Generative Materials Design

Rather than searching existing materials databases, EMHL begins with the desired functionality. We combine generative AI, first-principles simulations and device physics to design novel perovskite compositions, charge-transport materials and molecular additives tailored for specific optoelectronic, structural and stability targets. By adopting an inverse-design approach, we move beyond identifying existing candidates to proposing entirely new materials with properties optimised for next-generation photovoltaic technologies.

Candidate materials are evaluated using atomistic and electronic-structure simulations, integrated into realistic device models and prioritised according to their predicted performance, stability and synthesisability. The most promising designs are subsequently synthesised, characterised and refined through our autonomous experimental workflows, creating a continuous feedback loop between computational design, experimental validation and physical understanding. This tightly integrated framework accelerates the discovery of new photovoltaic materials while continuously improving both our predictive models and scientific understanding.