First Step Towards Planning of Syntheses in Solid‐State Chemistry: Determination of Promising Structure Candidates by Global Optimization

TLDR

The study presents a method to predict the existence and structure of (meta)stable solid compounds. The method employs adjustable modules that perform global optimization via stochastic simulated annealing, local minimization with stochastic quenches and gradient methods, and phase‑space analysis using a threshold algorithm, initially evaluating ionic compounds with empirical potentials and refining with Hartree–Fock calculations. Illustrative results include several unsynthesized but potentially kinetically stable compounds, such as additional alkali‑metal nitrides, Ca₃SiBr₂, and SrTi₂O₅.

Abstract

Abstract A method is presented here that allows, in principle, the prediction of the existence and structure of (meta)stable solid compounds. It is based on a set of adjustable modules that are applied to the study of the energy function of the chemical system of interest. The main elements are a set of routines for global optimization and local minimization, as well as algorithms for the investigation of the phase space structure near local minima of the potential energy, and the analysis and characterization of the structure candidates. The current implementation focuses on ionic compounds, for which empirical potentials are used for the evaluation of the energy function in the first stage, and a Hartree–Fock algorithm for refinements. The global optimization is performed with a stochastic simulated annealing algorithm, and the local minimization employs stochastic quenches and gradient methods. The neighborhoods of the local minima are studied with the threshold algorithm. The results of this approach are illustrated with a number of examples: compounds of binary noble gases, and binary and ternary ionic compounds. These include several substances that have not been synthesized yet, but should stand a fair chance of being kinetically stable, for example further alkali metal nitrides besides Li 3 N, as well as Ca 3 SiBr 2 or SrTi 2 O 5 .