Computational synthesis of 2D materials grown by chemical vapor deposition
Summary¶
Two-dimensional semiconductors are promising for flexible electronics, energy technologies, and quantum-adjacent device concepts, yet industrial adoption hinges on controllable, reproducible wafer-scale growth. Chemical vapor deposition is versatile but couples fluid mechanics, heat transfer, and surface kinetics across scales; subtle changes in inlet velocity, temperature, or precursor stoichiometry can flip morphology from contiguous monolayers to islanded films. Momeni, Ji, and Chen introduce a multiscale multiphysics model that couples continuum Navier–Stokes transport in a reactor to a phase-field description of two-dimensional film evolution on the substrate. They use WSe\(_2\) as a model material to relate macroscopic knobs that experimentalists control to mesoscale quantities such as surface diffusion and effective deposition rates, then to coverage of the as-grown film.
Methods¶
Continuum and mesoscale (CVD + phase field). This is not an atomistic MD or ReaxFF study. Momeni, Ji, and Chen present a multiphysics model for 2D film growth: Navier–Stokes (plus heat and mass transport) in a hot-wall CVD reactor with stated inlet, wall, and outflow boundary conditions, coupled to a phase-field description of 2D film evolution on the substrate. The gas-phase solution supplies local supersaturation to kinetic and order-parameter equations on the surface, with coefficients and WSe\(_2\)-relevant parameters taken from the cited kinetics/continuum literature. The implementation details, mesh/discretization choices, and case-study boundary values are in the J. Mater. Res. full text (pdf_path).
MD, force-field training, and standalone DFT. N/A — not part of this publication.
Reviews / non-simulation content. N/A as a “methods review” paper: it is a feature article with model equations and a WSe\(_2\) benchmark, not a literature survey without equations.
Findings¶
Outcomes. The coupled model shows that small changes in reactor operating parameters (e.g. inlet velocity and temperature in the case studies illustrated) can substantially change WSe\(_2\) film coverage and morphology, motivating simulation-informed process windows instead of only trial-and-error tuning. Synthesis rate and morphology in this level of theory are limited by gas-phase transport and the continuum kinetics that the phase field encodes, not by atomistic RMD in this manuscript.
Corpus positioning. The authors contrast this reactor+mesoscale framework with ab initio / reactive MD: the former is affordable at furnace scales, while defect incorporation and elementary barriers require DFT or ReaxFF-class studies elsewhere.
Limitations¶
Phase-field coefficients are effective fits; absolute growth rates should be validated against in situ probes such as reflectometry or Raman thermometry. This article does not use ReaxFF or other reactive force fields.
Coupled continuum–phase-field models also inherit uncertainties from precursor chemistry and surface reaction subsets that are not resolved atomistically; sensitivity studies in the paper should be reproduced from the primary text before using the model for quantitative reactor optimization. Mesh resolution and boundary-layer treatments can shift predicted supersaturation fields near the substrate.
Relevance to group¶
Peripheral to the ReaxFF corpus but relevant to 2D materials synthesis context at Penn State collaborators (L.-Q. Chen group).
Citations and evidence anchors¶
- J. Mater. Res. DOI 10.1557/s43578-021-00384-2 — model formulation and WSe\(_2\) case study.