Photoelectrochemical water splitting on hematite: Mechanistic insights from machine-learning-interatomic-potentials MD and microkinetic modeling
Simone Piccinin a
a CNR - Istituto Officina dei Materiali, Trieste, Italy
Proceedings of MATSUS Fall 2026 Conference (MATSUSFall26)
C1 Multiscale Insights into Solid–Liquid Interfaces for Sustainable Energy Technologies
Palma, Spain, 2026 October 26th - 30th
Organizers: Marco Fontana, Elena Magnano, Silvia Nappini and Francesca Risplendi
Invited Speaker, Simone Piccinin, presentation 227
Publication date: 22nd July 2026

The oxygen evolution reaction (OER) plays a crucial role in (photo)electrochemical devices that use renewable energy to produce synthetic fuels. While the mechanism of this reaction is still debated, recent measurements on semiconducting oxides like Fe2O3, BiVO4, TiO2 and WO3 [1,2] have shown that the dependence of the rate of OER on the surface hole density is a power law, suggesting a multihole mechanism via surface hole accumulation. This is reminiscent of the mechanism for OER promoted by the oxygen evolving complex in Photosystem II and in stark contrast with metallic oxides like IrO2 [3], where the dependence is exponential. Modeling this reaction requires all-atom simulations to capture the effects of the H-bond network of the solvent on the reactants, long molecular dynamics simulations to sample accurately the solvent degrees of freedom and enhanced sampling to model activated process like O-O bond formation. In this work, focusing on hematite (Fe2O3), we have used machine learning interatomic potentials (MLIP) trained on DFT energy and forces to achieve this goal. In contrast to what was previously assumed, we find that the reaction proceeds via direct coupling of oxygen adsorbates, triggered by the oxidation of terminal oxygen sites. This process is considerably faster than competing mechanisms like the nucleophilic attack of a water molecule or hydroxide ion. We tested the accuracy of these MLIP predictions, validating the kinetics of this process with DFT calculations. Microkinetic modeling was then used to predict the overall rate of the catalytic process and to relate the photocurrent so the surface charge accumulation. We find a power-law with a third order dependence, in agreement with experiments, and we assign its origin to the facile formation of the superoxo intermediate, a step whose activation energy is weakly dependent on the surface hole coverage.

We acknowledge funding from the Italian Ministero dell’Università e della Ricerca (MUR) through the program NEST - Network 4 Energy Sustainable Transition, Spoke 4: Clean hydrogen and final uses; funding from ICSC - HighPerformance Computing, Big Data and Quantum Computing Research Centre, Spoke 7: Materials and Molecular Sciences, both by funded by the European Union - NextGenerationEU; ISCRA for awarding this project access to the LEONARDO supercomputer, owned by the EuroHPC Joint Undertaking, hosted by CINECA (Italy).

© FUNDACIO DE LA COMUNITAT VALENCIANA SCITO
We use our own and third party cookies for analysing and measuring usage of our website to improve our services. If you continue browsing, we consider accepting its use. You can check our Cookies Policy in which you will also find how to configure your web browser for the use of cookies. More info