Zernike seminar: Professor Jacinto Sá "AI methods for producing colourful nanomaterials"

Developing scalable synthesis protocols for nanomaterials is a significant hurdle to their commercialization, prompting extensive research and development. In this lecture, I

introduce an automated machine-learning microfluidic platform to synthesize optically active nanomaterials based on target spectra derived from prior experiments, theoretical

models, or published data. The platform minimizes optimization time and the need for initial prior knowledge by employing unsupervised Bayesian optimization with

Gaussian processes. Using PTFE tubing and connectors allows for easy modification of reactor designs. This platform replaces the labour-intensive trial-and-error approach,

paving the way for standardization and large-scale synthesis and accelerating the commercialization of high-quality nanomaterials. As a proof-of-concept, protocols for

synthesizing Ag nanoplates and Prussian blue nanoparticles, along with their analogues, were optimized and validated for large-scale production.

I will conclude the seminar by presenting recent advances in the development of heterogeneous photo-electro-catalysis based on plasmonics. These advancements

leverage nanoparticles synthesized using the latest innovations from the AI-driven robot sensitizer.