Data-Driven Reactivity Predictions between a Solute and Solvent for Inorganic Crystal Growth in Solution

Liquid-phase crystal growth using molten salts as solvents (fluxes) is a unique technique to design functional crystalline materials. However, guidelines for the selection of fluxes are inadequate. This is problematic because certain fluxes cause unfavorable reactions with solutes, complicating the crystal design. Although the current phase diagrams are useful for flux selection, they are limited. Herein, we propose a quantitative model to predict the reactivity between solutes and fluxes, venturing beyond the limitations of the conventional approach. By identifying the dominant factors (e.g., Dietzel’s, acidity, and basicity parameters, and some ionization energies) from structural and physical properties, we could construct the model. This model is applicable to simple oxides as a solute and MoO3 or V2O5 as a flux. Although adaptive domains of our proposed dominant factors and numerical modeling remain unclear, these achievements would give novel insights into supporting the easy, high-speed development of various crystalline materials.