Skip to content

Atomistic insights into H\(_2\)O\(_2\) direct synthesis of Ni–Pt nanoparticle catalysts under water solvents by reactive molecular dynamics simulations

Summary

Direct synthesis of hydrogen peroxide from H₂ and O₂ over bimetallic catalysts is attractive industrially, but selectivity hinges on suppressing over-reduction and O–O cleavage pathways while keeping H₂ activation facile. Banisalman et al. simulate Ni–Pt nanoparticles spanning 1.5–3.5 nm in explicit water at compositions Ni₉₀Pt₁₀, Ni₈₀Pt₂₀, and Ni₅₀Pt₅₀ using ReaxFF MD, augmented by DFT for key elementary energetics. The study correlates facet and site-type statistics—especially NiNiPt-fcc and Ni–Ni bridge ensembles—with H₂ versus O₂ dissociation propensities and with H₂O₂ selectivity metrics emerging from trajectories.

Methods

MD application (ReaxFF, LAMMPS). The article uses LAMMPS with the published ReaxFF for Ni–Pt / water / H\(_2\) / O\(_2\) (force-field citation in the paper). Relaxation uses conjugate-gradient minimization plus MD in a microcanonical (NVE) ensemble as written; production MD extends to 400 ps with a 1 fs time step and Langevin-type temperature control (damping 10⁻¹ and 10 fs for Ni/Pt and O/H, respectively, in the text). Long-range electrostatics: PPPM for the Coulomb term. System class: fcc Ni\(_x\)Pt\(_{1-x}\) nanoparticles 1.5–3.5 nm in diameter and compositions Ni\(_{90}\)Pt\(_{10}\), Ni\(_{80}\)Pt\(_{20}\), Ni\(_{50}\)Pt\(_{50}\), in explicit water with >10⁴–10⁵-atom ReaxFF-MD cells (per the article). 3D PBC as in their non-bulk nanoparticle setup. Barostat — N/A (no NPT barostat in the stated protocol). Hydrostatic pressure as an MD knob — N/A in the same sense. Electric field, impact, umbrella — N/A in the documented runs.

Static / constrained QM. Complementary DFT (see the article) supplies selected barriers and mechanistic checks (e.g. O\(_2\) dissociation vs Pd-type surfaces) that are difficult in large solvated NP ReaxFF runs; full functional/basis lines are in the ACS AMI file + SI.

Force-field reparameterization. N/A (literature ReaxFF).

Findings

Catalytic performance and structure. In the ReaxFF-MD metrics reported, 3.0 nm NPs show a favorable balance of activity and H\(_2\)O\(_2\) selectivity even though catalytic response is not a simple function of surface area—site counts (NiNiPt-fcc, Ni–Ni bridge ensembles) co-vary with H\(_2\) dissociation and O\(_2\) dissociation propensities. The authors highlight Ni\(_{80}\)Pt\(_{20}\) at about 3 nm as a compositional optimum** in their simulation campaign.

Mechanism. Langmuir–Hinshelwood routes remain in play, but trajectories and supporting DFT also support water-mediated H\(_2\)O\(_2\)-forming pathways argued to be more accessible on Ni–Pt than on Pd-type surfaces.

Comparisons. N/A — direct laboratory kinetics in this computational paper; authors compare against prior DFT slab literature and ReaxFF+MD use cases. ReaxFF is noted as still parametrization-sensitive for oxides/hydroxide edge cases when predicting TOF-level quantities.

Sensitivity. Performance varies with NP diameter and Ni:Pt stoichiometry; concentration and water-mediated network structure enter the interfacial selectivity picture.

Limitations

ReaxFF remains parametrization-dependent for transition-metal/water interfaces; quantitative turnover frequencies require experimental calibration. Microkinetic interpretation should keep explicit water structure—double-layer fields and surface hydroxyl populations—in view when mapping trajectories to engineering metrics. Particle faceting may evolve under reaction conditions; the study’s frozen nanoparticle morphology is a modeling convenience that experiments may violate at long times. DFT segments in the paper are used to anchor selected barriers; treat ReaxFF selectivity trends as qualitative when extrapolating to industrial pressure or electrolyte pH outside the simulated water boxes.

Relevance to group

Reactive MD + DFT for bimetallic H\(_2\)O\(_2\) catalysis in explicit solvent—parallel themes to other ReaxFF catalysis entries.

Citations and evidence anchors

  • ACS Appl. Mater. Interfaces 13, 20123–20133 (2021); DOI 10.1021/acsami.1c01947.