Recipient organisationCardiff UniversitySource-published name: Cardiff University
Funding£594K
PeriodApr 2025 — Apr 2028
In plain English
AI plain-English summary
Computer simulations that predict how atoms and electrons behave are too slow for complex chemical reactions, so this Fellowship is building faster, smarter models that can handle bond-breaking and bond-forming without simulating every single electron. The problem is straightforward: today’s best simulation methods, like periodic density functional theory, become prohibitively expensive when applied to large systems such as catalytic reactions. This limits researchers’ ability to design new catalysts for renewable fuels or waste-to-fuel conversion. The Fellowship tackles this by embedding high-accuracy electronic structure methods into smaller, computationally cheaper “cluster” models, then accelerating them with machine learning. If successful, the work could speed up the discovery of catalysts for photo- and electro-catalytic hydrogen generation, and for turning waste into sustainable aviation fuel. That would directly affect energy infrastructure and industrial manufacturing processes—things most people never see but rely on daily. The methods will be released in open-source software packages (FHI-aims, ChemShell) used by academic and government labs worldwide, so the impact extends beyond any single application. The project also has a fundamental science component: it aims to reveal how the structure of catalytic active sites determines reactivity and selectivity. That knowledge, while not immediately commercial, underpins future catalyst design in the same way that early quantum chemistry eventually enabled modern drug discovery.
View original technical description
The combination of computer simulation with experiment is fundamental to achieving new understanding in chemistry, and to delivering advances that can address the most pressing societal challenges. The integration of computer simulation into research across the chemical sciences has been accelerated by the accessibility of high-performance computing infrastructure and tailored software that can harness the distributed architectures. New materials and chemical processes can be predicted by models of atoms and electrons using this infrastructure, with periodic density functional theory (DFT) at the forefront of the field of applied materials simulation. However, the efficacy of these modelling paradigms is proportional to the degrees of freedom in the system, which means that big models with lots of electrons, such as when considering catalytic processes, become very expensive to simulate. To address these shortcomings, this Fellowship looks to improve the capability and accessibility of methods that can provide high-level accuracy for electronic structure simulations, necessary for bond-breaking or bond-forming reactions, with reduced degrees of freedom, which means simulations can be performed quicker. This Fellowship is delivering new multiscale modelling paradigms, and the aim of this renewal is to make these paradigms more accessible through easier to use frameworks, and to extend our capabilities by integrating new machine-learning models into the simulation workflow, with the potential for acceleration in accurately resolving aspects of the system wavefunction. The new capabilities will continue to be developed in internationally leading software packages (FHI-aims, ChemShell) with collaborative partners distributed globally in academia and government research laboratories. The Fellowship will simultaneously look to demonstrate the potential of these new methods, with aims to resolve key mechanistic aspects of the synthesis of renewable fuel in collaboration with experimental partners in academia, notably at the host institution (Cardiff Catalysis Institute, Cardiff University) and via collaborations through the UK Catalysis Hub, as well as industry (Johnson Matthey, bp). The Fellowship aims to provide new knowledge of how the catalytic active site structure defines reactivity and selectivity in processes relating to photo- and electro-catalytic H2 generation; and also to explore how the structure of support materials influences thermally driven catalytic transformation of waste to sustainable aviation fuel. Finally, the Fellowship has complementary aims to support the transition of the research team from emergent researchers to influential and authoritative research leaders who can support the development of both new research domains and the next generation of researchers. The research team will be supported in developing, practising, and reflecting on their leadership activities, so they can deliver lasting impact in their sphere of influence.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know