Inorganic Polymerization Reactions. . . . . . . . . . mm MNTAL . C . O O O C O O O O O O O O O O O O O O O 0 Preparation of Reactants. . . . . . . . . . . . . . . Synthesis of Cyclohexyldiphenylsilanol . . . . . Preparation of Tricyclohexylsilanol. . . . . . . Preparation of Dicyclohexylsilanediol. . . . . . Preparation of p-Tolyldiphenylsilanol. . . . . . Preparation of Tri(l- naphthyl)silanol. . . . . . Preparation of Methyldiphenylsilanol . . . . . . Preparation of Hexaphenylcyclotrisiloxane and Octa- phenylcyclotetrasiloxane. . . . . . . . . . . Preparation of Hexaphenyltrisiloxane-l,5-diol. . Preparation of Chromyl Chloride. . . . . . . . . Preparation of Silylchromates . . . . . . . . . . . . Bis(triphenylsilyl)chromate. . . . . . . . . . . Bis(pftolyldiphenylsilyl)chromate. . . . . . . . Bis(cyclohexyldiphenylsilyl)chromate . . . . . Bis(tricyclohexylsilyl)chromate. . . . . . . . . Bis(trimethylsilyl)chromate. . . . . . . . . . . Bis(methyldiphenylsilyl)chromate . . . . . . . . Bis[tri(l-naphthyl)silyl]chromate. . . . . . . . Properties of Silylchromates . . . . . . . . . . Reactions of Diphenylsilanediol and Its Condensation Products with Chromium(VI) Compounds . . . . . . . Reaction of Diphenylsilanediol and Chromyl Chloride. Reaction of Diphenylsilanediol and Chromium(VI) Oxide. iii Page TABLE OF CONTENTS - Continued Page Characteristics of Compounds I and II . . . . . . . 55 Reaction of Cyclic Siloxane with Chromyl Chloride . . 60 Reaction of Baxaphenyltrisiloxane- -l ,5- diol and Tetraphenyldisiloxane-l, 3-diol with Chromium(VI) Compounds. . . . . . . . . . . . 62 Reaction of Diphenyldialkoxysilanes with Chromyl Chloride. 63 Reaction of Dicyclohexylsilanediol with Chromium(VI) Oxide . . . . . . . . . . . . . . . . 65 Spectroscopic Data . . . . . . . . . . . . . 68 Ultraviolet and Visible Spectra . . . . . . . . . . . 68 Infrared Spectra. . . . . . . . . . . . . . . . . 68 Discussion. . . . . . . . . . . . . . . . . . . . . . 73 Analytical Methods . . . . . . . . . . . . . . . . . . . . 76 Chromium Analysis . . . . . . . . . . . . . . . . . . 76 Silicon Analysis. . . . . . . . . . . . . . . . . . . 77 Molecular Weights . . . . . . . . . . . . . . . . . . 78 DISCUSSION. . . . . . . . . . . . . . . . . . . . . . . . . . . 82 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . 95 RECOMMENDATIONS FOR FUTURE WORK . . . . . . . . . . . . . . . . 98 APPENDIX 1. . . . . . . . . . . . . . . . . . . . . . . . . . . 100 APPENDIX 2. . . . . . . . . . . . . . . . . . . . . . . . . . . 102 APPENDIX 3. . . . . . . . . . . . . . . . . . . . . . . . . . . 119 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . . . . 125 iv LIST OF TABLES TABLE Page I Bond Lengths and Bond Angles in Chromyl Chloride . . . . . A II Solubility and Reactivity of Chromyl Chloride in Liquid Solvents . . . . . . . . . . . . . . . . . . 6 III Solubility of Solids in Chromyl Chloride . . . . . . . . . 7 IV Chromate Esters of Various Tertiary Alcohols . . . . . . . 15 V Metal Siloxy Compounds . . . . . . . . . . . . . . . . . . 35 VI Organosilyl Chromates Prepared . . . . . . . . . . . . . . A6 VII Analytical Data for Compound I from the Reaction of Di- phenylsilanediol with Chromyl Chloride . . . . . . . . . . 5h VIII Elemental Analysis of Compound II. . . . . . . . . . . . . 58 IX Infrared Absorptions of Silylchromates in 082 (O). . . . . 72 X Calibration of the Thermistor. ... . . . . . . . . . . . . 80 XI Interatomic Spacings for Bis(triphenylsilyl)chromate . . . 119 XII Interatomic Spacings for Bis(p—tolyldiphenylsilyl)chro— mate . . . . . . . . . . . . . . . . . . . . . . . . 120 XIII Interatomic Spacings for Bis(tricyclohexylsilyl)chromate . 121 XIV Interatomic Spacings for Compound II . . . . . . . . . . . 122 FIGURE 10. ll. 12. 13. 1h. 15. LIST OF FIGURES Differential Thermal Analysis of Compound I . . . . . Ultraviolet and Visible Spectra of Bis(triphenylsilyl)— chromate, Bis(p- tolyldiphenylsilyl)chromate, Bis(cyclo- hexyldiphenylsilyl)chromate, and Bis(trimethylsilyl)- chromate. . . . . . . . . . . . . . Ultraviolet and Visible Spectra of Compounds I and II o Ultraviolet and Visible Spectra of Bis(tricyclohexyl- silyl)chromate and Compound III . . . . . . . . . . Ultraviolet Spectra of Triphenylsilanol and Diphenyl- silanediol. . . . . . . . Infrared Spectrum of Bis(triphenylsilyl)chromate in Carbon Disulfide . . . . . . . Infrared Spectrum of Triphenylsilanol in Potassium Bro- mide Disc . . . . . . . . . . . . . . . . . . Infrared Spectrum of Bis(p- tolyldiphenylsilyl)chromate in Carbon Disulfide . . . . . . . . . . . Infrared Spectrum of p-Tolyldiphenylsilanol in Carbon Disulfide . . . . . . . . . . . . . . . . Infrared Spectrum of Bis(cyclohexyldiphenylsilyl)chro- mate in Carbon Disulfide. . . . . . . . . . Infrared Spectrum of Cyclohexyldiphenylsilanol in Car- bon Disulfide . . . . . . . . . . . . . . . . . . . Infrared Spectrum of Bis(tricyclohexylsilyl)chromate in Carbon Disulfide. . . . . . . . . . . Infrared Spectrum of Tricyclohexylsilanol in Potassium Bromide Disc. . . . . . . . . . . . . . . . . . . . . . Infrared Spectrum of Bis(trimethylsilyl)chromate in Carbon Disulfide. . . . . . . . . Infrared Spectrum of Trimethylsilanol in Carbon Disul- fide. . . . . . . . . . . . . . . . . . Page 57 69 7o 71 75 103 10h 105 106 107 108 109 110 111 112 LIST OF FIGURES - Continued FIGURE Page 16. Infrared Spectrum of Compound I in Carbon Disulfide . o a 113 17. Infrared Spectrum of Diphenylsilanediol in Potassium Bromide Disc. . . . . o . . . . . o . . . . . o . a o 0 11h 18. Infrared Spectrum of Compound II in Carbon Disulfide. . o 115 19. Infrared Spectrum of Tetraphenyldisiloxane-1,3-—diol in Carbon Disulfide . . . . . o . . . . o . o . o 116 20. Infrared Spectrum of Compound III in Carbon Disulfide . a 117 21. Infrared Spectrum of Dicyclohexylsilanediol in Carbon Disulfide . . a . . . . a . o . . . . . o . . . . a o 118 22. X— ~ray Diffraction Pattern of Bis(triphenylsilyl)chro— mate. a . . . . . . . . . . . . . . . a . . 123 23. X- -ray Diffraction Pattern of Bis(p- tolyldiphenylsilyl)- chromate. . . . . . . . . . . . . o . . o . . 123 2h. X- -ray Diffraction Pattern of Bis(tricyclohexylsilyl)- chromate. . . . . . . . . . . . . . . . . . . . . . 123 25. X-ray Diffraction Pattern of Compound I . . . . o . . . . 12h 26. X-ray Diffraction Pattern of Compound I} . . . . . ., . .. 121+ INTRODUCTION Research in high temperature polymers recently has developed be- cause of the growing need of industry and the military services for mater- ials of construction which will endure temperatures of 500°C and above. Thermosetting and thermOplastic materials, elastomers, adhesives, coat- ings, and hydraulic fluids based on a carbon skeletal framework undergo oxidation and bond rearrangement at high temperatures. The structural relationships between the metal silicates which compose the earth's crust and the man-made polysiloxanes (silicones) has led« many investigators (l) to conclude that the introduction of metals into the siloxane moiety may produce polymers stable at high tempera- tures. A wide latitude for variation in obtaining Optimal prOperties is available through the selection of functional groups which block valences of silicon and in the choice of metals for incorporation into the poly- siloxane. The choice of metals or metal-containing reactants is consid— erably restricted when the requirement of difunctionality is imposed. Further, the metal reactant must conform to the tetracoordination of the polysiloxanes to produce high molecular weight linear polymers. Higher coordination will produce cross-linked resins which may lead to brittle- ness. Several compounds of hexavalent chromium offer desirable proper- ties which could be incorporated into the siloxane linkage. Hexavalent chromium is a powerful oxidizing agent and thus is resistant to oxidation. Hexavalent chromium compounds are difunctional reactants and show tetracoordination. Chromyl Chloride and the chromates are known to undergo self-polymerization, and chromium(VI) oxide is com- posed of long tetrahedral CrOh chains in the solid state. Recent in- vestigations of these compounds show that their bonding is considerably covalent, which is also the bonding of the siloxanes. Consideration of these prOperties leads to the conclusion that an attempt should be made to incorporate hexavalent chromium into the siloxane linkage. The pri- mary reactants chosen for this study were chromyl chloride and chro- mium(VI) oxide. When this investigation was undertaken in 1958, no information was available on the chemistry of the hexavalent chromium-oxygen-silicon linkage. The large variety of known silicon intermediates which are commercially available or relatively easy to prepare provided an oppor- tunity to undertake a systematic study of the environmental effects on the chromium-oxygen-silicon linkage. The methods of preparation and characterization of organosiloxy Chromates, and also the properties of these compounds, are the subject of this investigation. HISTORY Chemistry of Hexavalent Chromium The chemistry of hexavalent chromium is marked by a few general characteristics. All the compounds of hexavalent chromium are strong oxidizing agents. In general, the ultimate reduction product is tri- valent chromium. Compounds of hexavalent chromium all exhibit approxi- mately tetrahedral coordination in solution and in the solid state. These compounds also have a tendency to form higher aggregates. Two examples of this are the formation of polychromates in strongly acidic chromate solutions and self-association of chromyl chloride in non- aqueous solutions. Chemistry of Chromyl Chloride. Chromyl chloride is a cherry—red transparent liquid having a boiling point of llS.7°C and a melting point of -96.5°C. The appearance and odor of this compound are similar to those of bromine. The compound was first prepared by Berzelius (2) by distilling a mixture of a chromate, sodium chloride, and sulfuric acid. Chromyl chloride is most conveniently prepared by the reaction of chromium(VI) oxide with concentrated aqueous hydrogen chloride in the presence of sulfuric acid as a dehydrating agent (3). Liquid chromyl chloride is easily separated and purified by distillation. Other methods are based on the reaction of chromium(VI) oxide, 8 chromate, or a di- chromate with an acid chloride followed by separation and purification by distillation. Hartford and Darrin in a recent review (4) have brought together the scattered information on chromyl compounds and report a large amount of unpublished information from their laboratory. Greater than half of the articles discussed in this review are from the nineteenth century. While much is known about the properties of the chromyl compounds, there is a definite need for more information on their chemistry. The structure of chromyl chloride was determined by Palmer (5) by electron diffraction measurements. These data (Table I) show the structure to be approximately tetrahedral. TABIE I BOND LENGTHS AND BOND ANGLES IN CHROMYL CHLORIDE (5) Bond Bond Length in A Bond Bond Angle Cr-O 1.571: 0.03 O-Cr-O 10506' i 1+0 Cr-Cl 2.12 t 0.02 Cl-Cr-Cl 113°16'i: 30 01-0 3.03 t 0.03 Cl-Cr-O 109°3u'i 3° 01-01 3.51 i 0.05 0-0 2.1+9 : 0.10 The symmetric structure is also substantiated by the low dipole moment (O.h7 D) and low dielectric constant (2.2-2.3) determined by Smyth, Grossman, and Ginsburg (6). walden (7) has also determined the dielectric constant and obtained a value of 2.6. The specific conduc- tivity is reported to be less than 10"10 ohm'1 (8). These data also suggest that the Cr-O and Cr-Cl bonds are nearly equivalent, which would account for the covalent and non-polar nature of this compound. The density of chromyl chloride obeys the following equation: (1 = 1.959 - 1.913 x lO'3t + 1.797 x 10'6t2 in which t is the temperature in degrees centigrade. This gives a value of 1.912 g/ml at 25°C. The vapor pressure may be calculated from 10gloP(mm) = 7.723 - 313.8; This gives a value of 8620 cal/mole for the heat of vaporization and a Trouton's constant of 22.1. The viscosity is about that of water (0.8 centipoise) at room temperature, but it is less temperature dependent. Chromyl chloride is highly soluble in most halogenated organic liquids and covalent halides such as stannic chloride. It is also solu- ble in common organic solvents, but chromyl chloride often reacts vio- lently with them. In general, it dissolves in most non-polar compounds and is insoluble in most polar compounds. Tables II and III give the solubility and reactivity (qualitative) of various substances with chromyl chloride. Several studies have been made of the molecular weight of chromyl' chloride in various solvents by cryos00pic and ebullioscopic means (9, 10, ll, 12). In nearly all cases, the values of the molecular weight indicate a degree of polymerization of chromyl chloride from 1.5-3. Ebullioscopic measurements in carbon tetrachloride and carbon disulfide give the calculated value. Cryoscopic measurements in stannic bromide, stannic iodide, antimony trichloride, antimony triiodide, and arsenic triiodide give values of one-half that calculated for chromyl chloride. The following is an example of the reaction proposed to take place in the latter solvents (l3): ,‘1 ++ : 0r02C12+ SnBrb, ——> 0102 + SnBrhC‘l2 The vapor density of chromyl chloride indicates the presence of only the Cr02012 molecule. TABLE II SOLUBILITY AND REACTIVITY OF CHROMYL CHLORIDE IN LIQUID SOLVENTS (h) Miscible; Miscible; Miscible, Rapid Essentially No Reaction Slow Reaction or Violent Reaction Immiscible carbon tetrachloride acetic anhydride acetyl bromide formic acid 1,2-dichloroethane acetic acid diethylsuccinate hydrofluoric acid 1,2-dibromoethane nitric acid (100%) pyridine phosphoric acid (85%) thionyl chloride n-hexane sulfur monochlor- sulfuric ide acid (70%) stannic chloride benzene toluene mercury acetyl chloride cyclohexane pfcymene methylene chloride nitrobenzene chlorosulfuric acid l,l,2,2-tetrachloro- carbon disulfide oleum ethane chloroform aniline Eggtfbutylamine triethanolamine methanol ethanol TABLE III SOLUBILITY OF SOLIDS IN CHROMYL CHLORIDE (h) Very Soluble Slightly Soluble Insoluble iodine (reacts) phenol (surface reaction) zinc chloride p—dichlorobenzene pentachlorophenol sodium chloro- chromate chromium(VI) oxide aluminum chloride sodium chromate naphthalene (reacts) ammonium hexachloro- stannate sulfur metals (surface reaction) Pure chromyl chloride may be stored in stainless steel, aluminum, or dark glass containers fitted with Teflon or Kel-F stoppers. In glass ‘ the material is stable if protected from light. When exposed to light, chromyl chloride decomposes into a black solid and chlorine gas. Schwab and Prakash (1%) studied the photochemical decomposition of chromyl chloride in carbon tetrachloride solution and proposed the fol~ lowing mechanism: Cr02012 + hv —) 01- + 0r0201- (initiation) 01' + Cr02012 -—> Cr02C1-+ C12 (propagation) Cr0201- + Cr02012-—) 2Cr02 + 012+ 01- (propagation) C1'-+ Cl' —————) C12 (termination) CrO2Cl- -——) CrO2 + 01- (termination) Chromyl chloride is thermally stable to 180°. Above 180°C a dark decomposition product begins to form, and at #0000 a magnetic oxide having the composition Cr509 results. On further heating, chromium(III) oxide is formed (h). Chromyl chloride is immiscible in water and hydrolyzes only slowly at the interface to form chlorides, chromates, and chlorochro- mates. It reacts with ammonia even in concentrated aqueous solution with incandescence to produce a brown solid which is apparently a mix~ ture of chromium(III) oxide and chromium nitride. Hydrogen sulfide and sulfur monochloride are oxidized by chromyl chloride. PhOSphorus trim chloride reacts explosively with liquid chromyl chloride but it forms an addition compound with composition CrOCl'P0013 when a carbon tetra- chloride solution of chromyl chloride is used. Phosphorus pentachloride reacts with the liquid to give phosphorus oxychloride, chromium(III) chloride, and chlorine. It forms an addition compound of composition GrOECle-P015 with phosphorus pentachloride in carbon tetrachloride. Chlorochromates are produced by the action of chromyl chloride on con= centrated chromate solutions. Chlorosulfuric acid and pyrosulfuric acid decompose chromyl chloride to produce a brown solid which has a composition corresponding to chromyl sulfate. Except for tin and antimony which are readily attacked, most metals are only superficially attacked by chromyl chloride. It is re- ported that zinc and sodium react with chromyl chloride vapor to pro- duce metallic chromium. Hydrogen reduces chromyl chloride to Cr203 (h) at elevated temperatures. Graphite reacts with chromyl chloride to form intercalary com- pounds containing 36.3-h0.h% Cr02012. The interplanar spacing of the ‘F graphite is increased from h.5 A to 9 A. When the compound is heated to 200-30000, the chromyl chloride is released and the graphite is ex“ foliated to many times its original volume (15). Up to 23% of the chro» myl chloride is retained at 30000. Clean white phosphorus reacts vigorously with chromyl chloride, whereas sulfur reacts with less vigor. The products of these reactions are unknown. The reactions of chromyl chloride with organic substances have received considerable attention (h). Historically, Etard was first to use this reagent to prepare benzaldehyde from toluene (16). In general, the Etard reaction involves the addition of chromyl chloride to an or» ganic molecule at one or more hydrogen atoms. The addition product is separated and hydrolyzed to give aldehydes, ketones, and quinones. The addition compound is formed by adding a solution of chromyl chloride in an inert solvent (carbon disulfide or carbon tetrachloride) to an excess of the organic compound in the same solvent. The insoluble addition compound is separated after the disappearance of the red color of chro- myl chloride. The addition compound is then decomposed with water or alcohol to give carbonyl compounds. When chromyl chloride is reduced in an alcoholic solution in the presence of an organic acid, a trivalent chromium complex of the acid is formed. The complex is soluble in organic solvents and is water re- pellent if the organic acid consists of a long carbon chain. If the acid contains only a few carbon atoms, the complex is a repellent of fats and oils. Highly fluorinated organic acids give complexes which are repellent to water and oils (8). 