Skip to content

Water Interactions with Nanoporous Silica: Comparison of ReaxFF and ab Initio Molecular Dynamics Simulations

Summary

Nanoporous silica exposes strained siloxane and defect-rich surfaces to water. The study compares two published ReaxFF Si/O/H parameterizations (Yeon et al., J. Phys. Chem. C 2016 and Fogarty et al., J. Chem. Phys. 2010) against DFT-based ab initio molecular dynamics (AIMD) for the same nanoporous models, emphasizing local structure, water dissociation paths, barriers, and diffusion—including nanoconfined water inside pores.

The corpus PDF is a journal proof/galley-style file (Rimsza_JPCC-water-silica-proof.pdf); treat the version of record DOI as authoritative for pagination and final wording.

Methods

1 — MD application (ReaxFF and AIMD). Nanoporous silica models with internal surfaces rich in strained siloxane and two-membered ring (2-ring) defects are equilibrated and sampled with ReaxFF using the Yeon et al. and Fogarty et al. Si/O/H parameterizations named in the abstract, and with DFT-based AIMD on the same structural classes. Compared observables include local hydration, water dissociation paths and barriers where reported, defect populations and lifetimes, and H versus O diffusion under nanoconfinement. Full protocol tables (atom counts, PBC, ensemble, timestep, thermostats, barostats, run lengths, electrostatics) are N/A — not transcribed from this proof/galley PDF; use pdf_path and cross-check the version-of-record page [[2016rimsza-venue-jp6b07939]]. Electric fields and enhanced sampling beyond standard MD are N/A — not summarized here.

2 — Force-field training. N/A — applies published ReaxFF sets in a validation study.

3 — Static QM / DFT. AIMD provides the QM reference; detailed DFT settings are in the article/SI, N/A — not duplicated on this note.

PBC supercells host explicit water in nanoporous Si/O/H frameworks. Molecular dynamics with ReaxFF and AIMD follows [[2016rimsza-venue-jp6b07939]] for engine (LAMMPS where applicable), NVT/NPT staging, timestep (fs), equilibration/production run lengths (ps/ns), thermostat and barostat/pressure control, and temperature (K); this proof/galley PDF should not be used alone for those numbers. Electric field driving and umbrella/metadynamics/replica-exchange sampling are N/A — unless the published SI states otherwise.

Findings

Outcomes. Pathways that eliminate high-energy 2-ring defects involve two distinct intermediate geometries whose lifetimes influence hydroxylation and 2-ring removal rates. Nanoconfinement lowers water diffusion and yields nanoconfined water behavior distinct from bulk-like water. Hydrogen diffuses about 10–30% faster than oxygen, consistent with H-hopping contributions under confinement.

Comparisons. Against AIMD, the Yeon et al. ReaxFF parametrization is reported to agree more closely on mechanisms, hydroxylation rates, defect concentrations, and activation energies in this benchmark; the authors recommend that parametrization for high-defect, highly strained water–silica interfaces such as complex nanoporous motifs.

Both ReaxFF sets reproduce parts of the AIMD picture, but defect-rich pores amplify differences; confirm numbers against [[2016rimsza-venue-jp6b07939]] and the publisher PDF when citing tables.

Limitations

  • Proof PDFs can differ slightly from the final typeset article; cite and read the VOR for exact numerical tables and supplementary cross-references.
  • Transferability of the recommended FF is stated for defect-rich, strained silica motifs; bulk-like silica or different chemistries may require separate validation.

Relevance to group

Co-authored by A. C. T. van Duin; directly benchmarks ReaxFF water–silica chemistry against AIMD for nanoporous, defective silica—an important validation setting for reactive silica simulations.

Citations and evidence anchors