Melika Javanmardi

Surface passivation is critical for the stability and performance of thin film solar cells, LEDs, and other energy technologies. Halide perovskite colloids, as a model surface-limited system, provide a great test bed to study these materials both experimentally and computationally. In my research, I collect experimental observations of single-particle emission from lead halide perovskites which have been systematically passivated with varying ligands. This experimental data is used alongside a large language model framework to generate descriptors. These descriptors can be used to train an AI agent that will predict ligands that would give the optimal passivation for the perovskite surfaces. The eventual outcome of this multiagent framework will help us make materials for not only photovoltaics and LEDs, but also new quantum computing systems.