10 lastly, it should be pointed out that if the heat evolved from the oxidation of organic material by chromyl chloride is not dissipated, the kindling temperature is achieved and a violent explosion may result. Thus care should always be exercised in the use of chromyl chloride, and it is always best to use an inert solvent to reduce the reactivity. Other Chromyl Compounds. Numerous other chromyl compounds have been described in the literature, but the chemistry of these compounds has not been studied as extensively as that of chromyl chloride. The chromyls that have been described are the fluoride, acetate, nitrate, perchlorate, bromide, thiocyanate (Isgfl), cyanate (Iso?), sulfate, pyrosulfate, and azide. Of these, only the fluoride and the nitrate have been studied extensively. The fluoride has been prepared in a very high state of purity by the reaction of anhydrous hydrogen fluoride and chromium(VI) oxide. This was purified by distillation or recrystallization in anhydrous hy- drogen fluoride (17). It may also be prepared by the reaction of iodine pentafluoride with chromium(VI) oxide, antimony pentafluoride, or fluorine with chromyl chloride and anhydrous hydrogen fluoride with potassium dichromate. The violet-red solid sublimes at 29.600 to a reddish-brown vapor. The chemistry of chromyl fluoride is similar to that of chromyl chloride. Chromyl fluoride exchanges its fluorine with metal oxides, forming boron trifluoride from boric oxide, silicon tetra- fluoride from glass, and arsenic fluorides from arsenous oxide (h). The mixed halide Cr0201F has been prepared by Flesch and Svec (18) by mixing chromyl chloride and fluoride. The reactivity and oxidizing strength of chromyl fluoride is considered to be greater than that of 11 chromyl chloride (h). This reagent may also be used for the Etard re- action. Chromyl nitrate is obtained as a deep red liquid by the reaction of dinitrogen pentoxide on chromyl chloride or chromium(VI) oxide (19). The liquid is unstable and slowly decomposes on standing even in the absence of light. It is considered to be a stronger oxidizing agent than either chromyl fluoride or chloride. It also may be used for the Etard reaction. Schmeissen has also reported (20) the preparation of chromyl per- chlorate from dichlorine hexoxide and chromium(VI) oxide. This compound is very unstable and its existence has been questioned (h). Chromyl bromide has been reported by Zellner (21) as the red— brown product of the reaction of chromyl chloride and anhydrous hydro- gen bromide at -70°C. The product was contaminated with chromyl chloride and decomposed at room temperature. Chromyl acetate has recently been prepared (22) by the reaction of chromium(VI) oxide with acetic anhydride in carbon tetrachloride in the absence of light and moisture. The red crystals can be distilled in high vacuum. The other chromyl compounds that have been mentioned are reported to be the products of metathetical reactions with chromyl chloride. Since these compounds were never characterized, their existence is questionable. Chromium(VI) Oxide, Chromates, and Polychromates. Chromium(VI) oxide is precipitated from alkali chromate or dichromate solutions by addition of a large excess of sulfuric acid. The dark red needles are —Y—v—y 1 12 washed with concentrated nitric acid and dried at 7000. The melting point is reported to be 197°C (23). The structure of chromium(VI) oxide has been determined by Bystrom and Wilhelmi (2h). Oxygen atoms form a distorted tetrahedron around each chromium atom, and the Cr-O distances are 1.79-1.81 A. The tetrahedra are linked into chains by sharing corners, but between CrO3 chains probably only weak van der waals bonds are operative. Many textbooks still report the erroneous structure of Brakken (25) which is based on Cr06 octahedra. Crystalline chromium(VI) oxide isgflezoelectric (26), indicating a permanent dipole exists in the crystal. Chromium(VI) oxide is also paramagnetic (27), as are the chro- mates. This paramagnetism is attributed to imperfect pairing of the electrons in the ground state. The differential thermal analysis of chromium(VI) oxide shows an endothermal reaction at 178-18700 which vasenin (28) attributes to fusion, and an exothermal reaction at 267-3h800 which is ascribed to decomposi- tion. Chromium(VI) oxide is a powerful oxidizing agent. It oxidizes alkali metals, sulfur, arsenic, phosphorus, hydrogen, ammonia, and hydro- gen sulfide. Alcohols may be oxidized by aqueous solutions of chromium(VI) oxide to aldehydes, ketones, or acids, depending on the conditions. These oxidations may also be carried out in glacial acetic acid or ace- tic anhydride in which chromium(VI) oxide is only slightly soluble. Chromium(VI) oxide also forms addition compounds with hetero- cyclic nitrogen bases such as pyridine (29, 30). The structure of the pyridine addition compound is considered to be trigonal bipyramidal with 13 the pyridine molecules at the apexes. These compounds are insoluble in non-polar organic solvents but are slightly soluble in polar solvents. They dissolve in the parent nitrogen base. All are photo-sensitive and decompose when heated above 100°C. The chemistry of the metal chromates is quite familiar. Alkali metal chromates are prepared commercially by fusion of chromite ore with alkali, followed by recrystallization. The principal uses of chro- mates are as oxidizing agents, in metal treatment, and for tanning leather. Acidic solutions of chromates have been shown to consist of the dichromate ion. Highly concentrated aqueous solutions of chromium(VI) oxide and highly acidic chrOmate solutions contain the trichromate and tetrachromate ions. The salts of these polychromates are prepared by the addition of chromium(VI) oxide to concentrated aqueous solutions of the metal dichromates. The color of the chromates passes from yellow for the monochromate to dark red for the tetrachromate. The following reactions illustrate aggregation of chromates in acidic solution: CrO): + H+ ——) (HCrOh') A 2(HCrOh') ‘_____ Cr207'-+ H20 = + ———A - Cr207 + 2H F— Cr3O:LO 4. H20 I «9—4 = Cr3010 + 2H ‘___ 0111013 + 1120 : +___§ nCrOh + 2_nH ‘__ (0r03)n_ + 31120 Chromate Esters. Gomberg (31) was the first to prepare a chro- mate ester. He prepared bis(triphenylcarbinyl)chromate by the reaction of silver chromate with triphenylmethyl chloride in benzene. Numerous other chromate esters have been prepared, but those which are insoluble are derived from tertiary alcohols only. 1h Several other methods have been described for the preparation of chromate esters. The most common method is to add chromium(VI) oxide to a solution of the alcohol in an inert solvent. The chromate esters may also be prepared by reaction of the alcohol with chromyl chloride in chloromethane solvent or sodium dichromate dihydrate in glacial acetic acid. When chromium(VI) oxide is added to a solution of a primary or secondary alcohol in an inert solvent such as benzene, a yellow to yellow-orange color is produced. Within a short period of time the solution becomes turbid and the yellow color fades. Attempts to iso- late these unstable chromates have been futile. Westheimer (32) has discussed the participation of a chromate ester in the oxidation of iSOpr0pyl alcohol by chromic acid. His mechan- ism for the oxidation involves two steps: formation of a chromate half ester followed by decomposition of the chromate half ester by removal of the d-hydrogen atom of the propyl group to form acetone. The latter decomposition is considered to be the rate determining step. The lack of stability of the primary and secondary alcohols may be related to the ease of removal of their c(-hydrogen atoms. The chromate esters which have been prepared, isolated, and characterized are listed in Table IV according to the alcohols from which they are derived. Other chromate esters have been prOposed, but they have not been isolated in pure form or characterized. Zeiss and Matthews (36) claim that the chromate esters may be ob- tained in pure form by a freeze-drying technique which removes all volatiles except the chromate ester. They also report that the chromate 15 esters may be stored indefinitely in the dark at ~6OOC. TABLE IV CHROMATE ESTERS OF VARIOUS TERTIARY AICOHOIS Alcohol Properties of Chromate Reference triphenylcarbinol m.p. l7h°C (decomp.) 31 tert-butanol liquid, decomp. 3h, 36 2-methylfenchol m.p. 109.500 33, 37 2-ethylfenchol decomp. lhOOC 33 2-methylborneol decomp. 105°C 33, 35, 37 cedrol decomp. 122°C 33, 35 l-octadecylcyclohexanol m_.p° 37-3800 (decomp.) 35 decahydro-Ba-napthol m.p. 7h-76OC (sealed tube) 35 gig-terpene hydrate decomp. 135°C 35 2,h-dimethyl-h—hexanol liquid, decomp. 36 3-ethyl-3-pentanol liquid, decomp. 36 Several studies have been made of the reactions of the chromate esters (35, 36, 37, 38, 39), particularly their hydrolysis and alcoholy- sis. In all cases, hydrolysis and alcoholysis yield the parent alcohol and a reduced chromium species which is claimed to be (H2Cr03). The hydrolysis of bis(t§rt-butyl)chromate and bis(triphenylmethyl)chromate has been carried out in water enriched in 018(38). These studies reveal 'that the oxygen-chromium bond is broken on hydrolysis in the case of ..._.._- l6 bis(t§rtfbutyl)chromate, and that the carbon-oxygen bond is broken in bis(triphenylmethyl)chromate. The hydrolysis or methanolysis of bis(2,h-dimethyl-h-hexyl)chromate derived from the dextrorotary alcohol gave the dextrorotary alcohol with retention of configuration, which substantiates chromium-oxygen bond,fission on hydrolysis or alcoholysis. Toxicity of Hexavalent Chromium. The toxic action of chromium is confined to the hexavalent state.(23). These compounds have an extreme- 1y irritating, corrosive, and, under some circumstances, toxic action on body tissue. The most common effects of chromium compounds on industrial workers are those which result from direct contact of the skin with chro- mates or chromium(VI) oxide. These skin reactions are of two different types: chrome ulcers and chrome dermatitis. Chrome ulcers are very persistent without prOper treatment. They are generally caused by contact only with the broken skin; however, even contact with the slightest abrasion of the skin may cause ulcers. Chrome dermatitis may result from contact of the skin with hexavalent chromium. This is easily avoided by proper washing after contact. Perforation of the nasal septum is a common result of inhalation of chromate dust.. The perforation is not disabling in any way, and it does not interfere with normal breathing, but inhalation may cause can- cer of the respiratory tract. Many studies have shown that the death rate due to respiratory cancer is very high among chromate workers. Also, prolonged contact with chromate dusts may cause chronic irritation of the respiratory tract, but there is no evidence that chromate workers are susceptible to tuberculosis. IIVII 17 Oral ingestion of hexavalent chromium compounds may lead to in- tense irritation of the gastrointestinal tract resulting in violent epigastric pain, nausea, vomiting, severe diarrhea, and hemorrhages. With large doses, circulatory collapse, unconsciousness, and death may follow rapidly. Other systemic effects from ingestion which have been reported in men or experimental animals include bronchi, serious cavi- ties, liver damage, hyperglycemia, glycosuria, marked fall in blood pressure, leukocytosis, anemia, muscle cramps, loss of weight, stagger- ing, convulsions, paralysis, and death. The fatal dose by ingestion has been determined to be one gram for chromium(VI) oxide and six grams for potassium dichromate. Symptoms . 3 of intoxication have been reported from ingestion of 0.5 g of potassium dichromate. It is also believed that exposure to chromates may increase susceptibility to infectious diseases. Nbrmal caution is the best personal protection against chromium poisoning. Safety glasses, rubber gloves, and protective clothing should be used when handling chromates for extended periods. The person work- ing with hexavalent chromium compounds should also shower promptly after : work. - Chemistry of Organosilicon Compounds The chemistry of the organosilicon compounds has been widely studied, largely because of the commercial value of the familiar sili- cones. The books by Rochow (ho), Andrianov (hl), and Gmelin”s Handbuch der Anorganischen Chemie (#2) are particularly good for their coverage 18 of organosilicon chemistry. These references have been used extensive- ly for the following discussion. Organosilicon compounds are prepared by two general methods. The first is a direct method in which a reaction of silicon or silicon compounds and organic reagents at high temperatures gives the organo- silicon compound (e.g., cyclohexane and silicochloroform to give cyclo- hexyltrichlorosilane). The second is the reaction of halo or alkoxy silanes with organometallic reagents (e.g., methylmagnesium iodide and ethylsilicate to give dimethyldiethoxysilane). The successive addition of organic groups to the silicon atom increases in difficulty with the number of groups bound to the silicon atom. The organochlorosilanes form the most important group of organo- silicon compounds since they are easily prepared and serve as intermed- iates in the preparation of other compounds. The disubstituted chloro- silanes are easily prepared by passing an organic chloride over silicon metal at 275-37500 in the presence of c0pper as a catalyst. The mono and trisubstituted chlorosilanes are prepared by the reaction of organo- metallic reagents with silicon tetrachloride. The organometallics of zinc, mercury, aluminum, sodium, lithium, and Grignard reagent are the common organometallic reagents used to prepare organosilicon compounds. The latter two reagents have been most widely used. The organosubstituted ethoxysilanes are prepared directly by the reaction of organometallic reagents with ethyl silicate or by the re- action of ethyl alcohol with a chlorosilane. Similar reactions are used to prepare the other alkoxy silanes. 19 Organosilanols or diols are prepared by the hydrolysis of the chlorosilane or alkoxysilane. The hydrolysis of organotrichloro or triethoxysilanes leads only to resinous polymers. The hydrolysis of the dichlorosilanes illustrates the differences in the chemistry of these compounds. When dimethyldichlorosilane is hydrolyzed, the product is octamethylcyclotetrasiloxane,[(033)2SiO]u, and other siloxanes. The preparation of dimethylsilanediol is only achieved by careful hydrolysis of dimethyldiethoxysilane (#3), but the diol slowly condenses on standing to siloxanes. Diethylsilanediol can be prepared by careful hydrolysis of the dichlorosilane at 0°C, and the direct hydrolysis of diphenyldichlorosilane gives only diphenylsilane- diol. Diphenylsilanediol is different from dimethylsilanediol in that it only undergoes condensation when heated above its melting point. It condenses in the presence of acids to hexaphenylcyclotrisiloxane, and in the presence of base to octaphenylcyclotetrasiloxane. The hydrolysis of dimethyldichlorosilane is the basis of the pre- paration of the polysiloxanes. Polysiloxanes are particularly important because of unique properties. They are heat resistant, moisture resis- tant, have good dielectric properties, and have considerable compressi- bility under high pressure. They form the basis of rubbers of high thermal stability, resins for plastics and electrical insulation, and water repellent liquids with a low temperature dependence of viscosity. The direct hydrolysis of dimethyldichlorosilane yields an oil Which is composed of a mixture of siloxanes. Distillation of the oil under reduced pressure separates a distillate composed largely of 20 octamethylcyclotetrasiloxane and lesser amounts of the cyclic pentamer, hexamer, and trimer. The residue is composed of higher cyclic siloxanes and high molecular weight linear siloxanes. This residue is treated with trimethylchlorosilane to produce methyl silicone oil. This treat- ment with trimethylchlorosilane is necessary to prevent further conden- sation of linear polymers. Silicone rubber is produced by further polymerization of the original hydrolysis product of dimethyldichlorosilane. Polymerization is accomplished by means of strong acids or bases which are capable of opening the cyclic rings to produce cations or anions which lead to high polymers by addition polymerization. Molecular weights as high as sev— eral million have been reported for elastomeric dimethylpolysiloxane. Silicone resins are produced by three different methods. The first is hydrolysis of dimethyldichlorosilane to form dimethylpoly— siloxane and then oxidation of the product with air to attain the de- sired CH3/Si ratio. The second method is based on the cohydrolysis of dimethyldichlorosilane and methyltrichlorosilane to produce the resin by co-condensation. Also, silicon tetrachloride may be methylated to the desired CH3/Si ratio and the reaction mixture hydrolyzed to the resin. The structure of polysiloxane fluids and elastomers is based on the linear disubstituted siloxy linkage [(R)2SiO]. Fluids and elastomers differ only in molecular weight. PolySiloxane resins consist of a three- dimensional structure with cross-linking between chains of linear silox- anes. Cross-linking is achieved by introduction of the (RSiOl.5) group. The siloxane cyclic trimers, (RESiO)3, have been shown to have a planar structure and may contain some bond strain (uh), whereas the 21 siloxane cyclic tetramers, (R28i0)h, have a puckered ring conformation and are free of strain. The higher cyclic siloxanes undoubtedly also have a puckered ring or crown configuration. Many other polysiloxanes with groups other than methyl are on the market today; those containing phenyl groups are more thermally stable than the aliphatic polysiloxanes. Phenyl-containing polysilox- anes, however, are not desirable since they set into brittle glasses. The similarity between the structures of organosiloxanes and the inorganic silicates has been observed in x-ray and electron diffraction studies on the siloxanes (hh). The silicon-oxygen distance in these 1 compounds is 1.61-1.67 A, which is about the same as that found in inor- ganic silicates, and it is considerably less than 1.83 A, which is the sum of the covalent radii. This indicates that there is a high degree of ionic character in the siloxane bond. As a result of this ionic character, siloxane bonds are quite non-directional. Rochow (MO) states, "The rigidity of silica and the silicates does not result from a rigidity of the bond angles of silicon and oxygen but from the high fields of force set up by all surrounding metal ions. Similarly, the silicon-oxygen tetrahedra of silica are uniform only because of the uniformity of the electric field pattern around them." There is also an apparent softness of these bonds and freedom of motion of the organo- silicon unit. This is substantiated by the flexibility of silicone rubber and the compressibility of siloxane fluids. The bond angles of the silicates and siloxanes in the solid state are also similar. The oxygen-silicon-oxygen bond angles in the siloxanes are close to the tetrahedral angle found in the silicates. 