Abstract
In recent years, computational methods have complemented a range of experimental techniques to enrich our understanding of chemical systems. This is particularly true in catalysis, as it is possible to gain atomic level insight into the active sites, adsorbate interactions, and the potential mechanisms occurring. Nanoparticles or nanoscale materials have become increasingly popular as catalysts in recent years, requiring additional computational considerations. In this review, we present an overview of existing and emerging computational methods in nanocatalysis and the unique catalytic behaviour that is revealed by these methods. We first discuss numerical tools used to calculate structures, energies and rates of catalytic reactions as well as different approaches to representing active sites of nanoparticle catalysts. We then focus on alloy and supported nanoparticle catalysts and explore how computational methods can reveal the more complex behaviours of these catalytic systems. Finally, we highlight some of the additional challenges in moving towards realistic systems, such as the dynamic nature of catalysts, the role of solvent, ensemble effects, the distribution of active sites and the influence of defects, which also have the potential to change the observed catalytic properties. We also present a discussion on emerging machine learning approaches as applied to nanocatalysis. Overall, this article aims to highlight the various methods available to model nanocatalysis and the catalytic insights that can be gained from a computational approach, using a range of examples from recent literature.