Computational Heterogeneous Catalysis
University of Otago, Dunedin, New Zealand
Our research is purely computational. We are primarily focussed on modelling catalysis of small molecule reactions on metallic catalysts. These are either extended metal surfaces or nanoparticles. We are also interested in structural determination of nanoparticles between ~100-5000 atoms in size. Please see below for more detail on some of our current and recent research projects.
We use a variety of software packages in our research projects. Click on the logos below to explore some of the tools that we use.
Below are some of our recent research interests.
The Haber-Bosch process produces NH3 by passing gaseous H2 and atmospheric N2 over a metal catalyst at very high temperatures and pressures. Due to these harsh conditions, studying the mechanism by which ammonia forms has been a challenge. However, the mechanism needs to be understood for further optimisation of the catalyst, to improve efficiency and output.
There are two possible mechanisms; termed associative and dissociative (see figure above for a graphical representation of these). Previous research has indicated the major contribution to the rate is from the dissociative mechanism.
Recent advances in the field have renewed interest in studying both mechanisms, most importantly, the increased interest in nanoparticle catalysts, and the ability to use computational techniques to model the reaction. Nanoparticle catalysts offer dual benefits over extended metal surfaces: i) they have a high surface area to bulk ratio, and ii) they offer a high density of undercoordinated ‘active’ sites, such as corners, steps, and edges, which can contribute differently to the rate than planar sites (which dominate extended surfaces). In this research effort, we have been investigating both the associative and dissociative mechanisms on several different active sites which are potentially present on a Ru nanoparticle.
See Publication 10.
Nanoparticles make ideal catalysts as they can show drastically different properties to the bulk, due to the high surface area available for reaction and the high density of undercoordinated sites, such as steps, edges and corner sites. These undercoordinated sites are usually significantly more active than flat active sites and often dominate the catalytic activity of the entire nanoparticle.
We have recently investigated the catalytic activity towards H2 formation of two different types of edge sites that may be present in an FCC Pt nanoparticle. We found striking results: these two types of edge sites differ significantly in reactivity, by around 5 orders of magnitude at room temperature! This indicates that to accurately model the catalytic activity of nanoparticles, we must understand the likely shape of the particles and the types of undercoordinated sites present.
Predicting nanoparticle shape corresponds to a global optimisation problem. For most nanoparticles of interest (>100 atoms) this is a prohibitively difficult task. Thus, the focus in our group is to conduct broad screening of nanoparticle shapes to attempt to (i) elucidate trends in shape dominance with size and (ii) hone in on size regions that are likely to yield nanoparticles that exhibit high catalytic activity. We are currently studying Au and Pt nanoparticles in collaboration with Prof. Richard Palmer of U. Birmingham as well as Cu and Pd nanoparticles with Dr. Andreas Pedersen of ETH Zurich.
The production of NH3 by the Haber-Bosch process operates at high temperatures and pressures and requires gaseous H2 as a reactant. In contrast, the enzyme nitrogenase produces NH3 at ambient conditions, using solvated protons and electrons, rather than H2(g). Given the mild operating conditions and possibility of using renewable electricity to drive the reaction, it is an alluring prospect to develop an analogous man-made electrochemical method of synthesising NH3.
We have recently been investigating transition metal nitrides as possible electrode materials for electrochemical NH3 formation. We have identified four potential nitride candidates that satisfy the following necessary conditions: (i) They form NH3 with a low applied bias and at a reasonable rate; (ii) They are stable with respect to nitrogen leaking from the material; (iii) They are stable with respect to poisoning in an electrochemical environment; (iv) They are stable with respect to decomposition with an applied bias; (v) They are unlikley to form oxides. Our promising nitrides are currently being grown and tested for electrochemical activity.