22 The silicon-oxygen-silicon angle in "unstrained" siloxanes is the same as that found in CX-quartz (lh20), which is greater than the tetrahed- ral angle. The carbon-silicon-carbon angle in the organosiloxanes is approximately tetrahedral, and the silicon-carbon bond distance is close to the sum of the covalent radii (1.90 A). In the organosiloxane polymeric materials the thermal stability of the siloxane bond is considerably weaker than in the silicates, and an increase in the number of organic radicals attached to the silicon atom decreases stability further. The molecular configuration (cyclic or linear) also has an influence on thermal stability. This is illus- trated by the cracking of linear dimethylsiloxanes at #0000 to form cyclic siloxanes. Cracking must involve breaking of the silicon- oxygen bond to form the more stable cyclic siloxane linkages. The chem- ical stability of the silicon-oxygen bond is very great and is ruptured only by heating with such reagents as concentrated hydrochloric acid, strong alkali, and fluorides. Chemically and thermally, the weak link in the polysiloxanes is the carbon-silicon bond. The bond energy of this linkage is about the same as the carbon-carbon bond, but it is believed that introduction of a silicon-carbon bond in organic compounds increases their thermal stability. The basis of this belief is the higher thermal stabilities of the organosilicon compounds compared with similar organic compounds. The silicon-carbon bond is subject to cleavage by air oxidation at tem- peratures above 300°C. The chemical reactivity to various reagents depends to a large extent on the group attached to the silicon atom, but generally vigorous conditions are necessary for cleavage of the 23 silicon-carbon bond. Bonding Considerations in Organosilicon Compounds and Chromates Silicon. The electronegativity difference between carbon and silicon is 0.85, which, according to Pauling (#5), indicates 16% ionic character in the bond. This difference in electronegativity suggests that the electron density constituting the bond between the two atoms is slightly larger nearest the carbon atom. Nucleophilic attack on the silicon atom should then depend on the influence of substituents on the carbon and silicon atoms. Rochow (#0) has discussed the nucleophilic attack of silicon by the hydroxide ion. The intermediate of the chemical reaction is be- 1ieved to be a pentacovalent Species which decomposes by elimination of a negative group forming an Si-OH bond. The ease of nucleophilic sub- stitution of the silicon-carbon bond may be reduced by placing positive substituents on the carbon (e.g., hydrogen atoms) and negative substi- tuents (e.g., oxygen atoms) on the silicon. This is found to be true since methylsiloxanes require strong alkali and high temperatures ‘ (200°) to liberate methane, while trichloromethyl groups are removed from silicon by water at room temperature. Thus it is reasonable to conclude that inductive effects play a large role in substitution re- actions of organosilicon compounds. The inductive effect of phenyl groups on silicon is withdrawal of electron density from the carbon attached to silicon, and the polarity of the bond is thereby enhanced. Sommer and his coworkers (#6, #7) have proposed a pentacovalent intermediate in the substitution reactions of asymmetric and bridgehead 2# silicon atoms. The high reactivity of substituents on a bridgehead sili- con atom, such as in l-silabicyclo(2.2.2)octane in which siliconium ions are unlikely and attack from the rear is impossible, required the postu- lation of a pentacovalent reaction intermediate. This was also prOposed for the substitution reactions of Optically active silicon compounds which gave pure retention of configuration. In these studies they fur- ther postulated that the intermediate addition complex was bound by use of one of the 3d orbitals of silicon. The configuration of the complex was suggested to be trigonal bipyramidal and the bonding to be the hybrid dsp3. Furthermore, the entering group and the group which is displaced need not occupy the apexes of the trigonal bipyramid. In- stead of an angle near 180°, the entering group may form an angle near 900 with the group to be displaced and the central Si atom. The concept of pentacovalent silicon was first introduced by Sidgwick (#8) in 1927. At that time he compared the stability of the carbon and silicon tetrahalides to hydrolysis and ammonolysis, and sug- gested that in the case of the silicon tetrahalides the reactivity resulted from coordination of the water or ammonia molecules prior to reaction. Pauling (#5) also suggested utilization of 3d orbitals to form ionic and double-bonded resonance structures which account for the abnormally short bond distances in some silicon compounds. Stone and Seyferth (#9) reviewed and interpreted the available literature concerning the use of d orbitals in silicon bonding. They point out that the coordination number of silicon is not always four, indeed, the fluorosilicate ion and siliconium acetylacetonate have co- ordination number six. The silicon tetrahalides also form addition 25 compounds with trialkylamines with an apparent coordination number of five and a reported coordination number as high as eight. The bonding in the octahedral complexes is believed to be 3Sp3d2 hybridization. In another example they cite the weak electron donor properties and planar structure of trisilylamine, (H3Si)3N, as an indication that the lone pair of electrons of the nitrogen are involved with the d orbitals of the silicon atom. On this basis they predicted that di- siloxane, (H3Si)20, should have a nearly linear structure, this was later observed by Lord, Robinson, and Schumb (50), and Curl and Pitzer (51) from infrared studies (Lord suggested 1800 and Pitzer estimated 1500). Dipole moment data on dimethyldisiloxane by Sauer and Mead (52) predict a siliconaoxygen-silicon bond angle of 160°. Dipole moment data on many organosilicon compounds have been interpreted to indicate d orbi- tal bonding. The acid strength of the trialkyl or triaryl silanols (53) com- pared to the carbinols suggests a removal of electron density on oxygen by silicon, resulting in a weaker oxygen-hydrogen bond which produces acids titratable in basic solvents. Finally, Craig, Maccoll, Nyholm, Orgel, and Sutton (5#) discussed the question of dfl--pm- bonding in a number of compounds, including silicon compounds, in some detail. From an evaluation of overlap inte- grals they conclude that highly electronegative ligands (F, O, N) are needed to form complexes with regular elements involving sp3d2 hybrid- ization, and that a strong‘n'Jbond may arise between an outer d orbital and a p orbital of an adjacent atom. In discussing the silicones in particular, they state that such bonding is possible between oxygen and 26 silicon atoms and is indicated by the short silicon-oxygen bond dis- tances, as compared to the bond distances expected from ionic radii. This would not affect the freedom of rotation about the bonds because there are five d orbitals available on each silicon atom which may form TT-bonds with oxygen no matter what the orientation of one silicon- oxygen bond to any other in the chain. Chromium. Little information is available on the nature of the bonding in hexavalent chromium compounds. Wlesberg and Helmholz (55) have made semiempirical molecular orbital calculations on the chromate ion. Agreement with the observed spectrum of the chromate ion was only fair. The most significant result of their calculation was the predic» tion of strong bonding of the TT-electrons of oxygen, which implies a double bond stronger than expected from valence bond theory. These re- sults have been criticized by Ballhausen and Liehr (56) as being based on an over-simplified calculation. Using a different model and energy lever scheme, they have made calculations which agree quite well with experiment. Their conclusion was that the charge transfer spectrum of the chromate ion cannot be interpreted on the basis of CY-bonding elec- trons alone, but that the oxygen Tr-molecular orbitals must also be included. Since the model used assumes tetrahedral symmetry for the chromate ion, one must conclude that the extent of TTebonding between the oxygen and chromium atoms is uniform for each bond. This is also substantiated by the force constants obtained from Raman data for the chromate ion in aqueous solution by Stammreich, Bassi, and Sala (57). Their model was based on the tetrahedral structure of the chromate ion with equal chromium-oxygen bonds, and they obtained a 27 value of 5.#8 mdyn/A for the stretching force constant of the chromium- oxygen bond. Their studies on the dichromate ion (58) based on a chromium-oxygen-chromium bridge model give two stretching force constants for the chromium-oxygen bond. These values were 3.27 and 6.07 mdyn/A. A comparison of the results of the two ions suggests that the chromate ion could be considered to be symmetrically bonded with double bond character in each bond, while the dichromate ion possesses non- equivalent chromium-oxygen bonds. The higher force constant for the dichromate ion compared to the chromate ion is expected since the double bond character is restricted to the outer chromium-oxygen bonds and the lower value is attributed to the chromium-oxygen single bond of the bridge. The concept of double bonding in chromates may also account for the Observed feeble paramagnetism of the chromates (27). That is, the the imperfect pairing suggested by Datar may be caused by contributions to TT-bonds. Helmholz, Brennan, and WOlfsberg (59) have extended semiempirical calculations from the chromate ion to the halochromate ion and chromyl chloride. The results of these calculations agree well with experiment. The conclusions from these calculations are particularly interesting. The first transition (555 mp) in chromyl chloride is assigned to 3b2——+ 5a, and the final excited orbital (5a) is strongly chromium- chlorine anti-bonding. The symmetry of this transition is B2 and involves a transfer of charge from chlorine to chromium which therefore leads to the dissociation Cr02012-—9.Cr02 + C12, as observed by Schwab and Prakash (l#). 28 The coefficients of the atomic orbitals in the molecular orbitals of chromyl chloride permit a rough calculation of the relative importance of fl'- and G 4bonding in the molecule. The C/Tr ratio for chlorine is 1.5, and for oxygen, 0.73, which indicates a decreased importance of TT-bonding for chlorine relative to oxygen. The stretching force con- stants of chromyl chloride calculated by Stammreich, Kawai, and Tavares (60) also substantiate this latter conclusion. The values ob- tained were 2.56 mdyn/A for the chromium-chlorine bond, and 7.17 mdyn/A for the chromium-oxygen bond. A comparison of the absorption frequencies of the symmetric chromium-oxygen stretching modes of the chromate ion, dichromate ion, and chromyl chloride (859, 90#, and 987 cm'l, respectively) shows that the extent of TT-bonding increases. This is also observed in the stretching force constants which increase (5.#8, 6.07, and 7.17 mdyn/A) for chromate ion, dichromate ion, and chromyl chloride. Inorganic Polymerization Reactions The concept of inorganic polymers--that is, high molecular weight compounds containing regular repeating units other than carbon--is rela— tively new, even though inorganic materials with a high state of aggre- gation have been known for some time. The importance of this subject is manifest in the increased activity in this field in the past few years. Inorganic reactions may lead to high molecular weight products by three means: condensation, addition, and coordination. Condensation 29 reactions involve the elimination of simple molecules between the reac- tants to form the aggregate. Addition polymerization reactions trans- form low molecular weight monomeric units into higher molecular weight substances by self-reaction of the monomers. Condensation of a metal ion with a polydentate‘ligand to form higher aggregates, incorporation of metals into polymers by coordination, and polymerization of monomers containing coordinated metal ions are all classes as coordination polymer- ization reactions. Sulfur is a familiar example of the many elements that exist in phases which can be classed as polymeric. When rhombic sulfur (SS) is heated to 187°C, a viscous mass is produced which, on slow cooling, be- comes a rubber-like, elastic material believed to be composed of linear chains of up to #0,000 sulfur atoms (61). Amorphous phases of selenium, tellurium, phosphorus, arsenic, antimony, and bismuth exist in states which may be classed as polymeric (62). The silicate minerals, particularly the pyroxenes and amphiboles, are composed of linear silicate chains and can be considered to be inor- ganic polymers. Amphibole asbestos is a good example of this type of inorganic polymer._ Sulfur trioxide vapor may be condensed at a temperature less than 25°C in a dry receiver to a modification composed mostly of cyclic tri- mers (63). Introduction of a trace of moisture causes addition polymer- ization to an asbestos-like form which has a melting point of 31.50C. This substance is composed of long chains of sulfur trioxide molecules. PhOSphonitrilic chloride is probably the best known of the syn- thetic inorganic polymers. A mixture of the cyclic trimer and tetramer 30 Of phosphonitrilic chloride is prepared by the action Of phosphorus pentachloride on ammonium chloride. If the mixture is heated to 300°C for several hours, the melt becomes viscous, and finally a rubber-like mass is Obtained. This "inorganic rubber" is composed Of chains of up to 200 units with the following structure: ” '“E It is thermally and hydrolytically unstable. A considerable effort is being made by many researchers (63) to modify the structure of phospho- nitrilic chloride in an attempt to improve its properties. Aggregation Of inorganic compounds by coordination to form poly- meric materials has been known for some time. Familiar examples of these coordination polymers are the dimers of aluminum and ferric chlorides, the planar chains of palladium(II) chloride, the linear tetrahedral chains of silicon disulfide, and the highly aggregated hy- drous oxides Of many metals. Attempts have been made (63) to prepare linear polymers by coordination Of bis(bidentate) ligands with tetra- coordinate metal ions, and bis(tridentate) ligands with hexacoordinate metal ions. The materials Obtained from such reactions have good ther- mal properties but are either amorphous powders or brittle glasses. Other attempts (6#) include incorporation of metal ions into polymeric Schiff base ligands and polymerization Of monomeric coordination com- pounds containing groups that undergo addition Or condensation. These compounds also have good thermal properties but are brittle. 31 Recently wagner (65) has reported the preparation of dimethyl- phosphinoborine polymers with a degree Of polymerization as high as 183. These polymers have been prepared by the pyrolysis of dimethylphOSphine borine in the presence of trialkylamine bases which act as blocking groups for the ends Of the linear chains. In the absence of these blocking groups, only the cyclic trimer and tetramer of dimethylphos- phinoborine are formed. The reaction for the preparation of these polymers is: n_(CH3)2PHBH3 + R3N __> [(033)2121332h . NR3 + EH2 These polymers show reasonable resistance to oxidation, hydrolysis, and heat. The preparation Of polyaluminoxanes (modified alumina polymers) has received considerable attention recently. Theobald (66) Obtained a brittle resin by heating isoprOpoxydiheptanoxyaluminum. This resin was soluble in toluene and had a molecular weight of 5730. The structure is believed to be that of polyheptanoxyaluminoxane,[- %1 - O i] , but OC7H15 2 no structural proof is Offered. Kugler (67) has prepared hard, brittle resins involving aluminum-oxygen linkages by the hydrolysis of an alum- inum chelate in the presence Of aluminum alcOhOlates. Andrianov and his coworkers (68) Obtained poly(triethylsiloxy)aluminoxanes by passing moist air through tris(triethylsiloxy)a1uminum at 165-17000. This poly- mer is a transparent, colorless, glassy substance. Certainly the most active field Of inorganic polymer research is the preparation of polymetallosiloxanes. The intention Of this research is to prepare high molecular weight heterOpOlymers with a metallosilox- ane backbone. It is felt by many active in this field of research that 32 alteration of the siloxane linkage to include metals in the backbone of the structure may produce polymers of high thermal stability (as the metal silicates) and good mechanical properties (as the polysiloxanes). Recently Andrianov (69) has reviewed the preparation of the poly- meric metallosiloxanes Of aluminum, titanium, boron, tin, antimony, phosphorus, and lead. The polymers containing these elements have not had particularly desirable properties. In the case of aluminum, titan- ium, tin, and antimony, brittle solids with a high silicon-tOemetal ratio were Obtained. Boron and phosphorus gave viscous fluids with poor thermal properties. More recently, Hornbaker and Conrad (70) have reported the pre- paration of polymeric metallosiloxanes of tin, lead, magnesium, and zinc. These polymers all have a very good silicon-tO-metal ratio, but they decompose rather easily to form siloxanes and metal oxides. Very little structural or molecular weight data is available for these com- pounds. Arsonosiloxanes have been prepared by Kary and Frisch (71). The products they have obtained range from transparent, rubbery materials to fine white powders. The authors suggest that these materials could be used for pest control products as well as for the products made from the conventional silicones. A wide variety of condensation reactions have been used to pre- pare the polymetallosiloxanes. Variation of the organic groups used to block valences and the metal-containing reactants has led to a large number of products (69). The groups most commonly attached to the metalloid compounds are methyl, ethyl, and phenyl. Alkoxy groups Of 33 the metals have been ethoxy, iSOprOpoxy, and nebutoxy. The reactions used to produce the polymers may be classed in the following way: 1. Cohydrolysis of a chlorosilane and metalalkoxide (69). R as H20 l l R2SiCl2 + T1(OR");+——) -Si-O-Ti-O- + HCl + RVOH heat I R OR“ 2. Cohydrolysis of a chlorosilane and metal chloride (69). R R9 H2O l | R 3101 ‘+ Rv SnCl -———+ -Si-O—Sn-O- +‘HCl 2 2 2 2 ' l R R9 3. Transesterification (69). R RESi(OR)2+ Sn(OAc)2 —-). -Si-O-Sn-O- + ROAc R 01‘ R OR R2Si(OAc)2+ Ti(OR)), ——> -Si-O-Ti-O- +ROAc A .R #. Condensation of a silanediol with a metal oxide (72). R R R23i(OH)2+ (R2SnO)X ___) O-di- O-S|n- + H20 R as. 5. Condensation Of a disodium silanolate with a metal chloride (70). R l R2Si(ONa)2 + SnCl2 ——; -Si-O-Sn-O- + NaCl R 3# 6. Reaction Of silanediols with metal alkyls (70). R l R23i(OH)2«+ ZnR"2 -————; -Si-O-Zn-O-‘+'R“H A 7. Condensation Of metallo acids with chlorosilanes (71). R R R2SiC12-+ R'Aso(OH)2 -———s> -éi-O-As-o- +-Hc1 A R A large number Of trialkyl or triarylsiloxy metal-containing com- pounds have been prepared. These compounds are tabulated in Table V. The condensation reactions used for the preparation of these compounds involve the following reaction combinations: 1. Metal and silanol. 2. Metal alkoxide and silanol. 3. Metal halide and silanol. #. Metal halide and sodium silanolate. 5. Metal oxide and disiloxane. 6. Metal oxide and silanol. 7. Metal acid and silanol. 8 Silver salt and chlorosilane. Metal acid and chlorosilane. \O 10. Metal acid and alkoxysilane. Other trialkyl and triarylsiloxy compounds are known, these in- clude the alkalimetal salts, sulfates (#2), phosphates (#2), perchlor- ates (93), and intermediate siloxy-titanium (9#, 95) and siloxy-vanadium (83) compounds. TABLE V METAL SILOXY COMPOUNDS Method of Melting or Boiling Compound Preparation Point in OC Reference (Me3SiO)3B 9, lo b.p. 9o 73, 7h (Et3SiO)3B 6 b.p. 152-#/3 mm 75 (Et3SiO)3A1 l m.p. 150 69 (Ph3SiO)3A1 (?) 2 m.p. 75—80 76 (Me3SiO)hTi 2, 3, h b.p. 6o/O.l mm 77, 78, 79 (Me2EtSiO)hTi 2 b.p. 86/O.1 mm 79 (MeEt2SiO)hTi 3 b.p. l2o/o.l mm 80 (Et3SiO)hTi 3 m.p. 110 80 (MegnPrSiO)hTi 2 b.p. 112/0 2 mm 79 (MegiPrSiO)hTi 2 b.p. 115/o.2 mm 79 (Me2PhSiO)hTi 3 b.p. 27h-5/7 mm 81 (MePh28i0)hTi 3 b.p. 37o-h/6 mm 81 (Ph3SiO)hTi # m.p. #80 (decomp.) 80 (Me3SiO)3VO #, 6 b.p. 118-20/18 mm 82, 83 (MeeEtSiO)3VO h, 6 b.p. 12#-6/7 mm 82 (MeEt2810)3VO #, 6 b.p. 169—70/5 mm 82 (Et3SiO)3VO #, 6 b.p. 198-201/13 mm 82 (EtgPhSiO)3VO h, 6 b.p. 263-8/0.l mm 82 Continued TABLE V - Continued 36 Method of Melting or Boiling Compound Preparation Point in 00 Reference (Ph3SiO)3VO u,6 m.p. 228 82 (Me3Si)20r04 5 b.p. 60/0.2 mm 8h, 85 (Ph3Si)2CrOh 6 m.p. l5#-5 86 (Ph3SiO)hGe u m.p. M72 (decomp.) 87 (Et3SiO)hSn h b.p. 200—2/h mm 88 (Ph3Sio)28nMe2 6 m.p. 165-6 72 (Ph3SiO)hSn # m.p. 322 (decomp.) 88 (Me3SiO)th u m.p. 152 89 (MegEtSiO)th u m.p. 105 89 (MeEtESiO)th h m.p. 30 89 (Et3SiO)th u b.p. 103/0.5 mm 89 (MeenPrSiO)qu # m.p. 60 89 (MegiPrSiO)th # b.p. llo/o.l mm 89 (Ph3SiO)qu h m.p #10 (decomp.) 87 (Ph3SiO)th # m.p. 383 (decomp.) 87 (Me3SiO)5Nb # ? 89 (Me2EtSiO)5Nb # ? 89 (Me3Si0)5Ta h m.p. 80 89 (Me2EtSiO)5Ta # m.p. 135 89 Continued 37 TABLE V - Continued r Method Of Melting or Boiling Compound Preparation Point in OC Reference (MeEtQSiO)5Ta # m.p. 180 89 (Et3SiO)5Ta u m.p. 210 89 (MeenPrSiO)5Ta u b.p. l70/0.l mm 89 (MeeiPrSiO)5Ta u m.p. 180 89 (Me3Si)ReOh 5, 8 m.p. 80 90 (Me3SiO)2Pb 6 2 91 (Et3SiO)2Pb (?) 6 ? 