Skip to content

Supplemental material: What drives metal surface step-bunching in graphene chemical vapor deposition?

Summary

This corpus entry stores the Supplemental Material PDF for Physical Review Letters 120, 246101 (DOI 10.1103/PhysRevLett.120.246101), which investigates Cu surface step bunching during graphene chemical vapor deposition (CVD). The SI bundles additional experiments (for example AFM measurements and tabulated growth conditions) and theory / molecular dynamics sections that quantify energetics of adatom diffusion, step interactions, and graphene-covered Cu geometries used in step-bunching simulations. The filename in papers/ reflects a Microsoft Word export artifact, but the registered bytes are the journal supplemental PDF. Scientific interpretation of why steps bunch and under which growth conditions should be taken from the main PRL letter; this page documents what the SI contains for reproducibility and retrieval.

Methods

The supplemental text details DFT (VASP) calculations using PBE with PAW potentials, a 400 eV cutoff, DFT-D2 dispersion correction, and climbing-image nudged elastic band (cNEB) barriers for selected diffusion processes, with force thresholds on the order of 1e-2 eV/Å as stated. Adatom diffusion beneath graphene is modeled with supercells cited in the SI (for example 2×2 graphene on 2×2 Cu(111) with five Cu layers in one referenced setup). Classical MD in LAMMPS uses EAM for Cu–Cu, AIREBO for graphene C–C, and a Lennard-Jones cross-term for Cu–C with parameters compared against VASP binding energies (Table S2 in the SI). Larger cells (10×10 graphene on 10×10 Cu(111) with 10 Cu layers, 1000 Cu atoms) support energy-tracking simulations of step bunching under NVT conditions at 1250 K with a Nose–Hoover thermostat, timestep 1.0 fs, and velocity-Verlet integration, as summarized in the computational details section. Periodic boundary conditions apply in-plane for the Cu(111) slab supercells described. Production MD segment lengths (cumulative ns or µs totals) appear in the SI figures tracking step energetics. Barostat / hydrostatic pressure: N/A — the excerpted NVT LAMMPS protocol does not impose bulk NPT pressure targets. The SI’s role is to show that the LJ Cu–C cross-term is not arbitrary: it is anchored to DFT binding trends for adspecies relevant to CVD-like conditions, which matters because step energetics can be dominated by subtle metal–carbon interactions under graphene coverages.

Findings

Mechanism / outcomes: This ingest is SI-only; atomistic diffusion / step-bunching kinetics are illustrated with extended LAMMPS setups, but headline mechanism claims remain in the main PRL text.

Comparisons: DFT cNEB barriers and EAM/AIREBO/LJ MD energies are compared to justify the Cu–C cross-term used in large-cell benchmarks relative to VASP references.

Sensitivity: Temperature (1250 K in the quoted NVT MD), graphene supercell size, and Cu layer count influence extracted step interaction energies.

Limitations / outlook: Quantitative conclusions about CVD window mapping belong to the letter plus any updated corrigenda; this PDF can lag figure numbering.

Corpus honesty: SI-only — cite Phys. Rev. Lett. 120, 246101 for assertions; use this slug for parameters and extended simulation traces (pdf_path).

Limitations

SI-only pages must be paired with the primary PRL text; the wiki filename does not affect DOI resolution.

Citations and evidence anchors