Autobot meets metal halide perovskites: building automated workflows for materials science

Materials Acceleration Platforms (MAPs) that integrate automated material synthesis and characterization with AI-driven data analysis provide new opportunities for materials optimization and discovery. Perovskite and perovskite-inspired metal halides are an emerging class of multidimensional materials (0D, 1D, 2D, and 3D) with widely tunable functional properties. Here, we will introduce the closed loop AutoBot platform to discuss two different workflows. In the first workflow, a multimodal data fusion approach is presented to predict synthesis–property relationships governing the optical quality of metal halide perovskite thin films across different relative humidities. In the second workflow, our approach to generate a large experimental dataset with thousands of thin film samples, spectroscopic screening to downselect samples, and synchrotron-based characterizations will be introduced.