Reactive molecular dynamic simulations of hydrocarbon dissociations on Ni(111) surfaces
Evidence and attribution¶
Authority of statements
Prose below summarizes the publication identified by doi, title, and pdf_path.
Summary¶
ReaxFF parameters for H, C, and Ni are developed using DFT training data on hydrogen and methane reactions with Ni(111). The fitted field is used in MD to study dissociative adsorption of methane, ethane, and n-butane on Ni(111). The work reports estimates for methane sticking (zero-coverage limit) and activation energy for the first C–H cleavage, elementary CH_x step rates via surface-species monitoring and a microkinetic construction, and qualitative networks for ethane / n-butane** decomposition.
Methods¶
QM training (DFT): Geometries and energies for Ni/C/H training data use a real-space numerical atomic orbital extended DFT code (PW91 exchange-correlation, DNP basis, semi-core pseudopotential for Ni with p valence, spin polarization). Convergence criteria and Monkhorst-Pack sampling are given in Section 2.2; surface reactions use a five-layer p(3x3) Ni(111) slab with the bottom two layers fixed, 2x2x1 k-points, and LST/QST transition-state searches with frequency confirmation.
ReaxFF development: Ni-H, Ni-C, and related valence terms are fit to the DFT set (parameters in SI tables); the article compares ReaxFF to DFT for Ni EOS and key adsorption geometries.
Reactive MD (GRASP): NVT simulations of methane, ethane, and n-butane on Ni(111) use a six-layer rectangular slab (720 Ni atoms in the main setup; a four-times larger 2880-atom slab for some methane runs), lateral periodicity, ~49.7 A vacuum normal to the surface, and up to 200 methane (or 60 ethane / 30 n-butane) molecules as stated. Alkane dissociation is studied from 700 K to 1500 K. Equilibration: 10^4 steps at Delta t = 0.25 fs with temperature rescaling; C-Ni/H-Ni reactive terms disabled during equilibration. Production: reactive terms enabled, Nose-Hoover thermostat (Q = 100), neighbor list every step, runs up to ~1 ns for sampling.
Microkinetics: Surface species counts from MD are coupled to a microkinetic model for methane stepwise CH4 -> CHx kinetics (Eqs. 5-6* in the paper).
MD application (GRASP reactive MD on Ni(111))¶
Engine / code: Reactive MD in GRASP (as stated in Section 2.3 of the article).
System size & composition: Six-layer rectangular Ni(111) slabs with 720 Ni atoms in the primary setup (and 2880 Ni atoms for selected larger methane runs), plus adsorbate loadings up to 200 CH\(_4\), 60 C\(_2\)H\(_6\), or 30 n-C\(_4\)H\(_10\) molecules as reported.
Boundaries / periodicity: Lateral periodic boundary conditions with ~49.7 Å vacuum normal to the surface.
Ensemble: NVT during production after equilibration.
Timestep: 0.25 fs (Δt) after equilibration.
Duration / stages: Equilibration: 10⁴ steps with temperature rescaling while C–Ni/H–Ni reactive terms are disabled; production: reactive terms enabled with runs up to ~1 ns for sampling.
Thermostat: Nosé–Hoover thermostat with thermostat mass parameter Q = 100 (units as defined in the article).
Barostat / pressure control: N/A — NPT barostat not used for these NVT surface simulations.
Temperature: 700 K to 1500 K range surveyed for alkane chemistry.
Pressure / stress: N/A — external hydrostatic pressure control not stated for these NVT slab runs.
Electric field: N/A — not used.
Replica / enhanced sampling: N/A — not used.
Force-field training (Ni–C–H ReaxFF)¶
Parent FF / elements: ReaxFF parameters for H, C, and Ni developed for Ni(111) hydrocarbon chemistry.
QM reference: Real-space numerical atomic orbital extended DFT with PW91 and DNP basis; spin-polarized Ni; Monkhorst-Pack k-sampling and convergence settings in Section 2.2.
Training set / reference data: DFT geometries/energies for a subset of hydrogen and methane reactions with Ni(111) surfaces; EOS and key adsorption motifs included in the fit.
Optimization: Parameters optimized to the DFT training set (implementation details and weighting in pdf_path).
Reference data used: DFT reaction and EOS data; ReaxFF vs DFT comparisons for Ni EOS and adsorption geometries as reported.
Findings¶
Methane: MD plus fitting give a zero-coverage sticking prefactor and an activation energy for dissociative adsorption in line with prior literature orders of magnitude; microkinetic rates fitted to the same trajectories reproduce surface CHx evolution, with CH decomposition rate-limiting at lower T and carbon buildup at 1500 K. Ethane / n-butane: trajectories illustrate multi-step decomposition networks on Ni(111) without the same quantitative kinetic closure as for methane.
Limitations¶
Empirical ReaxFF vs QM accuracy; finite surface models and coverage effects; extract is p1–2—full paper holds complete numerical tables.
Relevance to group¶
Ni hydrocarbon ReaxFF development and catalytic MD comparable in spirit to group metal/hydrocarbon reactive simulations.
Citations and evidence anchors¶
- DOI 10.1016/j.susc.2011.11.035 — Surface Science 606, 615–623 (2012).
- Extract:
normalized/extracts/2012liu-surface-scie-reactive-molecular_p1-2.txt.