Samuel A. French

QUASI: A general purpose implementation of the QM/MM approach and its application to problems in catalysis

Abstract We describe the work of the European project QUASI (Quantum Simulation in Industry, project EP25047) which has sought to develop a flexible QM/MM scheme and to apply it to a range of industrial problems. A number of QM/MM approaches were implemented within the computational chemistry scripting system, ChemShell, which provides the framework for deploying a variety of independent program packages. This software was applied in several large-scale QM/MM studies which addressed the catalytic decomposition of N 2 O by Cu-containing zeolites, the methanol synthesis reaction catalysed by Cu clusters supported on ZnO surfaces, and the modelling of enzyme structure and reactivity.

Computational approaches to the determination of active site structures and reaction mechanisms in heterogeneous catalysts

We apply quantum chemical methods to the study of active site structures and reaction mechanisms in mesoporous silica and metal oxide catalysts. Our approach is based on the use of both molecular cluster and embedded cluster (QM/MM) techniques, where the active site and molecular complex are described using density functional theory (DFT) and the embedding matrix simulated by shell model potentials. We consider three case studies: alkene epoxidation over the microporous TS-1 catalyst; methanol synthesis on ZnO and Cu/ZnO and C–H bond activation over Li-doped MgO.

Point defects in ZnO

We have investigated intrinsic point defects in ZnO and extended this study to Li, Cu and Al impurity centres. Atomic and electronic structures as well as defect energies have been obtained for the main oxidation states of all defects using our embedded cluster hybrid quantum mechanical/molecular mechanical approach to the treatment of localised states in ionic solids. With these calculations we were able to explain the nature of a number of experimentally observed phenomena. We show that in zinc excess materials the energetics of zinc interstitial are very similar to those for oxygen vacancy formation. Our results also suggest assignments for a number of bands observed in photoluminescence and other spectroscopic studies of the material.

Zinc oxide: A case study in contemporary computational solid state chemistry

Computational techniques have been applied to study a broad range of chemical and physical properties of zinc oxide. Both interatomic‐potential and density functional theory methods are used to investigate structural, thermodynamic, surface, and defect properties. We survey the structures and energies of nano‐particulate zinc oxide.

Hole localization in AlO4 (0) defects in silica materials

First-principles calculations based on cluster models have been performed to investigate the ground state and the optically excited states of the [AlO(4)](0) hole in alpha-quartz and in the siliceous zeolite ZSM-5. The structure and spectroscopic properties of this defect have been studied using the recently developed Becke88-Becke95 one-parameter model for kinetics (BB1K) functional of Zhao et al., [J. Phys. Chem. A 108, 2715 (2004)]. Our results show that the BB1K method is significantly more reliable and more accurate than the standard density-functional theory (DFT) functionals at reproducing the localized spin density on one oxygen atom and the hyperfine coupling constants associated with the hole. Furthermore, we find that the BB1K results are in close agreement with experiments, and with the self-interaction-free unrestricted Hartree-Fock (UHF) and unrestricted second-order Møller-Plesset perturbation theory (UMP2) calculations. For the first time, we present results of the ground-state paramagnetic properties of the Al defect in ZSM-5. Similar to the theoretical work for defective alpha-quartz, we find that the BB1K, UHF, UHFLee-Yang-Parr, and UMP2 calculations show a localized hole on one oxygen neighboring the Al, while even the best to date thermochemically derived hybrid generalized gradient approximation density-functional, B97-2, predicts a different model where the hole is distributed over two oxygen. We have further considered the optical transitions of the [AlO(4)](0) center in alpha-quartz and ZSM-5. In both systems, our BB1K time-dependent density-functional theory (TDDFT) and configuration interaction singles (CIS) calculations predict that the most likely transition involves electron transfer from the hole-bearing oxygen to other neighboring oxygen ions. This reinforces the experimental conclusions obtained for defective alpha-quartz. Notably, the two lowest, most dominant excitation energies calculated by BB1K-TDDFT (1.99 and 3.03 eV) show excellent agreement with experiment (1.96 and 2.85 eV [B. K. Meyer, J.M. Spaeth, and J.A. Weil, J. Phys. C: Solid State Phys. 17, L31 (1987)]) clearly outperforming the CIS method and other DFT calculations available in the literature.

Hydrogen dissociation and diffusion on Ni- and Ti-doped Mg(0001) surfaces

It is well-known, both theoretically and experimentally, that alloying MgH(2) with transition elements can significantly improve the thermodynamic and kinetic properties for H(2) desorption, as well as the H(2) intake by Mg bulk. Here, we present a density functional theory investigation of hydrogen dissociation and surface diffusion over a Ni-doped surface and compare the findings to previously investigated Ti-doped Mg(0001) and pure Mg(0001) surfaces. Our results show that the energy barrier for hydrogen dissociation on the pure Mg(0001) surface is high, while it is small/null when NiTi are added to the surface as dopants. We find that the binding energy of the two H atoms near the dissociation site is high on Ti, effectively impeding diffusion away from the Ti site. By contrast, we find that on Ni, the energy barrier for diffusion is much reduced. Therefore, although both Ti and Ni promote H(2) dissociation, only Ni appears to be a good catalyst for Mg hydrogenation, allowing diffusion away from the catalytic sites. Experimental results corroborate these theoretical findings, i.e., faster hydrogenation of the Ni-doped Mg sample as opposed to the reference Mg- or Ti-doped Mg.

