Skip to content

Simulation of titanium metal/titanium dioxide etching with chlorine and hydrogen chloride gases using the ReaxFF reactive force field

Evidence and attribution

Authority of statements

Sections below summarize the publication identified by doi, title, and pdf_path in the front matter.

Summary

Plasma-assisted etching of Ti and TiO\(_2\) with chlorine-containing gases underpins semiconductor and thin-film processing routes, but atomistic modeling requires a potential that captures Ti–Cl, Ti–O, and hydrocarbon/oxide coupling without prohibitive QM cost. This J. Phys. Chem. A article develops a ReaxFF parameterization for Ti/O/Cl/H aimed at etching simulations where Cl\(_2\) and HCl attack metal and oxide surfaces. The Ti–Cl parameters are trained against a QM dataset including dissociation curves, angular and dihedral scans, Ti + chlorine gas reaction energetics, and TiCl\(_x\) crystal heats of formation, then merged with an existing Ti–O–H description so that oxide and halide chemistry coexist in one reactive model. The abstract reports MD etching simulations comparing HCl versus Cl\(_2\) etching behavior to experiment with satisfactory agreement on relative performance in the metrics quoted. For process modeling, the contribution is a unified reactive framework that can simultaneously represent Ti metal, TiO\(_2\), and chlorine chemistry, enabling interface-limited etch pathways to emerge without ad hoc bond constraints. This is the type of parameterization that downstream plasma or ALD-adjacent studies can reuse when halogen chemistry dominates surface reactions. The article’s positioning is explicitly process-oriented: it targets etch simulations where both metal and oxide phases appear in one trajectory, avoiding stitched models that treat Ti and TiO\(_2\) with incompatible force fields. Any quantitative etch rate or barrier should be verified directly in papers/Kim_JPCA_TiO2_etching_2013.pdf, including any SI tables referenced there. Verify independently when quoting numbers.

Methods

The manuscript follows standard ReaxFF partitioning (bond/over/under/lone-pair/valence/Coulomb/vdW with EEM-style charges) and documents a successive one-parameter search strategy for Ti–Cl bonds and selected angles (Ti–Cl–Ti, Cl–Ti–Cl, and related terms) against QM references. Etching simulations place chlorine gases in contact with Ti and TiO\(_2\) slab models under protocols described in the Methods section (temperature, incident flux or gas-phase approximations, and analysis of removed atoms/etch products).

1 — MD application. Reactive molecular dynamics (RMD) for etching uses LAMMPS (or the engine named in the PDF) with PBC slab/supercell models, NVT or NPT as reported there, fs-scale timestep, ps/ns equilibration and production segments, thermostat/barostat settings, temperature (K), and optional pressure targets exactly as tabulated in papers/Kim_JPCA_TiO2_etching_2013.pdf—this wiki does not substitute missing table rows.

2 — Force-field training (when this paper fits ReaxFF parameterization):

  • Parent FF / elements: ReaxFF / bond-order reactive force field starting point as named in the article.
  • QM reference: DFT (functional, basis set, k-point mesh) as the authors report for reference data.
  • Training set: Training structures, reaction channels, and target observables (energies, barriers, EOS, etc.) enumerated in the PDF.
  • Optimization: Parameter optimization / least-squares or genetic-algorithm language as used by the authors’ fitting software.
  • Reference data / validation: QM, experiment, or benchmark sets used to validate the fitted potential.

Findings

The optimized parameter set reproduces key QM training metrics for small clusters and TiCl\(_x\) crystals in the benchmarks shown. Atomistic etching trajectories differentiate HCl versus Cl\(_2\) attack in ways that align qualitatively with experimental etch rate and selectivity trends summarized by the authors, supporting ReaxFF as a screening tool for halogen plasma chemistry on titania and titanium surfaces within the trained compositional domain. For downstream users, the important reproducibility note is to carry the same Ti/O/H baseline when merging Ti/Cl terms: mixed training sets can accidentally introduce compensating errors that look good on small clusters yet fail on slab etch fronts.

Sensitivity / outlook: Etch response depends on halogen concentration, temperature, and surface facet details in the MD setups—see Results in the PDF.

Corpus honesty: Protocol numerics belong to papers/Kim_JPCA_TiO2_etching_2013.pdf; quote tables/figures rather than this summary alone.

Limitations

Plasma chemistry includes excited states and ions not exhaustively represented by ground-state QM training; transfer to F/Br chemistries requires additional parametrization.

Relevance to group

Shows ReaxFF extension to Ti halide chemistry for semiconductor processing adjacent to oxide interface studies.

Citations and evidence anchors

  • Abstract and Secs. 1–2 (J. Phys. Chem. A 2013, 117, 5655–5663).