Modeling the sorption of gas molecules in densely packed bulk material can be a challenging task with standard molecular modeling methods like Molecular Dynamics (MD) or Monte Carlo (MC). This is especially true when the desired system also includes chemical reactions. However, when successfully applied, it can provide useful detailed information as it did in our application of thermochemical heat storage. For such a system, information on phase equilibria between different hydration levels of salts, at given vapor pressure and temperature, is essential for a good thermochemical heat storage design.
Recenatly a study combined the well-known Grand canonical Monte Carlo (GCMC) algorithm with reactive force field MD (ReaxFF-MD). While this combination has been used before, because the considered heat storage application requires the insertion of H2O molecules in dense bulk salt structures, when one only uses standard GCMC successful insertion of H2O molecules would be very limited to almost zero. This is due the physical volume of the water molecule and the limited available voids within the dense salt hydrate. As a consequence, chemical equilibria are hard to reach. To tackle this limitation, the ReaxFF-GCMC combination was biased with the computationally cheaper WCA potential. With this bias, each ReaxFF insertion is preceded by k (fast) insertion trials, and the most probable trials have the highest probability of being selected for a re-computation insertion trial with the more expensive ReaxFF potential.
By applying this WCA-ReaxFF-GCMC combination, the deliquescence hydration equilibrium of MgCl2.6H2O at given vapor pressure and temperature could be modeled successfully. Such a study would be computationally too demanding without the WCA bias. The WCA-ReaxFF-GCMC methodology can also be applied to other systems, includin reactions and sorption of molecules, such as carbonation, hydration, oxidation, etc.
K. Heijmans, I. C. Tranca, M.-W. Chang, T. J. H. Vlugt, S. V. Gaastra-Nedea, D. M. J. Smeulders, Reactive Grand-Canonical Monte Carlo Simulations for Modeling Hydration of MgCl2, ACS Omega 6, 32475-32484 (2021)Key conceptsEnergy storage materials science ReaxFF