Ferroan Anorthosite Suite Research Articles

Fragments of ancient lunar crust: Petrology and geochemistry of ferroan noritic anorthosites from the Descartes region of the Moon

Clasts of ferroan noritic anorthosite from Apollo 16 breccia 67016 have mineralogical and geochemical affinities with pristine ferroan anorthosites and therefore may represent primary igneous rocks from the lunar crust. The textures of these clasts document a complex history involving magmatic crystallization, brecciation, subsolidus recrystallization, and sulfide metasomatism. Nonetheless, the major, minor, and trace element compositions of their minerals, as determined by electron and proton microprobe, show that the clasts are essentially monomict and belong to the ferroan anorthositic suite of lunar highlands rocks. The low bulk Ni contents, low Ni Co ratios, and ancient SmNd isotopic age (≥4.5 Ga; Alibert et al., 1994) of the clasts are consistent with this interpretation. Strontium and gallium contents of plagioclase in the clasts show that their parental magma was depleted in Ga relative to Sr by a factor of about 30 relative to carbonaceous chondrites. This depletion of Ga, which is both moderately volatile and moderately siderophile, probably reflects volatile loss from the Moon rather than extraction of metal. The parental magma from which these clasts originally crystallized was enriched in Sr by about 4–6X relative to proposed bulk Moon compositions, which implies either that ferroan anorthosites crystallized late in the evolution of a lunar magma ocean, or that their parental magma was a basaltic partial melt. Bulk major and trace element compositions of these clasts are remarkably similar to that of the average upper crust of the Moon, which suggests that igneous ferroan noritic anorthosite such as that found in the North Ray Crater breccias and soils may be a significant primary component of the lunar crust. The ferroan noritic anorthosite clasts in 67016 are unusual among lunar samples for their abundance of troilite, and for the evidence of metasomatic introduction of volatile chalcophile elements such as Cu, Zn, Ni, Sb, and Se. Anhydrous COSCl vapors are the most likely transport agent for these metals, which would be more consistent with an indigenous lunar origin for the volatiles rather than a cometary or meteoritic source.

The ferroan‐anorthositic suite, the extent of primordial lunar melting, and the bulk composition of the Moon

The geochemical bimodality among compositionally pristine lunar rocks is strong evidence that a major component of the Moon's crust is fundamentally unrelated to the plagioclase flotation crust that accumulated over the primordial magma ocean. The ferroan anorthositic suite (FAS) has all the characteristics expected for a series of magma ocean plagioclase flotation cumulates, but we estimate, based on mass balance for Al2O3, that ∼ 30 wt% of the crust, and 20 wt% of the crust's total Al2O3 (or roughly 15% of the bulk‐Moon Al2O3 content, assumed to be 3–4 wt%), was emplaced by postmagma ocean, Mg‐suite “serial” magmatism. Remelting of plagioclase‐bearing magma ocean cumulates cannot plausibly account for the “late” emplacement of Al2O3, because the only plagioclase‐rich cumulates produced by a deep, ultramafic magma ocean would have mg ratios too low to engender the Mg‐suite. Conceivably, remelting of garnet‐rich magma ocean cumulates supplied the late Al2O3, but only if the bulk‐Moon Al2O3 content is substantially higher than generally assumed. High pressures in the magma ocean acted to enhance the py (pyroxene/(pyroxene+olivine)) ratio of the MO cumulates, and the Al2O3 content of whatever pyroxene formed. Remelting of pyroxene‐rich MO cumulates could conceivably have supplied the late Mg‐suite Al2O3 to the crust. The source region would have to encompass roughly 40 wt% of the mantle. Another possible source for late Al2O3 addition is a large portion of the deepest mantle that avoided more than a slight degree of partial melting, and thus remained Al‐rich, during formation of the magma ocean. The source region would have to encompass 20–30 wt% of the mantle. Ascent of diapiric masses from the lowermost mantle would likely coincide with (trigger, or be triggered by) a large‐scale overturn of the gravitationally unstable MO cumulate pile, with concomitant pressure‐release melting, in which case the Mg‐suite parental melts would be derived from a mix of deepest MO cumulates and the even deeper primitive mantle. The popular giant impact hypothesis of lunar origin implies an extremely hot beginning for the Moon, suggesting that the deep interior was either molten or depleted in Al2O3 (by differentiation in hot protomoons) as it formed. Origin of ∼ 30 wt% of the Moon's crust by postmagma ocean, serial magmatism is more consistent with coaccretion models that form the Moon without a giant, hot impact. In any scenario with a large fraction of the crust originating by partial remelting of deep magma ocean cumulates, mass balance for Al2O3 implies that the bulk‐Moon normative py ratio must be greater than roughly 0.5, i.e., significantly higher than generally estimated. In this respect, the Moon appears more compositionally similar to chondritic silicates than to the Earth's upper mantle.

