AI-Driven Fluidics for Autonomous Precision Synthesis of Colloidal Nanocrystals

Colloidal nanocrystals are pivotal in advancing applications across catalysis, photonics, and energy storage. Traditional synthesis methods, often labor-intensive and based on trial-and-error, limit the exploration of novel nanocrystal types. Recently, the concept of autonomous synthesis has emerged as a groundbreaking strategy for the synthesis of colloidal nanoparticles. In this talk, we will outline our method that incorporates automated synthesis through high-throughput fluidics, in-line characterization (including UV-Vis and SAXS), and machine learning (ML) algorithms to realize close-loop autonomous synthesis. Our platform facilitates rapid and efficient exploration of synthetic space, leading to the successful synthesis of targeted nanocrystals with a high degree of control over size and size distribution. We will also discuss our multimodal data analysis method using ML, which can predict high-fidelity nanoparticle properties from data collected by low-resolution, cost-effective, yet highthroughput techniques.