Accelerating Efficient Adsorption Energy Calculations Using Machine Learning

Design of catalysis is important for various sustainable purposes like enhancing reaction rates in industrial process, improving the overall yield of desired products, reducing harmful emissions in chemical manufacturing. Adsorption energy plays major role in understanding strength of catalyst by analysing reaction mechanism at molecular level. In this research adsorbML algorithm is used to identify low-energy adsorbate-surface configurations more accurately and efficiently. It uses fc100 dataset for catalyst analysis. This study demonstrates the low energy adsorbate-surface configurations of Au and Ni alloys by using ASE, LFGBS, algorithms for finding force values, and show the relaxation process over time. From this analysis Ni and Au adsorption energy is 2.28 eV and 2.00 eV at 100 and 79^th step. Similarly the minimum binding energy (adsorption energy) for any adsorbate-surface combination is found using this adsorbML algorithm.