92 Symbols: Me = methyl, Et = ethyl, nPr = n-propyl, iPr = iSOprOpyl, Ph = phenyl. The melting point is given for materials which are solid at room temperature and the boiling point is given for liquids. In general, the trialkylsiloxy compounds can be hydrolyzed to the silanol or disiloxane by water with varying difficulty (89). With the exception Of bis(trimethylsilyl)chromate which explodes on heating, all the compounds have better than average thermal stability. The two chromates in Table V are important because of later work. Schmidt and Schmidbaur (8h) have prepared bis(trimethylsilyl)chromate, but full characterization Of the compound is not reported. The patent Of Granchelli and walker (86) indicates the preparation Of bis(triphenyl- silyl)chromate based on analytical data. A more complete discussion Of these compounds will be found in a later section. EXPERIMENTAL Preparation of Reactants A large variety of organosilicon compounds are commercially available and have been utilized in this work. The major commercial sources are the Dow-Corning Corporation, Midland, Michigan, and Anderson Chemical Company, Weston, Michigan. A new organosilicon compound, cyclo- hexyldiphenylsilanol, has been prepared. , .. Synthesis of Cyclohexyldiphenylsilanol. Ph251012+ Li(cyclohex)-—)(cyclohex)Ph28101 1% (cyclohex)PhQSiOH Cyclohexyllithium was prepared by the method of Nebergall and JOhnson (96). To this reagent 25.36 g (0.1 mole) of diphenyldichlorosilane in 100 ml of petroleum ether was added drOpwise with constant stirring over one hour while the reaction mixture was cooled with an ice bath. After addition of the dichlorosilane, the reaction mixture was refluxed for five hours. Unreacted lithium was then removed by filtration, excess cyclohexyllithium was destroyed with hydrochloric acid, the petroleum ether layer was separated, and the solvent removed by evaporation. NO attempt was made to isolate the chlorosilane. When the resulting yellow- brown Oil was added to 200 ml Of boiling alcoholic potassium hydroxide, a precipitate of potassium chloride formed. The silanol was precipitated by adding 300 ml Of water to the alcoholic solution, and the crude prod- uct was separated by filtration. Recrystallization from benzene-alcohol 38 39 gave an Oily product from which the remainder of the oil was removed by a simple extraction method. A small portion Of the product was placed in the bottom Of a fluted filter paper supported by a short-stem funnel. The filter paper tip was immersed in about 1 mm of hexane, and the fun- nel was covered with a watch glass. Within three hours the Oil was carried to the top Of the filter paper. After recrystallization from hexane, the product had a melting point of l#3°C. The yield of the cyclohexyldiphenylsilanol was 7.1 g (25% of theory). Its molecular weight was found to be 290 (theory, 282). Analysis Of the compound gave 77.55% C (theory, 77.66%) and 7.9#% H (theory, 7.85%). Preparation Of Tricyclohexylsilanol. The method of Nebergall and JOhnson (96) was used to prepare tricyclohexylsilanol. Cyclohexyllith- ium was prepared by the reaction of cyclohexylchloride with lithium in petroleum ether. A solution of 17 g (0.1 mole) Of silicon tetrachloride in 50 ml Of petroleum ether was added to cyclohexyllithium at 0°C over a period of 1.5 hours. The mixture was allowed to stand at room tem- perature overnight and then was refluxed for an additional two hours. Excess lithium was separated by filtration, and excess cyclohexyllithium was destroyed with 6 M hydrochloric acid. The organic layer was separ- ated and evaporated to an Oil which was crystallized from glacial acetic acid. The white, sticky solid was added to 200 ml Of boiling 5% alco- holic potassium hydroxide, and 300 ml of water was added to precipitate the silanol. The product was separated by filtration and recrystal- lized from a mixture of acetone and petroleum ether (yield, 5.8 g crude product; m.p. l73-5°C). Recrystallization from petroleum ether gave #.6 g (16% of theory) of product melting at 176-800. Nebergall -. i4»... #0 and thnson report 52% yield, while Eaborn (97) reports 13% yield using the same method of preparation. Preparation of Dicyclohexylsilanediol. An attempt to prepare tricyclohexylsilanol by the method described above, except for the omission Of refluxing, gave 6.2 g (27% yield) of dicyclohexylsilanediol, m.p. l6#-5°C (reported, l6#-5°C). Dicyclohexylsilanediol was also pre- pared by use Of a four-to-One ratio Of cyclohexyl chloride and silicon tetrachloride. Preparation of-p-Tolyldiphenylsilanol. Speck”s method (98) Of treating excess of the Grignard reagent of pfbromotoluene with diphenyl- dichlorosilane in ether solution was used to prepare this compound. The chlorosilane was separated, distilled, and then hydrolyzed by shaking its ether solution with water. The silanol (yield, 21%) was purified by sublimation and recrystallization from petroleum ether. The melting point Of the silanol was 83-85°C (reported, 87-88°C). Preparation Of Tri(l-naphthyl)silanol. This compound was pre- pared by the reaction Of naphthyllithium with silicon tetrachloride as described by Gilman and Brannen (99). The chlorosilane was separated and recrystallized with a Soxhlet extractor and ethyl alcohol as sol- vent. Hydrolysis was carried out by adding water to the chlorosilane in 5% alcoholic potassium hydroxide. The resulting silanol was recrys- tallized from ethyl alcohol to give a product melting at 197-199°C (reported, 208-2090C). Preparation of Methyldiphenylsilanol. A mixture Of methyldi- phenylsilanol and s-dimethyltetraphenyldisiloxane was prepared by the hydrolysis Of methyldiphenylchlorosilane in aqueous ammonia. The #1 silanol was separated by distillation (b.p. 120-12200 at 1 mm). Preparation of Hexaphenylcyclotrisiloxane and Octaphenylcyclo- tetrasiloxane. An attempt was made to prepare hexaphenylcyclotri- siloxane by heating diphenylsilanediol to 180°C as described by Kipping (100), but only an oil was Obtained which apparently contained poly— siloxanes. The method of Burkhard (101) gave better results for the prepar- ation of hexaphenylcyclotrisiloxane. Ten grams of diphenylsilanediol was dissolved in 200 ml Of ethyl ether, and 5 ml of concentrated hydro- chloric acid was added. The reaction mixture was allowed to reflux on a steam bath for three hours, and the ether was then removed. The re- sulting mass was taken up in benzene-ethanol and filtered; on cooling, two crops of crystals separated. The crude product was recrystallized from acetone to give 3.8 g of hexaphenylcyclotrisiloxane, m.p. 188—89°C (reported, 188-890 and 190°C). This compound was also prepared by the reaction of 1 mole of diphenyldichlorosilane and 2 moles of diphenyl- silanediol in chloroform. The method described by Burkhard (101) was used for the prepara- tion of octaphenylcyclotetrasiloxane. Diphenylsilanediol was dissolved in boiling ethanol and a few drops Of aqueous potassium hydroxide were added. Within 15 minutes crystals separated from the boiling mixture. After cooling, the product was removed by filtration and recrystallized from benzene-ethanol. The needles Obtained had a melting point Of 201-201.5°C (reported, 201-2o2°c). Preparation of Hexaphenyltrisiloxane-1,5-diol. The method.of Kipping and Robison (102) was used to prepare hexaphenyltrisiloxane-l,5-diol .‘_—- ‘fi #2 in very poor yield. Diphenyldichlorosilane was hydrolyzed in concen- trated aqueous ammonia and the product was separated by recrystalliza- tion from a mixture of chloroform and petroleum ether. Hernbaker and Conrad (70) have reported that the major product of the reaction Of disodium diphenylsilanediolate with zinc chloride is hexaphenyltrisiloxane-l,5-diol. They also prepared hexaphenyltrisil- oxane-l,5-diol from the reaction Of diethylzincc with diphenylsilanediol. Hexaphenyltrisiloxane-l,5-diol was prepared in this investigation by the reaction Of twenty grams of diphenylsilanediol and twelve grams of fused zinc chloride in about 300 ml of acetone. The mixture was heated to reflux for three hours and the acetone was evaporated. The product was taken up in benzene, and after boiling and filtering the mixture, the solvent was removed from the filtrate and the Oily residue was taken up in chloroform. Petroleum ether was added to the chloro- form solution until it became turbid. The solution was separated from an Oil by decantation and was allowed to crystallize overnight. The white crystalline product was separated by filtration, giving 13 g of crude product (m.p. 9#-lOO°C). The product was recrystallized twice from chloroform and petroleum ether to give 5.2 g Of product (m.p. llO-lllOC; reported, 111°C). Preparation Of Chromyl Chloride. A mixture Of chromium(VI) oxide and concentrated hydrochloric acid was dehydrated with concentrated sul- furic acid at 0°C to give chromyl chloride (3). The chromyl chloride layer of the reaction mixture was separated and purified by distillation (b.p. 115°C). The distillate was stored in a brown glass bottle fitted with a Teflon lined cap. #3 Preparation Of Silylchromates Preliminary to the study of polycondensation reactions, an in- vestigation Of the synthesis and properties of compounds in the general class (R3Si)20r0u was undertaken. The methods developed for the pre- paration, isolation, purification, and characterization of these com- pounds were then used in studying the more complex reactions Of chromium(VI) compounds with the silanediols. Simple (R381)2Cr0h com- pounds also were used tO Observe the effects of different organic groups (R) on the prOperties Of silylchromates. Schmidt and Schmidbaur (8#) recently have reported the prepara- tion Of bis(trimethylsilyl)chromate from hexamethyldisiloxane and chro- mium(VI) oxide. This compound also was prepared as a part Of this in- vestigation. Granchelli and walker (86) have patented a method for the preparation Of bis(triphenylsilyl)chromate by the reaction of triphenyl- silanol with chromium(VI) oxide in glacial acetic acid or xylene. Pre- paration Of the compound was carried out independently in this investi- gation. The following general reaction was used for the preparation Of the silylchromates: 2R3SiOH + Cr03 -—§§——§5 (R3Si)2CrOh'+ HQO Reactions were carried out by adding 3 g Of chromium(VI) oxide to a solu- tion of about 6 g Of the silanol in 150-200 ml of methylene chloride and heating the mixture to reflux (#OOC) with stirring. The reactions were conducted in a 300 ml single-necked flask fitted with a reflux condenser. The reaction mixture was heated with a heating mantle and stirred by use Of an external magnetic stirrer. In general, the reaction was quite rapid. At the end of three hours the reaction mixture was ## filtered to remove excess chromium(VI) oxide and the solvent was re- moved with a rotary evaporator. The crude product, Obtained in 9#-97% yield (based on the silanol), was then purified by recrystallization from an appropriate solvent. Bis(triphenylsilyl)chromate. Three types of reactions were used under varying conditions to prepare bis(triphenylsilyl)chromate: (a) the reaction Of triphenylsilanol with chromium(VI) oxide, (b) the re- action Of triphenylsilanol with chromyl chloride, and (c) cohydrolysis of triphenylchlorosilane and chromyl chloride. The first general reace tion was studied most extensively because any excess chromium(VI) oxide is easily removed by filtration and a high yield (96-97%) Of crude product is Obtained regardless Of the solvent used or the presence Of base. Stirring the reaction mixture only seemed to shorten reaction time. The second reaction was carried out in refluxing carbon tetra- chloride with calcium oxide present tO remove the hydrochloric acid formed by the reaction and to destroy the slight excess Of chromyl chlo- ride used. By the reaction 2Ph SiOH‘+ CrO Cl 3 2 2 the crude product was Obtained in a 99% yield. Cohydrolysis Of triphenyl- —> (Ph3Si)2Cr0h + 2HCl chlorosilane and chromyl chloride was used in an attempt to prepare the silylchromate. This reaction was carried out by rapidly mixing a carbon tetrachloride solution of triphenylchlorosilane and chromyl chloride with a two-fold excess Of water. Only a small amount of the silylchro- mate was formed by this reaction; the major product was unreacted tri- phenylchlorosilane. The use Of a base did not improve the yield Of the silylchromate. Because Of the success of other methods, cohydrolysis #5 was not investigated further. Crude bis(triphenylsilyl)chromate from reactions (a) and (b) was recrystallized from methylene chloride and an 85% yield Of orange- colored crystals (m.p. 153.5-l5#°C) was Obtained. This compound decom- poses above its melting point. Several attempts were made to recrystal- lize bis(triphenylsilyl)chromate from diethyl ether, but the ether solu- tion turned black after a short time. Analytical data are recorded in Table VI. Bis(p-tolyldiphenylsilyl)chromate. This compound was prepared by the general method. After removal Of the solvent, a red, uncrystal- lizable oil was Obtained. This was dissolved in methylene chloride, hexane was added, and the solution boiled to remove methylene chloride. On cooling, a yellow crystalline product was Obtained. After recrystal- lization from hexane, the melting point was 98.5-99OC, decomposition occurs above the melting point. Analytical data are recorded in Table VI. Bis(cyclohexyldiphenylsilyl)chromate. The preparation of bis(cy- clohexyldiphenylsilyl)chromate was effected by the general method. After filtering Off excess chromium(VI) oxide, the solvent was evaporated from the filtrate. The residue was treated with 10 ml of pentane to dissolve the silylchromate, and the less soluble silanol was filtered Off. This extraction Operation was repeated five times, and the filtrate was evap- orated to give a red-orange solid. The crude product had a melting point Of 81-8300 and a molecular weight Of 61# (calculated for bis- (cyclohexyldiphenylsilyl)chromate, 6#7). This product was again ex- tracted with five portions of pentane and the filtered extract was #6 TABLE VI ORGANOSILYL CHROMATES PREPARED Composition (%) MOlecular Melting Compound Cr Si C H Weight Point (°C) bis(triphenylsilyl)- chromate 153.5-15# Calculated 8.19 8.8h 68.11 u.76 63# Found 8.20 8.91 68.0# h.80 626 $.10 bis(p-tolyldiphenyl- silyl)chromate 98.5-99 Calculated 7.85 8.#7 68.50 5.17 66h Found 7.76 8.55 68.51 5.22 6AA t 10 bis(cyclohexyldiphenyl- silyl)chromate 8#-85 Calculated 8.0# 8.68 66.8# 6.55 6#7 Found 7.72 9.25 66.56 6.52 633 t 10 bis(tricyclohexylsilyl)- chromate 125-126 Calculated 7.76 8.37 6h.#3 9.91 671 Found 7.68 8.110 6h .22 9.78 6113 i: 10 partially evaporated to crystallize out unreacted silanol. After evap- oration Of the solvent from the filtrate, the resulting orange solid was dried under vacuum at room temperature. Material so purified had a melting point Of 8#-85°C. Solid bis(cyclohexyldiphenylsilyl)chromate does not appear to decompose at its melting point but is unstable at 100°C. The yield after purification was 2.7 g (h7%). Analytical data are recorded in Table VI. The purity Of the silylchromate is 96% based on chromium content and 107% based on A7 silicon content. These data, along with the low molecular weight, in- dicate that separation of the silanol from the silylchromate was not complete. Since this product is highly soluble in common organic sol- vents and a good recrystallization technique for its purification could not be developed, further effort did not seem justified. Bis(tricyclohexylsilyl)chromate. This compound was also pre- pared by the general method described on page #3. It was recrystal- lized twice from methylene chloride and washed with petroleum ether. The red-orange product was dried 13 32222 at room temperature and had a melting point Of 125-6°C. It is unstable above its melting point and is unique in that it differs in color and in absorption spectrum (Figure #, page 71) from the other compounds in this series. Bis(trimethylsilyl)chromate. An explosion resulted when an at- tempt was made to follow Schmidt and Schmidbaur's procedure (8#) for the preparation of bis(trimethylsilyl)chromate by refluxing hexamethyl- disiloxane containing suspended chromium(VI) oxide. Therefore a solvent was used to avoid overheating. Hexamethyldisiloxane was refluxed with chromium(VI) oxide in methylene chloride, and after 12 hours the mix~ ture was filtered and solvent removed. The resulting red Oil was not distilled owing to its explosive character. Vacuum evaporation into a bath at -78°C was used to isolate the compound. The product of the distillation was stable when stored at -78°C in the dark, but at room temperature and in the absence of light it showed appreciable signs of decomposition within three hours. This compound was also prepared by the reaction of the silanol with chromium(VI) oxide in methylene chloride. In this case, the rate #8 of formation of the silylchromate is faster than when the disiloxane is used as a reactant. The infrared (Appendix 2, Figure 1#), visible, and ultraviolet (Figure 2, page 69) spectra were recorded for this compound in solution without isolation Of the compound; this was accomplished by preparing the silylchromate from the silanol and excess chromium(VI) oxide in the appropriate spectroscopic solvent. An attempt to determine the molecular weight of bis(trimethyl- silyl)chromate in benzene failed because of reaction Of the silylchro- mate with benzene. Bis(methyldiphenylsilyl)chromate. Methyldiphenylsilanol was allowed to react with chromium(VI) oxide according to the general method (page #3). The resulting red liquid, possibly bis(methyldiphenylsilyl)- chromate, decomposed at 85°C at 0.3 mm Hg pressure when an attempt was made to distill it. Vacuum evaporation into a trap cooled with liquid air was unsuccessful. Because of its instability, this product was not investigated further. BisI}ri(l-naphthyl)silyl]chromate. An attempt was made to pre- pare a chromate from tri(l-naphthyl)silanol. Brown noncrystalline mater- ials were Obtained. Crystallization from methylene chloride was attempted, but infrared spectra Of the fractions failed to show a decrease in com- plexity. Attempts at chromatographic separation were likewise unsuccess- ful. A model Of the compound suggests that steric hindrance might well prevent its formation. PrOperties Of Silylchromates. All of the silylchromates described in the previous section decompose into a black amorphous solid on #9 prolonged exposure to light. Bis(tricyclohexylsilyl)chromate is less light sensitive than the other silylchromates, while bis(trimethyl- silyl)chromate decomposes most rapidly on exposure to light. All except the aliphatic silylchromates may be safely stored in the dark at room temperatures. The aliphatic silylchromates and bis(cyclohexyldiphenylsilyl)- chromate are decomposed by concentrated nitric acid, and all are decom- posed by a mixture Of concentrated sulfuric and nitric acid. Both bis(trimethylsilyl)chromate and bis(methyldiphenylsilyl)chromate are de- composed by water. In the case Of bis(trimethylsilyl)chromate, the di- siloxane is formed on hydrolysis. All the Silylchromates are decomposed by homogeneous hydrolysis in an acetone-water solution. Homogeneous hydrolysis of bis(triphenylsilyl)chromate by addition of water to an acetone solution gives the silanol in a nearly quantitative yield, and about half Of the chromium remains in the hexavalent state, most likely as chromic acid. The silylchromates are very soluble in methylene Chloride, car- bon tetrachloride, chloroform, acetone, and ethanol. With the exception of bis(cyclohexyldiphenylsilyl)chromate and the aliphatic silylchro- mates, they are all only slightly soluble in cyclohexane and aliphatic hydrocarbons. Of the compounds prepared, thermal stability is greatest for bis(triphenylsilyl)chromate which decomposes above its melting point Of 153.5-15#°C. All of the organosilyl chromates decompose above their melting points except for bis(cyclohexyldiphenylsilyl)chromate which is stable to about 15° above its melting point Of 8#-85°C. 50 The presence Of electron donor groups in the silylchromates seems to lower stability. Thus bis(trimethylsilyl)chromate is explosive when heated, and bis(methyldiphenylsilyl)chromate and bis(pftolyldiphenyl- silyl)chromate decompose easily. Because Of the higher stability of bis(triphenylsilyl)chromate, polymerization reactions have been studied using diphenylsilanediol. Reacpions Of Diphenylsilanediol and Its Condensation Producpg With Chromium(VI) Compounds The reaction Of diphenylsilanediol with chromium(VI) oxide or chromyl chloride in methylene chloride or carbon tetrachloride has been studied. Although polysilylchromates with repeating Cr-O-Si linkages have not been Obtained owing to the tendency Of diphenylsilanediol tO undergo self—condensation, two interesting silylchromates have been pre- pared. Compound I’is a dark orange amorphous substance containing six silicon atoms for every chromium atom; compound II_is a yellow-orange crystalline solid having a silicon-tO-chromium ratio Of two-to-one. These compounds are: Ph Ph Ph 0 Ph Ph Ph I I I II I l I Ho-Si-O-Si-O-Si-O-Cr-O-Si-O-Si-O-si-0H (I) I Ph Ph Ph 0 Ph Ph Ph and Ph2\ /O \ /Ph2 //’S Si\\ O | I o = Cr = 0 0 = Cr = 0 (II) I I O 0 \\\ Si Si / 51 A study has been made of the reactions Of chromyl chloride with hexaphenylcyclotrisiloxane and with octaphenylcyclotetrasiloxane to de- termine if addition of the chromyl group to cyclic siloxanes is possible. Addition occurred in low yield with the trimer, but the tetramer did not react. The reaction Of hexaphenyltrisiloxane-l,5-diol and chromyl chloride gives a small amount of compound I_and largely hexaphenyle cyclotrisiloxane. Tetraphenyldisiloxane-l,3—diOl and chromium(VI) oxide gives compound 2;. ' 1 Reaction of Diphenylsilanediol and Chromyl Chloride. The reac- f tion Of chromyl chloride and