(673c) Bespoke Nanoparticle Synthesis and Chemical Knowledge Discovery Via Autonomous Experimentations

In this study, we present an innovative approach for optimizing the synthesis of metal nanoparticles (NPs) with specific optical properties. Traditional methods of exploring various synthetic variables are time-consuming and costly. Our solution involves an autonomous experimentation platform that integrates a batch synthesis module for metal NPs with a UV-Vis spectroscopy module, guided by AI optimization modeling. Using silver (Ag) NPs as a model system, we demonstrate the effectiveness of our approach, achieving precise control over the desired absorption spectra within a remarkably low number of iterations (200 iterations when optimizing among five different aqueous synthetic reagents) at room temperature. Furthermore, our analysis of synthetic variables uncovers a novel chemistry pertaining to the role of citrate in Ag NP synthesis. We find that the quantity of citrate plays a crucial role in modulating the competition between spherical and plate-shaped NPs, consequently influencing the shapes of the absorption spectra. Overall, our study showcases the platform's ability to expedite search processes and uncover new chemical insights through the analysis of data gathered from autonomous experiments.