Computer Simulations of Zeolite Formation

Abstract

Zeolites are functional nanoporous materials that are central to catalysis, separations, and several other energy-relevant technologies. Yet, despite hundreds of predicted frameworks, only a small fraction of these materials are reproducibly synthesized and used industrially. In part, this gap arises due to the complex interplay of many different phenomena that occur during nucleation, oligomerization, phase change, and crystal growth. To contextualize the progress made over the last few decades, this review explores seminal accomplishments spanning early quantum chemical and molecular dynamics studies to more recent examples using Monte Carlo methods, coarse-grained simulations, and enhanced sampling algorithms. These retrospective insights allow us to identify unresolved scientific questions, long-standing computational bottlenecks, and potential opportunities within this domain. In particular, we highlight five pressing challenges for the zeolite modeling community. These are: (1) shifting from method-driven to application-focused mindset; (2) building cost-efficient reactive force fields and/or machine-learning potentials for describing dynamic multicomponent systems; (3) expanding the use of rare-event and enhanced sampling methods; (4) integrating simulations with experiments for quantitative predictions; and (5) fostering interdisciplinary collaboration through shared data and open benchmarks. We anticipate that meeting these challenges will enable the rational, scalable synthesis of “designer” zeolites, which, ultimately, will transform our shared energy and chemicals infrastructure.

Graphical Abstract