UC Berkeley

This thesis investigates the existence and role of tetrahedral distortion of amorphous silica-supported titanol (≡Ti-OH) groups on their adsorptive and catalytic properties. Chapter 1 defines what distortion means and where it comes from and reviews the literature on distorted metal centers of silica-supported metal cations (SSMCs). We identify the significance of ≡Ti-OH groups and include a review of what has been studied specifically for ≡Ti-OH groups. We also review the literature on how SSMCs are currently modeled using computational chemistry (DFT) and identify development opportunities. Chapter 2 presents experimental and theoretical evidence of distorted ≡Ti-OH groups supported on SiO2 and examines the effect of distortion on the strength of adsorption of polar molecules. For the theoretical portion, we developed a model to represent isolated ≡Si-OH or ≡Ti-OH groups on the surface of amorphous silica. The ≡M-OH group is represented by a small cluster surrounded by a much larger cluster representing the surrounding amorphous silica. The small cluster's properties are described by high-level density functional theory (DFT) (i.e., the quantum mechanical (QM) region), whereas the large surroundings are represented by molecular mechanics (MM). We validated the QM/MM model by demonstrating that the predicted enthalpy of adsorption for seven polar molecules on ≡Si-OH groups agrees satisfactorily with experimentally measured values determined by microcalorimetry. We also found that enthalpies of adsorption on isolated ≡Si-OH groups determined from isotherms obtained by IR agree very well with the microcalorimetric values. We used IR spectroscopy to measure isotherms for pyridine adsorption to Lewis acidic Ti isolated ≡Ti-OH groups grafted to amorphous silica. The isosteric enthalpy of adsorption decreased in magnitude with increasing pyridine coverages up to a coverage of 15% and then remained relatively constant for higher coverages. Our QM/MM calculations made with our model revealed that the enthalpy of adsorption is proportional to the area of the triangular O-Ti-O facets of the ≡Ti-OH group to which pyridine is bound. ≡Ti-OH sites exhibiting facet areas consistent with those deduced from X-Ray Absorption Spectroscopy (XAS/EXAFS) measurements bind pyridine with adsorption enthalpies consistent with the observed plateau for pyridine coverage above 15%. Larger tetrahedral facet areas are required to explain the coverage-dependent adsorption enthalpies below 15% coverage, showing evidence of distorted ≡Ti-OH structures. Distorted ≡Ti-OH sites are also qualitatively consistent with reduced-intensity pre-edge X-Ray Absorption Near Edge Structure (XANES) measurements reported in this study. We carried out energy decomposition analysis (EDA) calculations to understand the underlying physical phenomenon governing the change in the enthalpies of adsorption pyridine as a function of tetrahedral facet area. Chapter 3 presents the effects of ≡Ti-OH site distortion in Ti/SiO2 on the kinetics and mechanism of gas-phase cyclohexene (C6H10) epoxidation to form cyclohexene oxide (C6H10O). We develop a steady-state microkinetic model (MKM) to investigate the effects of distortion on predicted kinetic observables (apparent activation energy (Ea) and reaction orders in the partial pressures of C6H10 and H2O2) and to establish whether a literature-proposed mechanism is consistent with observed kinetics. We discover that the predicted kinetics are inconsistent with experiments if the facet area distribution, given in Chapter 2 of this thesis, is the same facet area distribution under reaction conditions. This is due to the prediction of strong C6H10O adsorption and inhibition. We find that much smaller facets than in the as-prepared material are required for good agreement (<3.54 Å2) since C6H10O does not inhibit these facets. We find small facets are generated under reaction conditions when C6H10O adsorbs to ≡Ti-OH sites. This is because the facets opposite the one on which C6H10O adsorbs of the same ≡Ti-OH site become contracted. We find that these vacant, contracted facets are active for epoxidation. We also find that the predicted activity remains essentially constant with the level of ≡Ti-OH site distortion. This new mechanism shows quantitative agreement between experiments and our predictions for all rate parameters. In Chapter 4, we conduct a preliminary study of Isopropanol dehydration to propene and water. The mechanism occurs through an E2 (concerted) mechanism with more C-H bond-breaking character than C-O. We find that only 6% of active Ti sites are present, which was determined by combining the results of kinetic site-blocking measurements and in-situ FTIR spectroscopy. Theoretical calculations support the formation of Ti isopropoxide groups and show that these groups do not lead to appreciable propene formation rates. Further calculations reveal that sites must have tetrahedral facets larger than 4.2 Å2 for calculations to agree with experimentally observed partial pressure dependencies. Such large facets only occur for about 8% of ≡Ti-OH sites, reminiscent of the active fraction for IPA dehydration determined experimentally. This chapter shows that bulk spectroscopic techniques cannot identify the active sites for IPA dehydration since they are a minority species. Further research is needed to determine the active sites for this reaction. Chapter 5 presents a summary and conclusions of this work and offers suggestions for future investigations.