Single Metal Atom Catalyst Supported on g-C3N4 for Formic Acid Dehydrogenation: A Combining Density Functional Theory and Machine Learning Study

The development of an efficient formic acid dehydrogenation catalyst provides a solution for hydrogen storage and transportation. In this study, we systematically explored the catalytic performance of single atom catalysts (SACs) embedded on g-C3N4 (denoted as M@g-C3N4, M = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Zr, Mo, Ru, Rh, Pd, Ag, W, Os, Ir, Pt, and Au) for the formic acid dehydrogenation by the density functional theory (DFT) method and machine leaning (ML). First, the reaction energy of formic acid dehydrogenation on M@g-C3N4 is used as an indicator for screening the metal atom with favorable thermodynamic reaction activity, and then the energy barrier of the rate determination step (RDS) is calculated and used to find out the most optimal SACs. According to the DFT calculation results, the SACs of Rh-, Pd-, and Pt@g-C3N4 exhibit good performance in both thermodynamics and kinetics for formic acid dehydrogenation. Moreover, we found that the adsorption strength of formic acid on M@g-C3N4 largely determines the reaction energy of RDS of the formic acid dehydrogenation, which show a nearly linear correlation relationship. Finally, the machine leaning further indicates that the adsorption strength of formic acid on M@g-C3N4 is determined by the combination of the two key features, the electronegativity of metal atoms and the d-band center of metal atom. These findings can help us understand the intrinsic correlation between catalytic performance and the electronic structure of metal atoms.