A graph-based network for predicting chemical reaction pathways in solid-state materials synthesis

Matthew J Mcdermott,Kristin A Persson,Shyam S Dwaraknath

doi:10.1038/s41467-021-23339-x

Matthew J Mcdermott, Kristin A Persson + Show 1 more

Open Access

https://doi.org/10.1038/s41467-021-23339-x

Copy DOI

Abstract

Accelerated inorganic synthesis remains a significant challenge in the search for novel, functional materials. Many of the principles which enable “synthesis by design” in synthetic organic chemistry do not exist in solid-state chemistry, despite the availability of extensive computed/experimental thermochemistry data. In this work, we present a chemical reaction network model for solid-state synthesis constructed from available thermochemistry data and devise a computationally tractable approach for suggesting likely reaction pathways via the application of pathfinding algorithms and linear combination of lowest-cost paths in the network. We demonstrate initial success of the network in predicting complex reaction pathways comparable to those reported in the literature for YMnO3, Y2Mn2O7, Fe2SiS4, and YBa2Cu3O6.5. The reaction network presents opportunities for enabling reaction pathway prediction, rapid iteration between experimental/theoretical results, and ultimately, control of the synthesis of solid-state materials.

Highlights

Accelerated inorganic synthesis remains a significant challenge in the search for novel, functional materials
Miura et al demonstrated the synthesis of MgCr2S4 thiospinel via a metathesis route using novel precursors, which was shown to be thermodynamically favorable through computational phase diagram construction[12]
We describe a chemical reaction network framework for predicting and suggesting solid-state inorganic reaction pathways, which when combined with experimental efforts, aims to realize inorganic synthesis by design

Summary

Introduction

Accelerated inorganic synthesis remains a significant challenge in the search for novel, functional materials. As well as reaction network models derived from the atomistic potential energy surface (PES), have been successful in predicting chemical reaction pathways in molecular systems[14,15] but are much less developed for solid-state periodic systems, where monitoring each atom’s coordinates and interactions over the large time and spatial scales necessary rapidly becomes intractable. Despite these limitations, modeling of bounded solid-state reaction mechanisms at the atomistic level has been achieved in particular with molecular dynamics (MD)[16] and kinetic Monte Carlo (KMC)-based[17] approaches. We outline the methodology used to create the chemical reaction network from thermochemistry databases and demonstrate its capacity for solid-state reaction pathway prediction by applying it to several reported experimental syntheses, as well as to recommend chemical routes to a novel battery cathode material that has not been previously synthesized

Methods

Results

Conclusion