Skip to content

Development of ReaxFF Reactive Force Field for Aqueous Iron-Sulfur Clusters with Applications to Stability and Reactivity in Water

Summary

Iron–sulfur clusters appear across biochemistry, geochemistry, and energy technologies, yet their aqueous speciation and dynamics remain difficult to model at scale. This article develops a ReaxFF parametrization for Fe\(_x\)S\(_y\) clusters coordinated to water, trained on an extensive quantum chemical reference set that improves upon the earlier Shin et al. Fe–S parameters combined with biomolecular water models referenced in the paper. The new force field matches QM reference data for gas-phase species more closely than the prior ReaxFF description. Constant-temperature reactive MD in explicit water then explores stability and reactivity, reporting dynamic, temperature-dependent behavior that aligns with prior costly ab initio molecular dynamics benchmarks cited by the authors.

Methods

Force-field training (ReaxFF for Fe–S + aqueous O/H)

Starting from Shin et al.’s Fe–S parameters (developed largely for hydrocarbon oxidation on pyrite-bearing catalysts) and a recent water model for biomolecules, the authors augment the training with new QM data on small Fe–S hydrates and cluster isomers. DFT plane-wave training for PBE + DFT-D2, k-point Γ sampling in a 20 Å cell, and further cluster benchmarks are in §2.1 and SI of the article. Optimization of ReaxFF bond and Coulomb/EEM terms follows the standard ReaxFF fitting workflow. Representative gas-phase clusters such as FeS(H₂O)₃ and larger Fe\(_x\)S\(_y\) hydrates anchor the fit before embedding the same motifs in explicit water.

MD application (reactive Fe\(_x\)S\(_y\) in H₂O)

  • Engine / code: Reactive molecular dynamics with the new ReaxFF in a classical MD engine; QEq-style Coulomb/charge updates and EEM-consistent vdW terms are used as in the published parameterization. Implementation details: pdf_path.
  • System & composition: Aqueous Fe\(_x\)S\(_y\) clusters with H₂O solvent; total atom counts fall in the \(10^4\)–\(10^6\) atom class cited in the introduction for ReaxFF-scale sampling; the exact supercell for each trajectory is in pdf_path (N/A here to paste every build).
  • Boundaries / periodicity: 3D PBC in aqueous supercells (standard for bulk water-solvated clusters); if any slab or open boundary is used, it is in pdf_path (N/A in the short in-wiki list).
  • Ensemble, timestep, duration, thermostat, barostat, pressure, temperature: the abstract and Methods describe constant-temperature RMD in explicit water, i.e. NVT-class sampling with a thermostat; NPT isotropic pressure barostat N/A unless a NPT block is written for water density; copy time step in fs, equilibration and production run ps / ns, and K-scale set temperatures from pdf_path rather than this summary. Hydrostatic stress and 1 bar-class PBC-mean pressures can appear if NPT is used.
  • Electric field, MSST, umbrella, enhanced sampling: N/A in the public abstract unless SI documents them.

Static QM (DFT) — in this same article

Details are in §2.1; see bullets under Force-field training above and pdf_path for the PBE training functional, pseudopotential settings, and convergence criteria.

Findings

The abstract reports favorable comparison to reference QM calculations on gas-phase species and significant improvement over the previous ReaxFF parametrization for the targeted Fe–S aqueous chemistry. Aqueous trajectories exhibit dynamic behavior consistent with earlier ab initio MD studies referenced in the introduction, supporting use of the new parametrization for larger-scale reactive sampling than DFT allows.

Because Fe–S clusters participate in electron transfer, mineral nucleation, and prebiotic chemistry, the authors highlight temperature-dependent structural dynamics in explicit solvent as a key output: the force field reproduces the qualitative dynamical picture from costly ab initio MD benchmarks, enabling longer cumulative sampling for rare rearrangements that DFT cannot reach.

Relevance to group

Geochemical and bioinorganic Fe–S parametrization complements other aqueous ReaxFF developments in the knowledge base.

Citations and evidence anchors