Molecular simulations to estimate thermodynamics for adsorption of polar organic solutes to montmorillonite

Abstract

SummaryIdeally, one could use molecular mechanics or quantum mechanics to predict the magnitude of organic solute adsorption from water to soil minerals. Reproduction and/or prediction of mineral and interfacial structures remains challenging, but calculation of meaningful energy relations through computational chemistry techniques is even more difficult than structural calculations. This paper attempts to define the necessary and relevant components for an overall scheme that allows translation of computed interaction enthalpies to experimental adsorption enthalpies and vice versa. While the scheme could be applied to quantum calculations, we test it for the possibility of using empirical molecular mechanics to estimate relative energies for the adsorption of non‐ionic organic solutes in clay mineral‐water‐solute systems. We used molecular dynamics’ simulations to estimate relative clay–organic interaction enthalpies for a series of nitro‐aromatic solutes and hydrated, K‐saturated montmorillonite, for comparison with experimental adsorption isotherm data for the same clay‐nitroaromatic systems. The trend of computed interaction enthalpies (e.g. −234 ± 17 kJ mol−1 for trinitrobenzene and −154 ± 16 kJ mol−1 for p‐nitrobenzene) agreed modestly well with the trend of adsorption maxima from the experiments. Furthermore, we developed several variants on a thermodynamic cycle framework for comparing computed interaction energies with experimentally determined adsorption enthalpies. The algorithms, which include estimates for enthalpy changes both in bulk solution and in the clay interlayer, show promise: for p‐dinitrobenzene and for 1,3,5‐trinitrobenzene, the overall predicted adsorption enthalpies (e.g. −13 ± 22 and −67 ± 23 kJ mol−1, respectively) were in modest agreement with experiments (−18 ± 1 and −28 ± 4 kJ mol−1, respectively). We discuss shortcomings of the methods, in hopes of encouraging better estimates for the various energy terms, improvement of the algorithms, and more valid comparisons between quantum mechanical or molecular mechanical interaction energies and experimental enthalpies.