Skip to content

Diffusion-Controlled Epitaxy of Large Area Coalesced WSe2 Monolayers on Sapphire

Evidence and attribution

Authority of statements

Prose below summarizes the Nano Letters article identified by doi, title, and pdf_path.

Summary

Gas-source CVD on c-plane sapphire uses a multi-step diffusion-mediated recipe: a nucleation stage, H\(_2\)Se anneal to ripen oriented islands, then growth conditions that suppress new nucleation while laterally enlarging domains until full monolayer coalescence in <1 h. Post-growth STEM/LEED-style characterization (details in paper) shows epitaxial alignment with the substrate but antiphase boundaries where 0° and 60° domains merge. Domain ripening kinetics support 2D Ostwald-type scaling, used to estimate W-species surface diffusivity; lateral growth rate is weakly temperature-dependent over 700–900 °C, consistent with mass-transport control, while domain shape (triangular vs truncated) varies with Se/W ratio. The work positions separating nucleation, ripening, and lateral growth as a practical route to large single-crystal-like monolayers with fast coalescence, addressing persistent CVD challenges for TMD electronics.

Methods

  • CVD: W and Se precursors per gas-source configuration in §2; H\(_2\)Se exposure steps for ripening; temperature windows 700–900 °C for lateral growth.
  • Characterization: SEM/AFM (or equivalent) for domain size/density vs time; atomic-resolution imaging and diffraction for orientation and boundary structure; Raman/PL as cited to confirm monolayer character.
  • Kinetic analysis: Island-size vs time scaling interpreted via 2D ripening frameworks to extract effective surface diffusivities for W-containing species; lateral growth velocities analyzed for temperature dependence and mass-transport hypotheses.

Findings

  • Process: Separating nucleation, ripening, and lateral growth enables large single-crystal-like monolayers with controlled defect density and fast coalescence.
  • Kinetics: Island size vs anneal time follows 2D ripening, yielding diffusivity estimates for tungsten-containing surface species.

  • Growth regime: Nearly T-independent lateral velocity over the studied window implies mass-transport-limited kinetics; morphology still varies with Se:W adatom balance.

  • Limitations (as stated or implied): CVD recipes are reactor-specific; kinetic arguments are phenomenological (no atomistic MD in this paper); antiphase boundaries between 0°/60° domains may limit electronic quality even at full monolayer coverage.

  • Corpus honesty: Kinetic prefactors, SEM image statistics, and diffraction evidence should be read from papers/Others/Zhang_Choudhury_WSe2_sapphite_epitaxy_NanoLett_2018.pdf (and any publisher SI); this page is not a substitute for those tables.

Limitations

Empirical CVD windows are reactor-specific; atomistic reaction pathways are not simulated with MD here—kinetic arguments are phenomenological fits to SEM/AFM island statistics and diffraction data. Substrate miscut and precursor purity can shift nucleation outside the reported windows. Antiphase boundaries between 0° and 60° domains may degrade electronic quality even when monolayer coverage is complete—evaluate device relevance with local STEM metrics from the paper.

Relevance to group

Penn State (Redwing, Robinson, Alem, Crespi-adjacent 2DCC context) experimental WSe₂ CVD paper—useful for 2D TMD theme hubs and cross-links from computational growth reviews such as [[2020momeni-npj-computat-multiscale-computational]].

Citations and evidence anchors

  • DOI 10.1021/acs.nanolett.7b04521 — Nano Lett. 2018; papers/Others/Zhang_Choudhury_WSe2_sapphite_epitaxy_NanoLett_2018.pdf; extract normalized/extracts/2018xiaotian-zhang-nano-lett-20-diffusion-controlled-epitaxy_p1-2.txt.