Autonomous Data Acquisition Pipeline for High Throughput Statistical Analysis of Catalyst Nanoparticles at Elevated Temperature

Nanoparticles underpin a significant portion of the chemicals industry, with nanoparticle catalysts serving as some of the most extensively deployed technologies at scale. Most notably their performance can be directly linked to structural and compositional properties of the catalyst. Scanning transmission electron microscopy provides a comprehensive insight into a catalyst’s microstructure, however the technique is prone to manual and laborious data acquisition. As a result, it is difficult to perform this process over a statistically significant number of particles. This challenge grows when data must be acquired under industrially realistic operating conditions, such as at elevated temperatures and/or in oxidizing or reducing conditions. With recent developments in automation in the field, we present a framework for autonomous particle sizing and spectrum imaging acquisition which incorporates machine learning, programmable mask-based scanning, in-situ stimulus control, particle size distribution, and compositional analysis. This pipeline paves the way toward studying nanoparticle catalyst structure-property relationships both at scale and under operating conditions.