Liquid phase spectroscopic technique development for understanding solvent effects in acid catalyzed reactions
Date
2019
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Selecting the proper solvent is a major challenge in liquid phase catalysis, as predictive understanding of how the solvent affects reaction rates requires understanding of the reaction mechanism, possible transition state structures, and catalyst properties, such as acidity. Most experimental attempts to understand solvent effects involve screening multiple solvent choices via catalytic activity testing, which rarely leads to predictive insight of the optimal solvent for future reactions of interest. New or improved liquid phase experimental techniques are needed to provide measurements into how solvent affects substrate adsorption, transport, and phase equilibria, especially in non-ideal, porous materials where modelling techniques have difficulty capturing the complexity of the system. One such technique is attenuated total reflection infrared spectroscopy (ATR-FTIR), which is developed to quantitatively determine substrate and catalyst properties in the presence of solvent. ☐ In this thesis, probe molecule adsorption in FTIR is developed for liquid phase application. While this technique is commonly used to characterize acidity in vacuum, the effect of solvent properties on pore versus bulk-liquid phase equilibria, as well as subsequent protonation thermodynamics of zeolites are examined. Potential applications as well as limitations of probe molecule adsorption are first discussed in the context of a fundamental study in liquid water in which Brønsted acidity is detected over Na/Y, a zeolite with purely Lewis acidic properties when characterized in vacuum. Yet, despite detection of pyridine protonation in liquid water, Na/Y exhibits no catalytic activity for a Brønsted acid catalyzed reaction in aqueous solution, leading to an investigation of how probe molecule properties themselves can occasionally fail to predict behavior of reaction substrates. ☐ Subsequently, liquid phase probe molecule adsorption is applied to zeolites with intrinsic Brønsted acid sites to measure the effect of solvent on proton thermodynamics in zeolite pores. A method is developed involving the ATR-FTIR, in conjunction with a back-pressurized plug-flow reactor (PFR), for measuring temperature programmed desorption (TPD) of pyridine from Brønsted acid sites. As another technique that has only been utilized in vacuum until now, liquid pyridine TPD results provide insight into how the solvent stabilizes the proton during pyridine deprotonation. The properties of the solvent are shown to have a more significant impact on proton stability than the framework or silicon to aluminum ratio of the zeolite sample, with increasingly polar solvents leading to lower pyridine desorption temperatures over a common zeolite. In a common solvent, only minor differences are observed across four zeolite samples tested in this work. ☐ Further technique development in this work is directed toward increasingly quantitative measurements. To this end, a method is outlined to make the ATR-FTIR technique quantitative for probe molecule adsorption in the presence of solvent. The first liquid-phase FTIR extinction coefficients (ECs) for adsorbed pyridine on zeolites are shown to have little dependence on the zeolite aluminum content, but do depend on the solvent identity. The newfound quantitative spectroscopic technique is then applied to liquid-phase adsorption isotherms, to extract fundamental, quantitative, thermodynamic values pertaining to adsorption in micropores in solvent. These values include free energies, enthalpies, and entropies of transfer from the liquid to the zeolite phase (∆Gsolv, ∆Hsolv, and ∆Ssolv), pore-phase standard-state chemical potentials based on Henry’s Law ((𝜇○𝑍,𝐻), and zeolite pore-phase activity coefficients (γZ). Several liquid-zeolite phase equilibria problems are examined, including the effect of varying the pore diameter, differences between aqueous and non-aqueous solvents, using hydrophobic vs. hydrophilic zeolites, and the zeolite silicon to aluminum ratio (Si/Al). By comparing adsorption isotherms over purely siliceous and aluminum-containing zeolites of a common framework, equilibrium constants for adsorption into the pore phase (Kconf) are deconvoluted from those of proton transfer to pore-phase pyridine (Kprot), improving upon a common criticism of calorimetry measurements, in which heats of adsorption include van der Waals interactions with pore walls (Kconf), complicating acidity measurements. ☐ Finally, the techniques developed in this work are applied to a model, Brønsted acid catalyzed reaction, cyclohexanol dehydration. The solvent choice simultaneously affects multiple factors involved in the reaction including the chemical activity of cyclohexanol, the structure and stability of the solvent cluster stabilizing the pore phase proton, the pore-phase cyclohexanol concentration, the surface coverage of acid sites under reaction conditions, and the reaction order dependence on liquid phase cyclohexanol. The added complexities induced by the presence of solvent are discussed in detail, including the effects of solvent mixtures with water. For the pure solvents investigated in this work, the ranking order of their effect on proton stability exhibits an inverse correlation with reaction apparent activation energies, but not with turnover frequencies.
Description
Keywords
Characterization, Liquid phase, Probe molecule, Solvent effects, Zeolite