Eucrite versus Angrite: micrometeorites as natural forward melting experiments on chondrites

Early Solar System magmas are represented in meteorite collections by extrusive rocks exhibiting a rather limited number of compositions. The most common are the well sampled eucrites or other eucrite-like meteorites which are basalts consisting of plagioclase and pigeonite, probably sampling several different parent bodies. In contrast, a number of critically silica-undersaturated meteorites called angrites consist of olivine, anorthite, and Al-rich augite and sample a very different type of parent body which appears to be unique. The eucrite-angrite dichotomy is considered to reflect melting of chondritic starting materials under reduced or oxidised conditions respectively. This equates to transfer of Fe from metal to silicate in ordinary chondrite-like compositions, or melting of carbonaceous chondrite-like bulk compositions which already have a high FeO content. In CaO-MgO-Al2O3-SiO2-FeO space (CMAS+FeO), addition of FeO to ordinary chondrite brings the bulk composition away from the melting region which produces eucritic melts on the olivine-anorthite-pigeonite-liquid cotectic, and into a melting region which produces angritic melts on the olivine-anorthite-augite-liquid cotectic, on the low-silica side of a thermal divide (i.e. a tie-line) between anorthite and augite. Early angrite discoveries in Antarctica implicated a high Ca/Al source compared to the eucrite parent body (such as an enrichment of refractory material such as calcium-aluminium rich inclusions), but more recent angrite finds appear to be simple primary melts consistent with the model of melting chondrite with variable additional FeO.

We have investigated the mineralogy and geochemistry of a suite of more than one hundred Antarctic micrometeorite spherules, and have identified several with compositions equivalent to early Solar System extrusive magmas. At least five spherules contain glasses with compositions which could be regarded as basaltic in a general sense (~50 wt. % SiO2 and ~6-12 wt. % each of Al2O3 and CaO), with one very similar to natural eucrite compositions. However, these glasses differ from most eucrites in that they coexist with olivine in the spherule, and can therefore be regarded as more primitive portions of the fractionation sequence. In stark contrast to these objects, we have also identified a single spherule consisting of olivine, Fe3+ rich spinel, and a glass with an angrite-like composition (~40 wt. % SiO2 with ~12 wt.% Al2O3 and ~16 wt. % CaO). The angritic glass is very similar in terms of major oxide composition to the well studied D’Orbigny and Sahara 99555 angrites, which are postulated to be near-primary melts of an oxidised chondritic source. Genuine angrites have very low contents of NiO (less than ~100 wt. ppm in these cases), indicating that their parent body separated a Ni-(Fe, S) rich core. Our angritic spherule, on the other hand, contains substantial NiO in both olivine and Fe3+ rich spinel (~1.5 wt. %). In fact, the presence of Fe3+ rich spinel as a liquidus phase is enough to indicate that formation for our spherule occurred under even more oxidising conditions (i.e. in the atmosphere of the Earth) and precludes a genetic relationship with true angrites.

Nevertheless, complementary eucritic and angritic melts formed as fusion crust materials in the atmosphere of the Earth parallel the redox control which various chondritic compositions exert on magma evolution in early Solar System planetesimals. In micrometeorites, bulk composition mostly approximates alkali-free carbonaceous chondrite components, or an alkali-free ordinary chondrite with or without metal which is prone to oxidation during entry. First melting in such particles therefore occurs near the olivine-anorthite-pigeonite-liquid or olivine-anorthite-augite-liquid cotectics (and importantly in the angritic case, on the low-silica side of the thermal divide, away from eucrites), the difference being whether exposure of Fe-metal to air occured during atmospheric entry. In the event that Fe-metal in the micrometeoroid was shielded from air or physically separated, melting should produce a silicate assemblage of olivine and a eucritic liquid which quenches to glass. Exposure of Fe-metal to air at high temperature during entry leads to oxidation and repartitioning into the silicate melt, producing an angritic liquid saturated in olivine and spinel. As previously noted, the formation of angritic melts may be assisted by incorporation of refractory calcium-aluminium rich inclusions.

There exists a small region in CMAS+FeO along the ordinary chondrite+FeO vector which results in melting along the olivine-anorthite-augite-liquid cotectic, but on the high-silica side of the thermal divide, nearer the eucrites (i.e. between eucrites and angrites). This region was not explicity discussed in previous consideration of the phase equilibria of angrites, but high degrees of melting of this composition will result in magmas equivalent to terrestrial mid-ocean ridge basalts (MORB) in terms of major element composition and their sequential saturation of olivine, anorthite, and augite during cooling (±spinel which is sensitive to minor trivalent components such as Cr3+ and Fe3+). The absence of MORB-like extraterrestrial rocks formed during low pressure melting of planetesimals with intermediate redox states, and probably their absence from our micrometeorite survey as well, indicates that this region is very small. It is bound, on the one side, by a tie-line between anorthite and the olivine-anorthite-augite-pigeonite eutectic, and on the other side by a tie-line between anorthite and the thermal divide. Thus, despite the complementary nature of eucrites and angrites, it is in fact MORB which is the nearer relation of both – at least in terms of phase equilibria.