Coarse-grained molecular dynamics of the CTAB/water/silicate system (Chien et al., J. Phys. Chem. C 2017) showed that at high surfactant concentration a preformed surfactant liquid crystal collapses when silica monomers are added and reforms as an ordered silica-surfactant mesophase only once oligomeric silicates exist, supporting the cooperative templating mechanism for MCM-41 and MCM-48 — but the authors could not rule out liquid-crystal templating (LCT) for the lamellar MCM-50 phase at very high surfactant concentrations. The same study identified what keeps the question open: silica condensation chemistry was never simulated on the fly (monomers and dimers were predefined species), the coarse-grained ion model distorts phase boundaries near the top of the accessible concentration range, full equilibration was computationally out of reach, and on-lattice and off-lattice modeling paradigms disagree. The open problem is to build a chemically resolved, quantitatively reliable simulation of surfactant/water/silicate systems at high concentration in which Si-O-Si condensation occurs in situ, to map the phase behavior over temperature, concentration, and degree of condensation, and to decide whether LCT can operate for MCM-50. The gap is not yet closed elsewhere: the reactive sticky-MARTINI model (Carvalho et al., 2022) demonstrated micelle-catalyzed silica polymerization but left higher-concentration mesophase simulations for future work, and the most recent model re-parametrization (Stavert et al., 2025) still uses pre-formed oligomers, validates only at low concentration, and leaves the MCM-50 question unresolved.

Background

Periodic mesoporous silicas of the M41S family are ordered porous materials built around surfactant templates: MCM-41 consists of a hexagonal array of cylindrical pores, MCM-48 of a bicontinuous cubic pore network, and MCM-50 of stacked silica sheets (a lamellar, layered phase). They are synthesized under basic conditions from a soluble silica source in the presence of the cationic surfactant CTAB (cetyltrimethylammonium bromide). Above its critical micelle concentration, CTAB self-assembles in water into micelles and, at sufficiently high concentration, into lyotropic liquid-crystal mesophases — ordered surfactant phases (hexagonal arrays of rods, bicontinuous cubic networks, and lamellar sheets) whose geometry mirrors the final pore architecture of the material.

Two idealized mechanisms have long been debated for how the silica wall forms. In liquid-crystal templating (LCT), a preformed surfactant liquid crystal acts as a static template and silica simply condenses around and within the already-ordered surfactant structure. In cooperative templating (CTM, also called cooperative self-assembly), there is no long-lived preformed template: anionic silicate oligomers and the cationic ammonium head groups co-assemble, with electrostatic charge matching jointly organizing silica and surfactant into the mesophase. Distinguishing the two matters for rational synthesis design, because they make different predictions for how synthesis conditions (surfactant concentration, pH, silica condensation rate) map onto pore structure.

Molecular simulation is a natural probe, but the process spans many scales: Si-O-Si bond formation between silicic-acid-derived species (condensation) is atomistic, while mesophase ordering is mesoscale. Coarse-grained (CG) molecular dynamics with the MARTINI force field — which maps roughly four heavy atoms onto one interaction bead, enabling microsecond simulations of large systems — offered a workable compromise. Chien et al. (2017) used such a CG model to extend earlier low-concentration studies (Jorge et al. 2007; Pérez-Sánchez et al. 2013, 2016) into the high-concentration regime. They first reproduced the experimental CTAB/water phase diagram qualitatively, then introduced silicate species as predefined monomers or dimers. The preformed hexagonal liquid crystal was found to collapse upon addition of silica monomers and to reform as a phase-separated ordered silica-surfactant mesophase only when oligomeric silicates were present, supporting CTM for MCM-41 and arguing against LCT for MCM-48. For the lamellar phases — precursors to MCM-50 — the mesophase instead remained stable with bromide, monomers, and dimers alike at very high surfactant concentrations, so LCT could not be ruled out for MCM-50.

The authors were explicit about what keeps this account incomplete. Silica condensation chemistry was not simulated on the fly: monomers and dimers were inserted as fixed, predefined species, so the simulations cannot speak to how condensation kinetics couple to mesoscale ordering. The MARTINI representation of bromide as a single bead carrying six implicit water molecules caps the accessible surfactant concentration near 77 wt% and compresses the high-concentration phase diagram, and decoupling the individual effects of the CG approximations (no directional hydrogen bonding, simplified electrostatics, coarse charge distribution) was described by the authors as very challenging and in some cases not currently possible. Equilibration at low temperature was out of reach — a replica-exchange MD attempt (an enhanced-sampling method that runs many copies of the system at different temperatures and swaps them, from 60 replicas over 300–403 K) did not converge after roughly 6 microseconds — and a discrepancy with on-lattice Monte Carlo models that include reactive tetrahedral silica (Jin et al. 2013, where monomers promote ordering) remains unreconciled; the lattice sites may effectively mimic a higher degree of silica condensation. The open problem below asks for the simulation capability and the mechanistic answer that these limitations currently prevent.