diphenylsilanediol has been carried out under various conditions. The principal product Of this reaction is compound I, with smaller amounts Of compound I; and hexaphenylcyclotri- siloxane. The variation Of reaction time, method of addition, and sol- vent only varies the yield of the above products. Direct mixing of the reactants for thirty minutes, followed by isolation Of products, was found to be the most convenient reaction method. The following is a typical reaction. A slurry Of 30 g of diphenylsilanediol and 200 ml of methylene chloride was added portion-wise to 12 ml Of chromyl chloride in 100 ml of methylene chloride contained in a single-neck 500 ml flask. The flask was fitted with a condenser and heated to reflux for 30 minutes. During this period, the mixture was stirred by a magnetic stirrer. Ex- cess chromyl chloride was destroyed by adding about 30 g Of anhydrous sodium bicarbonate in small portions to the reaction mixture. The re- action mixture was then filtered through a fluted filter to remove the 52 bicarbonate. The filtrate was reduced to about 50 ml, cooled, and then filtered to remove crystalline compound II. The filtrate was evapor- ated to near dryness and then treated with a mixture of methylene chlo- ride and petroleum ether and filtered to remove compound II, This Operation was repeated once more. The yield of compound II was 5.5 g (16% based on the diol). The filtrate was dried IE_X§EEQ at room temperature to give 17.5 g Of crude compound I (58% based on the diOl). The molecular weight Of this product was 1090. This material was once more treated with methylene chloride and petroleum ether and filtered. The filtrate was dried Ip_z§gpg_for two days. The resulting product had a molecular weight of 1330 and contained 12.97% silicon, #.93% chromium, 65.80% carbon, and #.80% hydrogen. The structure proposed for compound I corresponds to a molecular weight Of 1307 and to the composition Si, 12.88%; Cr, 3.98%; C, 66.13%; and H, #.78%. The high results for chromium and the molecular weight may be accounted for by a small amount of compound II which is not very soluble in benzene and which has a high chromium content. The above reaction was varied by use Of carbon tetrachloride as a solvent, addition Of the diol by use Of a Soxhlet extractor, longer reaction times, and use Of benzene to destroy the excess chromyl chlo- ride. The use Of carbon tetrachloride as a solvent gave compound I in a slightly lower yield and enabled a better separation Of compound II because Of its lower solubility in carbon tetrachloride. The addition Of the diol by use Of a Soxhlet extractor required a longer reaction time, and the Soxhlet thimble was attacked by chromyl chloride which distills with the solvent. Longer reaction time and use Of benzene to 53 destroy the excess chromyl chloride resulted in a reduction Of the yield Of compound I to 18%, along with an increase in the amount Of decomposi- tion products. TWO other methods were used to purify compound I. Partition chromatography using columns Of cellulose, sugar, and anhydrous sodium bicarbonate resulted in only partial removal of impurities. Absorption chromatography using columns Of activated alumina and charcoal was un- successful because compound I decomposed on these columns. The best method for purifying compound I was a mixed solvent technique. The product was dissolved in a minimum amount Of methylene chloride and re- precipitated by addition Of hexane. The hexane mixture was boiled to remove methylene chloride and increase the solubility of the impurities, after which the solution was decanted from the separated Oil. It was also possible to Obtain compound I_by cooling the hexane solution which was decanted from the Oil. Treatment by this method reduced the yield Of compound I considerably. This method, however, removes all Of compound II and as high as 2 g Of hexaphenylcyclotrisiloxane (based on 30 g of the diol). The analytical data for compound I is given in Table VII. Re- peated purification lowers the chromium content Of the compound since compound II, with a higher chromium content, is removed. Reaction Of Diphenylsilanediol and Chromium(VI) Oxide. The re- action Of diphenylsilanediol and chromium(VI) oxide has been carried out in methylene chloride, carbon tetrachloride, and glacial acetic acid. The principal product isolated was compound II, with lesser amounts of compound I except when glacial acetic acid was used as a solvent. TABLE VII ANALYTICAL DATA FOR COMPOUND I FROM THE REACTION OF DIPHENYLSILANEDIOL WITH CHROMYL CHLORIDE 5# MOlecular Run Solvent % Si % Cr % C % H Weight 1 0C1, . ... .... .. . .... 1029 Repeated purification ..... .... .. .. 113# Repeated purification 13.22 #.16 ... .. . 1127 2 Gel, 13.#8 #.00 . .. ... 10#7 Repeated purification .... . .. ..... ... 1303 Repeated purification 13.63 3.89 ..... . 128A 3 CCl l2.#7 7.12 ..... . 1281 Repeated purification ..... .... ... ... 1278 Repeated purification 12.92 5.#8 ..... .... 1153 # CCl 13.25 #.95 .. . 1315 Repeated purification 13.02 #.73 ... . .. .... 5 CH'2012 13.5# 3.82 . .. .... ... 6 CH'2C12 ..... .... .. . ... 1090 Repeated purification 12.97 #.93 ..... .... 1330 7 CH2012 13.10 #.35 65.80 #.80 1190 8 CH2Cl2 13.38 3.## 66.25 #.75 1110 9 CH2C12 13.29 #.20 65.9# #.68 1180 Calculated for compound I 12.88 3.98 66.13 #.78 1307 The reaction of the diol and chromium(VI) oxide in methylene chloride served as the best method fOr the preparation of compound II. A slurry Of 30 g of diphenylsilanediol and 200 ml of methylene chloride was added in small portions to about 15 g of chromium(VI) oxide covered 55 with 200 ml Of methylene chloride. During the addition, the mixture was brought to reflux temperature by heating with a mantle; magnetic stir- ring was used. The reaction was allowed tO continue for 30 minutes after the addition of the diol. The reaction mixture was filtered while hot to remove excess chromium(VI) oxide. Solvent was removed from the fil- trate to give a total volume of about 50 ml. This was then cooled and filtered to remove a brown mass. The filtrate was further concentrated, cooled, and filtered twice more to remove the brown material. This brown mass was washed with two 25 ml portions of precooled methylene chloride to give 18.6 g (5#% based on the diol) Of a yellow-orange crystalline material, compound II. Compound II is easily recrystallized from methylene chloride to give a yellow-orange solid. The filtrate from which compound II was separated was evaporated to give 5.# g Of compound I (18% yield). Table VIII (page 58) gives the elemental analysis Of compound II. When carbon tetrachloride is used as a solvent, the yield of com- pound I is increased and that Of compound II diminished. Compound II is not very soluble in carbon tetrachloride. The reaction Of the diol and a solution Of chromium(VI) oxide in glacial acetic acid gave a very small amount (3.3% yield) of compound II and a large amount of green re- duction products (not Characterized). Characteristics of Compounds I and II. Numerous determinations Of molecular weight and elemental analysis Of compound I have been made and are reported in Table VII. Although complete removal Of impurities (hexaphenylcyclotrisiloxane, compound II, and solvent) is difficult, the most nearly pure product has a silicon-chromium ratio approaching 56 six-to-one and a molecular weight Of about 1300. These data are con- sistent with structure I (page 50). The infrared spectrum Of this compound (Appendix 2, Figure 16) shows weak 0H, strong Si-O-Si, and strong Cr=0 absorptions. The pre- sence Of the OH group was determined only by the OH absorption. A strong 0H absorption was Obtained by using more concentrated solutions or pressing a sample between sodium chloride plates. Several attempts were made to prepare the acetyl and benzoyl derivatives Of this com- pound by reaction with acetyl chloride and benzoyl chloride, but decom- position Of compound I took place; the only product isolated and iden- tified was hexaphenylcyclotrisiloxane. A solution of this compound in anhydrous benzene did not react with sodium metal. Compound I is repellent to cold water and slowly attacked by hot water. It can be hydrolyzed completely by adding water to an acetone solution. In this case, the only product isolated was hexaphenylcyclo- trisiloxane. This compound is sticky when freshly prepared and is easily drawn into fibers. 0n standing, it hardens into a yellow-brown glassy sub- 1 stance. It is quite stable at room temperature in the absence Of light. In the presence Of light, it slowly decomposes to a black mass from which hexaphenylcyclotrisiloxane has been isolated. Even in the glassy state, the compound is easily melted with the heat Of the hand. The differential thermal analysis Of this compound (Figure 1) shows that it undergoes fusion from room temperature to 275°C, where it begins to decompose. The decomposition reaches a maximum at #6000. The decom- position products appear tO be chromium(III) oxide and Silicon dioxide. 57 com M vasomaoo e0 mflmmama< HmanmgB HmflpcmaOMMHm ocmamflpcmo mcosmcc cm chapmaomama cos com . om a _ .H Ossmflm IEWJQQlOPUE Temaaqioxa 58 The x-ray diffraction pattern Of a fiber showed broad diffraction at 10.3 and #.5 A, which indicates a very low order Of crystallinity. The intrinsic viscosity in 3% benzene solution was determined to be 0.0192 dl/g. Compound I is highly soluble in carbon disulfide, benzene, tol- uene, methylene chloride, carbon tetrachloride, and chloroform. It is only slightly soluble in aliphatic hydrocarbon solvents. I appears to be partially decomposed by acetone. The last traces of the soluble solvents are very difficult to remove from the compound. Drying IE vacuo is facilitated by continually turning up new surfaces since dif- fusion Of the solvent to the dried surface does not take place. The analytical data for compound II is given in Table VIII. TABLE VIII ELEMENTAL ANALYSIS OF COMPOUND II % Calculated Element (Structure II) % Found Carbon 58.05 58.16, 58.1# Hydrogen #.06 #.17, #.29 ' Chromium 10.#8 10.#2, 10.A7, 10.60, 10.#o, 10.37, 10.#9 Silicon 11.30 11.58, 11.59. 11.27, 11.36, 11.51, 11.53 The molecular weight of compound II using diphenyl as a cryoscopic solvent (m.p. 71°C, Kf 8.#6) was 1155, 1180, and 1170 i 100 in separate 59 determinations. The molecular weight in benzene was found to be 1030, t 1030, and 1170 i 100. The low accuracy of these determinations is due to low solubility Of compound II in the cryoscopic solvents. The value 1170 may be in error since a slight turbidity of the solution was de- tected (this would give a high value). Higher values Of the molecular weights were Obtained, but in each case not all the solute had dis- solved in the solvent. A search for more satisfactory solvents failed. The solute is either insoluble or reacts with the common solvents (12) used for molecular weight determinations. The elemental analysis indicates a ratio of two silicon atoms for every chromium atom in this compound and is consistent with the basic unit, [(Ph2Si0)2Cro3]. The molecular weight Of this unit is #96.5, suggesting that two units are present in compound II. Compound II may be hydrolyzed by adding water to its acetone solu- tion. Tetraphenyldisiloxane-l,3-diol has been isolated from the hydroly- sis products in 8#% yield. This information indicates that the disiloxy group exists in compound II. The infrared spectrum (Appendix 2, Figure 18) of this compound . shows no 0H absorption, medium Si-O-Si absorption characteristic of strain-free siloxanes, and Cr=0 absorptions. These data are consistent with structure II (page 50). The molecular weight of this compound is 993; its elemental com- position is given in Table VIII. A rough model of this structure using bond angles and distances of strain-free cyclic siloxanes (##) indi- cates that the above structure is plausible and probably has a "crown" 60 conformation: O 0 \/\/°\/\/ Cr Si Si Cr This compound has a sharp decomposition point at 169°C which is unchanged by recrystallization. It slowly decomposes to a brown solid when exposed to light and is decomposed in hot acetone, ethanol, and benzene. The compound is sparingly soluble in cold methylene chloride, chloroform, carbon tetrachloride, carbon disulfide, and acetonitrile. The infrared spectrum Of this compound does not decrease in com- plexity on continued recrystallization from methylene chloride. The x-ray powder diffraction pattern is given in Appendix 3. Compound II in anhydrous benzene does not liberate hydrogen when sodium metal is added. This compound is also repellent to water but can be hydrolyzed from homogeneous solution. Compound II does not re- act further with chromium(VI) oxide in methylene chloride, but does react with excess diphenylsilanediol to give a product with the char- acteristics Of compound I. Reaction Of Cyclic Siloxanes With Chromyl Chloride. The addition Of chromium(VI) compounds tO cyclic siloxanes has been studied because Of its relationship to the complex reaction Of diphenylsilanediol with chromium(VI) compounds. It was found that diphenylsilanediol dissolves in methylene Chloride in the presence Of a trace Of chromyl chloride to give hexaphenylcyclotrisiloxane and an oil with a degree of polymer- ization Of 3.6. Diphenylsilanediol is only Slightly soluble in methylene chloride even at reflux, while its condensation products are soluble in 61 this solvent. This indicates that condensation of the diol takes place under the influence of chromyl chloride. The reaction Of hexaphenylcyclotrisiloxane with chromyl chloride and a small amount Of water was carried out several times in methylene chloride solution in the manner described for diphenylsilanediol using a one-tO-One ratio Of siloxane and chromyl chloride. The length Of the reaction was varied between 12 and 30 hours. The reaction mixture was filtered, evaporated, and treated with hexane to remove unreacted hexa- phenylcyclotrisiloxane. The treatment of the filtrate with hexane was repeated several times, but complete removal Of the siloxane is very difficult. The products isolated had molecular weights from 690 to 990. The molecular weight of hexaphenylcyclotrisiloxane is 598, and compound I is 1307. The yield of the product was less than 10%. In general, the yield of the product increased with increased reaction time and with the use Of sodium hydroxide to supply water and destroy excess chromyl chloride. The purified product is similar in appearance and has the same x-ray diffraction pattern as compound I. The lower molecular weight is probably due to incomplete removal of the siloxane. Final removal Of the siloxane probably could be effected by the above puri- fication procedure, but the amount Of chromium-containing product is decreased drastically. Hexaphenylcyclotrisiloxane does not react with chromyl chloride in moist carbon tetrachloride; this may be due to the insolubility of the siloxane in carbon tetrachloride. It also does not react with chromyl chloride under anhydrous conditions. 62 Apparently addition takes place, but only to a very small ex- tent; it is catalyzed by the presence Of a base such as water or sodium hydroxide. A comparison of the ease Of formation of compound I in the reaction Of diphenylsilanediol with chromyl chloride and the difficulty Of addition Of chromyl chloride to hexaphenylcyclotrisiloxane leads to the conclusion that the latter reaction is not involved in the reaction Of the diol with chromyl chloride. The reaction Of octaphenylcyclotetrasiloxane with chromyl chlo- ride and water has also been attempted in the same manner as described for hexaphenylcyclotrisiloxane, but the tetrasiloxane was recovered largely unchanged. A very small amount (#.5%) Of a brown product with an apparent molecular weight of 875 was Obtained; the molecular weight Of octaphenylcyclotetrasiloxane is 793. It is therefore concluded that the tetrasiloxane is not significantly attacked by chromyl chloride. This may be due to the high stability Of this compound. Reaction Of Hexaphenyltrisiloxane-l,5-diol and Tetraphenyldi- Siloxane-I,3-diol with Chromium(VI)_COppounds. The reaction of hexa- phenyltrisiloxane-l,5-diOl and chromium(VI) oxide was carried out in methylene chloride solution for three hours. The methylene chloride- hexane method was used to separate 65% of the starting material from a small amount Of black amorphous material. Compound I was not formed. A product with the characteristics of compound I was Obtained from the reaction Of 11 g Of hexaphenyltrisiloxane-1,5-diOl with excess chromyl chloride in methylene chloride. The reaction was carried out for three hours, after which the excess chromyl chloride was destroyed with sodium bicarbonate and the mixture was filtered. Three grams Of 63 the product was separated from three grams (yield, 28%) of hexaphenyl- cyclotrisiloxane by the methylene chloride-hexane method. The infra- red spectrum of the product was similar to that of compound ;_but the molecular weight was determined to be 8hl, which is lower than the value of 1307 required of compound 1. The reaction of tetraphenyldisiloxane-l,3-diol and chromium(VI) oxide was carried out for one hour in methylene chloride solution. Compound I; was recovered in 68% yield from the reaction mixture. No other products were isolated. The yield of compound T; in this case is comparable to that obtained in the reaction of diphenylsilanediol with chromium(VI) oxide. Reaction of Diphenyldialkoxysilanes with Chromyl Chloride The reaction of chromyl chloride and dialkoxysilanes was under- taken to determine if condensation could take place through evolution of ethyl chloride. The results of the study of this reaction indicate that chromyl chloride attacks the alkoxy group to form an fitard complex which, when decomposed with water, gives largely polysiloxanes. The reaction of chromyl chloride and diphenyldiethoxysilane in methylene chloride serves as an example of the general reaction. When 13 ml of chromyl chloride were added to a solution of 30 g of diphenyldiethoxysilane in 300 ml of methylene chloride, 3 reaction took place immediately. This reaction was almost violent, but it sub- sided to give a red-brown mixture. water was added to decompose excess chromyl chloride, the methylene chloride layer was separated, and the 6h solvent was removed to leave a viscous, brown oil. By dissolving the product in methylene chloride and reprecipitating it with hexane, two grams of hexaphenylcyclotrisiloxane was separated and 13 g of product remained. Its molecular weight was determined to be 707. Further treatment with methylene chloride and hexane removed less than one gram of hexaphenylcyclotrisiloxane from the product and raised its molecular weight to 718. Analysis gave 13.90% silicon and 0.70% chromium, which corresponds to one chromium atom for every thirty-seven silicon atoms. This ratio is not reasonable in a pure compound with a molecular weight of 718. The calculated silicon content for diphenylsiloxane, (Phesiojn, is 1h.15%. - The observed silicon content and that calculated for diphenyl- siloxane are close enough to conclude that the product is composed mostly of polysiloxanes. The infrared spectrum of this product shows strong Ph-Si, Si-O-Si, and Si-O absorptions as expected for polydiphenyl- siloxane. Several attempts were made to distill the reaction product. In all cases a yellow oil with a molecular weight of 653 was separated by distillation. Hexaphenylcyclotrisiloxane (molecular weight 598) was the major compound of the distillate. The glassy residues of the dis- tillations contained small amounts of chromium and had molecular weights as high as 2000. The reaction of chromyl chloride and diphenyldiethoxysilane was carried out in carbon tetrachloride and, with inverse addition of re- actants, in methylene chloride. The same product was obtained except that the content of hexaphenylcyclotrisiloxane was higher. 