(383at) Accelerated Nanomaterials Discovery with a Multi-Robot Self-Driving Lab

The development of next-generation semiconductors is essential to meet the growing demands for sustainable and efficient energy technologies. However, the discovery and optimization of advanced energy materials remain constrained by traditional experimental paradigms. Conventional synthesis methodologies often rely on manual, sequential exploration of single parameters, which are not only time- and resource-intensive but also prone to batch-to-batch variability and limited throughput. Furthermore, the lack of integrated, data-informed decision-making and the disjointed nature of synthesis, characterization, and experimental planning significantly hinder progress in both fundamental understanding and practical deployment of novel materials.

To overcome these limitations, we developed a multi-robot self-driving laboratory (SDL) designed to autonomously accelerate the synthesis and optimization of solution-processed semiconducting materials. The SDL integrates multiple robotic modules for reliable and reproducible material synthesis and characterization, enabling a fully automated, high-throughput experimental workflow. By miniaturizing batch reactors, the platform significantly reduces material consumption while generating scalable insights. The integration of a fine-tuned artificial intelligence (AI) algorithm facilitates efficient navigation of the complex, mixed-variable design space—rapidly identifying optimal combinations of discrete and continuous parameters that yield target material properties.

Through continuous closed-loop experimentation, the developed SDL autonomously explored undercharacterized regions of the synthesis space, revealing new structure–property relationships and reducing the time-to-solution by over 350-fold. The platform's adaptability was further demonstrated through its successful application to the retrosynthetic exploration of multiple classes of metal halide perovskite nanocrystal examples, highlighting its generalizability and potential for broader use in solution-phase materials discovery.

This work establishes a transformative approach to materials research by seamlessly combining robotics, automation, and AI within a unified SDL. The SDL offers a scalable and generalizable strategy for accelerating synthesis science and advancing the discovery of next-generation energy materials.