Skip to content

Sulfur removal from petroleum coke during high-temperature pyrolysis. Analysis from TG-MS data and ReaxFF simulations

Evidence and attribution

Authority of statements

Prose sections below (Summary, Methods, Findings, etc.) are curated summaries of the publication identified by doi, title, and pdf_path in the front matter above. They are not new primary claims by this wiki.

For definitive numerical values, reaction schemes, and interpretations, use the peer-reviewed article (and optional records under normalized/papers/ when present)—not this page alone.

Summary

Thermogravimetry–mass spectrometry (TG-MS) on petroleum coke samples quantifies sulfur rejection versus temperature, particle size, and source, showing that high temperature (≳1673 K) drives large desulfurization extents (often ~80%) for fine fractions, with more variable behavior for large particles in some feeds. ReaxFF MD on a macromolecular coke model with thiophene-like sulfur probes atomistic elimination pathways at 3000–4000 K in NVT runs, rationalizing volatile sulfur species (e.g. SO₂, CS₂, COS) observed or inferred from experiment. Adri C. T. van Duin and Jonathan P. Mathews are Penn State coauthors with Zhong and collaborators in China. The paper frames the combination as a dual-scale story: macro mass-loss and evolved-gas signatures paired with microscopic bond-breaking sequences that would be inaccessible at calcination-relevant timescales without accelerated reactive MD.

Methods

Experiment (TG-MS). Thermogravimetry–mass spectrometry tracks mass loss and evolved gases for petroleum coke samples across particle sizes and sources under controlled heating programs, with temperature windows tied to H₂O, volatiles, CO₂/H₂, CO/SO₂, and trace CS₂ signals summarized in the abstract.

ReaxFF MD (§2.2.2, J. Anal. Appl. Pyrolysis 132 (2018) 134–142). Engine / code: ReaxFF molecular dynamics is run in ADF with the ReaxFF–CHONSSi parameterization cited in the article. Model / composition: a Qingdao petcoke-inspired macromolecule C₁₉₂H₉₆O₇N₃S₆ sits in a cubic 64 × 64 × 64 Å PBC cell (Fig. 1b). Ensemble: NVT; atomic valency and torsion cutoffs for bond-order assignment are 0.3 Å and 0.001 as stated in §2.2.2. Timestep: 0.25 fs (tests at 0.1 fs gave similar S decomposition kinetics; 0.25 fs preferred for S tracking per §2.2.2). Duration / temperature: 250 ps at each T in 1000–4000 K, with an additional 250 ps if transformations stall. Thermostat: Berendsen with 100 fs damping. Barostat / pressure: N/A — NVT (constant volume). Electric field / enhanced sampling: N/A — not used.

Analysis alignment. Simulation temperature is an acceleration device to populate elementary elimination and rearrangement events, interpreted alongside qualitative TG-MS product trends rather than as a literal furnace replica.

Findings

  • Gas evolution features include H₂O (low-T), volatiles, then CO₂/H₂, CO/SO₂ at higher T, and trace CS₂; COS/H₂S are absent or below detection under the reported conditions.
  • ReaxFF trajectories suggest a multi-step S removal sequence from thiophenic sulfur toward small S-bearing intermediates and ultimately SO₂ / CS₂-class products, consistent with the TG-MS picture.
  • Together, the datasets support the idea that thiophenic sulfur can be liberated through routes that pass through small oxygenated and sulfur-bearing fragments before appearing as stable gas-phase monitors such as SO₂.

Sensitivity and outlook. Experiment emphasizes temperature and particle size as levers on desulfurization extent; simulation highlights how ultrahigh T changes reaction accessibility within 250 ps windows. Limitations as authored include the illustrative single macromolecular model vs real petcoke heterogeneity (see ## Limitations). Corpus honesty: definitive mechanistic sequences and any literature benchmarks beyond the abstract bullets require the PDF at pdf_path.

Limitations

  • ReaxFF temperatures are far above industrial calcination to access reactions in nanoseconds; mechanistic insight is qualitative for absolute yields.
  • Petcoke structural heterogeneity means any single macromolecular model is illustrative, not sample-specific.

Relevance to group

Companion JAAP study to paper:2018qifan-combustion-a-reaxff-simulations on petcoke desulfurization, with van Duin / Mathews reactive MD involvement.

Citations and evidence anchors

  • DOI: https://doi.org/10.1016/j.jaap.2018.03.007 (papers/Zhong_PetCoke_JAAP_2018.pdf).