Computationally guided synthesis of carbon coated mesoporous silica materials

Classical micelle assembly, boosted ReaxFF silica condensation, and high-temperature pyrolysis of several carbon precursors inside pores are combined in a corrected-proof Carbon article on mesoporous silica with in-pore carbon healing.

Non-primary article PDF

This file is a corrected proof PDF for the same study as paper:2024dasgupta-carbon-221-2-computationally-guided (version-of-record PDF). Use the VOR page for DOI-backed citation; keep this page for corpus provenance of the proof.

Summary¶

Mesoporous silica synthesis and pore healing via in-pore carbonization are explored with non-reactive MD of Pluronic L64 self-assembly and bond-boosted ReaxFF MD of silicic acid condensation, followed by high-temperature carbonization of several precursors inside pores. The workflow emulates a computationally guided synthesis loop in which soft-matter self-assembly sets mesoscale confinement, reactive silica chemistry locks in pore order, and pyrolytic carbon deposition is compared across feedstocks that differ in oxygen content and aromaticity.

Methods¶

Self-assembly (classical MD)¶

Pluronic L64 micelles in water; H-bond partitioning (hydrophilic blocks vs water).

Silica condensation (ReaxFF + bond boost)¶

ReaxFF bond boosting at 300 K; 1500 K unboosted runs; Si(OH)\(_4\) addition on micelle templates.

Confined carbonization (ReaxFF, high T)¶

2200–2600 K PE, lignite, sucrose, PET-like precursors in silica pores—rings, H\(_2\), turbostratic signatures; mirror thermostat/boost/ramp from paper:2024dasgupta-carbon-221-2-computationally-guided/VOR text.

Observables in high-T trajectories. The production runs monitor carbon ring formation, H\(_2\) release bursts, sp\(^2\) content proxies, and whether tars occlude pore openings—metrics reported quantitatively in the Carbon article and duplicated here for proof-PDF tracking. When citing exact percentages, prefer the VOR page [[2024dasgupta-carbon-221-2-computationally-guided]] to avoid proof/final text drift.

MD application (classical + ReaxFF, multi-temperature). LAMMPS (and classical MD for Pluronic): periodic cells for self-assembly; NVT/NPT as in main text; bond boost+ReaxFF for 300 K condensation and 1500 K unboosted and 2200–2600 K pyrolysis segments—time step (fs), ns-scale (or as stated) durations, Nosé–Hoover thermostat, and N/A—NPT Parrinello barostat only if a pressurized segment is documented. N/A—external E-field; bond boost stands in for umbrella-style rare-event sampling as an acceleration method. Atom/stoichiometry in paper:2024dasgupta-carbon-221-2-computationally-guided.

Findings¶

~80% of H-bonds involve hydrophilic micelle segments and water in the non-reactive model. >60% of silicic acid additions yield periodic mesoporous silica order in the boosted ReaxFF runs. 1500 K unboosted condensation doubles polymerization rate versus 300 K boosted runs. Polyethylene and high-rank lignite best form turbostratic graphene-like carbon; PE shows strong H₂ release and sp² content, whereas sucrose yields the least planar carbon. A PET-derived tar can coat silica pores, blocking silicic acid ingress in the illustrated trajectory. The combined metrics are presented as screening criteria for which carbon precursors heal defects without occluding continued silica condensation.

Limitations¶

Prefer 2024dasgupta-carbon-221-2-computationally-guided for bibliographic citation; this proof PDF may differ slightly from the final Carbon layout.

Relevance to group¶

ReaxFF-driven silica condensation and pyrolytic carbon healing from Dasgupta/Mao/van Duin collaboration.

Citations and evidence anchors¶

See paper:2024dasgupta-carbon-221-2-computationally-guided for DOI 10.1016/j.carbon.2024.118891.

reaxff-family

Proof duplicate of [[2024dasgupta-carbon-221-2-computationally-guided]]; cite the VOR for DOI-backed bibliography.