(116g) Calculations of Catalytic Activity of Metal Nanoclusters | AIChE

(116g) Calculations of Catalytic Activity of Metal Nanoclusters

Authors 

Garden, A. L. - Presenter, University of Iceland
Skúlason, E., University of Iceland
Jónsson, H., University of Iceland



Metal nanoclusters with up to a few thousand atoms are known to exhibit interesting and novel properties, which differ drastically from those of the bulk crystal. The catalytic activity of such clusters is often very high and strongly dependent on the cluster size. In addition to size effects, it is known that catalytic activity on metal surfaces can be dominated by steps and defects in the surface. Thus it follows that, to accurately predict the catalytic activity of a cluster, it is necessary to determine the structure of the cluster, including a full description of edges, corners and any defects.

The task of determining the lowest energy arrangement of atoms in nanoclusters is a serious challenge already for a cluster containing a hundred atoms. As the size of the system grows, the number of minima on the potential energy surface increases exponentially and a thorough sampling becomes difficult. In the current work, an optimization algorithm based on the adaptive kinetic Monte Carlo (AKMC) algorithm is employed to determine low energy configurations of metal nanoclusters [1,2]. In this algorithm, the system is advanced from one local minimum to another via first-order saddle points on the potential energy surface. An empirical EMT potential is used in the AKMC calculations but the low energy structures are then refined and catalytic activity predicted using density functional theory calculations.

The current optimization scheme has been utilized to determine and analyze the structure of gold nanoclusters near the magic sizes in the range from 55 to 561 atoms.  These clusters exhibit distorted icosahedral or decahedral structure, the implications of which will be discussed within the context of catalysis, in particular CO oxidation. 

We have also applied the same approach to copper nanoclusters. Copper is a unique catalyst in that it shows high efficiency for reduction of CO2 to hydrocarbons. The catalytic activity can be expected to increase further when in nanocluster form. The density functional theory calculations are performed to predict the catalytic activity of the various facets, edges and corners present in the most stable structures of the nanoclusters.

References:

[1] G. Henkelman and H. Jónsson, ‘Long time scale kinetic Monte Carlo simulations without lattice approximation and predefined event table’, J. Chem. Phys.,115, 9657 (2001).

[2] A. Pedersen, J-C. Berthet and H. Jónsson, ‘Simulated Annealing with Coarse Graining and Distributed Computing’, Lecture Notes in Computer Science, 34, 7134 (2012).

Topics