Researchers at PNNL have developed a new approach to streamline the synthesis of targeted particles of materials using data science and machine learning (ML) techniques. The study, published in the Chemical Engineering Journal, details how the researchers addressed two main issues: identifying feasible experimental conditions and foreseeing potential particle characteristics for a given set of synthetic parameters.
The ML model they developed can predict potential particle size and phase for a set of experimental conditions, helping identify promising and feasible synthesis parameters to explore. This innovative approach represents a paradigm shift for metal oxide particle synthesis and has the potential to significantly economize the time and effort expended on ad hoc iterative synthesis approaches.
By training the ML model on careful experimental characterization, the approach demonstrated remarkable accuracy in predicting iron oxide outcomes based on synthesis reaction parameters. Additionally, the search and ranking algorithm used revealed the previously overlooked importance of pressure applied during the synthesis on the resulting phase and particle size.
For more information, Juejing Liu et al’s study, “Machine learning assisted phase and size-controlled synthesis of iron oxide particles,” can be found in the Chemical Engineering Journal (2023) with the DOI: 10.1016/j.cej.2023.145216.