Methods to quantify framework Lewis acidic Sn4+ sites in zeolite Beta (Sn-Beta) with four Lewis base titrants (pyridine, deuterated acetonitrile, n-propylamine, ammonia) were developed using infrared (IR) spectroscopy or temperature programmed desorption (TPD). Integrated molar extinction coefficients (E, cm μmol-1) were measured for IR bands of pyridine adsorbed to Lewis acidic Sn sites (E(1450 cm-1) = 1.42 ± 0.30) and of CD3CN adsorbed to open (E(2316 cm-1) = 1.04 ± 0.22) and closed (E(2308 cm-1) = 2.04 ± 0.43) Sn sites, and differ from analogous E values for Lewis acidic Al sites by up to 3.6×. TPD of Sn-Beta samples after saturation with NH3 or n-propylamine (NPA) and purging to remove physisorbed species also enabled quantification of Lewis acid sites, which is seldom performed despite analogous methods that use such titrants to quantify Brønsted sites in solid acids. These four Lewis bases titrated Lewis acidic Sn sites with equimolar stoichiometry and counted similar numbers of Sn sites on low-defect Sn-Beta samples synthesized in fluoride media (Sn-Beta-F) with high Sn content (Si/Sn <150). NPA binding on residual silanol defect sites, however, caused overestimation of Sn sites by TPD on Sn-Beta-F samples with low Sn content (Si/Sn >175) or high defect Sn-Beta samples prepared via post-synthetic insertion of Sn atoms into framework vacancy defects of dealuminated Beta zeolites (Sn-Beta-OH). Molar ratios of open-to-closed Sn sites varied widely (0.29-1.64) among the eight Sn-Beta samples studied here. Open Sn sites have been proposed as the dominant active sites for aqueous-phase glucose-fructose isomerization via intramolecular 1,2-hydride shift, consistent with successive poisoning of Sn-Beta samples with pyridine prior to measurement of initial glucose isomerization rates that suppressed reactivity at pyridine uptakes similar to the number of open Sn sites counted ex-situ by CD3CN. Measured first-order glucose isomerization rate constants in water (per open Sn, 373 K), which reflect free energy differences between isomerization transition states and two coordinated water molecules at Sn sites, were ∼50× higher on hydrophobic Sn-Beta-F than on hydrophilic Sn-Beta-OH zeolites. The characterization methods reported herein enable normalization of initial glucose isomerization turnover rates on Sn-Beta zeolites by their number of open Sn sites, as required prior to interpreting the catalytic consequences of structural heterogeneities introduced by differences in sample preparation or treatment history.
Bibliographical noteFunding Information:
We acknowledge the financial support provided by the Purdue Process Safety and Assurance Center (P2SAC), a 3M Non-Tenured Faculty Grant , and a Kirk Endowment Exploratory Research Recharge Grant . We thank Jason Bates for helpful discussions and comments on this manuscript and Viktor Cybulskis for developing the IR cell design. We also thank Dr. Yury Zvinevich for assistance in constructing a custom-built active site titration unit, Cindy Tavera for experimental assistance with sample preparation, Jacklyn Hall for experimental assistance with isotopic tracer and liquid NMR experiments, and John Degenstein and Dr. Mike Detwiler for assistance with parameter estimation and spectral peak deconvolution. Finally, we thank Profs. John Morgan and Julie Liu for access to the lyophilizer, and Dr. John Harwood (Purdue Interdepartmental NMR Facility) for assistance with NMR.
© 2016 Elsevier Inc. All rights reserved.
- Glucose isomerization
- Integrated molar extinction coefficient
- Lewis acid zeolite