65 The reaction of chromyl chloride and diphenyldimethoxysilane under the same conditions as those for the diethoxysilane gave a very high yield of hexaphenylcyclotrisiloxane. It can only be concluded that the reaction of chromyl chloride and diphenyldialkoxysilanes under the conditions described produces polysiloxanes. This reaction probably proceeds through the formation of an fitard complex between the chromyl chloride and the aliphatic groups of the alkoxide. The addition of water decomposes the fitard com- plex and cleaves the carbon-oxygen bond of the dialkoxysilane to pro- duce polysiloxanes. Because of the acidic conditions, hexaphenylcyclo- trisiloxane is the principal product. The evidence for the formation of the Rtard complex is the sus- pension of a brown solid (characteristic of fitard reactions) in the reaction mixture and the highly exothermic reaction when water is added to the reaction mixture. Reaction of Dicyclohexylsilanediol with Chromium(VI) Oxide The reaction of dicyclohexylsilanediol with chromium(VI) oxide was carried out in methylene chloride. Excess chromium(VI) oxide was added to a solution of 5.39 g of dicyclohexylsilanediol in 300 ml of methylene chloride and the reaction mixture was stirred at reflux tem- perature three hours even though the reaction appeared to be quite rapid. The mixture was then filtered to remove excess chromium(VI) oxide, and the solvent was removed from the filtrate to give a bright red, viscous oil. The oil was dissolved in pentane and filtered to remove unreacted 66 dicyclohexylsilanediol. The pentane was partially evaporated from the filtrate and cooled. This was then filtered to remove very little of the diol. The pentane was removed from the filtrate and the product was dried in_v§gug_overnight. The resulting product (6.h8 g) was a dark red, amorphous solid. The molecular weight of this product was 1690 i 50. Analysis gave the following results: Si, 9.h5%, 0r, 15.23%; c, h8.l9%; H, 7.50%. The ratio of silicon to chromium is 1.15:1, and the ratio of carbon to silicon is 11.9:1. The following structure was found to fit these data (Ch = cyclohexyl): Ch 0 Ch 0 Ch 0 Ch 0 Ch 0 Ch I H I II ,| II I II I II I HO-Si-O-Cr-O-Si—O-Cr-0-Si-0-Cr-0-Si-O-Cr-O-Si-O-Cr-O-Si-0H (III) III III III I” In I Ch 0 Ch 0 Ch 0 Ch 0 Ch 0 Ch Structure 111 requires a molecular weight of 1780 and the follow- ing elemental composition: Si, 9.h6%; Cr, 1h.61%; C, h8.50%; H, 7.58%. The ratio of silicon to chromium is 1.20:1 and that of carbon to silicon is 12:1. The yield of the above reaction based on structure 1;; was 93%. Two previous runs of this reaction with less extensive purification have given a similar product in 100% yield. In one case, the crude prod- uct was isolated and analyzed without purification. The molecular weight was 1h20, and the content of silicon and chromium was 9.26 and 13.37% respectively. The second reaction product was treated once with pentane and filtered to give a product with a molecular weight of 1700, a silicon content of 9.18%, and a chromium content of 15.60%. The infrared spectrum (Appendix 2, Figure 20) of this compound shows weak OH absorption, strong Cr=0 absorption, and a medium absorption 67 at 9.3a , which may correspond to either Si-O;Si or cyclohexyl absorp- tion. The possible presence of Si-O-Si absorption is not consistent with structure III. The possibility of the compound being a mixutre of siloxanes and chromates is precluded by the high molecular weight of the product and the unlikelihood of polydicyclohexylsiloxanes. The for- mation of an Si-O-Si linkage consistent with the 1.15 silicon-to- chromium ratio and the molecular weight is also unlikely. Compound III decomposes on standing at room temperature in the dark to form a black, brittle, amorphous solid. The product shows con- siderable signs of decomposition within four days after preparation. Extraction with pentane and methylene chloride gives a product with a molecular weight of 1220 and the following composition: Si, 11.06%; Cr, 8.78%; C, 56.h6%; and H, 9.08%. The ratio of silicon to chromium is 2.34:1, while the carbon-to-silicon ratio remains 11.9:1. The high silicon-to-chromium ratio is due to the presence of cyclohexylsilicon compound in the extracted product. The residue of the extraction is a black solid which is probably high in chromium content. The formation of a mixture of products rather than a single com- pound is to be expected in a polycondensation reaction such as that which leads to compound III. Thus a mixture of larger and smaller chains than compound III may contribute to the number average molecular weight. Because of the low degree of polymerization involved, it is also possible that under the conditions of the reaction only compound III is formed; that is, bond breaking and bond making may be in equi- librium at the chain length corresponding to compound III. 68 Spectroscopic Data Ultraviolet and Visible Spectra. The ultraviolet and visible absorption spectra (Figures 2, 3, h) of the major products described previously were measured in methylene chloride solution with a Cary Mbdel 11 SpectrOphotometer. The compounds which contain phenylsub- stituents” all show similar absorption in this region, compounds with only cyclohexyl substituents on the silicon atom are similar in absorption particularly in the near ultraviolet region. The phenyl- containing silylchromates all show an absorption plateau in the 300-350 mllregion and phenyl absorptions in the 250-270 mpgregion. Bis(tricyclohexylsilyl)chromate has a distinct maximum absorption at #07 my; compound III (page 66) has a less distinct maximum absorption at #00 my. Both show absorption in the ultraviolet which is probably due to chromate since cyclohexyl groups, unlike phenyl, do not absorb in this region. Bis(cyclohexyldiphenylsilyl)chromate has greater ab- sorption in the visible region than the other phenyl-containing silyl- chromates, and bis(p-tolyldiphenylsilyl)chromate has the most intense absorptions of the silylchromates. Bis(trimethylsilyl)chromate is simi- lar in absorption to the phenyl-containing silylchromates in the 320- #50 m+lregion but is distinctly different in the far ultraviolet. The extinction coefficients (Figure 2) for this compound are arbitrary. Infrared Spectra. The infrared spectra of the compounds pre- pared in this investigation are compiled in Appendix 2. These spectra were recorded in carbon disulfide, carbon tetrachloride solutions, and in potassium bromide pellets where apprOpriate. The absorptions present 69 mpmaoagoAHhHHmahgpoaHapvmflm cam .opmSoagoAazafimammogmficamxogoaomoVmflm .opmsoagoAahaflmamnogmflcazaopumvmfim .opmaoanoAahaflmHthnmflapvmfim a0 mapoomm oHQHmH> dam PoHOfl>mapHD .m oasmflm mcoaofiaflaafls QH Spwqoao>m3 ems cos 0mm com 0mm If. . / I:.....nuu.n.u.un.n.nnwu.uu.. /././ I'll/ll /./.1/ xxx / I/ a .o o I /a 1/ no /. , / I ./. a. x. ” opmaoasoAHaHsmHahposaapvmas / ,a/ - l. I I opmsossoAaaaamaacohmfloaaxosoaoaoVmas .-:;::;:s:::;;;. x . o/ol opmsoagoAHEHHmH%Qo£mH@H%HOPLNVmHQ ........ ..uil / opmsoanoAHhHflmHmaosmHapvmflp IIIIIIIIIIII JJGOO uotloutixe eroN Ln 0 (T_m0) 17.0I X 1UGIOI O.H mm was m mpgdomsoo mo mapoomm wapfimfl> was PoHOfl>mapHD .m madmflm 70 mQOMOflEHHHHS CH Spmnoao>m3 Om: co: . 0mm com 0mm MM vasomaoo r a H pndomaoo lllllllllllll / 7x ” _ _ - Jgeoo notioutixa IBION O IOI (T_wo) 17_o‘[ x due I O H HHH padomaoo was oPmSOHQOAdhaflmamxogoao%0Hapvam mo wapoomm oHQHmH> was poaow>maPHD .: madmflm ll 7 onaoflefiHHHE QH flpmdoao>m3 Om: 00: 0mm 00m 0mm j J . J a _ . s + q q . a a a _ — . . q _ O 1 m . o A039 hp // po©fl>flp pcofiOHM%moo soapoqflpxo amaoav HHH dadoQEOQ lllllll / / cacaosho 33315830 Eta B / - / / / .: o.H / / / - x / / u Jgeoo notioutlxe IBTOW (I_mo) +7_o‘[ x duets: 72 in the silylchromates but not present in the corresponding silanol are tabulated in Table IX. TABLE IX INFRARED ABSORPTIONS OF SILYLCHROMATES IN 082 (p) Cr=O Cr=O Asymmetric Symmetric Si-O-Cr Compound Stretch Stretch Unassigned Unassigned Stretch (Me3Si)20rOh 10.1h (s) 10.29 (s) 10.7 (m) 12.1 (vs) (Ph Si) CrO 10.1h (s) 10.26 (s) lO.h (m)* 10.9 (m)* 11.u (vs) 3 2 u broad (p-TolPhESi)2Cr0u 10.15 (3) 10.27 (s) 10.5 (m)* 11.5 (vs) ‘ broad (PhQChSi)2Cr0h 10.13 (s) 10.26 (s) lo.h (m)* 11.1 (vs) (Ch3Si)2Croh 10.20 (5) 10.30 (s) 11.6 (vs) Compound I (masked)** 10.22 (m) 11.5 (s) broad Compound II 10.10 (s) 10.25 (s) 11.0 (vs) Compound III 10.15 (vs) 10.30 (vs) 11-12 (vs) broad Intensities: (3) strong, (m) medium, (vs) very strong * shoulder ** masked by Ph-Si Absorptions at 10.1 and 10.311 correspond to the asymmetric and symmetric stretching frequencies of Cr=O recently reported for chromyl chloride (103). There is evidence (lou, 105, 106, 107) that the Si-O-M (M = metal) linkage absorbs in the lO-ll.5fL region. Silylchromate ab- sorptions in the ll.h-ll.511 region may tentatively be assigned to the Si-O-Cr linkage; however, Smith (108) assigns absorptions in this region 73 to the asymmetric stretching frequency of Si—O. This region is diagnos- tic for silanols, while no absorptions are present for cyclosiloxanes or alkali metal salts of the silanols. Discussion. The ultraviolet absorption of bis(trimethylsilyl)- chromate, bis(tricyclohexylsilyl)chromate, and compound III is due to the chromate group since the other groups of these compounds do not absorb in this region and other chromium(VI) ions and compounds do abs sorb in this region. The spectra of bis(tricyclohexylsilyl)chromate and compound III 'are similar in the visible region (maxima h07 and #00 mp, respectively), but compound III exhibits a broad absorption similar to the phenyl- containing silylchromates and longer extinction coefficients. The ab- sorption spectra of these compounds are markedly different from the other silylchromates except bis(cyclohexyldiphenylsilyl)chromate, which has a broader absorption in the visible and could be considered an ad- mixture of the spectra of bis(triphenylsily1)chromate and bis(tricyclo- hexylsilyl)chromate. The spectra of bis(2-methylfenchyl)chromate (39), which has a maxima at 393 and 287 m+b and chromyl acetate (22) with maxima at hOO and 283 mu, indicate a similarity in chromate absorption with the cyclohexyl-containing silylchromates in the visible region. The phenyl-containing silylchromates and bis(trimethylsilyl)- chromate have similar absorption in the near ultraviolet. This absorp- tion is due to the chromate group since this region is characteristic of the absorption of chromates, and the silanols do not absorb in this region. The broadening and shifting to lower wavelengths of the chromate absorption is suggestive of interaction of the chromate and silyl groups. 71. The phenyl absorptions of diphenylsilanediol and triphenylsilanol (Figure 5) are shifted to higher wavelengths, increased in intensity, and decreased in fine structure when compared to benzene (109). These effects are not as pronounced as in phenol or aniline, but do suggest that interaction of the phenyl groups and the silicon atom does take place. This interaction has been used by previous workers to explain the dipole moment, substitution reactions, and ionization constants of trimethylsilyl substituted phenyl, phenol, and aniline compounds (#9). Hewever, except for compound I, the phenyl-containing silylchromates show a decrease in fine structure beyond the parent silanol. This re- duction in fine structure by replacing hydrogen with chromium(VI) sug- gests interaction of the metal with the phenyl group, for if they were perfectly insulated from one another, the fine structure reduction would not be observed. Once again this effect may not be too large since it is not accompanied by a shift of the absorption of the phenyl groups to higher wavelengths. The presence of chromate absorption in the phenyl region prohibits an intensity study. The exceptional phenyl fine structure of compound I is explained by its structure. It is not composed extensively of Cr-O-Si bonds as are the other phenyl-containing silylchromates, and it possesses fine structure similar to the diphenyl- silanediol. A comparison of the extinction coefficients of the silylchromates in the 300-350 my region gives no correlation among the compounds. No experiment was performed to determine if Beer's law is obeyed by these compounds. 75 triphenylsilanol diphenylsilanediol AH 1809 muoa x PQoH0H%wooo cowposflpxo amaoz 500 wavelength in millimicrons 250 Ultraviolet Spectra of Triphenylsilanol Figure 5. and Diphenylsilanediol 76 Infrared spectrosc0py has been useful in the determination of the structures of the compounds prepared in this work. The detection of OH group in compounds I and III and the absence of OH absorption in compound II was achieved by a study of these compounds in concentrated solutions. In each case, the difference between observed absorption of the compound and the blank of the sodium chloride cells confirmed or denied the presence of the OH group. The strong infrared absorption of the Si-OeSi linkage has also been used for confirmation of structure. Analytical Methods The determination of silicon and chromium in silylchromates is troublesome because complete decomposition of the compounds and total oxidation of chromium is difficult. The following methods were found to be satisfactory for the determination of silicon and chromium in silylchromates. Carbon and hydrogen analyses in this investigation were carried out by Spang Microanalytical laboratory, Ann Arbor, Michigan. All silylchromates prepared except compound II are highly solu- ble in benzene; molecular weights were therefore determined cryoscop- ically in that solvent. Chromium Analysis. The sample was weighed into an iron crucible (low in chromium) and mixed with sodium carbonate as a protective flux. The crucible was then heated for about two hours on a hot plate at high heat to decompose the silylchromate. After cooling the crucible, sodium peroxide was added and the covered crucible was heated with a Maker burner. If the preliminary decomposition was not done carefully, an 77 instantaneous reaction sometimes blew off the cover. Fusion tempera- ture was maintained for three to four minutes. The cooled sample was washed into a beaker and the solution boiled for one-half hour to re- move any peroxide. The solution was acidified with sulfuric acid and chromate was titrated with a recently standardized ferrous sulfate solu- tion. The end point of the titration was determined amperometrically using a rotating platinum anode and a calomel cathode at a potential difference of one volt (110). The detection of a diffusion current by a sensitive galvanometer in the circuit indicated the equivalence point. If an overtitration occurred, the equivalence point could be determined by extrapolation of the galvanometer readings to zero cur- rent. Titration to the diphenylamine sulfonate end point was found to be unsatisfactory because of its sensitivity to acid concentration, particularly phosphoric acid. A dead stOp amperometric titration using similar platinum electrodes came to equilibrium too slowly to be used for rapid titrations. Silicon Analysis. A method of silicon analysis based on the pro- cedure of McHard, Servais, and Clark (111) was used in this investigation. The silylchromate sample was weighed into a platinum crucible and two mililiters of fuming sulfuric acid was added cautiously. TWO mililiters of fuming nitric acid were added to the cooled mixture if needed to facilitate decomposition. The digestion and evolution of sulfur tri- oxide was carried out with a low Bunsen burner flame. The decomposition takes from one to two days. When the sulfur trioxide fumes ceased, the crucible was gradually heated to redness for about one-half hour and then heated at 800°C in a pre-heated muffle furnace for one hour. The 78 crucible was cooled in a desiccator and weighed. The weight of the - residue was taken as chromium(III) oxide and silicon dioxide. Two mililiters of h6% hydrofluoric acid and one mililiter of sulfuric acid were cautiously added and the hydrogen fluoride and sulfur trioxide fumes were removed with the low heat of a Bunsen bur- ner. The crucible was ignited to redness for one-half hour after fumes of sulfur trioxide ceased, and then at 800°C for one hour. The weight loss of the cooled crucible represents silicon dioxide. The weight of the residue of chromium(III) oxide served as a check on the chromium analysis. Molecular Weights. The molecular weights of all the compounds ‘ I prepared in this investigation were determined by freezing point de- é pression of benzene solutions. An apparatus based on temperature measurement with a thermistor was constructed for the measurement of the freezing point of these solutions. A thermistor*, two 1000 ohm resistors ( O.1%)**, and two resis- tance boxes in series were used to construct a Wheatstone bridge cir- cuit. The current for the Wheatstone bridge was supplied by a 1.5 volt . . dry cell battery. Bridge balance was determined with a Scalamp Gal- vanometer***. The thermistor was attached to a cap which could be screwed onto a two ounce, narrow-mouth, amber bottle containing the solution. The *Victory Engineering Corporation, Union, New Jersey. **Shallcross Company, Collingdale, Pennsylvania. ***W. G. Pye, Ltd., Cambridge, England 79 bottle was fitted into a cylinder surrounded by ice, and the solution was stirred by a magnetic stirrer. The thermistor was calibrated using the resistance values for the freezing points of conductivity water (0.00000) and benzene which had been purified by recrystallization twice and distillation (5.h93°c) (20). Intermediate temperature values were calculated from the exponential equation of resistance: QT - Tb)B . a log 3_ T0T(2.71828) R0 The constant B is characteristic of the thermistor and was cal- culated using the freezing points of water and benzene. The resistance of the thermistor, R, at temperature T (OK) was calculated using the re- sistance R0 and the temperature T6 for the freezing point of water. The calibration was extended to 5.75000 so that nitrobenzene could be used as a cryosc0pic solvent (Table X). The sample and benzene for the molecular weight determination were weighed directly into the sample bottle, the magnetic stirring bar and thermistor were inserted, and the bottle was placed into the ice bath. Resistance was measured at intervals as the solution cooled and froze. The cooling and freezing curves were then extrapolated to their intersection at the freezing point. The freezing point in ohms was then converted to degrees centigrade by interpolation. The differ- ence between the freezing point of benzene (prior determination) and the solution gave the depression of the freezing point of the solution. The molecular weight was then calculated from the freezing point depres- sion equation Mgw, s Kth.solute(lOOO) AT Wt . solvent CALIBRATION OF THE THERMISTOR TABLE X Temperature in OC \J'I\I‘I U1\J'I\J'I\J'l mmm .000 .100 .200 .300 .hOO h h h h h h. h h h h 500 .600 .700 .800 .900 .000 .050 .100 .150 .200 .250 .300 .350 .hOO Resistance in ohms 5559» 553h- .6 5h86. 5h62. 5h38. 5510 5hlh. 5390- 53670 53h3. 5320. 5308. 5297- 5285. 527k. 5262. 5250. 5239- 5228. 1 7 47-" U0 UL) U.) U1 0 Continued 81 TABLE X - Continued Temperature in OC Resistance in ohms 5.h50 5216.6 5.500 5205.3 5-550 5193.9 5.600 5182.5 5.650 5171.3 5.700 5160.0 . 5.750 51h8.8 where M.W., Wt.solute, Wt.solvent, and ZlT are the molecular weight, weight of solute, weight of solvent, and freezing point depression re- spectively. K? is the freezing point depression constant which is given as 5.069 (20). A check on the use of this constant was made using high purity pftolylsulfonylmethylnitrosamide. The observed molecular weight was 231.2 (calculated, 231.3) using the above constant. The error in the determination iS'i:0.h ohms, which is equivalent to :t 0.002%. DISCUSSION Prior to this investigation, two silylchromates have been re- ported in the literature. Bis(trimethylsilyl)chromate was prepared from hexamethyldisiloxane and chromium(VI) oxide and bis(triphenylsilyl)- chromate was reportedly prepared by the reaction of triphenylsilanol with chromium(VI) oxide (86). In this investigation bis(pftolyldi- phenylsilyl)chromate, bis(cyclohexyldiphenylsilyl)chromate, and bis(tricyclohexylsilyl)chromate have been prepared and characterized, bis(triphenylsilyl)chromate was prepared and characterized independ- ently. In addition, further information was obtained on unstable bis(trimethylsilyl)chromate and bis(methyldiphenylsilyl)chromate, and the preparation of a new silanol, cyclohexylsilanol, was achieved. Numerous investigators (69, 70, 71, 72) have reported the pre- paration of polysiloxy-metal compounds but none have been character- ized. No known heterOpOlycompound with an A-B repeating unit of mod- erate molecular weight has previously been prepared. The preparation of compound III (page 66) is the first instance of the formation of such a heteropolymer. Cyclic siloxanes containing metals have been re- ported by Kbenig (72) and by Zeitler and Brown (112). Koenig obtained the following compound from the reaction of dibutyltinoxide and diphenyl- silanediol in dioxane: 83 Zeitler and Brown have reported the preparation of the following com- pound from the reaction of titanium tetrachloride and diphenylsilane- diol: 0 —Si si—o I \o\ o/ l Phg—Si /Ti Si.