ReaxFF molecular dynamics simulation and experimental validation of chemical reactions of water and alcohols on SiC surfaces
Filename vs venue
The slug contains nat but the publication is Ceramics International (not Nature-branded).
Summary¶
Silicon carbide’s hardness and chemical inertness make wafer processing challenging; chemical mechanical polishing (CMP) is widely used, yet the molecular-level oxidation chemistry of SiC in water versus alcohol-based slurries remains disputed. Chen et al. combine ReaxFF molecular dynamics for solvent films on 6H–SiC (001) with CMP-style experiments, atomic force microscopy, and X-ray photoelectron spectroscopy. The reactive simulations follow bond making and breaking as water and alcohols approach and react with the surface. Experiments compare material removal rates for aqueous versus alcoholic polishing environments and relate morphology and surface chemistry to the simulated reactivity ordering. The introduction motivates alcohol solvents by citing prior fixed-abrasive polishing work where methanol and ethanol improved material removal rate and surface finish relative to aqueous slurries, framing an open question about how hydroxyl-bearing solvents alter SiC surface chemistry compared with water-dominated routes.
Methods¶
ReaxFF interface MD (B)¶
6H–SiC (001) with explicit water and alcohol solvents at temperatures stated in Ceram. Int.; trajectory stages follow undercoordinated Si/C approach, adsorption, bond cleavage, and oxygen incorporation consistent with the abstract’s three-stage oxidation narrative.
Experiments (CMP + characterization)¶
Removal-rate measurements under controlled slurry chemistry; AFM morphology; XPS oxide/hydroxide speciation.
Reproducibility checklist (primary text)¶
ReaxFF citation, supercell/termination, solvent coverage, T/P, timestep, duration, and analysis scripts—see Ceramics International 50, 4332–4349 and DOI 10.1016/j.ceramint.2023.11.070 (aligned with [[2023chen-nat-reaxff-molecular]] for scientific content).
Interface chemistry stages. The reactive trajectories are organized into the three-stage oxidation narrative stated in the abstract: undercoordinated Si/C attack by solvent species, Si–C bond weakening, and growth of Si–O–Si-rich oxidized networks. Alcohol solvents alter H-bond donation and alkoxy-mediated pathways versus water-only films, which the authors connect to differences in removal rate between CMP conditions.
MD application (ReaxFF, 6H–SiC (001)). LAMMPS+ReaxFF for 6H–SiC (001) slabs wetted by H\(_2\)O and alcohol films: NVT equilibration and production at temperature setpoints in Ceramics Int.; periodic PBC in the surface plane, N/A for exact lateral supercell sizes and atom counts on this page. Time step (≈0.25 fs typical of ReaxFF; N/A—exact fs in SI), Nosé–Hoover/Berendsen thermostat settings, equilibration and ns-scale (or as stated) trajectories—see the version-of-record article. N/A—NPT barostat; N/A—external E-field; N/A—metadynamics; Hydrostatic pressure N/A for NVT slab films.
Findings¶
Reactivity vs polishing rates¶
Simulated solvent–surface reactivity ordering matches experimental removal-rate ordering.
Mechanism¶
Undercoordination attack → Si–C weakening → Si–O–Si network growth via OH/H migration—supported by AFM/XPS trends.
Alcohol vs water¶
Frames why alcohol slurries can outperform water in some fixed-abrasive CMP settings (tribochemical contribution, not abrasion alone).
Limitations¶
Real CMP involves particles, fluid flow, and tribological contact pressures beyond flat-surface solvent MD. ReaxFF barriers are approximate relative to quantum methods.
Relevance to group¶
External ReaxFF application to wide-bandgap semiconductor aqueous and alcoholic processing—methodologically comparable to group oxide reactive MD studies.
Citations and evidence anchors¶
- DOI:
10.1016/j.ceramint.2023.11.070