Inorg. Chem. zyxwvu 1991, 30, 4403-4408 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM 4403 zyxwvutsrqpo Contribution from the Department of Chemistry, Princeton University, Princeton, New Jersey 08544- 1009, and Department of Chemical Engineering, Princeton University, Princeton, New Jersey 08544-5263 zyxwvutsrqpo Surface Cristobalite Formation by Mild Hydrothermal Treatment of Silica Gel and Its Effect on the Deposition of Tris(ally1)rhodium and Subsequent Reactivity of (silica)Rh( allyl)* H. Eric Fischer,t Steven A. King,t John B. Miller,t Jackie Y. Ying,* Jay B. Benziger,l and Jeffrey Schwartz**t Received zyxwvutsrqp April 18. 1991 Titrimetric analysis of surface silanol sites with methyllithium measures total surface -OH group content. This correlates with simple drying but does not predict organorhodium complex deposition loadings onto the oxide. It is suggested that methyllithium does not discriminate among chemically different types of silanol sites as does tris(ally1)rhodium. Furthermore, there is no correlation between simple silica gel drying procedures and the subsequent reactivity of the supported organometallic. A sequence of mild hydrothermal treatment (hydration), followed by drying of the silica, enhanced organorhodium complex deposition yields and enabled reproducibility of supported complex reactivity. Transmission and photoacoustic FTIR analysis of silicas treated by hydration-drying showed new absorptions in the isolated -OH and Si-0-Si ring deformation regions of the spectra, demonstrating surface cristobalite formation. No change in silica surface area was detected for this process. Introduction Hydroxylated oxides are common supports for organometallic complex deposition which can occur by protolytic reaction with a surface hydroxyl group. These oxides are usually pretreated to diminish batch-to-batch aliquot variation by processes ranging from simple drying in vacuo' to multistep ones? including washing and oxidation. The surface of silica contains hydrogen-bonded -OH group (Figure 1, I1 and 111) and isolated -OH ones (I).+' Assignments of -OH groups have been made by IR spectrosco- p~;'.~.' physisorbed water (IV) may also be present, hydrogen- bonded to Si-OH or Si-&Si functionalities. We note here that although titrimetric analysis of silica gel surface silanol sites with methyllithium does correlate with simple drying conditions, it is inadequate to predict organorhodium complex deposition yields on aliquots of gels variously treated before final drying. However, a sequence of mild hydrothermal treatment prior to final drying (hydration-drying) gave silica aliquots showing enhanced deposition yields and reproducible supported complex reactivity. Infrared analysis of these aliquots showed the appearance of new bands in the isolated -OH and Si-0-Si group vibration regions. Experimental Section General Procedures. Analysis was done on a Hewlett Packard 5840A GC instrument quipped with a flame ionization detector and a direct- injection system. Mass spectral identification of GC components was obtained by using a Hewlett Packard 5992B GC/MS instrument equipped with a variable-voltage ionization gun and a jet separator op- erating with I/, in. packed glass columns. All air- or moisture-sensitive materials were handled in a nitrogen-filled Vacuum Atmospheres glovebox (residual water content of less than 1.0 ppm). Rhodium tri- chloride hydrate was purchased from either Engelhardt or Johnson Matthey and was used as received. Allyl chloride (Aldrich Gold Label) was zyxwvutsrqpon pasxd through basic alumina just prior to use. Celite, used as a filter aid with moisture-sensitive complexes, was prepared by heating in a standard furnace tube at IO-' Torr, at 280 OC, for 24 h. Diethyl ether and T H F were distilled from sodium benzophenone ketyl. Other solvents were dried by distillation from calcium hydride under dry nitrogen in Teflon sleeve-sealed stills. Octane (olefin-free grade, R u b ) was predried over lump sodium, and toluene (spectral grade, Baker) was predried over KOH. Following distillation, the solvent was cannula-transferred into a flask and ca. 70% of the material was passed into a vacuum line valve equipped vessel by bulbto-bulb distillation. This step was taken to ensure grease-free, degassed solvent. Hydrogen and CO (UHP grade, MG Scientific and Scott, respec- tively) were used as received. Methane (UHP grade, Matheson). used for GC calibration of methyllithium titrations, was found by GC and GC/MS to contain water and t r a m of C2 and C, hydrocarbons. It was passed through a '/2 in. X 3 ft column packed with 12-mesh silica gel 'Department of Chemistry. :Department of Chemical Engineering. 