Predicting the Synthesizability of Double Perovskite Halides via Interface Reaction Pathfinding

Abstract

Recent advancements in high-throughput screening and data mining have significantly expedited the discovery of new multicomponent materials, replacing the traditionally time-consuming trial-and-error methodologies. However, accurately predicting their synthesizability remains a formidable challenge, primarily due to discrepancies between theoretical predictions and experimental processes. Theoretical predictions are focused on the stability of the final crystal structure, like energy above hull and structural factors. Experimental evolution process has complex conditions: temperature, pressure, and reaction mechanics like interface reaction. This study demonstrates that incorporating reaction pathways markedly enhances the synthesizability prediction accuracy for double perovskite halides. We predict intermediates and synthetic pathways through a detailed analysis of interface reaction mechanisms and chemical reaction networks. Specifically, the formation of the A₃B′₂⁽³⁺⁾X₉ intermediate is predicted with a high driving force during the precursor’s interface reaction. Subsequently, the residual Gibbs free energy of formation necessary for the transition from the A₃B′₂⁽³⁺⁾X₉ intermediate to double perovskite halides is shown to be crucial in determining the synthesizability. This approach surpassed existing structural factor-based approaches in accuracy, enabling us to predict synthesizable double perovskite halides such as Cs₂AgYCl₆ and Cs₂KInCl₆ more effectively. These findings show the critical role of incorporating reaction mechanisms into synthesizability predictions, thereby facilitating the discovery of new multicomponent materials.