This paper introduces a robotic platform designed to synthesize colloidal nanocrystals with controlled morphologies, addressing the limitations of traditional trial-and-error methods. Gold and double-perovskite nanocrystals are used as proof-of-concept materials. The platform integrates data mining of synthesis parameters from literature, automated synthesis and in situ characterization, and machine learning for inverse design. The platform successfully demonstrates controllable synthesis and morphology-oriented inverse design, opening avenues for data-driven nanocrystal synthesis.

Haitao Zhao, Wei Chen, Hao Huang, Zhehao Sun, Zijian Chen, Lingjun Wu, Baicheng Zhang, Fuming Lai, Zhuo Wang, Mukhtar Lawan Adam, Cheng Heng Pang, Paul K. Chu, Yang Lu, Tao Wu, Jun Jiang, Zongyou Yin, Xue-Feng Yu