Problem Statement

Construct and validate a molecular-simulation description of the cationic-surfactant/water/silicate system at high surfactant concentrations — the regime where pure surfactant solutions form lyotropic liquid-crystal phases — in which silica condensation (formation and hydrolysis of Si-O-Si bonds, growth beyond monomers and dimers) occurs on the fly within the simulation rather than through predefined oligomer species. The simulation must be quantitatively reliable in this regime (in particular not limited by implicit ion-solvation artifacts that cap the accessible concentration and distort high-concentration phase boundaries) and its sampling must be sufficient to distinguish equilibrium phases from kinetic trapping. Use it to map the phase behavior over temperature, surfactant concentration, and degree of silica condensation, and to settle the mechanism question the 2017 simulations left open: can the liquid-crystal-templating mechanism operate in the synthesis of MCM-50 at very high surfactant concentrations — that is, does a preformed lamellar surfactant phase survive the addition of silica and convert, by in-situ condensation, into an ordered layered silica-surfactant composite — or does it dissolve and reform through cooperative silicate-surfactant assembly, as observed for the hexagonal (MCM-41) and bicontinuous (MCM-48) phases? The same body of work should test whether the cooperative-templating conclusions for MCM-41 and MCM-48 survive explicit condensation chemistry, and should clarify the mapping between on-lattice (tetrahedral-silica) and off-lattice model representations to the extent needed to interpret the remaining discrepancy between the two paradigms. Quantitative reliability and adequate sampling are judged operationally, by convergence of the identified phases with respect to sampling length — equilibrium phases must be separated from kinetic traps rather than assumed. The central question is whether LCT can operate for MCM-50, and answering it depends on constructing and validating the simulation and on mapping the phase behavior over temperature, surfactant concentration, and degree of silica condensation at least across the range that spans the lamellar phase and its transformation. Whether the cooperative-templating conclusions for MCM-41 and MCM-48 survive explicit condensation chemistry, and how the on-lattice and off-lattice representations map onto each other, further constrain that interpretation.

References

  1. Szu-Chia Chien, Germán Pérez-Sánchez, José R. B. Gomes, M. Natália D. S. Cordeiro, Miguel Jorge, Scott M. Auerbach, Peter A. Monson, 'Molecular Simulations of the Synthesis of Periodic Mesoporous Silica Phases at High Surfactant Concentrations', The Journal of Physical Chemistry C 2017, 121(8), 4564–4575, doi:10.1021/acs.jpcc.6b09429, https://doi.org/10.1021/acs.jpcc.6b09429
  2. Miguel Jorge, José R. B. Gomes, M. Natália D. S. Cordeiro, Nigel A. Seaton, 'Molecular Simulation of Silica/Surfactant Self-Assembly in the Synthesis of Periodic Mesoporous Silicas', Journal of the American Chemical Society 2007, 129(50), 15414–15415, doi:10.1021/ja075070l, https://doi.org/10.1021/ja075070l
  3. Germán Pérez-Sánchez, José R. B. Gomes, Miguel Jorge, 'Modeling Self-Assembly of Silica/Surfactant Mesostructures in the Templated Synthesis of Nanoporous Solids', Langmuir 2013, 29(7), 2387–2396, doi:10.1021/la3046274, https://doi.org/10.1021/la3046274
  4. Lin Jin, Scott M. Auerbach, Peter A. Monson, 'Simulating the Formation of Surfactant-Templated Mesoporous Silica Materials: A Model with Both Surfactant Self-Assembly and Silica Polymerization', Langmuir 2013, 29(2), 766–780, doi:10.1021/la304475j, https://doi.org/10.1021/la304475j
  5. Germán Pérez-Sánchez, Szu-Chia Chien, José R. B. Gomes, M. Natália D. S. Cordeiro, Scott M. Auerbach, Peter A. Monson, Miguel Jorge, 'Multiscale Model for the Templated Synthesis of Mesoporous Silica: The Essential Role of Silica Oligomers', Chemistry of Materials 2016, 28(8), 2715–2727, doi:10.1021/acs.chemmater.6b00348, https://doi.org/10.1021/acs.chemmater.6b00348
  6. Miguel Jorge, Andrew W. Milne, Olivia N. Sobek, Alessia Centi, Germán Pérez-Sánchez, José R. B. Gomes, 'Modelling the self-assembly of silica-based mesoporous materials', Molecular Simulation 2018, 44(6), 435–452, doi:10.1080/08927022.2018.1427237, https://doi.org/10.1080/08927022.2018.1427237
  7. Shingo Urata, An-Tsung Kuo, Hidenobu Murofushi, 'Self-assembly of the cationic surfactant n-hexadecyl-trimethylammonium chloride in methyltrimethoxysilane aqueous solution: classical and reactive molecular dynamics simulations', Physical Chemistry Chemical Physics 2021, 23(26), 14486–14495, doi:10.1039/d1cp01462e, https://doi.org/10.1039/d1cp01462e
  8. Johanna R. Bruckner, Jessica Bauhof, Jacqueline Gebhardt, Ann-Katrin Beurer, Yvonne Traa, Frank Giesselmann, 'Mechanisms and Intermediates in the True Liquid Crystal Templating Synthesis of Mesoporous Silica Materials', The Journal of Physical Chemistry B 2021, 125(12), 3197–3207, doi:10.1021/acs.jpcb.0c11005, https://doi.org/10.1021/acs.jpcb.0c11005
  9. André P. Carvalho, Sérgio M. Santos, Germán Pérez-Sánchez, José D. Gouveia, José R. B. Gomes, Miguel Jorge, 'Sticky-MARTINI as a reactive coarse-grained model for molecular dynamics simulations of silica polymerization', npj Computational Materials 2022, 8, 49, doi:10.1038/s41524-022-00722-w, https://doi.org/10.1038/s41524-022-00722-w
  10. Tom Stavert, Siddharth V. Patwardhan, Miguel Jorge, 'Development of a coarse-grained model for the early stages of ordered mesoporous silica formation', Molecular Simulation 2025, 51(3), 188–205, doi:10.1080/08927022.2025.2467834, https://doi.org/10.1080/08927022.2025.2467834
  11. Da Zheng, Junfeng Wang, Hongqiang Cui, Dinglin Zhang, Guohui Li, 'Combining Coarse-Grained Reactive Molecular Dynamics with an Enhanced-Sampling Method for MFI Zeolite Crystallization', Journal of Chemical Theory and Computation 2026, 22(3), 1409–1425, doi:10.1021/acs.jctc.5c01745, https://doi.org/10.1021/acs.jctc.5c01745