Skip to content

ReaxFF study of surface chemical reactions between α-Al₂O₃ substrates and H₂O/H₂ gas-phase molecules

Second galley PDF byte variant for the Zhang et al. J. Phys. Chem. C sapphire surface chemistry article; same DOI and science as 2024zhang-venue-manuscript.

Evidence and attribution

Authority of statements

Prose follows the publication identified by doi and the primary curated page. This file exists for hash-level provenance of an alternate galley export.

Summary

An Al/O/H ReaxFF is trained to α-Al₂O₃ surface energies (A, C, R, M facets; Al- and O-terminated models), hydrolysis and hydrogen-diffusion targets on (0001), and step–terrace dehydration data, then applied in large-scale MD of H₂O and H₂ on α-Al₂O₃(0001) and related terminations with Berendsen thermostats and barostats mimicking vacuum and MOCVD-like conditions. Section 4 of the article documents NPT then NVT segments, mixed thermostats for gas versus substrate, 0.15 fs timestep, and isothermal holds over hundreds of picoseconds to nanoseconds depending on case, including H₂ protocols with optional modified H–H σ bond energy to accelerate dissociation in controlled tests.

Methods

Force-field training: The Al/O/H ReaxFF is built with successive single-parameter refinement against DFT reference data (formation energies, (0001) and stepped surface slabs, 50% Al-termination hydrolysis paths, step–terrace dehydration targets, bulk cell parameters). The parent scope and QM level follow 2024zhang-venue-manuscript; this duplicate page does not re-tabulate the training tables.

MD application: Reactive MD in LAMMPS (see primary page) studies H₂O and H₂ on α-Al₂O₃(0001) (ordered and random Al terminations, variable loadings). The article’s Methods report NPT then NVT stages, Berendsen thermostats and barostats, 0.15 fs timestep, and isothermal holds including ~300 K and ~350 K temperature settings in the spray scenarios. ps–ns production duration and equilibration are tabulated in the VOR. 3D PBC periodic cells for slabs. N/A — electric field; N/A — umbrella / metadynamics (not used). Full atom counts and thermostat damping: 2024zhang-venue-manuscript and PDF.

Findings

The model reproduces DFT surface energetics and hydrolysis and dehydration training targets within the reported figures. Water networks accelerate hydroxylation; 100% Al-terminated (0001) hydroxylates more readily at 350 K than 50% termination in the explored sprays, while full dehydroxylation is not achieved under those conditions. Random Al distributions on (0001) can be more reactive than ordered surfaces at comparable coverage. H₂ on O-rich surfaces shows enhanced dissociation when oxygen coverage is high; parameter-modified H–H tests accelerate O removal while leaving some hydroxyls after the simulated segments.

This duplicate galley hash exists for provenance only; keep the narrative in lockstep with 2024zhang-venue-manuscript so numbers and conditions do not drift between sibling PDF exports.

Limitations

Duplicate galley artifact; prefer the version-of-record PDF for citation, pagination, and final figures. Kinetic rates are classical ReaxFF estimates; σ(H–H) modification is a diagnostic, not a general gas-phase H₂ model.

Relevance to group

Group-authored ReaxFF for sapphire and 2D-growth substrate chemistry; this slug tracks a second galley hash.

Citations and evidence anchors

Reader notes (navigation)