Plagioclase Flotation Research Articles

Tirhert and Aouinet Legraa: Rare unbrecciated eucrite falls

AbstractTirhert and Aouinet Legraa are the only documented unbrecciated eucrite falls in Africa. Aouinet Legraa fell in Algeria on July 17, 2013. Tirhert's fall occurred about a year later in Morocco, on July 9, 2014. Both meteorites are covered by a black and glossy fusion crust as is typical of eucrites. Tirhert has a poikilitic texture with remnant subophitic pockets, and consists of millimeter‐sized grains of plagioclase (An87‐91), pigeonite (Mg# 42) with augite exsolution lamellae, and interstitial opaque minerals. Aouinet Legraa has a subophitic texture, and it is dominated by plagioclase laths (An82‐89) enclosed by pigeonite (Mg# 37), with exsolution lamellae of augite. Remnant Ca zoning in pyroxene is observed in both rocks, although it is more abundant in Aouinet Legraa than Tirhert. The presence of exsolved pyroxenes suggests that these meteorites have undergone thermal metamorphism. Equilibration temperatures estimated from pigeonite and augite pairs using the QUILF program are ∼931 °C in Tirhert and ∼758 °C in Aouinet Legraa. This indicates that these rocks had distinct thermal histories. Aouinet Legraa has trace element abundances similar to the typical main group eucrite Juvinas, confirming its origin as a main group eucrite. The trace element abundances of Tirhert fall between those of cumulate and main group eucrites. Its rare earth element pattern is flat with a positive Eu anomaly. This likely suggests that Tirhert is a partial cumulate of plagioclase from a main group magma, or a flotation cumulate formed by flotation of plagioclase in a subvolcanic chamber or by scavenging crystals during eruption.

Experimental Crystallization of the Lunar Magma Ocean, Initial Selenotherm and Density Stratification, and Implications for Crust Formation, Overturn and the Bulk Silicate Moon Composition

AbstractEleven isobaric experimental series simulate the fractional crystallization of 1,150 km deep lunar magma ocean. Crystallization begins at 1,850°C with olivine (to 32 per cent solidified, pcs), followed at 1,600°C by olivine + opx ± Cr‐spinel (to 62 pcs), at 1,210°C cpx + plagioclase ± olivine ± Ti‐spinel (to 97 pcs) and at 1,060°C quartz + cpx + plagioclase + Ti‐spinel, leaving 1.8 wt% residual magma that crystallizes minor K‐feldspar and apatite in addition. Melt compositions remain near 45 wt% SiO2, while FeO increases from 11 to 26 wt%, TiO2 peaks at 4 wt% at Ti‐spinel saturation. The available experimental liquid lines of descent yield an overall fractional crystallization sequence of olivine→opx→cpx + plagioclase→quartz→FeTi‐oxide. Plagioclase appears concomitantly with cpx, a result of the low magma ocean floor pressures (≤1 GPa) after 66%–76% of olivine + opx‐fractionation. A few wt% of FeTi‐oxides form mostly once the quartz + plagioclase + cpx‐cotectic is reached, cumulate densities remain ≤3,740 kg/m3. Scaled to a full magma ocean, plagioclase appears at 210–120 km depth, mainly as a function of bulk Al2O3. As buoyancy driven plagioclase‐cpx separation is likely limited, these depths may correspond to the primordial lunar crustal thickness. Allowing for complete plagioclase flotation to the quartz + plagioclase + cpx + FeTi‐oxide ± olivine cotectic yields 95–70 km primordial crust of anorthosite and quartz‐gabbro, far in excess of the 35–50 km observed. This supports an overturn of primordial layers, remelting of dense gabbroic cumulates in the harzburgitic cumulate mantle leading to further mixing and differentiation. We posit that such complex density induced convection led to a lunar marble cake mantle with primitive and fairly evolved reprocessed cumulates next to each other.

Garnet stability in the deep lunar mantle: Constraints on the physics and chemistry of the interior of the Moon

High-pressure, high-temperature experiments have been conducted at deep lunar mantle conditions to constrain the garnet stability field. Using the Taylor Whole Moon composition, garnet is found to be stable at pressures above 3 GPa and temperatures below 1700°C, yielding a smaller stability field than previously suggested on the basis of thermodynamic calculations. Experimental data are used to model equilibrium crystallization in a ‘two-stage’ model of lunar magma ocean (LMO) crystallization starting from a fully molten Moon. In the first stage, isothermal (1600°C) equilibrium crystallization of the LMO would produce garnet-bearing lherzolite cumulates (containing up to ∼20 wt.% garnet) in the lowermost lunar mantle. Garnet in the deep lunar mantle would significantly decrease the Al2O3 content of the residual LMO and impact HREE/MREE fractionation. Numerical modeling of the second stage (residual LMO fractional crystallization) shows a delay in plagioclase saturation compared to models of single-stage fractional crystallization of a whole-Moon LMO of the Taylor Whole Moon composition, thinning the anorthositic crust from 95 km to 75 km. To reach the upper limit of current estimates of the average lunar crustal thickness (∼45 km), the two-stage scenario needs to be accompanied by a total of 10% liquid trapped in cumulates and 70% efficiency of plagioclase flotation. We also conduct trace element evolution modeling and reproduce a REE pattern identical to high K, REE, and P (KREEP) compositions after 99.8% solidification, when starting with a CI chondritic REE abundance. The density of a garnet-bearing deep lunar mantle is significantly higher than the density of olivine/orthopyroxene mixtures without garnet. The present-day lowest mantle in the Moon could therefore be characterized by chemical interactions between the earliest (garnet-bearing) and latest (ilmenite-bearing) products of LMO crystallization.

