UC Berkeley

The development of reliable, unambiguous, and useful relationships between the properties of solid Brønsted acids (composition, acid strength, structure, site proximity, etc.) and their catalytic function is a critical step for the improvement of solid Brønsted acid catalysts. Structure-function relations for solid Brønsted acid catalysis are developed here using Brønsted acids of known structure and a broad range of acid strengths (Keggin polyoxometalate (POM) clusters) and a set of prototypical, but industrially important, Brønsted acid catalyzed reactions (alkane hydroisomerizations and alcohol dehydrations) that are accessible to detailed mechanistic studies and theoretical calculations. Keggin POM clusters exhibit well-defined atomic structures amenable to reliable theoretical estimates of deprotonation energies (DPE) as rigorous descriptors of acid strength, diverse chemical compositions that provide a wide range of acid strengths, and relatively high stability. These properties make them ideally suited for elucidating reaction mechanisms, developing structure-function relationships, and comparing experiments with theoretical estimates of chemical dynamics and thermodynamics. The identity of central atoms (X) are systematically varied in Keggin POM (H8-nXn+W12O40) clusters (X = P5+, Si4+, Al3+, or Co2+) to examine their effects on the reactivity of these clusters for hydroisomerizations reactions of C6 aliphatic alkanes and methylcyclohexane and dehydration reactions of ethanol (EtOH). Kinetic and thermodynamic constants for kinetically-relevant steps and surface intermediates in these reactions are obtained from mechanism-based interpretations of experimentally measured turnover rates or from free energies derived from density functional theory (DFT) calculations. Active sites are counted by in operando titrations with 2,6-di-tert-butyl pyridine during catalysis. Experimental constants are compared with ones from DFT-derived free energies and correlated with catalyst DPE values through structure-function relations.

Isomerization reactions and dehydration reactions occur more rapidly on stronger Brønsted acids (with lower DPE values) because the ion-pair transition states (TS) that mediate kinetically relevant steps in these reactions benefit from their more stable conjugate anions. Electrostatic interactions between the cationic moieties and POM anions at such TS offset the large energetic costs for separating charge and attenuate differences in conjugate anion stabilities on activation barriers. The amount and distribution of charge at cationic moieties determines the extent of charge separation, which dictates the fraction of the additional energy required to deprotonate weaker acids that is recovered by the TS through the electrostatic interaction of the ion-pair. Cationic moieties with either larger amounts of or more delocalized charge result in ion-pair TS with greater extents of charge separation. Reactions mediated by TS structures with either small partial charges or localized proton-like charges are thus least sensitive to acid strength, because they recover a large fraction of the ionic and covalent components of DPE. Intrinsic selectivities in alkene isomerization, ring contraction, and EtOH dehydration reactions are independent of acid strength because the TS mediating the kinetically relevant steps for forming each product in these reactions have cationic moieties with similar amounts and localization of charge, which benefit similarly from electrostatic interactions with the conjugate anion.

Although intrinsic selectivities among isomerization or ring contraction products are independent of acid strength, selective product formation may be obtained by exploiting the effects of diffusion-enhanced secondary reactions. Alkane hydroisomerizations reactions are carried out on bifunctional catalysts containing both acid and metal functions. Metal functions serve to equilibrate alkanes with their alkenes and H2; this provides a low and known concentration of alkenes at acid sites where they undergo rapid isomerization reactions, which limits the build of product alkenes due to the thermodynamic preference for alkanes at the conditions of these reactions. The size and diffusion properties of the acid domains in bifunctional catalysts, and the concomitant metal-acid site proximity, influence rates and selectivities by changing the Thiele moduli for primary and secondary isomerization reactions within acid domains. Measured product selectivities do not rigorously reflect the intrinsic formation rates of each isomer, because subsequent isomerization events of product alkenes occur at rates comparable to their diffusion out of acid domains. Such secondary reactions are more consequential for measured selectivities on stronger acids because their larger rate constants amplify the effects of acid site density on Thiele moduli.

Alcohols can eliminate water on Brønsted acid sites via the formation of monomolecular or bimolecular products (alkenes and ethers, respectively). EtOH is the simplest alcohol to probe both pathways, as it forms both ethylene (EY) and diethyl ether (DEE) on Brønsted acids at conditions relevant to the practice of dehydration catalysis. Measured product formation rate ratios were inconsistent with conventional dehydration mechanisms that propose only monomolecular elementary steps for EY formation and bimolecular steps for DEE formation, because such mechanism overlook low energy bimolecular pathways for EY formation which become relevant at the high EtOH pressures of practical applications. Experiments and theory combine to indicate that both direct and sequential (ethoxide-mediated) routes contribute to DEE formation on Brønsted acids and that EY predominately forms through sequential routes. The kinetically-relevant steps for routes leading to DEE and ethoxide formation are mediated by SN2-type substitution TS and those for EY formation are mediated by monomolecular and bimolecular syn-E2-type elimination TS; all of these TS benefit similarly from the more stable conjugate anions in stronger acids making the relative rates of their associated steps independent of acid strength.

Through these studies, it is shown how fundamental properties of solid acid catalysts, such as their acid strengths or site proximities, influence reactivities and selectivities in acid-catalyzed reactions. The former affects the stabilities of relevant intermediates and ion-pair transition states according to the amount and locale of their charge, and the latter dictates how diffusion-enhanced-secondary reactions influence measured rates and selectivities.