Iron oxide particle synthesis has traditionally been a time-consuming and inefficient process, with researchers relying on intuition or trial-and-error methods. However, a new approach developed by researchers at PNNL using data science and machine learning (ML) techniques has the potential to streamline synthesis development for iron oxide particles.
The study, published in the Chemical Engineering Journal, details the researchers’ innovative approach to 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 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.
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.