—.-Ph2 I ,0 \o I O—Sii \Si—o I Ph2 Ph2 Compound II (page 50) is unique in that it is a large ring compound con- taining two metal atoms and is the first reported compound of this type containing chromium. Compound I (page 50) is also unique for chromium and is also the first characterized structure of this type. Compounds similar to compound I may have been obtained in previous works but they have not been isolated or well characterized. Bis(triorganosily1)chromates were synthesized by condensation of the appropriate silanol with chromium(VI) oxide in chlorinatedemethane solutions in which the silanols are soluble but chromium(VI) oxide is not. Solutions of the silanol immediately took on a yellow-orange color when chromium(VI) oxide was added, and in all cases the reaction mixtures were highly colored within thirty minutes. The products of these reactions were in general easily isolated and purified. Charac- terization of the products was based upon analytical and spectrophoto- metric data. Since chromium(VI) oxide is insoluble in the solvent, the initial step of the reaction for the formation of the silylchromate is probably absorption of the silanol on the surface of chromium(VI) oxide. Surface 8h reactions sites probably are defects which contain chemisorbed water-- that is, sites containing chromic acid. The absorption complex may be attacked by the silanol to give the silylchromate or by water to form the half chromate ester. The half chromate ester (R3SiOCr03H) may be freed from the surface by the solvent and react immediately with a second silanol molecule to form the silylchromate: REACTION SCHEME A SiR SiR I 3— 3 q_ SiR3 0 55 /H O ' s 0 03)) 31% : ' - R i H : O‘Qf’o .3EIL. O‘flg’o___ .;i__s, _I:~c /'___ II "" ““ II N 0 0 0 Ié;>\\\ l 0 ll R3SiOH (R3Si)2Cr0h.e_R3SiO - fire” 3 - H «<————-R3Si00r03H (R3Si)2Cr0h _./ 0 0\ . SlR3 The proposed formation of a half ester is based on a similar pro- posal in the oxidation of alcohols by chromic acid (32), and retention of oxygen by the silicon atom is based on studies involving organochro- mate esters (36, 38) and the silicon-oxygen bond strength. Chromic acid or hydrated chromium(VI) oxide is justified by the observed en- hancement of the reaction when slightly moist chromium(VI) oxide is used as a reactant. The homogeneous reaction of triphenylsilanol and chromyl chloride probably proceeds by a similar two-step mechanism in solution with elimination of hydrogen chloride. 85 The rapid reaction of diphenylsilanediol with either chromium(VI) oxide or chromyl chloride is surprising. Formation of a mixture of com- pounds I and II in each case was accomplished within thirty minutes, whereas longer reaction times (10-30hours) reduced the yields of products and increased the yield of decomposition products. no decom- position products were obtained from the thirty-minute reaction. Failure to obtain an A-B silylchromate with alternate chromium- oxygen-silicon-oxygen bonds from the reaction of diphenylsilanediol and chromium(VI) oxide or chromyl chloride is attributed to the tendency of the diol to undergo rapid self-condensation which is either acid or base catalyzed. This is illustrated by the observation that diphenyl- silanediol is not soluble in pure methylene chloride, but in the pre- sence of a trace of chromyl chloride it dissolves and forms hexaphenyl- cyclotrisiloxane and higher siloxanes. The condensation of diphenyl~ silanediol to form hexaphenylcyclotrisiloxane apparently does not par- ticipate in the overall formation of compound I, Compound I ispro- duced only in poor yield by the reaction of hexaphenylcyclotrisiloxane and chromyl chloride or by the reaction of chromium(VI) oxide or chro- myl chloride with hexaphenyltrisiloxane-1,5~diol. Compound II is not obtained at all from these reactions but is produced in good yield by the reaction of tetraphenyldisiloxane-l,3-diol and chromium(VI) oxide. Also, compound II undergoes a rapid reaction (15 minutes) with diphenyl- silanediol in methylene chloride to give a product with the character- istics of compound I. A possible mechanism for the formation of compound II from the reaction of diphenylsilanediol with chromyl chloride in methylene chloride i I "I i 86 is a reaction of the diol with chromyl chloride by elimination of hydro— _ gen chloride to form Ph2(OH)SiOCr(O2)C1. This intermediate may react with a second molecule of the diol to form Ph2(OH)SiOCr(02)OSi(0H)Ph2, which may in turn undergo condensation with a similar molecule to form compound II. Compound II could also be produced by the formation of tetraphenyldisiloxane-l,3-diol by self-condensation of the diol, fol- lowed by successive reactions with chromyl chloride. These two plausible mechanisms for the formation of compound II are illustrated by the following schemes: REACTION SCHEME B Ph Ph 0 I /H...__ -HCl | || HO - Si--.-._Q_ 0.1 ——> HO-Si-O-Cr-Cl I ‘b'sclr=0 I H Ph 01 Ph 0 khZSflOHb, ~HCl o B I I‘ 1'>h2 o/Il \o Ph2 _ Si \ Ph 0 Ph l I bimolecular H (II) 0 0 condensation HO-Si-O-fiP-O-Sf-OH | _ /Si 0 Si ( 2320) Ph 0 Ph / Ph2 0\|l /0 Ph2 87 REACTION SCHEME C Ph Ph bimolecular Ph Ph 0 Ph Ph -H2O | l condensation | I H I I 2Ph Si OH “——" HO-Si—O-Si-OH ‘__ ’;: HO-Si-0-Si-0-Cr-0-Si-O-Si-0H 2 I )2 I I 33333 I I II I I Ph Ph Ph Ph 0 Ph Ph Cr02C12, (-HCl) / \ / \ Ph 0 H 0 Ph Phg H 0 Ph2 2“Si// 0 \\S 2’ 2 \\Si// o ‘\Si// | -HC1 I I (II) 0 0 ‘+————’ 0 p I l Si 5 Si Si / \ / \ / \ O '\ Hg “GI/fl m5 Ha oq,fl.mo Hg r Cr 3H fi ..\ 0 0 Cl Compound I may be formed from the reaction of 2 molecules of di- phenylsilanediol with compound II (Reaction Scheme D), hexaphenylcyclo- trisiloxane with chromyl chloride, or hexaphenyltrisiloxane-l,5-diol with chromyl chloride (Reaction Scheme E). Only the first and last reactions seem plausible. The reaction of diphenylsilanediol with chromium(VI) oxide prob- ably proceeds in a manner similar to Scheme A. Intermediates analogous to those of Schemes B and C may give compound II. Reaction Scheme D is favored over a reaction scheme analogous to E since combination of hexa- phenyltrisiloxane-l,5-diol with chromium(VI) oxide failed to give compound I. The higher yield of compound II over compound I in the reaction of 88 REACTION SCHEME D PI PI Cr Cr /” \ /I| \ ph 0 o p Ph2 /,0 0 0 /,Ph 2\ Sli/ O \Sli/ 1'12 \Ti \fol 2 Ph Si(OH) 2 (H) O ‘I ‘2’ (I) I’ I . Si Si Si Si-Ph2 Ph/ \0 I). 0 \Ph Ph/ \OIC? 0 2 \Cr/ 2 2 \Cr I \ IOI IOIOH ISi—Ph2 OH lPh2Si(OH)2, .-H2CrOI, Ph Ph Ph fi Ph Ph Ph | l | HO-Si-O-Si-O-Si-O-Cr-O-Si-O-Si-O-Si-OH (I) I I l H l I Ph Ph Ph 0 Ph Ph Ph REACTION SCHEME E Ph Ph Ph ~2H20 3Ph28i(OH)2 -————s>- HO-Si-O-Si-O-Si-OH Ph Ph Ph lor02012, (-2HC1) Ph Ph Ph fi Ph Ph Ph HO-Sfi-O-Sfi-O-Sfi~O-Cr-O-Sfi-O-Si-O-Si-OH (I) II ' Ph Ph Ph 0 Ph Ph Ph 89 diphenylsilanediol with chromium(VI) oxide may be due to the fact that reactions of Scheme E do not occur with chromium(VI) oxide, or to a slower reaction due to the insolubility of chromium(VI) oxide in methyl- ene chloride. The formation of compound III from the reaction of dicyclohexyl- silanediol and chromium(VI) oxide can only be envisaged as resulting from consecutive chain-building reactions such as are suggested by Scheme A. The termination of the reaction of structure III could be due to an equilibrium of bond making and breaking which prevents the formation of longer chains. It is also possible that a cyclic inter- mediate is formed and that this terminates the reaction. Compound III might then be formed by the reaction of the cyclic structure with the diol in a manner similar to Scheme D. It is also possible that compound III is largely a mixture of compounds with an A-B structure but with silicon-to-chromium ratios of 5zh, 6:5, and 7:6 which are formed in a reaction similar to that of Scheme A. Self-condensation of dicyclohexylsilanediol takes place only under conditions which are probably not attained in the heterogeneous reaction of the diol with chromium(VI) oxide in methylene chloride. The thermal stability of the bis(triorganosilyl)chromates follows a definite pattern. The least stable is bis(trimethylsi1y1)chromate which eXplodes when heated to about 80°C; the most stable is bis(tri- phenylsilyl)chromate which decomposes as it melts at 153.5-15h00. The effect of the presence of donor groups on the stability of the silyl- chromate is illustrated by the reduced stability of bis(petolyldiphenyl- silyl)chromate (m.p. 98.5-9900, decomp.) or bis(methyldiphenylsilyl)chromate 90 (decomp. 85°C) compared to bis(triphenylsilyl)chromate. Of the solid silylchromates prepared, only bis(cyclohexyldiphenylsilyl)chromate was stable above its melting point. Compound I is the most thermally stable compound prepared in this investigation. It has no definite melting point but passes from a brittle glass into a viscous fluid by the heat of the hand. Differen- tial thermal analysis of this material shows it begins to decompose at 275°C. Compound II decomposes very sharply at 169°C without melting. Compound III decomposes on standing, and no thermal investigation was made on the freshly prepared oil. The high stability of compound I. may be due to the presence of the linear chains of siloxy groups in the molecule; linear polysiloxanes containing phenyl groups are stable to hOO-SOOOC (MO). Incorporation of chromium(VI) into the siloxane moiety apparently reduces its stability. Compound II, like compound I, pos- sesses siloxy groups in its structure, however, its lower stability may be due to thermal cracking of the cyclic structure. Even though their structures are different, a comparison of the stabilities of com- pounds I, II, and III suggests that the lower the silicon-to-chromium ratio, the lower the stability of the product. This further implies that a high temperature stable heteropolymer containing repeating oxygen- chromium(VI)-oxygen-silicon linkages cannot be prepared under the con- ditions discussed in this work. The spectra of the silylchromates (Figures 2, 3, h) suggest that the nature of the bonding in these compounds is complex. The phenyl- containing silylchromates as well as bis(trimethylsilyl)chr0mate have a similar absorption spectrum in the near ultraviolet. The chromate ion 91 has a Sharp absorption at 379 mytwhile bis(2-methylfenchyl)chromate and chromyl acetate absorb at 393-MOO mg. The near ultraviolet absorption in the silylchromates is unquestionably due to the chromate group. The broadening and shifting of this absorption to lower wavelengths may be due to interaction of the chromate group with the substituents of the silicon atom. The failure to obtain any correlation of the extinction coefficients for this absorption and the reduction of the fine structure of the phenyl absorptions in the ultraviolet further substantiate this hypothesis. The similar near ultraviolet absorption of bis(trimethyl~ silyl)chromate and the phenyl-containing silylchromates indicates a similar effect is Operative. The silicon and oxygen atoms of silanols are known to interact to form a partial double bond involving a vacant d orbital on the sili- con atom and a filled p orbital on the oxygen. This bond lowers the electron density on the oxygen atom forming R~Si-=O*. In phenyl- containing silicon compounds there is also evidence that theTT-electrons of the benzene ring may partially overlap with the vacant d orbitals of silicon. Thus in the phenyl-containing silanols the oxygen atom is partially in conjugation with the phenyl group. The removal of charge from the oxygen atom and the conjugative effect should tend to stabilize the chromium-oxygen bond since the chromium atom is highly electrophilic. A similar stabilization in the case of the methyl group could be brought about by hyperconjugation of the protons of the methyl group. These effects may only be operative to stabilize the excited states brought on by the absorption of near ultraviolet radiation rather than 92 stabilization of the ground state. The definite inductive effect of the trimethylsilyl group (h9) should increase the electron density on the oxygen atom and this in turn should facilitate charge transfer to the electrophilic chromium atom. Bond fission thereby becomes more likely. Thus bis(trimethylsilyl)chromate should have lower thermal sta- bility as observed. The exceptional near ultraviolet Spectrum of the cyclohexylsilyl- chromates may be due to two separate effects. The cyclohexyl group is unable to undergo the complex interactions possible with the methyl and phenyl groups and therefore would not stabilize the excited state as prOposed. Steric hindrance is always a consideration of cyclohexyl compounds, since it is a bulky group. This effect may be particularly Operative in the tricyclohexylsilyl group since tricyclohexylsilane fails to react even with methyllithium (96) under forceful conditions. Thus if Operative in bis(tricyclohexylsilyl)chromate, steric hindrance should destabilize the ground state and bring about absorption at higher wavelengths. This is partially substantiated by the broad absorption spectrum of compound III in which steric hindrance is less likely. The infrared spectra of the silylchromates (Table IX or Appendix 2) all show absorptions at 10.1 and 10.3ti. These absorptions are iden- tical to those assigned to asymmetric and symmetric stretching fre- quencies of the chromium-oxygen double bond in chromyl chloride (60, 103). A comparison with the higher wavelength infrared absorptions of the chromate and dichromate ions (60), where pure double bonds are not involved, suggests that in the silylchromates as in chromyl chlor- ide the absorptions are due to the presence of a chromium-oxygen double bond. 93 Tentatively the infrared absorptions of the silylchromates in the 11.5}; region have been assigned to the chromium-oxygen-silicon linkage, and the shift of this absorption to 12.1}; in bis(trimethylsilyl)chro- mate might be taken as cOrroborating evidence for the existence of hyperconjugation effects. This same absorption occurs at 11.0rL in compound II which may represent Cr-O-Si in strain-free cyclic systems. All of the Silylchromates prepared in this work are sensitive to light. Photochemical decomposition did not seriously hamper this investigation, since it is slow enough to allow the compounds to be exposed to light for short periods. An example of this is the decom- position Of compound II in light. The bright yellow-orange crystals Of this compound fade to a brown color within two-three hOurs of ex- posure to room light; on prolonged exposure only an unidentified black, amorphous solid remains. The photochemical decomposition of the silylchromates may be analogous to that of chromyl chloride (1h, 59), since the internal symmetry about the central chromium atom is the same in each case. This decomposition would involve absorption of radiation in the near ultra- violet which gives rise to charge transfer to the central chromium atom and leads to the dissociation: Since the products have not been isolated, this can only be considered as a possible mode of decomposition. 9h The toxicity of all known hexavalent chromium compounds (page 16) warrants a word of caution concerning the silylchromates. The physio- logical prOperties of these compounds have not been investigated and more than normal care should be observed in handling them. This inves- tigator has had contact between the skin and these compounds with no adverse effects; however, in each case of contact the area was thor- oughly cleansed. SUMMARY AND CONCLUSIONS Condensation of silanols and silanediols with chromium(VI) oxide andchromyl chloride proceeded easily to give several interesting prod- ucts. Triorganosilanols react with chromium(VI) oxide to give bis(tri- organosilyl)chromates in good yields, while the reaction of diorgano- silanediols with chromium(VI) oxide or chromyl chloride gave more com- plex products. The products Obtained were the first reported A-B heteropolymer containing silicon and chromium, the first cyclic siloxane containing chromium, and a linear siloxy-chromium compound. The com- pounds were easily prepared and characterized, but isolation of the product in a pure state varies since final removal of impurities was difficult in some cases. The reactions of triorganosilanols with chromium(VI) oxide gave a series of interesting compounds which were easily isolated, purified, and characterized. The stability and spectra of these compounds were related to the organic substituents of the silicon atom. Reaction of either chromium(VI) oxide or chromyl chloride with diphenylsilanediol in chlorinated methane solvents gave a mixture of two siloxy-chromium compounds. Compound I, a linear siloxy chromate, was the most difficult to Obtain in a pure state because of contamina- tion by compound II, a cyclic siloxy chromate, and by condensed silox- anes. These compounds contain the siloxane (Si-O-Si) linkage due to the tendency of diphenylsilanediol to undergo condensation under the 95 96 reaction conditions. Studies of the reactions of the condensation prod- ucts of diphenylsilanediol with chromyl chloride or chromium(VI) oxide enabled the postulation of reaction schemes for the production of com- pounds I and II. Self-condensation of dicyclohexylsilanediol is unlikely when the diol undergoes reaction with chromium(VI) oxide in methylene chloride solution. The product of this reaction, compound III, was assigned a structure containing regular A-B repeating units of chromium-oxygen- silicon-oxygen on the basis of analytical data. This product decomposed on standing to a black, amorphous solid. Attempts to prepare silylchromates by addition of chromyl chlor- ide to cyclic siloxanes was unsuccessful. Also, the condensation of dialkoxysilanes with chromyl chloride was unsuccessful due to the probable reaction of chromyl chloride with the aliphatic groups. All of the silylchromates prepared in this investigation slowly decompose in the presence of light. The thermal stability of these compounds is dependent on the substituents of the silicon atom and the structure of the compound. In general the phenyl-containing compounds were the most stable, while the least stable silylchromates were those containing aliphatic groups. Also, it was found that the larger the number of siloxy groups in the compound, the greater the stability. Spectral studies of the silylchromates reveal that interactions of the silicon substituents and the chromate group are Operative. The observed shifts of chromate absorption to lower wavelengths may be due to stabilization of the excited state rather than the ground state. The anomalous spectrum of the cyclohexyl-containing silylchromates was 97 tentatively explained on the basis of steric hindrance or inability of interaction. Further studies of related compounds are necessary to in- terpret fully the nature of the proposed interactions. Infrared spectrOSCOpy was important as a means of product iden- tification. All the groups present in the silylchromates participate in absorptions in the infrared. The absorptions of the OH, Si-OoSi, and tentatively assigned Cr-O-Si linkages were particularly important for structural assignments. The original intention, to prepare high temperature stable poly- mers containing chromium and silicon, was not achieved. A comparison of the stabilities of compounds I, II, and III in which the silicon-to- chromium ratio varies from 3:1 to 1.2:1 indicates that an increased chromium(VI) content in the siloxane moiety lowers its thermal stabil- ity; this comparison neglects structural differences which also may effect the stability of the compound. The thermal decomposition may be due to fission of the chromium-oxygen bond as in the thermal decom- position of chromium(VI) oxide. The stability trends and proposed interactions may be particularly important in view of their relation to siloxy compounds of other transition metals. RECOMMENDATIONS FOR FUTURE WORK 1. Compound II has been Obtained as prismatic needles of almost a centimeter in length when slowly crystallized from methylene chloride solutions. Single crystal xeray diffraction studies should be carried out to determine its structure in the solid state. 