(453h) Accelerating Catalysis Screening: Advances in Dftb for Electronic Structure Prediction

Catalysis is crucial in the transition to renewable energy, yet the quest for innovative materials capable of driving carbon-neutral fuel and chemical production reactions with optimal efficiency and selectivity remains an open challenge. Theoretical methods such as density functional theory (DFT) have provided invaluable insights into the mechanisms underlying heterogeneous catalytic reactions. However, the computational demands of traditional DFT calculations, particularly for screening new materials, often confine studies to simplified systems. To address this limitation, we turned to density functional tight binding (DFTB), a method that offers electronic band structure information at a significantly reduced computational cost compared to DFT. Leveraging DFTB, one achieves a 2-3 order of magnitude acceleration in computation.

Our work focused on the identification of spin-coupling constants accounting for magnetism, together with the development of Slater-Koster (SK) parameterizations optimized on electronic band structures of metallic transition-metal (TMs) catalysts. Such parameters allow the accurate prediction of electronic descriptors such as band centers and edges, which play a key role in high-throughput screening studies on large systems otherwise not accessible to DFT. With this contributing work, we will further present applications of our approach on adsorption energy trends of O, N, H on TM catalytic surfaces, showcasing matching parity plots of both DFT and DFTB.