Phase equilibria of ultramafic compositions on Mercury and the origin of the compositional dichotomy

Abstract

a b s t r a c tMeasurements of major element ratios obtained by the MESSENGER spacecraft using x-ray fluorescence spectra are used to calculate absolute element abundances of lavas at the surface of Mercury. We discuss calculation methods and assumptions that take into account the distribution of major elements between silicate, metal, and sulfide components and the potential occurrence of sulfide minerals under reduced conditions. These first compositional data, which represent large areas of mixed highreflectance volcanic plains and low-reflectance materials and do not include the northern volcanic plains, share common silica-and magnesium-rich characteristics. They are most similar to terrestrial volcanic rocks known as basaltic komatiites. Two compositional groups are distinguished by the presence or absence of a clinopyroxene component. Melting experiments at one atmosphere on the average compositions of each of the two groups constrain the potential mineralogy at Mercury's surface, which should be dominated by orthopyroxene (protoenstatite and orthoenstatite), plagioclase, minor olivine if any, clinopyroxene (augite), and tridymite. The two compositional groups cannot be related to each other by any fractional crystallization process, suggesting differentiated source compositions for the two components and implying multi-stage differentiation and remelting processes for Mercury. Comparison with high-pressure phase equilibria supports partial melting at pressure o10 kbar, in agreement with last equilibration of the melts close to the crust-mantle boundary with two different mantle lithologies (harzburgite and lherzolite). Magma ocean crystallization followed by adiabatic decompression of mantle layers during cumulate overturn and/or convection would have produced adequate conditions to explain surface compositions. The surface of Mercury is not an unmodified quenched crust of primordial bulk planetary composition. Ultramafic lavas from Mercury have high liquidus temperatures (1450-1350 1C) and very low viscosities, in accordance with the eruption style characterized by flooding of pre-existing impact craters by lava and absence of central volcanoes.

Sulfur solubility in reduced mafic silicate melts: Implications for the speciation and distribution of sulfur on Mercury

“…Assuming that Mercury differentiated through the formation of a magma ocean (Brown and Elkins-Tanton, 2009;Charlier et al, 2013;Chabot et al, 2014), early differentiation of the planet did set the sulfur content of the core and the lherzolitic mantle. Partial melting of the mantle then produced the secondary crust Namur et al, 2016).…”

“…It is made of a very large core (∼65 wt.% of the planet; ) and a thin mantle (420 ± 30 km; Hauck et al, 2013;Padovan et al, 2015) dominated by olivine, orthopyroxene, clinopyroxene, ± spinel and feldspar Charlier et al, 2013;Namur et al, 2016;Vander Kaaden and McCubbin, 2016). Building blocks of Mercury could be compositionally close to enstatite chondrite or bencubbinite chondrite mete- The data for the Mercurian crust were calculated using the MESSENGER data presented by Weider et al (2015).…”

Melting processes and mantle sources of lavas on Mercury

“…However, our model has the advantage that a single process, adiabatic decompression melting during the progressive cooling of Mercury's mantle, can explain the temporal evolution of lava compositions, their melting extents and temperatures. Charlier et al (2013) proposed that the mantle source of the low-Mg IcP-HCT lavas was lherzolitic while the source of the highMg IcP-HCT lavas was harzburgitic, suggesting that several mantle sources may have contributed to Mercury's magmatic activity. After more data on Na 2 O of surface compositions became available (Peplowski et al, 2014, it was apparent that most surface compositions for IcP-HCT as well as lavas from NVP and SP contain a significant and variable Na 2 O content which we have included in this new set of experiments.…”

“…It is therefore likely that variations in the source regions also involve variations in the CaS and MgS components. Contrasting mantle lithologies on Mercury may have formed by lateral and/or vertical compositional stratification of the mantle during crystallization of a magma ocean (Charlier et al, 2013), by vertical migration of heterogeneous mantle parcels due to convection of mantle cells (Tosi et al, 2013), or by progressive evolution of mantle sources by melt extraction through time. The most likely source material for Mercury is an EH chondrite.…”

“…We considered an average Na/Si ratio of 0.20, leading to ∼7 wt.% Na 2 O in our NVP composition (Peplowski et al, 2014). IcP-HCT lavas span a larger compositional range than NVP and may have a broadly bimodal distribution (Charlier et al, 2013;Weider et al, 2015). We prepared a starting composition corresponding to the average composition of the High-Mg group of IcP-HCT lavas.…”