Decoding Complexity in Chemical Vapor Deposition Processes of Two-Dimensional Materials via Atomistic Modeling

Abstract

Chemical vapor deposition (CVD) has emerged as a scalable route for preparing high-quality, large-area two-dimensional (2D) materials. However, achieving reproducible control over film morphology, crystalline quality and yield remains challenging due to the cascade of coupled atomic-scale events. In this review, guided by scientific questions and from an atomistic-simulation perspective, we elucidate the fundamental scientific challenges and the corresponding computational strategies faced in each stage of 2D material CVD. These stages include: (i) gas phase-involved chemical reactions, (ii) adsorption of active species, (iii) surface diffusion of adatoms, (iv) nucleation of stable 2D domains, and (v) subsequent growth and coalescence of these domains. We highlight how atomic-resolution computational methods, such as density functional theory (DFT), molecular dynamics (MD), kinetic Monte Carlo (kMC), and related machine learning algorithms, can be applied to address specific mechanistic problems in each stage. These techniques are used to clarify reaction pathways, energy landscapes, and dynamic structural evolution, in light of precursor decomposition, selective adsorption, diffusion barriers, critical nucleation, preferred nucleation sites, growth modes, edge attachment kinetics, and grain boundary formation. This review aims to construct a systematic framework that links stage-specific scientific challenges to targeted computational solutions across the CVD process, thereby deepening our understanding of atomic-scale dynamics and providing theoretical guidance for the controlled synthesis of high-quality 2D materials. In addition, we discussed current limitations in bridging different spatiotemporal scales and integrating simulations with in-situ experimental observables, and outlined new opportunities for interoperable toolchains, standardized ontologies, and interpretable machine learning models to accelerate the predictive design and scalable production of next-generation 2D materials.