Understanding photoelectrocatalysis across scales: Computational studies towards enlightenment and innovation

Photoelectrochemical (PEC) catalysis has attracted considerable attention in the context of sustainability, enabling opportunities for solar powered processes such as water splitting to hydrogen, CO 2 reduction to hydrocarbons (fuels), or conversion of nitrogen-based species (e.g. nitrates) to ammonia. The PEC catalytic systems considered here entail light absorption, charge generation, bulk charge-carrier transport, recombination, interfacial charge transfer and surface reactions, inherently spanning multiple length and time scales. Such processes have traditionally been developed through chemical intuition and trial-and-error experimentation, but computational methods are increasingly proving useful in guiding innovation in this field. Yet, while electronic-structure methods provide indispensable microscopic descriptors, and continuum models address device-level transport, a persistent gap remains in the handoffs through the mesoscopic regime, where spatial heterogeneity, stochasticity, and non-equilibrium driving forces collectively determine catalytic performance. In this Review, we examine modelling approaches for PEC catalysis across multiple scales. We first discuss recent advances in electronic-structure methods for describing excited states, band alignment, charge localisation, polarons and excitons. We then proceed to explore the atomistic modelling approaches for treating nonadiabatic dynamics, thermally assisted charge transfer, polaron hopping and elementary reaction mechanisms. Finally, we consider mesoscale kinetic and multiscale transport-reaction models, including kinetic Monte Carlo (kMC) and related stochastic approaches, emphasising how different choices of mesoscale state variables enable descriptions of charge transport, reaction kinetics, material evolution and reactor-scale coupling. Building on this multiscale perspective, we discuss emerging challenges specific to PEC catalysis and conclude by outlining future directions towards integrated multiscale modelling frameworks that connect parameter transfer and atomistic reaction energetics with stochastic reaction dynamics and macroscopic PEC performance.