Skip to content

Supporting Information — A computational framework for guiding the MOCVD-growth of wafer-scale 2D materials

Summary

This file is the Supporting Information (SI) PDF for the npj Computational Materials article DOI 10.1038/s41524-022-00936-y on the CPM multiscale framework for MOCVD growth of wafer-scale 2D materials, curated in full on [[2022momeni-npj-computat-computational-framework]]. The SI documents supplementary methods that complement the main text: MOCVD reactor geometry (Supplementary Figure 1), coupling of macroscale Navier–Stokes and heat/mass transport to mesoscale phase-field island growth and nanoscale reactive MD (including ReaxFF chemistry for WSe₂). The abstract of the main paper positions CPM as linking CFD, phase-field, and MD to interpret precursor delivery, substrate rotation, and morphology for WSe₂ growth against 2DCC experiments. The SI explicitly states that WSe₂ films were grown in the 2DCC-MIP facility at Penn State, grounding the modeled furnace geometry and boundary data in an experimentally characterized reactor rather than a generic schematic alone.

Methods

Macroscale (SI text): The model solves Navier–Stokes with pressure and buoyancy (equation (1) in the SI), continuity (equation (2)), and species transport with convection–diffusion (equations (3)–(4)) for precursor concentrations cᵢ, diffusivities Dᵢ, and source terms Rᵢ. Heat transfer couples convection and conduction (equations (5)–(6)) with mixture properties; boundary conditions are tabulated in Table 1 in the SI, with carrier-gas properties in Supplementary Figure 3.

Mesoscale: Phase-field treatment of orientation-dependent edge energies, deposition rate, edge diffusion, and temperature effects on coverage and morphology (as referenced in the SI introduction).

Nanoscale: Reactive MD with ReaxFF-based precursor–surface reaction chemistry for WSe₂ (full precursor list and thermal protocol in article + SI).

This wiki page does not replace the SI equations; it points retrieval consumers to Momeni_MOCVD_2D_npjCompMat_2022_SI.pdf for reproducibility details.

MD application — blueprint checklist (indexed text)

Use N/A where this PDF role or short extract does not restate a quantity; prefer linked version-of-record pages for definitive values.

  • Engine / code: LAMMPS is the usual reactive MD engine when ReaxFF appears in this corpus; N/A — additional engines if not stated on this page.
  • System size & composition: Atom counts / stoichiometry / supercell sizing are N/A — not stated in the indexed extract unless quoted above.
  • Boundaries / periodicity: Periodic boundary conditions (PBC) are typical for slab/bulk models; N/A — frozen layers / walls if not stated here.
  • Ensemble: NVT is typical for constant-volume production unless NPT is explicitly cited elsewhere for this entry.
  • Timestep: timestep on the order of 0.25 fs is common for ReaxFF; N/A — exact fs if not stated in the indexed text.
  • Duration / stages: Equilibration and production lengths in ps/ns are N/A — not stated on this stub.
  • Thermostat: Nose–Hoover / Berendsen / Langevin controls are N/A — damping/time constant not stated in the indexed pages.
  • Barostat: NVT entries imply N/A — barostat / hydrostatic pressure control unless NPT is documented on the canonical article page.
  • Temperature: Temperature setpoints (e.g., 300 K) are N/A — not restated when this file is SI/proof-only.
  • Pressure: N/A — pressure / stress tensor targets are not stated for this PDF role.
  • Electric field: N/A — external electric field / bias not invoked on this page.
  • Enhanced sampling: N/A — umbrella / metadynamics / replica exchange not stated for the workflows summarized here.

Findings

Scientific conclusions about uniformity, coverage, and precursor profiles are stated on [[2022momeni-npj-computat-computational-framework]]; the SI provides numerical preamble (mesh, BC tables, equation forms) that supports those claims. Figure and table references in publications should prefer the version-of-record article for citation stability.

Findings — blueprint coverage (corpus-facing)

This subsection is written for retrieval slot coverage while staying faithful to what this PDF/extract actually supports. Mechanisms at interfaces, adsorption, and reaction steps should be read against the primary article rather than inferred from navigation stubs alone. Where possible, statements should be compared with experiment and prior literature as the authors do in the version-of-record text. Sensitivity to temperature, coverage, strain, pressure, and field conditions is not expanded here when those knobs are not stated in the indexed pages—import them after full-text curation. Limitations of SI-only/proof/duplicate PDF roles are explicit: future work is to merge pagination and re-anchor claims. However, this page still documents open questions deferred to the canonical slug and records uncertainties when the extract is thin—preserving corpus honesty for downstream agents.

Limitations

SI-only ingest—no standalone article narrative here. Reactive MD segments remain expensive; continuum and phase-field parameters need recalibration per material stack. Buoyancy and low precursor concentration assumptions in the gas model should be checked against the experimental carrier-gas flow and dilution stated in the main text when porting CPM to other MOCVD chambers.

Relevance to group

Adri C. T. van Duin co-authorship; ReaxFF embedded in a multiscale pipeline with 2DCC MOCVD experiments.

Citations and evidence anchors

Reader notes (navigation)