Ages of pristine noritic clasts from lunar breccias 15445 and 15455

Rb-Sr and Sm-Nd isotopic studies were undertaken of two Apollo 15 breccias, 15445 and 15455, collected near Spur Crater on the Apennine Front. Seven pristine lithic clasts including four norites, one anorthositic norite, one troctolitic anorthosite and one spinel troctolite, as well as a matrix sample of 15445, were analysed. In addition, a K-Ar age determination was also made for a plagioclase separate from a different anorthositic norite clast. Sm-Nd isotopic data of a pristine Mg-norite sample, 15445,17, from the large white Clast B of breccia 15445 yielded a precise internal isochron age of 4.46 ± 0.07 Ga for λ( 147 Sm) = 0.00654 Ga −1 and an initial 143Nd 144Nd of 0.506058 ∓ 0.000078 (normalized to 148Nd 144Nd = 0.24308) corresponding to an initial ϵ Nd of +0.71. However, Sm-Nd isotopic results of another Mg-norite sample, 15445,247, chipped from the same Clast B, about 1 cm away from 15445,17, yielded a significantly younger isochron age of 4.28 ± 0.03 Ga and a higher initial 143Nd 144Nd of 0.506246 ∓ 0.000037 corresponding to an initial ϵ Nd of −0.35. The Sm-Nd data indicate that Clast B is heterogeneous and contains at least two similar lithologies. The Rb-Sr isotopic systems were severely disturbed for both norites. No useful isochron ages were observed. The Sm-Nd age and ϵ Nd data for A-15 Mg-norites 15445,17 and 15445,247, A-17 Mgnorite 78236, and two Mg-gabbronorites 73255,27,45 and 67667 are clearly resolved from each other. Four distinct parental magmas are required for derivation of these two major groups of Mg-suite noritic rocks. The Rb-Sr isotopic system of an anorthositic norite clast, 15455,228, was slightly disturbed; and isotopic data defined an isochron age of 4.59 ± 0.13 Ga for λ( 87Rb) = 0.0139 Ga −1 or 4.55 ± 0.13 Ga for λ( 87Rb) = 0.01402 Ga −1 . The corresponding initial 87Sr 86Sr for the rock is 0.69899 ∓ 0.00006 and is in satisfactory agreement with the LUNI value of Nyquist et al. (1974). The Sm-Nd system is more disturbed than the Rb-Sr system for this rock. A tie-line between whole-rock samples and a pyroxene separate gives a Sm-Nd age of 4.53 ± 0.29 Ga, which is in nominal agreement with the Rb-Sr isochron age. The 39Ar 40Ar age of a plagioclase separate of the rock has a weighted average value of ~ 3.83 Ga, suggesting that the age was reset by the Imbrium impact event. Probably, the Sm-Nd and Rb-Sr systems of the clast were also partly reset by this event. Our age results indicate that some Mg-suite rocks are as old as ferroan-anorthosite-suite rocks. Furthermore, age data of three major crustal rocks (a Mg suite, a ferroan-anorthosite suite, and an evolved suite) show that they all have variable ages. These observations are difficult to explain by the “magma ocean” hypothesis alone. Petrogenetic processes such as production of Mg-suite rocks by “serial magmatism” ( Walker, 1983), of ferroan-anorthosite-suite rocks by “multiple diapiric intrusions” ( Longhi and Ashwal, 1985), and evolved-suite rocks by “continuous magmatism” ( Meyer et al., 1988, 1989) probably all played a role. These processes seem to be able to accommodate the wide ranges of ages and different origins observed for lunar crustal rocks.

