Heterogeneous catalyst design using stochastic optimization algorithms

We describe the application of two stochastic optimization algorithms to he terogeneous catalyst design. In particular, we discuss the optimal design o f a two-component catalyst for the diffusion limited A + B --> 0 and A + B- 2 --> 0 reactions in which each of the reactant s are adsorbed specifically on one of the two distinct catalytic sites. The geometric arrangement of t he catalytic sites that maximizes the catalyst activity is determined by th e use of a genetic algorithm and a simulated annealing algorithm. In the ca se of the A + B --> 0 reaction, it is found that the catalyst surface with the optimal active site distribution, that of a checkerboard, is approximat ely 25% more active than a random site distribution. A similar increase in catalytic activity is obtained for the A + B-2 --> 0 reaction. While both t he genetic and simulated annealing algorithms obtain identical optimal solu tions for a given reaction, the simulated annealing algorithm is shown to b e more efficient.