2. Kinetic studies on the formation and decomposition of the silylchromates prepared in this work might give considerable informae tion on the stability of chromium-oxygen-silicon linkage. The formation of the bis(triorganisilyl)chromates could easily be followed by spec- trOphotometric techniques, as could their photochemical decomposition. The direct hydrolysis of bis(trimethylsilyl)chromate with water enriched in 018 would enable a determination of the relative stability of the bonds in this compound. HOmogeneous hydrolysis of the other silylchro- mates in acetone-water solutions could also be studied using 018. 3. The reaction of cyclic compound II with diphenylsilanediol has only been studied qualitatively. From these observations compound I apparently is easily produced in good yield. The reaction of the diol and compound II should be studied more thoroughly and the products characterized. This reaction should give compound I’in a higher pur- ity since the diol is easily separated from compound I. The reaction might also be carried out using other diols or trisubstituted silanols to give a whole series of derivatives which would have varying prOper- ties. 98 99 h. The further polymerization of compounds I, II, and III should be studied. The terminal OH groups of compounds I and III should under- go further condensation with chromyl chloride or chromium(VI) oxide to give higher molecular weight products. The cyclic structure of compound II suggests that it may undergo addition polymerization in the presence of a catalyst such as Iawis acids or alkali. 5. The failure to produce an A-B heteropolymer from the reac- tion of chromyl chloride and diphenylsilanediol may be due to the re- action conditions. Perhaps at lower temperatures (-hO°C) condensation of the diol and chromyl chloride may give the A-B heterOpOlymers. This same reaction might also be tried in a solvent like acetic anhy- dride in which both the diol and chromyl chloride are soluble. 6. Resolution of the nature of the interactions predicted from the spectral studies of the silylchromates might be accomplished by studies of silylchromates containing allyl, vinyl, and isopropyl groups. ThelT-electrons of the vinyl group reportedly interact with Silicon in vinylsilanes, but theTT-electrons of the allyl group do not interact with silicon. The isopropyl group has about the same inductive effect as a cyclohexyl group but is less subject to steric hindrance. APPENDIX 1 REACTIONS OF CHROMYL CHLORIDE At the beginning of this investigation a search was made for po- tential co-monomers for reaction with chromyl chloride to give polymeric materials. This search was unsuccessful; however, some of the informa- tion should be reported. Chromyl chloride reacts with the following compounds in methylene chloride to give hygroscOpic, brown addition compounds which dissolve in water to give green solutions: succinic acid terephthalic acid ethylene glycol dimethyltindichloride dibutyltindiacetate glycol di-mercaptoacetate The chromyl chloride addition products of these materials may probably be considered to be Etard complexes. Ethylenediamine reacts with chromyl chloride in methylene chlor- ide to give a red-brown solid which dissolves in water to form a dark red solution and may be a complex of chromium(III). Silico-oxalic acid gives silicon dioxide when it reacts with chromyl chloride. Sulfur re- acts with chromyl chloride in carbon disulfide to give a brown precipi- tate which decomposes on addition of water to give sulfur and a green solution. 100 101 The reaction of the Etard complex of benzene and chromyl chlor- ide with ethylene glycol gave a hard, green solid which is soluble in water. The cohydrolysis of dimethyldichlorosilane and chromyl chloride in refluxing methylene chloride gave orange needles when pyridine was used as a base. In the presence of other bases and lower temperatures only brown addition compounds were obtained. The orange needles slowly decomposed into a brown solid; they may be the methyl analog of com- pound II. Cohydrolysis of dimethoxydichlorosilane and chromyl chloride gave only brown solids which were easily hydrolyzed to silicon dioxide. Addition of chromyl chloride to a glacial acetic acid solution of potassium chromate gave potassium chlorochromate as bright orange platelets. The hydrolysis of very dilute carbon tetrachloride solue tions of chromyl chloride at different temperatures indicates that the hydrolysis is slow and the rate decreases with decreasing temperature. APPENDIX 2 INFRARED SPECTRA 102 m©H%HSmHm mopamo CH mpdfioyfloAHhHHmahcmamflhpvam mo ESMPommm commHMGH .w ohzwflm mGOHQHE EH Spmcmam>m3 i ma ma 3 OH m m we m in a) A. ll. . _ 4 . . fl a . 0 no 1. z ‘1 Umppflao «Awuw QOHpmHoQO pcm>aom n om c J OOH uotssrmSUBJi queoxea 101+ :H ma omflm mwwsonm afiflmmmpom QH HommHHmHhcmsmHHB Mo Esapommm cosmamcH .P maswflm NH HH mmoaofle SH Spmsmam>m3 OH m w pm u ‘ _ dmppflao «Awnm uofimma l 0: ow OOH uotsstmsuexi iueoxsa 105 '. m©HMHSmflQ monamo ca umSOHSOAahaflmamcmmmflcahaomevmHm wo Esapommm wmhmamcH .w madmflm mQOMOHE QH npwqmam>m3 i 3 ma 3 0H m m e. m m "3?!" dmppflao 4+>um QOHPQHoQO pcm>aom ON 3 om ow OOH uoIsstmsueJi iueoxag m©H%H5mHQ dopamo 2H HocmaflmahconmflwamaoELm yo Esapommm copmsqu .m madmflm mflOpowa CH npmcmam>m3 :H ma ma Ha OH 0 w >_m m 4 _ n _ a . . |q _ a O 6 _ no .l pmpPHSo «Ipuw GOHPQMOQO pcm>Hom L ON 104 1 00 -8 OOH morsstmsuexq iueoxeg 107 m@HMH5mHQ monmmo QH umSOHQoAHhHHmHhcmanUHhmeOHohoVmflm mo 539%ommm wmhmmmsH .OH msstm MH NH maOHOHS SH npmcmHm>m3 OH m w % m s m 4 _ _ Umppflao «‘Nuw QOHPQHOQO pdm>Hom ‘ - 0: ow OOH uotsstmsuaxi ierxaa 108 mUHMHSmHO consmo QH HoquHmHhcmanUHhmeOHoho mo Esaeocgm UmamaMcH NH HH muonoHa cH npwamHm>m3 OH m m is m in .Ha mssmnm ‘ Umppflao «Awum coapmaomnm pcm>Hom ON 0: om OOH uotsstmsusxi iueolea mdflngmflw copnmo QH wmeOHaoAHhHHmHhmeOHohoHapvam mo Edapummm wmhmpwsH .NH maswflm muonofla CH memmHm>m3 S S d 2 a w L s m s 109 - _ H _ . _ _ a _ wmppflao «bwnw QOHPQMOmDm pmm>Hom uorssrmsuaiq qusoxsa omHO meanm ESHmmmpom cH HommHHmHhchOHohoHaE mo asapomgw mommamsH .MH madem muonoHS QH npmcmHm>m3 110 :H H NH HH OH m w Lo m _ _ dcpPHfio «‘PIO QOHmmp ON 0: ow OOH q iueoxea uorssrmsusx 111 i mUHMHSmHQ sonhmo QH mpmsopnoAHhHHmthpcSthvam mo asapommm wmmmame .qH madem mQOMOHS SH :pwcmHm>m3 S E. S on a m aw m s cmPpHEo «nwnw QOHpgaompm PQm>Hom ON om OOH uctsstwsuexi iueoxaa :H mUHHHSmHO conhmo QH HoquHmHhsmeHaE mo SSMPommm wchmamnH .mH mHSmHm mdonoHE GH QHmmmHm>m3 3 ma 2 0H m m w 112 _ _ . a . . . UmPpHEo afiwum QOHHQMoQO pum>Hom ON c 4- 1g iueoxea ow uorssrwsue ow OOH 113 m©HMHSmHQ monamo 2H AOm cmmgo m dadomaoo Mo adhpommm ccamammH .OH mssme mehoHE QH gpmmmHm>m3 fl ma ma 3 S m w L m LIN ..d‘ @mppHso .LNIO SOHPQHome PGm>Hom ON 0: om OOH uotsstmsuemi iueoxeg :H 3 omHm mUHanm nBHmmmpom SH HOHcmmmHHmHhccgonHQ .Ho 83.50ch concamcH SH madem mQOHOHE GH gpwcmHmhrms ma 3 0H m w s_ m m s 11h mePHSo iwuw QOHMPH ON on 0 KO uotssrmsu'ex OOH q, iueoxed 115 mGOHOHE 2H npmsmHm>mB OH 0 w memHSmHQ copamm QH AOm mmmgv mm ondomsoo mo Edhpocmw Umamsqu .wH cadem N_w m m Nmo m5mhm> Nmo* @prHEo «*wuw moHpmHome pmm>Hom UOISS IUIS ERIC). CLUQDlaa 116 mOHMHdeQ conpmo SH HOHOIM.HucmmeHHmHthmmnmmhpcB mo asapommm wmhmhmmH .mH cpzmHm muoaoHE :H npmquc>m3 :H MH NH HH OH m m N O m a - _ _ - _ . _ a . O wmppHao afiwuw GOHPQHOmDm pdm>Hom I ON 10: low low OOH i iusoxsa IIOTS S IUIS UPI 117 wUHwHSmHO mopsmo CH Aww cwmmg HHH Undomaoo mo Edhpocmm UmMmHMcH .ON mastm mdopoHE :H prsmHm>m3 3 NH i. on m w L m m s wmppHEo «bwuw QOHPmaompm pcm>Hom ON c ..:t’ i iusoxea O \O uotssrmsuex ow OOH 118 mewHSwHO Sonamo SH HOHOmSmHHthXcSOHohoHO mo adapocmm dogmaMSH .HN isme mSOSOHS SH SpmSmHm>m3 ma ma 1 OH m w L o m in UmppHao (fiwnw SOHpmaoQO pSm>H0m OOH uotsstwsusxi iusoxsd APPENDIX 3 X-RAY POWDER DIFERACTION DATA TABLE XI INTERATOMIC SPACINGS FOR BIS(TRIPHENYLSILYL)CHROMATE 6 Relative Line (degrees) d’hkl (A) Intens ity Chromium K Radiation (%.= 2.2909 A) 1 5.58 11.79 5 2 7.51 8.77 8 3 8.58 7.68 1 h 9-15 7.20 5 r 5 9.611 6.814 1 6 10.29 6.u2 1 3 7 11.23 5.88 2 ' 8 12.09 5.h8 2 9 13.76 n.82 *3 10 1h.11 h.70 1 11 15.0u 1.12 10 12 16.10 1.13 3 Copper K Radiation (>.= 1.5h18 A) 13 10.95 u.06 1 1h 11.38 3.91 1 15 11.69 3.81 1 16 12.8u 3.h7 1 17 13.20 3.38 0.5 18 13.51 3.30 1 19 11.31 3 11 1 20 1h.80 3 02 1 21 15.h6 2 90 1 22 16.05 2 79 1 23 18.5u 2.h7 1 119 120 3 TABLE XII INTERATOMIC’ SPACINGS FOR BIS(P_-TOLYLDIPHENYISILYL)CEROMATE Relative Line (degrees) dhkl (A) Intensity Chromium K Radiation (x: 2.2909 A) 1 5.59 11.75 h 2 7.00 9.39 5 3 7-5h 8.73 5 u 8.1M 8.09 7 5 8.59 7.67 5 6 9.86 6.68 10 7 11.15 5.92 3 8 12.18 5.43 3 9 12.86 5.15 5 10 13.58 u.88 1 11 1A.10 1.70 1 12 1h.71 h.51 9 13 15.0h u.ui 10 1A 15.31 A.3h 1 15 15.79 1.21 1 16 16.30 u.o8 2 17 17.05 3.91 1 18 17.h5 3.82 2 19 19.2h 3.u8 1 20 19.83 3.38 1 21 20.30 3.33 1 22 21.23 3.16 1 23 21.60 3.11 1 2A 22.23 3.03 1 25 23.29 2.90 1 26 2u.13 2.81 1 27 2h.96 2.72 1 28 25.75 2 6h 1 29 26.23 2 60 1 30 26.6h 2.55 1 31 28.20 2.u2 1 121 TABLE XIII INTERATOMIC SPACINGS FOR BIS(TRICYCIOHEXYLSILYL)CHROMATE Relative Line (degrees) dhkl (A) Intensity Chromium K Radiation (l= 2.2909 A) 1 5.29 12.h1 h 2 7.08 9.30 h 3 8.08 8.16 A h 9.00 7.33 6 5 9.39 7.03 2 6 9.95 6.6a 1 7 10.60 6.2A 1 8 12.33 5.37 2 9 13.3h u.97 1 10 13.7h h.83 10 11 1h.3u n.63 1o 12 15.19 n.38 1 13 15.51 h.28 1 1A 16.36 4.07 2 15 18.A3 3.62 1 16 19.00 3.52 1 17 19.61 3.h1 1 18 20.21 3.33 1 19 21.00 3.20 1 20 21.66 3.10 1 21 22.35 3.01 1 22 22.93 2.9a 1 23 23.66 2.85 1 Copper K Radiation (1.: 1 5A18 A) 2h 16.66 2 69 1 25 17.99 2 50 1 26 18.15 2 AB 1 27 18.6h 2 A1 1 28 22.u6 2 02 1 29 26.83 1 71 1 122 TABLE XIV INTERATOMIC SPACINGS FOR COMPOUND E 0 Relative Line (degrees) dhkl (A) Intensity Chromium K Radiation ().= 2.2909 A) 1 5.65 11.62 7 2 6.78 9.70 7 3 7.AA 8.8A 10 A 8.85 7.AA 2 5 9.63 6.8A 1 6 10.A6 6 31 1 7 11.36 5 81 1 8 12.A3 5 32 1 9 13.68 A 8A 10 10 15.06 A A1 10 11 15.66 A.2A 8 12 15.96 A 17 3 13 16.38 A 07 1 1A 17.08 3 90 2 15 17.79 3 75 l 16 18.20 3 67 2 17 19.70 3 A0 1 18 20.25 3 31 1 19 20.70 3 2A 1 20 21.25 3 16 1 21 21.5A 3 12 1 22 21.99 3.06 1 23 23.1A 2.92 1 2A 23.8A 2.83 1 25 26.5A 2.56 1 26 27.51 2.A8 1 27 30.79 2.2A 1 28 31.58 2.18 1 29 32.3A 2.1A 1 123 Figure 22. X-ray Diffraction Pattern of Bis(triphenyl- silyl)chromate, Chromium K“ Radiation Figure 23. X—ray Diffraction Pattern of Bis(petolyl- diphenylsilyl)chromate, Chromium Kg Radiation Figure 2%. X-ray Diffraction pattern of Bis(tricyclo— hexylsilyl)chromate, Chromium Kg Radiation Figure 26. Figure 25. 12A X—ray Diffraction Pattern of Compound II, Chromium K4 Radiation X-ray Diffraction Pattern of Compound I, Chromium Kg Radiation -—— ....‘o—_ . . 10. 11. 12. 13. 1A. 15. 16. BIBLIOGRAPHY Conference on High Temperature Polymer and Fluid Research, Wright Air Development Center, Dayton, Ohio (May 26-28, 1959) J. J. Berzelius, Ann. Physik Chem. I, 3A (182A) H. H. Sisler, Inorganic Synthesis, Vol. II, pp. 205—7, McGraw- Hill Book Company, Inc., New York (19A6) W. H. Hartford and M. Darrin, Chem. Revs. 5g, 1-61 (1958) K. J. Palmer, J. Am. Chem. Soc. 69, 2360-9 (1938) g C. P. Smyth, A. J. Grossman, and S. It Ginsburg, J. Am. Chem. Soc. 92. 192—5 (19Ao) P. Walden, Z. physik. Chem. (Leipzig) Ag, 180 (1903) Solvay Process Division, Allied Chemical and Dye Corporation, Data Sheet Pd-C 118, "Chromyl Chloride" E. Moles and L. Gomez, Z. physik. Chem. (Leipzig) @9, 513-30 (1911) E. Moles and L. Gomez, Anales soc. espaH. fis. y quim. Ig, 1A2-5h (191A); C. A. 9, 318A (1915) G. Oddo and E. Serra, Gazz. chim. ital. 22, ii, 318-29 (1899) International Critical Tables, McGraw-Hill Book Company, Inc., New York (1933) E. Moles and L. Gomez, 2. physik.Chem. (Leipzig) 99, 596, 602 (1915) G. M. Schwab and s. Prakash, z. physik. Chem. (Frankfurt) 6, 387- 93 (1956) R. C. Croft, Nature 112: 725—6 (1953); R. C. Croft and R. G. Thomas, Nature I68, 32-3 (1951) A. Etard, Compt. rend. 85, 127-9 (1877) A. Engelbrecht and A. V. Grosse, J. Am. Chem. Soc. 15, 5262-h (1952) 125 18. 19. 20. 21. 22. 23. 2A. 25. 26. 27. 28. 29. 30. 31. 32. 33. 3A. 35. 36. 38. 39- A0. 126 G. D. Flesch and H. J. Svec, J. Am. Chem. Soc. 69, 3189-91 (1958) M. Schmeissen and D. Lutzow, Angew. Chem. 66, 230 (195A) M. Schmeissen, Angew. Chem. 61, A93-501 (1955) H. Zellner, Monatsh. Chem. 89, 317-29 (l9A9) H. L. Krauss, Angew. Chem. 76, 502 (1958) M. J. Udy, Chromium, Vol. I, Reinhold Publishing Corp., New York (1956) A Bystrom and K. A. Wilhelmi, Acta. Chem. Scand. A, 1131-A1 (1950) H. Brakken, z. Krist. 76, A8A (1931) V. A. Keptsik, Izvest. Akad. Nauk S.S.S.R., Ser. Fiz. 29, 219-25 (1956), C. A. 29, 117A2 (1956) D. s. Datar and s. s. Datar, Nature I58, 518-19 (19A6) F. I. Vasenin, Zhur. Obshchei Khim. 11: A50-6 (19A7), C. A. Ag, AA2 (19A8) H. H. Sisler J. D. Bush, and O. E. Accountius, J. Am. Chem. Soc. 19. 3827 (1938) H. H. Sisler, W. C. Loh Ming, E. Mettler, and F. R. Hurley, J. Am. Chem. Soc. 72, AA6-8 (1953) M. Gomberg, Chem. Ber. 32, 2397 (1902) F. H. Westheimer, Chem. Revs. A2, A19-5l (l9A9) H. Weinhaus, Chem. Ber. A7, 322-31 (191A) H. Weinhaus and w. Treibs, Chem. Ber. §6§, l6A8-53 (1923) L. F. Fieser and G. Ourisson, Bull. soc. chim. France I223, 1152—5 H. H. Zeiss and C. N. Mathews, J. Am. Chem. Soc. 76, 169A-8 (1956) H. H. Zeiss and D. A. Pease, J. Am. Chem. Soc. 76, 3182-8 (1956) M. Anbar, I. Dostrovsky, D. Samuel, and A. D. Yoffe, J. Chem. Soc. 195A, 3603 H. H. Zeiss and F. R. Zwanzig, J. Am. Chem. Soc. 12, 1733-8 (1957) E. G. Rochow, An Introduction to the Chemistry of the Silicones, second edition, John Wiley and Sons, Inc., New York (1951) A1. A2. A3. AA. A5. A6. A7. A8. 1‘9' 50. 51. 52. 53- 5A. 55- 56. 57- 58. 127 K. A. Andrianov, Organic Silicon Com ounds, English translation, 59-11239, U. S. Department of Commerce, Office of Technical Ser- vices, Washington, D. C. (1959) Gmelin's Handbuch der Anorganischen Chemie, Silicium, Teil C, Organische Silicium Verbindungen, Verlag Chemie, Weinheim/Berg- strasse, Germany (1958) J. F. Hyde, J. Am. Chem. Soc. 72, 2166-7 (1953) Tables of Interatomic Distances and the Configuration in Molecules and Ions, The Chemical Society, London (1958) L. Pauling, The Nature of the Chemical Bond, Cornell University Press, Ithaca, New York (19A8) L. H. Sommer and O. F. Bennett, J. Am. Chem. Soc. 19, 1008-9 (1957); J. Am. Chem. Soc. 6;, 251-2 (1959) L. H. Sommer and C. L. Frye, J. Am. Chem. Soc. 6;, 1013 (1959) N. V. Sidgwick, The Electronic Theory of Valency, Oxford University Press, Iondon (1927) F. G. A. Stone and D. Seyferth, J. Inorg. & Nuclear Chem. I, 112- 118 (1955) R. C. Lord, D. W. Robinson, and W. C. Schumb, J. Am. Chem. Soc. 16, 1327-32 (1956) R. F. Curl and K. s. Pitzer, J. Am. Chem. Soc. 66, 2371-3 (1958) R. 0. Sauer and D. J. Mead, J. Am. Chem. Soc. 66, 179A-7 (1956) R. West and R. IL Baney, J. Inorg. & Nuclear Chem. 1, 297-8 (1958) D. P. Craig, A. Maccoll, R. S. Nyholm, L. E. Orgel, and L. E. Sutton, J. Chem. Soc. 195A, 332-353 M. Wolfsberg and L. Helmholz, J. Chem. Phys. g9, 837-A3 (1952) C. J. Ballhausen and A. D. Liehr, J. Mbl. Spectroscopy 2, 3A2-6O (1958) H. Stammreich, D. Bassi, and o. Sala, Spectrochim. Acta Ag, A03-5 (1958) H. Stammreich, D. Bassi, O. Sala, and H. Siebert, Spectrochim. Acta 11: 192-6 (1958) 59- 60. 61. 62. 63. 6A. 65. 66. 67. 68. 69. 70. 71. 72. 73- 7A. 75- 128 L. Helmholz, H. Brennan, and M. Wolfsberg, J. Chem. Phys. 22, 853- 8 (1955) H. Stammreich, K. Kawai, and Y. Tavares, Spectrochim. Acta 1959, A38-A7 K. H. Meyer, Natural and Synthetic High Polymers, Interscience Publishers, Inc., New York (1950) H. Krebs, Angew. Chem. 76, 615-638 (1958) D. B. Sowerby and L. F. Audrieth, J. Chem. Educ. 31, 2-10, 86-91, 13A-37 (1960) J. C. Bailar and coworkers, Polymerization Through Coordination, W.A.D.C. Technical Report 57-391, Part II, wright Air Development Center, Dayton, Ohio (1958) R. I. wagner, Conference on High Temperature Polymer and Fluid Re- search, Wright Air Deve10pment Center, Dayton, Ohio (May 26-28, 1959) C. w. Theobald, U. 8. Patent No. 2,7AA,07A (May 1, 1956) V. thler, J. Polymer Sci. 22, 637-A5 (1958) K. A. Andrianov, A. A. Zhanov, N. A. Khrasheva, and V. G. Dulova, Dokaldy Akad. Nauk S.S.S.R. 113: 1050-2 (1957) K. A. Andrianov, Uspekhi Khim. 26, 905-16 (1957) (translation cour- tesy of Dow-Corning Corporation, Midland, Michigan) E. D. Hornbaker and F. Conrad, J. Org. Chem. 2A, 1858-61 (1959) R. M. Kary and K. C. Frisch, J; Am. Chem. Soc. 72, 21Ao (1957) P. E. Kbenig, Conference on High Temperature Polymers and Fluids, W.A.D.C. Technical Report 57-657, pp. 50-55, wright Air Development Center, Dayton, Ohio (January. 1958) British-Thompson-Houston 00., British Patent No. 6A3,298; C. A. i2. 7819 (1951) Dow-Egrning, Ltd., British Patent No. 69A,526, C. A. A6, 10765 195 B. I. Yakoulev and N. V. Vinogradova, Zhur. Obshchei Khim. 22, 695-6 (1956) 76. 77. 78. 79- 80. 81. 82. 83. 8A. 85. 86. 87. 88. 89. 90. 91. 92. 93- 9A. 95. 129 C. F. Gibbs, H. Tucker, G. Shkapenko, and J. C. Park, Development of_Inorganic Polymeg Systems, W.A.D.C. Technical Report 55-A53, wright Air Development Center, Dayton, Ohio (May, 1956) w. D. English and L. H. Sommer, J. Am. Chem. Soc. 71, 170 (1955) v. A. Zeitler and C. A. Brown, J..Am. Chem. Soc. 72, A616-21 (1957) B. N. Dolgov and N. F. 0rlov, Doklady Akad. Nauk S.S.S.R. 1112 617-18 (1957); c. A. 62, 8996 (1958) B. N. Dolgov and N. F. 0rlov, Izvest. Akad. Nauk S.S.S.R., Otdel. Khim. Nauk 1957, 1395-6, C. A. 62, 7139 (1958) K. A. Andrianov and N. V. Delazari, Doklady Akad. Nauk S.S.S.R. £13. 393-6 (1958) N. F. 0rlov, B. N. Dolgov, and M. G. Voronkov, Doklady Akad. Nauk S.S.S.R. Igg, 2A6-9 (1958), C. A. 63, 2076 (1959) M. Schmidt and H. Schmidbaur, Angew. Chem. 11» 220 (1959) M. Schmidt and H. Schmidbaur, Angew. Chem. 76, 70A (1958) E. w. Abel, z. Naturforsch..166,\S7y(1960)., F. E. Granchelli and G. B. walker, U. S. Patent No. 2,863,891 (December 9, 1958) V. Cutman and A. Meller, Monatsh. Chem. 2;, 519-22 (1960) K. A. Andrianov and A. A. Zhdanov, Izvest. Akad. Nauk S.S.S.R., Otdel. Khim. Nauk 1958, 779-80; C. A. 52, 19916 (1958) D. C. Bradley and I. M. Thomas, J. Chem. Soc. 1959, 3AOA-ll M. Schmidt and H. Schmidbaur, Chem. Ber. 92, 2667-9 (1959) w. Patnode and F. C. Schmidt, J. Am. Chem. Soc. 61, 2272-3 (19A5) K. A. Andrianov, A. A. Zhdanov, and E. A. Kashutina, Zhur. Priklad. Khim. 32, A63-A (1959); C. A. 63, 11079 (1959) V. wannagot, F. Brandmair, W. Liehr, and H. Niederprum, Z. anorg. u. allgem. Chem. 302, 185-98 (1960) K. A. Andrianov and V. V. Astakhin, Doklady Akad. Nauk S.S.S.R. 121. 101A-5 (1959) K. A. Andrianov and V. G. Dulova, Izvest. Akad. Nauk S.S.S.R., Otdel. Khim. Nauk 1958, 6AA-6 96. 97. 98. 99. 100. 101. 102. 103. 10A. 105. 106. 107. 108. 109. 110. 111. 112. 130 w. H. Nebergall and 0. H. Johnson, J. Am. Chem. Soc. :3, AO22-A (l9A9) C. Eaborn, J. Chem. Soc. I222, 28AO-6 S. B. Speck, J. Org. Chem. I6, 1689-1700 (1953) H. Gilman and C. G. Brannen, J. Am. Chem. Soc. 13, A6AO-AA (1951) F. S. Kipping, J. Chem. Soc. I2I2, 2125-A2 C. A. Burkhard, J. Am. Chem. Soc. 61, 2173-A (19A5) F. S. Kipping and R. Robison, J. Chem. Soc. I2IA, A8A A. Miller, G. L. Carlson, and W. B. White, Spectrochim. Acta 59) 709’16 K. A. Andrianov, N. P. Gashnikova, and E. Z. Ashovich, Izvest. Akad. Nauk S.S.S.R., Otdel. Khim. Nauk 1960, 857-62; C. A. 6A, 22008 (1960) V. A. Zeitler and C. A. Brown, J. Phys. Chem. 6I, ll7A-7 (1957) P. M. Dupree and C. Theis, private communication w. s. Tatlock and E. c1 Rochow, J. Org. Chem. I7, 1555 (1952) A. L. Smith, Spectrochim. Acta 1960, 87-105 R. A. Friedel and M. Orchin, Ultraviolet Spectra of Aromatic Com- pounds, JOhn Wiley and Sons, Inc., New York (1951) I. M. KOlthoff and D. R. May, Ind. Eng. Chem., Anal. Ed. I9, 208-9 (19A6) J. A. MCHard, P. C. Servais, and H. A. Clark, Anal. Chem. 29, 328 (19h8) V. A. Zeitler and C. A. Brown, J. Am. Chem. Soc. Z2, A618-2l (1957) ““17717171777771717171711“