Surface Orientation and Temperature Effects on the Interaction of Silicon with Water: Molecular Dynamics Simulations Using ReaxFF Reactive Force Field
Summary¶
Silicon wafer processing, MEMS fabrication, and chemical–mechanical polishing rely on understanding how water interacts with low-index silicon surfaces. This study applies ReaxFF molecular dynamics in LAMMPS to compare water adsorption, dissociation, and oxidation on Si(100), Si(110), and Si(111) at 300 K and 500 K, using periodic slab models with explicit water films so facet-resolved speciation can be compared at matched thermal conditions. Periodic slab models expose crystalline surfaces to water layers, sampling molecular versus dissociative adsorption, hydrogen versus hydroxyl termination, and insertion of hydroxyl oxygen into Si–Si backbonds to form Si–O–Si bridges. The work positions results as qualitatively consistent with many experimental termination trends cited in the article.
Methods¶
Molecular dynamics (reactive). ReaxFF molecular dynamics simulations study H₂O reacting with crystalline Si(100), Si(110), and Si(111) slabs at 300 K and 500 K, exposing each facet to an explicit water film so adsorption, dissociation, and oxidation can be compared on equal thermal footing. The excerpted Computational details section documents the reactive force-field energy decomposition (bond, vdW, Coulomb, hydrogen-bond terms) and cites the parent ReaxFF training literature for Si/O/H chemistry. Periodic boundary conditions are implied by the slab-on-water setup described in the article; exact supercell sizes, water column thickness, timestep (fs), thermostat choice, NVT/NPT staging, equilibration versus production duration (ps/ns), and any barostat parameters should be taken from the full J. Phys. Chem. A PDF (pdf_path) because the local normalized/extracts snippet ends early. Electric fields and metadynamics/umbrella sampling are not indicated in the indexed pages.
Force-field fitting. N/A — this work consumes a published Si/O/H ReaxFF parameterization trained to quantum chemical reference data (cited in-section); it does not report a new fit.
Static QM / DFT. N/A — trajectories are reactive MD, not AIMD.
Review scope. N/A — primary research article (duplicate galley bytes tracked separately as [[2017wen-venue-research]]).
Findings¶
Outcomes and mechanisms. Water adsorbs molecularly and dissociatively on each facet; molecular adsorption dominates on (100) and (110), whereas dissociative adsorption dominates on (111). Si(100) trends toward H-terminated regions, while (111) becomes more OH-rich. Hydroxyl oxygen can insert into Si–Si backbonds, forming Si–O–Si bridges that oxidize the surface. Raising temperature from 300 K to 500 K increases dissociation and overall oxidation extent, matching the abstract’s statement of qualitative agreement with many experiments.
Comparisons. The discussion ties simulated terminations to experimental surface spectroscopy cited in the introduction, positioning ReaxFF as a large-scale complement to static DFT studies of isolated water adducts.
Sensitivity / design levers. Facet orientation and temperature are the explicit comparative axes; both shift the balance between molecular water, dangling H, and OH coverages that control hydrophilicity and friction in MEMS/CMP contexts.
Limitations / outlook. Finite cells and nanosecond-scale trajectories limit rare-event sampling; electrolyte chemistry beyond neutral water is out of scope.
Corpus honesty. Full numerical protocol beyond the excerpted Computational details header lives in the PDF; use [[2017wen-venue-research]] only when you must cite the duplicate galley ingest explicitly.
Limitations¶
Finite cells and short timescales limit rare-event sampling; electrolyte chemistry beyond neutral water is outside the scope.
Corpus notes¶
Pair this page with [[2017wen-computationa-atomistic-mechanisms]] when building CMP-oriented reading lists: overlapping authorship and topic make them natural siblings for retrieval even though DOIs differ.
For benchmarking, note that ReaxFF water dissociation barriers on silicon are force-field dependent; if a downstream project requires quantitative sticking coefficients, compare against experiment or higher electronic structure theory on small cluster models before scaling to large slabs.
Slab thickness and water column height in the published setups influence field effects and hydrogen-bond networks; reproduce those dimensions when comparing to other silicon oxidation papers in the corpus.
Relevance to group¶
Foundational van Duin-group ReaxFF + water + silicon reference tied to manufacturing contexts.