S. Gelin | arXiv Science

arXiv2026

Unified Mapping of Multi-Site Electrocatalytic Activity Using a Single Descriptor

We present a precise and general method to map the activity of electrocatalysts across multiple sites. Starting from a mean-field statistical mechanics model, we introduce an effective adsorption free energy descriptor that explicitly incorporates lateral adsorbate-adsorbate interactions, enabling the construction of coverage-consistent volcano relationships. Extending this approach, we show that adsorption energetics and interaction strength define a two-dimensional activity landscape that gives rise to a "volcano ridge" that captures the coupled influence of binding and interactions on catalytic performance. For multi-site systems, we demonstrate that the inherently nonlinear coupling between distinct adsorption environments leads to multi-peaked activity trends that cannot be represented by conventional single-site descriptors. To address this, we introduce a reduced descriptor mapping that projects the multidimensional activity landscape onto a single effective coordinate while preserving the underlying physics of site heterogeneity and lateral interactions. The resulting framework generalizes Sabatier-type analysis to complex alloy catalysts and provides a physically interpretable route for screening electrocatalytic materials of arbitrary compositional complexity.

cond-mat.mtrl-sci↗

arXiv2024

Cluster expansion by transfer learning for phase stability predictions

Recent progress towards universal machine-learned interatomic potentials holds considerable promise for materials discovery. Yet the accuracy of these potentials for predicting phase stability may still be limited. In contrast, cluster expansions provide accurate phase stability predictions but are computationally demanding to parameterize from first principles, especially for structures of low dimension or with a large number of components, such as interfaces or multimetal catalysts. We overcome this trade-off via transfer learning. Using Bayesian inference, we incorporate prior statistical knowledge from machine-learned and physics-based potentials, enabling us to sample the most informative configurations and to efficiently fit first-principles cluster expansions. This algorithm is tested on Pt:Ni, showing robust convergence of the mixing energies as a function of sample size with reduced statistical fluctuations.

cond-mat.mtrl-sci↗