Data-Driven Optimization of Au Nanoparticle Synthesis with Automated Flow Microreaction System

Abstract

We developed a flow microreaction system automated by a LabVIEW-based program enabling a control of flow rates of four pumps, an in-line UV–visible spectroscopic monitoring, and a process data logging including the pump pressures. The system was applied to the synthesis of gold nanoparticles (NPs) by a reduction of hydrogen tetrachloroaurate (III) (HAuCl4) with sodium tetrahydroborate (NaBH4) in the presence of trisodium citrate (Na3C6H5O7). Concentrations of the three reagents were controlled by changing the flow rates and were optimized so that the spectrum of the NPs gave higher peak at shorter wavelength. A gradual change in the concentrations from the dilute condition according to decision-tree algorithm as well as a careful monitoring of the pump pressures were shown to be effective for the optimization without clogging of the particles in the reactor.

Keywords

• Nanoparticle
• Data driven
• Machine learning
• Flow microreaction
• Micromixer