Abstract
Oxygen evolution/reduction reaction (OER/ORR) play pivotal roles in reversible oxygen electrodes for rechargeable metal-air batteries, electrolyzers and fuel cell systems. However, the development of efficient and cost-effective electrocatalysts remains a significant challenge. In this work, the catalytic performances of 252 single-atom catalysts (SACs) supported on 2D transition metal dichalcogenides (TM@VX/Y-WXY (X, Y = S, Se, Te)) for OER/ORR are systematically investigated through density functional theory (DFT) calculations. It is demonstrated that Pt@VS-WS2 exhibits exceptional bifunctional catalytic performances, with remarkably low overpotentials of 0.37/0.31 V for OER/ORR. The sine-function relationship between the number of d-electrons (Nd) and OER/ORR overpotentials demonstrates that moderate intermediate adsorption is most favorable for the reaction process. The d-band center and charge redistribution of Pt@VS-WS2 result in optimal adsorption strength between Pt and intermediates, thus leading to high catalytic activity. Constant potential simulations reveal that Pt@VS-WS2 exhibits superior catalytic performance under acidic conditions and at higher applied potentials. The trained LightGBM model simultaneously predicts the adsorption free energies of *OH, *O and *OOH with an R2 of 0.935 on the test set. By integrating DFT calculations with interpretable machine learning approaches, this work provides novel insights for rational design of bifunctional electrocatalysts.
| Original language | English |
|---|---|
| Article number | 164976 |
| Journal | Applied Surface Science |
| Volume | 718 |
| DOIs | |
| State | Published - 15 Feb 2026 |
Keywords
- Constant potential
- Machine learning
- Oxygen evolution reaction
- Oxygen reduction reaction
- Single atom catalysts
Fingerprint
Dive into the research topics of 'Data-driven screening of single-atom catalysts supported on 2D transition metal dichalcogenides for oxygen evolution/reduction reactions (OER/ORR)'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver