How Much Hydrothermal Hydrogen Might We Find in Enceladus’ Plume?

Introduction: Geochemical [1,2] and geophysical [3,4] data obtained by the Cassini spacecraft strongly suggest the existence of a liquid water ocean [5] in contact with a rocky core on Saturn’s satellite Enceladus. This leads to the possibility of hydrothermal activity that could support life [6]. Indeed, it has been proposed that there are hydrothermal vents inside Enceladus to explain the formation of silica nanoparticles that are derived from Enceladus [2]. It has also been suggested that Enceladus has an alkaline ocean as a result of serpentinization reactions between water and rocks [7]. If there are serpentinizing hydrothermal systems on Enceladus that are currently active, then we should search for other clues in the plume to confirm their existence. In this respect, a smoking gun may be molecular hydrogen (H2), which is abundant in hydrothermal systems on Earth (in particular Lost City) that may be analogous to those on Enceladus [8]. During previous Cassini flybys of Enceladus, the Ion and Neutral Mass Spectrometer (INMS) detected counts at mass channel 2 in closed source neutral mode that are attributed to H2 [9]. The signal was enhanced at faster flyby velocities as a result of impact-induced chemistry in the antechamber of the instrument [10], but up to ~15% H2 was still detected consistently during the slowest flybys [9]. At present, it is unclear if this H2 is native to the plume or an artifact of highspeed sampling of the H2O-rich plume [11]. In an attempt to resolve this question, a search for H2 was performed using the open source neutral beam mode of INMS during the E21 flyby, for which the data are being analyzed [12]. To assist in the interpretation, we have made three theoretical estimates of how much hydrothermal H2 could be present for different geochemical/geophysical scenarios. Estimate based on redox mass balance: We have constructed a mass balance model of serpentinization and H2 production on Enceladus. In this model, H2 production from H2O is stoichiometrically coupled to the oxidation of iron, which may be the most abundant and strongest reductant in rocks on Enceladus (sulfide S and organic C are likely to be present, but may not be as important to H2 production). The geochemical model quantifies phase transformations in the Mg-Si-Fe-S-O-H system. We estimate the abundance of iron by assuming a solar composition of rock-forming elements in the core [13], and scaling it to the internal structure model of [14]. We then calculate the amount of H2 that can be produced, which depends on the evolution of the oxidation state (Fe/Fe/Fe) of the core. Here, we consider a two-stage scenario of redox evolution: (1) silicates are hydrated and metallic iron is oxidized to ferrous serpentine as a result of low-temperature water-rock interaction during differentiation; (2) a more prolonged period of hydrothermal oxidation follows that is driven by heating of the core and subseafloor fluid flow [15]. In the latter stage, which may be occurring today [2], H2 generation is coupled to the transformation of ferrous serpentine to magnetite. We find that the theoretical yield of H2 is large. The geochemical mass balance indicates that ~10 moles of H2 can be produced during the differentiation stage. If this process were to occur over 100 Myr, then the mean production rate would be ~10 mol/yr. This is large when compared to the present rate of H2O emission in the plume (~3.5×10 mol/yr; ref. 16). For the later stage of redox evolution, we find that up to ~5×10 moles of H2 can be produced, depending on the unknown extent of reaction progress (0-100%). The mean production rate over 4500 Myr could be as large as ~10 mol/yr in the limit of reaching 100% progress today. This rate can be converted to an upper limit for hydrothermal H2 in the plume of ~3%. This is an upper limit for three reasons. First, it seems unlikely for stage two to just be finishing. Instead, it could have ended in the past [17], in which case there would be very little H2 in the plume; or stage two may be only partially complete. Second, the H2 production rate may not be constant but may decrease through time, as an oxidation front migrates downward into the core. Reactant rocks would be deeper and less accessible to ocean water at later times, so the average rate may be an overestimate for the present. Third, not all of the H2 may be lost via the plume – there could be diffuse emissions. These effects are difficult to quantify, but provisionally we suggest that a possible range for hydrothermal H2 in the plume may be ~0.1-1%. Estimate based on hydrothermal energy balance: In addition to H2 and other chemical species, hydrothermal fluids would deliver heat from the core to the ocean on Enceladus [15,18]. The heat flux provides a geophysical constraint on hydrothermal H2 in the plume. The rate of delivery of H2 can be expressed as