Skip to content

Modulation Effect of Substrate Interactions on Nucleation and Growth of MoS2 on Silica

Evidence and attribution

Authority of statements

Prose below summarizes the J. Phys. Chem. C article identified by doi, title, and pdf_path. This slug uses the online / VOR-class PDF in the corpus (Nayir_MoS2_silica_JPCC_2023_online.pdf).

Summary

Large-area MoS\(_2\) growth by chemical vapor deposition on insulating SiO\(_2\) supports is sensitive to surface hydroxylation: silanol groups provide reactive anchors for oxide precursors, whereas dehydroxylated silica can be comparatively passive. This study integrates ReaxFF reactive molecular dynamics, density functional theory benchmarks, and experimental CVD trials to show how substrate hydroxylation modulates MoO\(_3\) precursor chemistry and subsequent sulfurization toward MoS\(_2\). The paper argues that hydroxylated surfaces thermodynamically and kinetically favor reactions that nucleate molybdenum oxide species on the support, enabling growth pathways that remain suppressed on more dehydroxylated silica under comparable feeds. Together, the results frame surface engineering (tuning silanol coverage) as a handle on nucleation density, effective growth temperature, and film continuity for MoS\(_2\) on oxide supports.

The computational strand is intentionally interface-first: by constructing oxide slabs with controlled silanol density, the authors isolate how precursor adsorption and reduction change when the support can participate in acid–base-like interactions and Mo–O–Si coupling, as opposed to treating SiO\(_2\) as an inert scaffold.

Methods

Force-field (block 2 — preface). The JPCC work develops a trainable ReaxFF for CVD of MoS\(_2\) from MoO\(_3\) and S\(_2\) or H\(_2\)S (and related steps); FF development and QM training set details are in the SI.

1 — MD application (ReaxFF MD with ADF/Amsterdam Modeling Suite)

  • Engine / code: ReaxFF molecular dynamics (RMD) with the ADF/ReaxFF simulation software stack cited in the article; reactive snapshots processed with OVITO and VESTA (JPCC Methods).
  • Time integration / timestep: Verlet (velocity) integration with 0.25 fs timestep; PBC in all three directions; NVT ensemble; Berendsen thermostat with 100 fs damping (as stated in J. Phys. Chem. C for this work).
  • System construction — amorphous silica and hydroxylation (Figure 1 path): 150 SiO\(_2\) “molecules” in a 50 Å cubic box, heated to 1000 K at 0.005 K/ns, then 1 ns equilibration; resulting a-SiO\(_2\) is further exposed to 200 H\(_2\) in a 50 Å box at 1000 K for 1 ns to make surface —OH and dehydroxylated (non-passivated) motifs as in the figures. N/A here for every atom count of each subfigure not copied from the same paragraph in the main text.
  • MoO\(_3\)-only and chalcogen exposure: multiple protocols at ~1000–2000 K with ~2.5 ns equilibration segments, ~30 Å boxes and 50–200-molecule gas loads of MoO\(_3\), S\(_2\), H\(_2\)S as in §2.2.1; includes heating at 0.005 K/ns to 1000 K + 1 ns equilibration for the 50 MoO\(_3\)/silica “silica only” case and two-stage MoO\(_3\)+S (or H\(_2\)S) mixed anneals per the article. N/A for every variant not recited in one paragraph—full table is in the paper and SI**.
  • Barostat / pressure / electric field / enhanced sampling: NVT-style gas-phase/oxide handling as described; N/A — no NPT or barostat called out in the excerpted Methods lines above; N/A — no E-field; N/A — no metadynamics/umbrella; ts-search for one H-bonded path uses bond restraints in MD to reach a DFT product structure (§2.2.1, H-bonded O–H / Mo–O/Si–O restraints as printed).

2 — DFT (static, benchmark)

  • Code / setup (non-exhaustive): Jaguar-based (nonperiodic) and periodic examples as in §2.3; B3LYP (and other settings) in the JPCC Methods—reproduce from the VOR PDF/SI for functional, basis, and k-sampling, because this wiki only summarizes headlines.

3 — Experiment (CVD, §2+)

  • CVD growth of MoS\(_2\) on treated fused silica and comparative substrates with O\(_2\)/Ar and H\(_2\)-anneal/oxidation steps—imaging links nucleation contrast to Ga⁺-beam-style treatments in Figure 1; N/A for a full table of T, P, flow ssccm here (see article Experimental and SI for numbers).

Corpus / KB honesty. The wiki paragraphs above are checked against the JPCC online PDF; reactor and HRTEM conditions for all panels are in the full article.

Findings

Substrate chemistry effect

Dehydroxylated silica is comparatively inert to MoO\(_3\) in emphasized conditions; hydroxylated surfaces support Mo oxide chemistry and nucleation.

Growth facilitation

OH promotes precursor anchoring and sulfurization toward MoS\(_2\); authors argue lower effective growth T vs less reactive supports.

Qualitative consistency: microscopy shows where islands form; models rationalize OH-rich MoO\(_x\) binding—barriers/coverage in primary text.

Limitations

ReaxFF for Mo–S–O–Si chemistry inherits training limitations; experiments provide qualitative rather than atom-resolved operando mapping of every intermediate. Long-timescale coalescence and grain-boundary evolution are not fully captured in atomistic runs.

Relevance to group

2DCC / Penn State collaboration linking ReaxFF to TMD CVD on oxides with Joan Redwing.

Citations and evidence anchors

Reproducibility and corpus locators

This note documents where to find primary evidence in-repo; it does not add new scientific claims beyond the cited publication.

Normalized layer. When present, normalized/papers/{slug}.json mirrors manifest hashes, bibliography fields, and extraction pointers; if pdf_path or PDF bytes change, follow AGENTS.md and docs/PHASE3_RUNBOOK.md to re-profile rather than editing PDFs in place.

Authority chain. For numerical settings (cutoffs, timesteps, ensembles, kinetics), use the peer-reviewed PDF (and publisher Supporting Information) as the authoritative source; this wiki summarizes for navigation and retrieval.