(337d) Modeling Complex Reactions in Zeolites: Effects of Acid Site Location, Framework, and Reagent Structure on Methanol-to-Hydrocarbon Reactions

Hibbitts, D., University of Florida
Kravchenko, P., University of Florida
Nystrom, S. V. Jr., University of Florida
Deluca, M., University of Florida
Hoffman, A., University of Florida
Acid-catalyzed reactions within proton-form zeolites are mediated by cationic transition states which (along with their precursors) are stabilized by electrostatic interactions with the conjugate base and are ‘solvated’ by the zeolite framework through van der Waals interactions. The diversity of zeolite frameworks has sustained decades of research into determining and synthesizing optimal zeolite materials for many reactions, including the conversion of methanol into hydrocarbons (MTH). The size and topology of zeolite pores can shift selectivity during MTH from light olefins (in TON zeolites) to gasoline-range hydrocarbons (in MFI zeolites), but zeolites relevant to industrial practice for MTH processes rapidly deactivate. Acid site locations within zeolites are often poorly controlled/understood, leading to difficulties in modeling zeolite-catalyzed reactions with theoretical tools, because of the configurational complexity of such reactions. Here, we investigate reactions within methanol-to-hydrocarbons reactions and show how they are affected by the local zeolite environment (by altering acid site location within a framework or by changing the framework entirely). Ultimately, the methods outlined here will enable rigorous screening of zeolite materials to identify synthesis-targets for controlling heteroatom location and new framework synthesis.

Acid site location directly impacts molecular adsorption energies, reaction energies, and activation barriers. Here, we examine the effects of acid site location using novel methods that rigorously examine molecular adsorption and reaction pathways at all acid sites within zeolite frameworks. Methanol adsorption free energies, calculated using dispersion-corrected energies within a periodic model of MFI, for example, vary from −126 kJ mol−1 to −16 kJ mol−1 at the 48 Al–O site pairs which define acid sites within MFI. Intrinsic free energy barriers for methanol dehydration to form surface methoxy species (CH3-Z) similarly vary from 80 kJ mol−1 to 150 kJ mol−1. Varying acid site location within MFI has the same impact on intrinsic rates as changing the entire zeolite framework, indicating the importance of NMR and synthesis techniques which attempt to locate and control the location of heteroatoms within zeolite frameworks. In this work, we will present methanol and ethanol dehydration pathways within all acid sites of TON, CHA, and MFI and use these data to demonstrate that rates, selectivities, and even mechanisms can be affected by the local zeolite environment by changing the framework or acid site location.

Methanol-to-hydrocarbon reactions are also impacted by the local zeolite environment, by certain environments permitting the formation of aromatic species which constitute a “hydrocarbon pool” of species which co-catalyze further reaction cycles—but can also grow into coke precursors, leading to catalyst deactivation. This “hydrocarbon pool” consists of methylated benzenes (from benzene to hexamethylbenzene, 13 species) and each of these species can 1) by methylated by methanol, 2) undergo intramolecular C–C bond rearrangements to form alkyl substituents, and 3) dealkylate to form light alkene products. The total number of elementary steps involved is enlarged by the large set of possible co-catalysts. Here, we use structural refinement methods to examine reaction energetics all co-catalysts for each elementary step. The free energy barrier for methylating an aromatic ring, for example, increases from 80 kJ mol−1 to form toluene to 140 kJ mol−1 to form hexamethylbenzene, indicating that methylation becomes more difficult as reagents become more sterically-crowded by methyl-substituents. These effects of co-catalyst structure, along with kinetic Monte Carlo methods, allow us to report not only the mechanism of light alkene formation during MTH, but also the hydrocarbon pool species predominantly responsible for light alkene production.

These two related topics focus on computational methods and strategies for examining reactions solvated by diverse zeolite environments (within one and across multiple frameworks) and for examining the effects of reagent structure through systematically varying –H and alkyl substituents in hydrocarbon pool species which facilitate methanol-to-hydrocarbon reactions. These approaches lay the groundwork for examining the complete MTH pathways in all known and predicted zeolite structures to ultimately use computation to direct zeolite synthesis efforts by identifying promising novel materials in which the hydrocarbon pool can be formed without concomitant coke buildup and catalyst deactivation.