FORMATION of the LUNAR HIGHLANDS Mg-SUITE
as told by SPINEL
Finally, two competing hypotheses suggest lunar Mg-suite parental melts formed: (1) by shallow-level partial melting of a hybridized source region (containing ultramafic cumulates, plagioclase-bearing rocks, and an evolved incompatible element enriched component), producing a plagioclase-saturated, MgO-rich melt, or (2) when plagioclase-undersaturated, MgO-rich melts experienced magma-wallrock interactions within the anorthositic crust. I performed a series of phase equilibria experiments to test each model above in "Formation of the Lunar Highlands Mg-suite as told by Spinel."
Fig. 1. Liquid lines of descent for each starting composition with respect to the % normative anorthite in the melt (similar to a generalized Fo-An binary phase diagram). Symbol legend provided with green, light green, light blue and blue symbols representing compositions A, B, C, and D respectively. Estimated phase boundaries are drawn with solid and dashed black lines. Cr# of spinel listed next to experiments containing spinel. Melt compositions starting with greater than 55% normative An are Pl-saturated whereas melts with < 55 %An are Ol-saturated. b) 3D perspective view with % normative Qtz plotted on the z-axis, similar to Fo-An-Qtz ternary space. Phase surfaces (forsterite = green, spinel = pink, anorthite = blue – shown beneath the spinel stability field) and isotherms (green dashed lines on forsterite surface and black dashed lines on the spinel surface) estimated from experimental data.
Experimental and modeling results suggest the hybridized source model (1) is inconsistent with observed Mg-suite mineralogy. Instead, results indicate Al2O3-poor and MgO-rich Mg-suite parental melts formed chromite (FeCr2O4) in lunar troctolites, whereas pink spinel within the PST can be used as an indicator for magma-wallrock interactions within the lunar crust (a mechanism that increases the Al2O3 contents of initially Al-poor Mg-suite parental melts).
If PSA is used as a proxy for Mg-suite (Prissel et al. 2014), results from this study suggest remote detections of Mg-spinel represent areas of turbulent or replenished Mg-suite magmatism (e.g., pulsed injections, stoping, fracturing) that interacted strongly with the crust. If so, the concentration of PSA detections within the nearside southern highlands suggests this is the most promising region to study Mg-suite magmatism (Prissel et al. 2016).
If PSA is used as a proxy for Mg-suite (Prissel et al. 2014), results from this study suggest remote detections of Mg-spinel represent areas of turbulent or replenished Mg-suite magmatism (e.g., pulsed injections, stoping, fracturing) that interacted strongly with the crust. If so, the concentration of PSA detections within the nearside southern highlands suggests this is the most promising region to study Mg-suite magmatism (Prissel et al. 2016).
Fig. 2. The potential vertical distribution of spinel anorthosite lithologies within the lunar crust as a function of magma composition. Shown are various magmas ponding at the base of the crust, with subsequent dike propagation/intrusion. Mg# of magmas are listed within the legend boxes (Mg-suite = Mg-suite parental liquid (Longhi et al., 2010); A15C GG = Apollo 15C green glass; A15 YG = Apollo 15 yellow glass; A15 RG = Apollo 15 red glass) (Delano, 1986). Colored dashed lines correspond to potential minimum depths of formation of the respective magma based on spinel stability (e.g. Prissel et al., 2012 has shown that spinel production during magma–rock reactions between A15C green glass and anorthosite may be restricted to crustal depths near 10 km or greater). Spinel formation during lunar melt–rock reactions can proceed to shallower depths with Mg-suite parental melts. The potential for a compositional diversity of spinel anorthosites is discussed in the text. (After Head and Wilson, 1992).