Skip to content

Coke resistant catalyst for hydrogen production in a versatile, multi-fuel, reformer

Summary

The authors report a potassium-promoted Ni–Pt catalyst on a pillared clay–alumina support for oxidative steam reforming of several liquid fuels (methanol, gasoline, diesel). Near 5 wt% K the system stays on stream for about 42 h under start-up/shut-down cycling with similar conversion and H\(_2\) productivity across fuels. Characterization ties coke resistance to stronger metal–support interaction, higher basicity (including CO\(_2\)-TPD), and K–Al–Si–O phases in the support. ReaxFF MD in the ADF suite (Cu/zeolite ReaxFF) anneals a K\(_2\)O–Al\(_2\)O\(_3\)–SiO\(_2\) mixture, with g® compared to DFT reference KAlSiO\(_4\) and KAlSi\(_3\)O\(_8\) structures to support assignment of K–Al–Si–O environments.

Methods

1 — MD application (atomistic dynamics)

  • Engine / code: ReaxFF reactive MD in the ADF Modelling Suite (noting the manuscript’s Section 2.6 text); coordinates visualized in VMD as stated.
  • System & composition: A mixed K\(_2\)O/Al\(_2\)O\(_3\)/SiO\(_2\) system with elemental ratios tied to the experimental support (Table S1 / discussion); reference DFT structures of Al\(_2\)O\(_3\), SiO\(_2\), K\(_2\)O, KAlSiO\(_4\), KAlSi\(_3\)O\(_8\) from the Materials Project for property comparison. Catalyst activity and TGA/XPS/BET/TPR/SEM-class data are from experiment in Sections 2–3, not from MD.
  • Boundaries / periodicity: NPT 3D cell (simulation of bulk mixed-oxide annealing as described).
  • Ensemble: NPT for the annealing trajectory.
  • Timestep: 0.25 fs for T < 2200 K; 0.1 fs for T > 2200 K (as stated in Section 2.6).
  • Duration / stages: 300 K for 100 ps; then 4 K/ps ramp to 2800 K; 2800 K held 200 ps; then 4 K/ps cool to 300 K; 100 ps at 300 K. Separate 100 ps NPT reference runs for comparison structures at 300 K.
  • Thermostat: Berendsen with 100 fs damping.
  • Barostat: Berendsen barostat with 1500 fs damping in the NPT protocol.
  • Temperature: 300 K holds; ramp to 2800 K; cooling; 4 K/ps rate during ramps.
  • Pressure: NPT (hydrostatic) as implemented with the Berendsen barostat; exact target pressure follows the Journal of Catalysis text (isotropic NPT in ADF per standard practice unless SI states otherwise).
  • Electric field: N/A.
  • Enhanced sampling: N/A (accelerated anneal via high-T MD rather than umbrella/hyperdynamics); N/A for metadynamics.

Experiment (reforming): Pillared support synthesis, Ni/Pt impregnation, reforming of commercial fuels, characterization as in the main text and ESI; N/A in the MD slot sense for atomistic table—handled as laboratory protocol in the article’s experimental sections.

2 — Force-field training

N/A — the article uses the ReaxFF Cu/zeolite field from the literature [33] in ADF for this mixed-oxide anneal; the authors do not re-fit parameters in this work. DFT (Materials Project) structures are used to check density to within ±5% before/around the anneal narrative.

3 — Static QM

DFT reference geometries from the Materials Project for K–Al–Si–O crystal benchmarks; the MD section compares the annealed structure’s RDFs to those of KAlSiO\(_4\) and KAlSi\(_3\)O\(_8\) from literature/MP sources. This supports short-range order assignment; N/A as a peer study of reforming on Ni surfaces.

4 — Review

N/A — primary experiment + targeted ReaxFF support study.

Findings

  • Reforming: Optimum near 5 wt% K balances activity and stability; lower/higher K loadings shift Ni sintering, side reactions (methanation, RWGS, coking), and H\(_2\) yield in the direction summarized in the article’s Figs. 1–2.
  • Support chemistry: K raises BET and pore volume to intermediate K; CO\(_2\)-TPD basicity ties to reverse Boudouard-style CO\(_2\) chemistry that reduces carbon buildup.
  • ReaxFF: Simulated g® of the annealed K–Al–Si–O mixture aligns K–Si correlations more closely with KAlSiO\(_4\) than with KAlSi\(_3\)O\(_8\); Al–Si motifs match aluminosilicate-like order, supporting XRD/XPS evidence for KAlSiO\(_4\)-type phases. N/A — the anneal is kinetic and need not reach full crystallization; authors state limits explicitly.

Limitations / outlook (authored): ReaxFF explores kinetic accessibility of local structures, not full thermodynamic equilibrium of industrial supports. TGA-level coke assignments remain as in the main text. Comparisons: K promotion routes (during support synthesis vs co-impregnation) and fuel-to-fuel conversion plateaus. Corpus / KB honesty: protocol numbers are from Journal of Catalysis 402 (2021) 177–193; confirm any SI-only tables locally.

Limitations

ReaxFF annealing is a finite-time, high-temperature model of mixed-oxide order; it complements but does not replace long-range crystallinity from XRD. External validation of the MD force field to K–Al–Si chemistry is as stated in the paper.

Relevance to group

ReaxFF MD alongside van Duin-coauthored reforming and K–aluminosilicate characterization.

Citations and evidence anchors

  • Reforming and characterization: J. Catal. 402 (2021) main text and ESI.
  • ReaxFF: Section 2.6 and Fig. 10 (discussing RDFs vs KAlSiO\(_4\)/KAlSi\(_3\)O\(_8\)).