Thermodynamic and Kinetic Trapping of NaCl in Ice VII

High-pressure ices (ice VI and ice VII) are believed to be the major constituents of the deep interiors of icy satellites and water-rich exoplanets. Incorporation of the impurities is a central problem as it alters the physical and chemical properties of high-P ices and thus influences the interiors of planets. However, the solubility of salt in ice VII remains poorly constrained. Different experiments have reported different estimates. Here, we address this discrepancy from a thermodynamic perspective.

We first developed a machine-learning interatomic potential based on the r²SCAN functional, covering a P-T range of 5–30 GPa and up to 1600 K. The predicted ice VII melting curve matches two recent experimental determinations across the investigated pressure range. Free-energy calculations indicate that the equilibrium solubility of NaCl in ice VII is limited to sub–mol% levels, substantially lower than several previously reported experimental estimates.

Deep-supercooling simulations of homogeneous saline liquids reveal rapid three-dimensional nucleation and growth of ice VII. During this process, the crystallization front advances much faster than solute transport in the liquid, leading to efficient solute trapping and incorporation of salt at concentrations far above the equilibrium limit. We further performed interfacial simulations near solid–liquid coexistence conditions, which show that solute diffusion in the solid remains strongly limited even close to the liquidus.

These results imply that salt retention in high-pressure ice is highly sensitive to the thermodynamic path by which the solid forms. The extremely low diffusivity of salt in the solid also suggests that kinetically produced, supersaturated “salty” high-pressure ice can persist over long timescales at low temperatures.