Modelling metal centres, acid sites and reaction mechanisms in microporous catalysts.

Alexander J O'Malley,A J Logsdail,A A Sokol,C R A Catlow

doi:10.1039/c6fd00010j

Alexander J O'Malley, A J Logsdail + Show 2 more

Open Access

https://doi.org/10.1039/c6fd00010j

Copy DOI

Abstract

We discuss the role of QM/MM (embedded cluster) computational techniques in catalytic science, in particular their application to microporous catalysis. We describe the methodologies employed and illustrate their utility by briefly summarising work on metal centres in zeolites. We then report a detailed investigation into the behaviour of methanol at acidic sites in zeolites H-ZSM-5 and H-Y in the context of the methanol-to-hydrocarbons/olefins process. Studying key initial steps of the reaction (the adsorption and subsequent methoxylation), we probe the effect of framework topology and Brønsted acid site location on the energetics of these initial processes. We find that although methoxylation is endothermic with respect to the adsorbed system (by 17-56 kJ mol(-1) depending on the location), there are intriguing correlations between the adsorption/reaction energies and the geometries of the adsorbed species, of particular significance being the coordination of methyl hydrogens. These observations emphasise the importance of adsorbate coordination with the framework in zeolite catalysed conversions, and how this may vary with framework topology and site location, particularly suited to investigation by QM/MM techniques.

Highlights

Computational methods are very widely used in catalytic science and are increasingly powerful in obtaining an understanding of catalysis at the molecular level, where they yield models for both the structure and mechanism that assist and complement the interpretation of experimental data
The majority of contemporary computational studies in catalysis use electronic structure methods, especially Density Functional Theory (DFT) which is very widely applied, employing methods based on periodic boundary conditions (PBCs)
We present a comparison of the deprotonation, methanol adsorption, and methoxylation energies in H-ZSM-5 and H-Y zeolite frameworks, in an attempt to investigate the effect of framework topology and active site location on reactive sorbate behaviour

Summary

Introduction

Computational methods are very widely used in catalytic science and are increasingly powerful in obtaining an understanding of catalysis at the molecular level, where they yield models for both the structure and mechanism that assist and complement the interpretation of experimental data. The majority of contemporary computational studies in catalysis use electronic structure methods, especially Density Functional Theory (DFT) which is very widely applied, employing methods based on periodic boundary conditions (PBCs) Such methods have enjoyed considerable success and have become increasingly predictive. An alternative approach is to use “QM/MM” or embedded cluster methods, which have been extensively used in computational chemistry and biomolecular sciences and in modelling localised states in solids Such methods describe a nite region (which in the case of catalysis includes the active site, the reacting species and surrounding atoms) by a quantum mechanical method, while describing more distant regions of the solid by a more approximate method, o en based on interatomic potentials. They have a number of advantages which will be discussed in greater detail ; but it can be argued that they are inherently more appropriate for modelling a localised state, such as an active site on a surface or in an enzyme, as they focus the computational effort on the region containing the active site rather than the whole molecule or solid

Methods

Results

Conclusion