Defect-Tolerant Materials for Energy

Defects are at the heart of modern electronic materials. Our ability to precisely design and control defects has advanced computing, energy generation, catalysis, and hundreds of other applications contributing over €4 trillion to the global economy. It is therefore surprising that our capacity to a priori rationalise the presence and behaviour of defects is severely limited. While computational tools can now predict defect properties with high accuracy, they are constrained in their widescale application due to the enormous computational costs involved. This presents a major bottleneck in the design, optimisation, and adoption of next-generation electronics, impacting our ability to create new renewable energy devices for the net zero target.MATERIALISE will discover new energy materials at the forefront of the EU’s innovation strategy including solar absorbers, thermoelectrics, and transparent conductors. This will be achieved by developing and applying a novel method to capture the impact of defects in semiconductors rapidly and accurately. Based on proof-of-principle work where I demonstrated state-of-the-art machine learning approaches can accelerate and circumvent the use of costly computational defect calculations, MATERIALISE, will enable a 4-5 order-of-magnitude increase in the number of materials that can be accurately screened for energy applications. Crucially, I will identify defect design principles that can yield optimisations across the semiconductor sector more broadly.MATERIALISE will adopt an integrated approach to materials discovery, extending the boundaries of what is possible in computational design through electronic-structure calculations of bulk, surface, interfaces, transport, and synthesisability. My international network of collaborators will validate and build devices for the most promising candidates. This new approach to designing defects in materials will establish my group at the forefront of computational materials science.