Computational study of the factors controlling enantioselectivity in ruthenium(II) hydrogenation catalysts

The reduction of prochiral ketones catalyzed by Ru(diphosphine)(diamine) complexes has been studied at the DFT-PBE level of theory. Calculations have been conducted on real size systems [trans-Ru(H)2(S, S-dpen)(S-xylbinap) + acetophenone], [trans-Ru(H)2(S, S-dpen)(S-tolbinap) + acetophenone] and [trans-Ru(H)2(S, S-dpen)(S-xylbinap) + cyclohexyl methyl ketone] with the aim of identifying the factors controlling the enantioselectivity in Ru(diphosphine)(diamine) catalysts. The high enantiomeric excess (99%) in the hydrogenation of acetophenone catalyzed by trans-Ru(H)2(S, S-dpen)(S-xylbinap) has been explained in terms of the existence of a stable intermediate along the reaction pathway associated with the (R)-alcohol. The formation of this intermediate is hindered with the competitive pathways, which consequently increases the activation energy for the hydrogen transfer acetophenone/(S)-phenylethanol reaction. For the [trans-Ru(H)2(S, S-dpen)(S-tolbinap) + acetophenone] system, the lower enantioselectivity (i.e. 80%) is rationalized by the smaller differences in the activation energy between the competitive pathways which differentiate between the two diastereomeric approaches of the prochiral ketone. The DFT-PBE results suggest that this reaction is driven to the (R)-product only by the process of binding the acetophenone to the active site of the trans-Ru(H) 2(S, S-dpen)(S-tolbinap) catalyst. For the hydrogenation of cyclohexyl methyl ketone catalyzed by trans-Ru(H)2(S, S-dpen)(S-xylbinap), the low performance in the enantioselective hydrogenation of the dialkyl ketone (i.e. 37%) is again explained by the small differences in the activation and binding energies which are the factors which could effectively differentiate between the two alkyl groups.

Assignment of the complex vibrational spectra of the hydrogenated ZnO polar surfaces using QM/MM embedding

Hydrogenated zinc oxide gives rise to complex vibrational spectra with many prominent features that remain unexplained. Our calculations have unambiguously shown that the presence of vacant oxygen and zinc interstitial surface sites is the only way to rationalize the observed spectra, notably the 1710 cm−1 zinc hydride stretching mode. The large number of such sites, which expose low-coordinated surface ions, are inherent at ionically reconstructed polar surfaces. The thermal stability of the sorbed hydrogen and the infrared activity of the resulting species are correlated with site coordination and coverage.

The nature of the oxidation states of gold on ZnO

The interaction between gold in the 0, i, ii and iii oxidation states and the zinc-terminated ZnO(0001) surface is studied via the QM/MM electronic embedding method using density functional theory. The surface sites considered are the vacant zinc interstitial surface site (VZISS) and the bulk-terminated island site (BTIS). We find that on the VZISS, only Au(0) and Au(i) are stable oxidation states. However, all clusters of i to iii oxidation states are stable as substitutionals for Zn2+ in the bulk terminated island site. Au(OH)(x) complexes (x= 1-3) can adsorb exothermically onto the VZISS, indicating that higher oxidation states of gold can be stabilised at this site in the presence of hydroxyl groups. CO is used as a probe molecule to study the reactivity of Au in different oxidation states in VZISS and BTIS. In all cases, we find that the strongest binding of CO is to surface Au(i). Furthermore, CO binding onto Au(0) is stronger when the gold atom is adsorbed onto the VZISS compared to CO binding onto a gas phase neutral gold atom. These results indicate that the nature of the oxidation states of Au on ZnO(0001) will depend on the type of adsorption site. The role of ZnO in Au/ZnO catalysts is not, therefore, merely to disperse gold atoms/particles, but to also modify their electronic properties.

A computational modelling study of oxygen vacancies at LaCoO3 perovskite surfaces

Atomistic computational modelling of the surface structure of the catalytically-active perovskite LaCoO(3) has been undertaken in order to develop better models of the processes involved during catalytic oxidation processes. In particular, the energetics of creating oxygen ion vacancies at the surface have been investigated for the three low index faces (100), (110) and (111). Two mechanisms for vacancy creation have been considered involving dopant Sr(2+) cations at the La(3+) site and reduction of Co(3+) to Co(2+). For both mechanisms, there is a general tendency that the smaller the cation defect separation, the lower the energy of the cluster, as would be expected from simple electrostatic considerations. In addition, there are clear indications that oxygen vacancies are more easily created at the surface than in the bulk. The results also confirm that the presence of defects strongly influences crystal morphology and surface chemistry. The importance of individual crystal surfaces in catalysis is discussed in terms of the energetics for the creation of oxygen vacancies.