(169d) Computational Alchemy to Drive Searches for Catalysts through Materials Space

First principles Kohn-Sham density functional theory (KS-DFT) enables the search for novel heterogeneous catalysts, but accurate and reliable calculations demand significant amounts of computational time that restrict the scope of searches through materials space. In general, computational effort needed for screening typically scales linearly with the number of systems. Clearly, expansive searches for novel catalysts will require developments in computational methods with significantly better scaling while preserving universality and reliability compared to KS-DFT. Here, we demonstrate the potential and limitations of computational alchemy for rapid predictions of descriptors for heterogeneous catalysis. We will present data showing the performance of computational alchemy with the oxygen reduction reaction on oxides, nitrogen reduction on nitrides, and hydrogen evolution reactions on carbides. We will critically evaluate the strengths and limitations of computational alchemy as a general method for catalyst screening through materials space.