The composition of impurities in ice controls the stability of liquid water and thus the distribution of potential aqueous habitats. We present a framework for modeling the brine volume fraction in impure water ice as a polynomial function of temperature and bulk ice salinity, inspired by models originally developed for sea ice. We applied this framework to examine the distribution of brine within the thermally conductive layer of Europa’s ice shell, considering binary (NaCl and MgSO4) and multi-ion “analog” (Cl-dominated and SO4-dominated) endmember impurity compositions. We found the vertical extent of brine in a conductive ice layer, expressed as a fraction of the total layer thickness, to be <12% for NaCl, <2% for MgSO4, and <18% for both the analog endmember impurity compositions, suggesting that the depth where brine is stable in an ice shell is more sensitive to composition when only two ionic species are present. For the same temperature and bulk ice salinity, the brine volume fraction is higher in a Cl-dominated ice shell than a SO4-dominated ice shell. Pressure, governed by the ice thickness, was found to have only a minor effect on the vertical extent of brine within an ice shell, relative to temperature and bulk salinity. The minimum stable bulk ice shell salinity formed through freezing of an ocean was found to be insensitive to composition and ultimately governed by the magnitude of the assumed percolation threshold.

Accreted ice retains and preserves traces of the ocean from which it formed. In this work we study two classes of accreted ice found on Earth—frazil ice, which forms through crystallization within a supercooled water column, and congelation ice, which forms through directional freezing at an existing interface—and discuss where each might be found in the ice shells of ocean worlds. We focus our study on terrestrial ice formed in low temperature gradient environments (e.g., beneath ice shelves), consistent with conditions expected at the ice-ocean interfaces of Europa and Enceladus, and highlight the juxtaposition of compositional trends in relation to ice formed in higher temperature gradient environments (e.g., at the ocean surface). Observations from Antarctic sub-ice-shelf congelation and marine ice show that the purity of frazil ice can be nearly two orders of magnitude higher than congelation ice formed in the same low temperature gradient environment (~0.1% vs. ~10% of the ocean salinity). In addition, where congelation ice can maintain a planar ice-water interface on a microstructural scale, the efficiency of salt rejection is enhanced (~1% of the ocean salinity) and lattice soluble impurities such as chloride are preferentially incorporated. We conclude that an ice shell which forms by gradual thickening as its interior cools would be composed of congelation ice, whereas frazil ice will accumulate where the ice shell thins on local (rifts and basal fractures) or regional (latitudinal gradients) scales through the operation of an “ice pump”.

Brine systems in Europa’s ice shell have been hypothesized as potential habitats that could be more accessible than the sub-ice ocean. We model the distribution of sub-millimeter-scale brine pockets in Europa’s ice shell. Through examination of three habitability metrics (water activity, ionic strength, salinity), we determine that brine pockets are likely not geochemically prohibitive to life as we know it for the chloride and sulfate-dominated ocean compositions considered here. Brine volume fraction is introduced as a novel habitability metric to serve as a proxy for nutrient transport and recycling—because of its role in governing permeability—and used to define regions where nutrient-open, nutrient-closed, and relict habitats are stable. Whereas nutrient-closed habitats could exist wherever brine is stable, nutrient-open habitats are confined to meter-scale regions near the ice-ocean interface where freezing is occurring. This classification scheme can help guide future life-detection missions to ocean worlds.