Navigating Phase Diagram Complexity to Design More Efficient Solid-State Synthesis Recipes

An emerging goal in the <i>ab initio </i>materials design community is to predict efficient synthesis recipes to novel functional materials. Precursor selection plays a major role in designing effective recipes, as there are many examples where a reaction can work from one set of precursors but not another. Here, we present a conceptual strategy to navigate high-dimensional convex hulls in the search of reactive precursors for more-efficient materials syntheses. The overarching strategy is to determine pairs of precursors where the reaction to a target material has large driving force and few competing phases. Using a high-throughput robotic synthesis laboratory, we design novel precursors for a diverse set of 32 quaternary oxide materials, and show that our DFT-guided precursors are substantially better at synthesizing the target oxides than naïve traditional precursors. This enables us to guide precursor selection for solid-state synthesis using information that is largely-available in high-throughput materials databases like the Materials Project.