(653c) Influence of Confining Environment Polarity and Active Site Structure on Ethanol Dehydration Catalysis By Lewis Acid Zeolites
AIChE Annual Meeting
2018
2018 AIChE Annual Meeting
Catalysis and Reaction Engineering Division
Catalysis with Microporous and Mesoporous Materials IV: Conversion of Renewables, Natural Gas, and Petroleum
Thursday, November 1, 2018 - 1:10pm to 1:30pm
Jason S. Bates and Rajamani Gounder*
Davidson School of Chemical Engineering, Purdue University, West Lafayette, IN 47907
The different reactivity of Lewis acid sites (M) in zeolite frameworks, when confined within non-polar (hydrophobic) or polar (hydrophilic) secondary environments, can arise from differences in competitive inhibition by solvents [1], solvent-mediated mechanisms [2], and extended solvent structures [3]. Framework Lewis acid centers also adopt open ((HO)-M-(OSiâ¡)3) and closed (M-(OSiâ¡)4) configurations that show different reactivity for Baeyer-Villiger oxidation [4], glucose isomerization [3], and aldol condensation [5]. Here, we interrogate the reactivity of Lewis acidic Sn centers isolated within Beta zeolites using bimolecular ethanol dehydration to diethyl ether (404 K), which is a gas-phase probe reaction sensitive to the hydrophobic character of their confining microporous voids.
Dehydration turnover rates (404 K, per Lewis acidic Sn, 0.5â35 kPa C2H5OH, 0.1â50 kPa H2O) measured on ten low-defect (Sn-Beta-F) and high-defect (Sn-Beta-OH) zeolites were described by a rate equation derived from mechanisms identified by DFT calculations [6], and simplified using microkinetic modeling to identify kinetically-relevant pathways and intermediates. Polar hydroxyl defect groups located in confining microporous environments preferentially stabilize reactive (ethanol-ethanol) and inhibitory (ethanol-water) dimeric intermediates over monomeric ethanol intermediates. As a result, equilibrium constants (404 K) for ethanol-water and ethanol-ethanol dimer formation are 3â4Ã higher on Sn-Beta-OH than on Sn-Beta-F, consistent with insights from single-component (302 K) and two-component (303 K, 403 K) ethanol and water adsorption measurements. Intrinsic dehydration rate constants (404 K) were identical among Sn-Beta-OH and Sn-Beta-F zeolites; thus, measured differences in dehydration turnover rates solely reflect differences in prevalent surface coverages of inhibitory and reactive dimeric intermediates at active Sn sites. The confinement of Lewis acidic binding sites within secondary environments of different defect density confers the ability to discriminate surface intermediates on the basis of polarity, providing a design strategy to accelerate turnover rates and suppress inhibition by water.
Sn sites with distinct spectroscopic signatures, typically associated with open (2316 cm-1) and closed (2308 cm-1) configurations, were quantified from infrared spectra of adsorbed CD3CN before and after reaction, which indicated these sites convert to structurally similar intermediates during ethanol dehydration catalysis (404 K) yet revert to their initial configurations after regenerative oxidation treatments (21% O2, 803 K). Such recovery of initial site identities indicates that two distinct Sn site configurations are associated with specific structural locations within zeolite frameworks. In-situ pyridine titrations during ethanol dehydration catalysis (404 K) indicate that 10â35% of the total Sn atoms in Sn-Beta-F and Sn-Beta-OH samples are dominant active sites, in reasonable quantitative agreement with the number of sites associated with the 2316 cm-1 CD3CN IR peak. Catalytic data and IR spectra of CD3CN on Sn sites located within Beta, MFI, BEC, CHA, and STT frameworks are used to determine the structural characteristics of Sn sites that give rise to CD3CN IR peaks at 2316 cm-1, which are dominant active sites in several classes of reactions.
[1] Conrad et al., ChemCatChem. 9 (2017) 175â182.
[2] Li et al., Catal. Sci. Technol. 4 (2014) 2241â2250.
[3] Harris et al., J. Catal. 335 (2016) 141â154.
[4] Boronat et al., J. Catal. 234 (2005) 111â118.
[5] Lewis et al., ACS Catal. (2018) 3076â3086.
[6] Bukowski et al., J. Catal. (2018) submitted.