ABSTRACT Unfrozen water content is a key factor governing the physical properties of frozen soils, and variations in mineral surface properties play a critical role in determining this content. However, the fundamental physical mechanisms of how different mineral surfaces affect soil water freezing remain poorly understood. This study employs the molecular dynamics method to investigate the influence of mineral surface on soil water freezing, taking kaolinite as an example. By constructing atomic models of kaolinite, liquid water, and an initial ice nucleus and simulating the cooling process, the temperature dependence of unfrozen water on two different surfaces of kaolinite is obtained and compared with freezing simulation data on mica. Results show that the unfrozen water film thickness decreases in the order: kaolinite alumina surface (Al‐S) > mica surface (Mica‐S) > kaolinite silica surface (Si‐S). By employing two order parameters and analyzing hydrogen bonds (H‐bonds), the study reveals that surface hydroxyl groups on Al‐S promote strong water molecular ordering and extensive H‐bonds, conferring superior antifreeze properties. In contrast, the Si‐S exhibits limited capacity for H‐bonds formation, resulting in low water molecular ordering and poor antifreeze performance. The surface cations on Mica‐S increase the probability of H‐bonds formation between water molecules and the mineral surface, consequently rendering its surface water more ordered than that on the Si‐S and improving its antifreeze capability. This work provides atomic scale insights into soil water freezing mechanisms, aiding the interpretation of soil freezing behaviors.
Building similarity graph...
Analyzing shared references across papers
Loading...
Wang et al. (Wed,) studied this question.
synapsesocial.com/papers/69b3aba202a1e69014ccca3f — DOI: https://doi.org/10.1002/nag.70280
Yijie Wang
Zhen‐Yu Yin
Pierre‐Yves Hicher
International Journal for Numerical and Analytical Methods in Geomechanics
Centre National de la Recherche Scientifique
Hong Kong Polytechnic University
École Centrale de Nantes
Building similarity graph...
Analyzing shared references across papers
Loading...