Paired lunar meteorites MAC88104 and MAC88105: A new “fan” of lunar petrology

New lunar meteorite MAC88104/5 represents an exciting new opportunity to study a potentially unsampled region of the Moon. We have analyzed six thin sections by electron microprobe and three bulk samples by Instrumental Neutron Activation (INA) in order to determine the chemical characteristics of this new lunar sample. Lunar meteorite MACS 8104/5 is dominated by lithologies of the ferroan anorthosite (FAN) suite and contains abundant granulitized highland clasts, devitrified glass beads of impact origin, and two small clasts which appear to be of basaltic origin. One of these “asaltic” clasts (clast E in MAC88105,84) is probably mesostasis material, whereas the second larger clast (clast G) may be similar to the Very Low-Ti (VLT) or low-Ti/high-alumina mare basalts. Impact melt clasts MAC88105,69 and 72 have major and trace element compositions similar to the bulk meteorite. There is little evidence of any LKFM (Low-K Fra Mauro or low-K KREEP) contribution to this meteorite, as MAC88104/5 and other brecciated lunar meteorites are Fe-rich and poor in the incompatible elements relative to Apollo 16 regolith and feldspathic breccias. While the exact site of origin for the lunar meteorites cannot be pinpointed, it is evident that they were derived from a relatively KREEP-free ferroan anorthosite terrain.

The MacAlpine Hills lunar meteorite and implications of the lunar meteorites collectively for the composition and origin of the Moon

The MAC88104/MAC88105 meteorite is a lunar highlands regolith breccia even more anorthositic than previously available samples of highlands regolith. Clasts studied include two unusual pristine rocks. One, a 2.5-mm, slightly granulitic clast rated as probably pristine, contains extraordinarily Fe-rich (Fo 40) olivine. The other, a 5-mm clast with clear vestiges of a poikilitic cumulate texture, has silicate compositions that extend the range of the Mg-suite in the direction of the high-mg ∗ end of the ferroan-anorthositic suite. The pyroxene of the latter clast is relatively Ca-rich and poorly equilibrated by lunar cumulate standards, suggesting that it may have formed in an uncommonly shallow intrusion. The four lunar highlands meteorites are all regolith breccias (or, in one case, a marginal type between regolith breccia and “ordinary” polymict breccia), and they probably represent three to four separate lunar source craters. The proportion of regolith breccias is higher among lunar highlands meteorites than among rock samples from the classic highlands site, Apollo 16 (~9% regolith breccias). For the majorelement composition of the crust, the highlands meteorites confirm that the two low-Th central nearside sites, Luna 20, and especially Apollo 16, are approximately representative. The consistently high-Al 2O 3 composition indicated for the upper crust supports the magmasphere hypothesis. For the trace-element composition of the crust, the highlands meteorites indicate that the central nearside Apollo and Luna sites are in several respects grossly unrepresentative. Concentrations of siderophile elements are far lower in highlands-meteoritic regolith breccias than in their central nearside counterparts. The high overall siderophile levels and hyperchondritic Ni Ir and Au Ir ratios characteristic of highlands materials from Apollo 16 and Apollo 14 are evidently idiosyncracies of the central nearside. Concentrations of incompatible elements, including K, Th, and U, are far lower in the highlands meteorites than in regolith samples from the central nearside. This trend implies that certain lower limits on the bulk-Moon content of U (and associated refractory lithophile elements) should be relaxed. Models of lunar origin implying large enrichments of refractory lithophile elements are not favored by the new constraints from these meteorites.

