Understanding Soil-Water Interaction Mechanisms via Molecular Simulations

Soil-water interaction is important to understand various soil behaviors. However, the theoretical understanding of soil-water interaction mechanisms is still incomplete especially in the adsorptive regime of soil water retention curves. Molecular simulation techniques were introduced herein to fill historical knowledge gaps in classical soil mechanics, to provide a novel angle to characterize soil behaviors and to promote the theoretical understanding of soil-water interaction. First, a general framework was developed to determine the contact angles of soil minerals. This procedure was employed to simulate the contact angles of α-quartz, orthoclase and muscovite. The simulated contact angles showed good agreement with the reported experimental or numerical results and thus substantiated the feasibility and accuracy of the proposed method. Second, molecular dynamics was employed to investigate the freezing and melting of water confined in nano-size pores. A general framework based on molecular dynamics simulations was developed to investigate the phase transition behavior of water confined in nano-size pores. It is found that the Young-Laplace equation may not apply in the low temperature range. An unfreezable threshold was identified in the phase composition curves and found to correspond to a pore diameter of 2.3±0.1 nm. Third, molecular dynamics was utilized to understand stress states of porous materials. A general framework was developed to extract effective stress from molecular simulations. Numerical simulations showed that the intergranular stress is different from the effective stress. Then, another simulation procedure was established to characterize nanoscale liquid bridges. Comparisons between macroscopic solution and simulation results revealed that the interplay of adsorption and capillarity is substantial at the nanoscale. At last, a general theoretical framework based on metadynamics was proposed to determine the lowest matric potential. The matric potential was derived from partial volume free energy and can be further calculated by the adsorption free energy. The lowest matric potential was determined as -2.00 GPa. In all of these studies, molecular simulations have shown strong potentials to resolve critical issues in soil behavior involving multiple phases. It is expected that the proposed general frameworks will open a new window to characterize soil-water interaction