0020-1669/91/1330-4403%02.50/0 and condensed into a liquid-nitrogenaoled trap. The condensed gas was then allowed to warm, and the boil-off was pumped into a lecture bottle which had been purged with "zero" grade helium followed by evacuation. The pressure in the lecture bottle was brought to ca. 5 atm (approxi- mately half the methane which had been condensed into the trap initially; the remaining methane in the trap was vented). The gas was assayed by GC and by GC/MS and found to be free of detectable contaminants. Butane (CP grade, Matheson or AGL), used as an internal standard for methyllithium titrations or for propene assay, had many impurities and was purified in a manner similar to that described for methane, except it was condensed into a trap cooled with a dry ice/acetone slurry and Toepler-pumped into a slurry-cooled sample bulb. The contents of the sample bulb were first frozen in liquid nitrogen. Following proper purging and pumping of the system, the bulb containing the frozen bu- tane was opened to the calibrated sample bulb which was to be filled with standard. The liquid-nitrogen bath was removed, and the storage bulb was gently warmed. When the pressure began to rise, no further heat was applied. When the desired sample pressure was achieved, the storage bulb was sealed off and returned to the liquid-nitrogen bath. The pressure was allowed to equilibrate for a few seconds and then recorded from the mercury manometer. The calibrated sample bulb was subse- quently sealed off from the system, and the butane in the system was condensed back into the storage bulb. The system and storage bulb were closed off, and the storage bulb was removed from the system. Material prepared in this manner was found to be zyx >99% n-butane by GC and GC/MS analysis. To make standard gas mixtures containing one or more components which had substantial vapor pressure at liquid-N2 temperatures (such as methane, p = 80 Torr at -196 "C), for which the final pressure of the sample was less than 800 Torr, measured gases were pumped from sample bulbs into a trap or common vessel. The mixture was then 'stirred" by circulating in a loop accomplished by Toepler- pumping for 20-60 min. Gas mixtures in which all components could be condensed at liquid-N, temperatures and for which final pressures totaled up to 2 atm were made by condensing measured samples from a calibrated sample bulb into a common vessel. Care was taken that the common vessel had a volume relative to the combined volumes of the two calibrated bulbs such that a total pressure of 2 atm was not exceeded upon thawing the common vessel. Silica Support Treatments. Mild Hydrotberml Processing (Hydra- tion). Mild hydrothermal processing of silica gel (hydration) was done in an adaptation of a Schlenk flask fitted with a ball joint, a high-vacuum stopcock, and a 7-mm O-ring connector with a small sample bulb clamped to the side arm (shoulder flask). The vessel was sealed with a ball-to-ball adapter during the hydration run. The flask was usually Schwartz, J.; Ward, M. D. J. Mol. Coral. 1980, 8, 465. Foley, H. C.; Danio, S. J.; Tau, K. D.; Chao, K. J.; Onuferko, J. H.; Dybowski, C.; Gates, B. C. J. Am. Chem. Soc. 1983,105,3074. McDonald, R. S. J. Am. Chem. SOC. 1957, 79, 850. (4) Iler, R. The ChemistryOfSilicrr; John Wiley & Sons: New York, 1979; Chapters 5 and 6. Lochmiiller, C. H.; Kersey, M. T. hngmuir 1988, 4. 572. (5) McDonald, R. S. J. Phys. Chem. 1958,62, 1168. (6) Sindorf, D. W.; Maciel, G. E. J. Phys. Chem. 1983, 87, 5516. (7) Fripiat, J. J.; Uytterhoeven, J. J. Phys. Chem. 1962, 66, 800. I , , zyxw 0 1991 American Chemical Society