Mineralogical model of the mantle of Mercury

Mercury possesses the second, highest density (5.23 g/cm³) in the Solar System after Earth. This high density is likely the result of the presence of a large inner core, composed of iron-light elemental alloys, overlayed by a relatively thin silicate shell, comprising the crust and the mantle [1].  The mercurian crust has been analyzed by the Messenger spectroscopic suite of instruments, which included, among others, the XRS (X-ray Spectrometer) and GRS (Gamma-Ray Spectrometer) spectrometers, capable of detecting the elements present on Mercury’s surface. The surface mineralogy of Mercury is dominated by enstatite and plagioclase, with small amounts of sulfides (oldhamite, CaS), the presence of which is a strong clue of the extremely reducing conditions which have led to Mercury’s accretion and differentiation [2]. The mercurian crust has been found to be very thin with estimates ranging between 26  ± 11 km and 35 ± 18  km [3,4].  Moreover, the mercurian mantle is also thin, thinner than other terrestrial planets' mantles, with an estimated thickness between 300 km – 500 km [5]. In addition,  the mantle shows a great lateral heterogeneity in mineral compositions, as indicated by the local, abrupt chemical changes in crustal chemistry [6].

Mercury’s large metallic core, likely partially molten and making up to 42% of its volume, combined with surficial observations (which have revealed a very small FeO concentration), and the peculiar position occupied by Mercury in the solar nebula, lead us to hypothesize a very reduced geochemical environment as its birthplace [7]. In literature, chondrites belonging to CB and enstatite chondrites (EN) have been considered the best precursor materials for Mercury’s composition [6, 8, 9, 10], sharing many analogies both in geochemistry and thermal evolution.

In light of the above, we chose a CB-like bulk composition to model the mineral assemblage of the mercurian mantle.

We reconstruct the evolution of the mercurian mantle starting from a CB chondrite-like bulk silicate composition, at thermodynamic equilibrium, as a function of temperatures and pressures estimated for Mercury’s mantle employing the Perple_X algorithm (6.9.1 version) [11]. We describe a dry scenario because the water abundance estimated for the bulk composition of Mercury silicate shells is quite low (0.3wt%, [12]) and due to the high-temperature ranges included in the model.

We predict that the peculiar geochemical environment where Mercury may have originated is characterized by a very low oxygen fugacity, which would result in a very reduced mineral assemblage for the mantle, dominated by pyroxenes and silica polymorphs, as shown in [9]. We expect that significant mantle phase transitions are unlikely due to the relative thinness of the mantle and the consequent low-pressure ranges (always <10 GPa) [13].

In conclusion, contrary to the terrestrial mantle, olivine is not predicted to be stable in our model. In effect, the low fO₂ results in stabilizing pyroxenes relative to olivine [9], producing mineral assemblages quite different from terrestrial peridotites.

Acknowledgments

G.M. and C.C. acknowledge support from the Italian Space Agency (2017-40-H.1-2020).

References

[1] Solomon, S. C., et al., (2018). Mercury: The view after MESSENGER (Vol. 21). Cambridge University Press. [2] Weider, S. Z., et al., (2012)., J. Geophys. Res. Planets, 117(E12).[3] Sori, M. M. (2018). Earth & Planet. Sci. Lett., 489, 92-99. [4] Padovan, S., et al., (2015). Geophys. Res. Lett., 42(4), 1029-1038. [5] Tosi P. et al. (2013), J. Geophys. Res. Planets, 118(12), 2474-2487.[6] Charlier, B. et al., (2013). Earth & Planet. Sci. Lett., 363, 50-60. [7] Cartier, C., and Wood, B. J. (2019). Elements,15(1), 39-45. [8] Stockstill‐Cahill, K. R., et al., (2012). J. Geophys. Res. Planets, 117(E12). [9] Malavergne, V. et al., (2010). Icarus, 206(1), 199-209. [10] Zolotov, M. Y., et al., (2013). J. Geophys. Res. Planets, 118(1), 138- 146. [11] Connolly, J. A. (2005). Earth & Planet. Sci. Lett., 236(1-2), 524-541. [12] Vander Kaaden, K. E., & McCubbin, F. M. (2015). J. Geophys. Res. Planets, 120(2), 195-209. [13] Riner M. A.,et al., (2008). J. Geophys. Res. Planets, 113(E8).