The H2O-MgSO4 system up to 1 GPa: implications for deep oceans in Ganymede and Titan.

Abstract:

Geophysical data from the Galileo and Cassini-Huygens missions are consistent with the presence of aqueous subsurface oceans in several icy satellites of the solar system. The generic structural model of ocean-harboring icy moons has stood for decades: in the H₂O unary system, the density contrasts between the liquid and the ice polymorphs imply that an aqueous ocean will sink relatively to an ice Ih surface layer but will be buoyant against high-pressure ices. However, high-pressure experiments in water-salt systems of interest to outer solar system bodies (e.g. the H₂O-MgSO₄ and H₂O-NaCl systems) illustrated different density contrasts between brines and ices, thus questioning this structural paradigm [1,2]. To determine whether an ocean may be stable below a specific ice layer, composition-density data alone are not sufficient and must be coupled to solubility constraints. At any brine-ice boundary, the density of the liquid may only exceed that of the ice phase if its contaminants are soluble enough, i.e. if the eutectic composition of the water-salt system at the specific P-T conditions of the boundary allows the critical density to be achieved [3]. However, the exploration of liquid-ice equilibria in water-salts systems of interest is still incomplete between 0 and 1 GPa, the pressure range expected in the hydrospheres of Ganymede and Titan. To address this problem, we developped a quantitative Raman spectroscopy approach to explore the liquid-ice equilibria in the H₂O-MgSO₄ system. The acquired dataset complete previous works in this binary system in the 0 – 1 GPa range by (a) extending the exploration of the ice VI solidus and (b) giving reliable PTx datapoints on the liquidii of ices at low (~7 wt. %) and high (~15 wt. %) aqueous MgSO₄ concentrations. With available data constraining the density of these solutions, these new results now make possible a reliable description of the H₂O-MgSO₄ system to be established up to 1 GPa. This background will be used to illustrate that the formation of deep high-pressure oceans may be a necessary consequence of the cooling of large differentiated icy moons such as Ganymede and Titan.

[1] Vance and Brown, 2013, GCA 110, 176-189. [2] Journaux et al., 2013, Icarus 226, 355-363. [3] Bollengier et al., 2014, Workshop on the Habitability of Icy Worlds, 4036.