Petrology and chemistry of hyperferroan anorthosites and other clasts from lunar meteorite ALHA81005

Thirteen previously undescribed clasts from lunar meteorite Allan Hills A81005 include one that is a cataclastic, “hyperferroan” noritic anorthosite. Mafic minerals in this clast (low‐Ca pyroxene with mg′ 52 and olivine with Fo 41) have very low mg compared to most lunar anorthosites described before the discovery of 81005. Siderophile element concentrations are low and indicate that the clast is compositionally pristine. Rare earth elements (REE) are 2.5–10 times more abundant and slightly more fractionated than in previously described ferroan anorthosites. Another probably pristine clast is a granulitic gabbroic norite that has low‐Ca pyroxene with mg′ 53. Hyperferroan anorthosites are a significant component of 81005. Extreme iron enrichment and high REE abundances indicate that these rocks formed late in the evolution of the magma from which the ferroan anorthosite suite formed. Plagioclase was not enriched in Na even at these late stages. Another clast is a cataclastic ferroan anorthosite (Fo 66) that might be pristine. Of the remaining 10 clasts, nine are granular to cataclastic rocks (primarily anorthositic troctolites or troctolitic anorthosites), and two are impact melt breccias, all of which probably represent mixtures of pristine rock types. The former have REE that are essentially unfractionated relative to chondrites and range in abundance from 1.5–3×Cl, and the latter are similar to bulk 81005 (REE 5–8×Cl). Siderophile element abundances are high and indicate the presence of a significant meteoritic component. Our data indicate that 81005 has only a small KREEP (potassium, rare earth elements, phosphorus) component. This is consistent with previous results.

Meteorite ALHA81005: Petrology of a new lunar highland sample

Meteorite ALHA81005 is a shock‐compacted lunar highland regolith breccia. Besides glasses (fragments, spherules, matrix glass), chondrules (of ANT composition), and mineral fragments this new lunar sample contains principally three lithologies: 1) A pristine ferroan anorthosite suite consisting of anorthosites, norites, and one plagioclase peridotite. 2) A high‐Mg granulitic breccia suite representing variable mixtures of troctolitic and noritic precursor rocks. 3) A melt‐rock suite consisting of feld‐spathic basalts and metabasalts of a composition intermediate between (1) and (2). Most glass compositions and the chondrule compositions also project into this group. All lithologies and their derivatives are characteristically poor in alkalis. Neither a KREEP nor a mare basalt composition could be identified with certainty. ALHA81005 apparently comes from a highland area far from any mare basalt and KREEP source.

Rock 67015: A feldspathic fragmental breccia with KREEP‐rich melt clasts

Breccia 67015, one of the feldspathic fragmental breccias (FFBs) from the rim of North Ray Crater, contains a wider range of clast lithologies than has been described previously in FFB's. Our petrologic and INAA studies have revealed two components of major petrologic and geologic significance. One of these is a dark, vesicular, KREEP‐rich melt rock of VHA (very high alumina) basalt composition, which occurs in clasts up to 6 cm long. Melt rocks of this type are common at the Apollo 16 site, but in previous years had not been discovered as prominent constituents of the feldspathic fragmental breccias. Their apparent absence led to hypotheses, no longer valid, that the KREEPy melt rocks were younger than the anorthositic breccia matrices and that the two lithologies might represent Imbrian and Nectaris ejecta, respectively. The second lithology we discovered is a large clast of medium‐grained ferroan anorthositic norite with a composition that falls among the most magnesian members of the ferroan anorthosite suite of pristine cumulate rocks. This material constitutes a previously ‘missing’ member of certain mixing models designed to account for the North Ray Crater melt rocks and bulk breccias.