Structure and Activity of Potential-Dependent Pt(110) Surface Phases Revealed from Machine-Learning Atomic Simulation

Ya-Hui Fang,Dan-dan Song,Hui-xia Li,Zhi-Pan Liu
DOI: https://doi.org/10.1021/acs.jpcc.1c02222
2021-01-01
The Journal of Physical Chemistry C
Abstract:Electrodes can undergo significant surface structure reconstruction under electrochemical conditions, which in turn significantly affects its electrochemical performance. This complex phenomenon raises continuous interests in both science and industry for understanding the structure-activity dynamics under electrochemical conditions. Here we report the first theoretical attempt, by combining the machine-learning-based global optimization (SSW-NN method) and modified Poisson-Boltzmann continuum solvation (CM-MPB) based on first-principles calculations, to elucidate the potential-dependent structure evolution on a stepped Pt surface and its catalytic activity, in which the cyclic voltammetry (CV) curves are simulated to compare with experiment. We selected four types of structure domains on Pt(110), namely, Type-Ia: ordered Pt(110)-(1 x 1); Type-Ib: disordering of Pt(110)-(1 x 1); Type-II: ordered Pt(110)-(1 x 2) (the missing-row reconstruction); and Type-III: reconstructed Pt(110)-(1 x 4), and reveal the potential dependence of the coverage of H atom adsorbate, the total passed charge, and the CV of these Pt(110) surface domains. In particular, the surface reconstruction from Type-Ia to -Ib occurs at +0.20 V vs NHE when the average H coverage is above 0.60 ML, which produces the key five-coordinated Pt (Pt-5c) sites. The Pt-5c sites exhibit the superior activity for hydrogen evolution reaction (HER) and are the key species responsible for the high HER activity of Pt electrode.
What problem does this paper attempt to address?