The Extent, Nature, and Origin of K and Rb Depletions and Isotopic Fractionations in Earth, the Moon, and Other Planetary Bodies

Abstract Moderately volatile elements (MVEs) are depleted and isotopically fractionated in the Moon relative to Earth. To understand how the composition of the Moon was established, we calculate the equilibrium and kinetic isotopic fractionation factors associated with evaporation and condensation processes. We also reassess the levels of depletions of K and Rb in planetary bodies. Highly incompatible element ratios are often assumed to be minimally affected by magmatic processes, but we show that this view is not fully warranted, and we develop approaches to mitigate this issue. The K/U weight ratios of Earth and the Moon are estimated to be 9704 and 2448, respectively. The 87Rb/86Sr atomic ratios of Earth and the Moon are estimated to be 0.072 5 and 0.015 4, respectively. We show that the depletions and heavy isotopic compositions of most MVEs in the Moon are best explained by evaporation in 99%-saturated vapor. At 99% saturation in the protolunar disk, Na and K would have been depleted to levels like those encountered in the Moon on timescales of ∼40–400 days at 3500–4500 K, which agrees with model expectations. In contrast, at the same saturation but a temperature of 1600–1800 K relevant to hydrodynamic escape from the lunar magma ocean, Na and K depletions would have taken 0.1–103 Myr, which far exceeds the 1000 yr time span until plagioclase flotation hinders evaporation from the magma ocean. We conclude that the protolunar disk is a much more likely setting for the depletion of MVEs than the lunar magma ocean.

Lunar feldspathic meteorite Dhofar 081: Petrochemical constraints on petrogenesis

Two alternative petrogenetic models, plagioclase flotation and serial magmatism have been proposed to explain the origin of the lunar anorthositic crust, covering ~80% of the lunar highland. In this study, we re-examine the possible relict igneous texture present in an inferred lunar highland breccia clast (area ~1 mm2) in the Dhofar 081 meteorite. Our new petrographic and in-situ mineral microprobe chemical data on this clast show this coarse grained (average grain size ~0.5 mm) clast preserves relict igneous texture where subhedral, prismatic low-Ca pyroxene has intergrown with anhedral anorthitic plagioclase, suggesting its eutectic crystallization from its parent silicate magma. Absence of maskelynite and similarity of Na, K contents of plagioclase with the FAN assemblages negate the possibility of crystallization of the studied relict clast from an impact melt. The mineral-chemical data of Dhofar 081 suggest it is FAN (Ferroan anorthosite) in composition (after Warren in Annu. Rev. Earth Planet. Sci. 13:201–240, 1985). Hence, intergrown crystallization of minerals in the present relict igneous clasts and other reported FAN samples argues against a cumulate origin of the lunar anorthosite. The orthopyroxenes present in the unbrecciated portion of this meteoritic clast include bimodal low- and high-iron geochemical sub-groups. The application of orthopyroxene and plagioclase thermobarometry (after Gasparik in Contrib. Mineral. Petrol. 96:357–370, 1987) on our new microprobe data, and also two-pyroxene thermometry (after Lindsley in Am. Mineral. 68:477–493, 1983; Putirka in Rev. Mineral. Geochem. 69(1):61–120, 2008) on our new microprobe data and synthesis of literature data constrain the pressure and temperature of crystallization of lunar anorthosite parent magma close to 8 kbar and 1050°C, respectively. Application of Fo–An–Q experimental phase diagram at high pressure (up to 20 kbar) negates the possibility of generation of lunar anorthosite from a lherzolite source, the parent magma of these anorthosites probably lie on or close to Fo–An join of this phase diagram close to the spinel field.

Effect of crustal porosity on lunar magma ocean solidification

The lunar ferroan anorthosites, formed by plagioclase flotation from the crystallization of the lunar magma ocean, have an age span of over ~ 200 Ma. However, previous thermal models predicted a much shorter time range. We propose that a much smaller thermal conductivity of anorthositic crust due to its high porosity may have delayed the solidification of the lunar magma ocean. Our thermal simulation results, using the thermal conductivity of porous lunar crust, show that crystallization of a 1000 km deep magma ocean could be prolonged to tens of millions of years, and up to 180 Ma under some extreme conditions. The porous crust alone can’t explain the large crustal age span, however. Other circumstances must be taken into consideration, such as a thick lunar soil.

Evidence for anorthositic crust formed on an inner solar system planetesimal

doi: 10.7185/geochemlet.1921 During the first million years of solar system history, planetesimals experienced extensive melting powered by the radioactive decay of 26 Al (Lee et al., 1977). To date, the only known anorthositic crust on a solar system body is that of the Moon, formed by plagioclase flotation on top of the magma ocean (Wood et al., 1970). Here we show evidence from the ungrouped achondrite meteorite Northwest Africa (NWA) 8486 that an anorthositic crust formed on a planetes-imal very early in solar system history (<1.7 Ma). NWA 8486 displays the highest anomalies in Eu and Sr found in achondrites so far and, for the first time, this characteristic is also identified in clinopyroxene. Elemental modelling , together with calculated timescales for crystal settling, show that only the melting of an anorthosite can produce NWA 8486 within the first 5 million years of solar system history. Our results indicate that such a differentiation scenario was achievable over short timescales within the inner solar system, and must have contributed to the making and elemental budget of the terrestrial planets.

Crystallization of the lunar magma ocean and the primordial mantle-crust differentiation of the Moon

We present crystallization experiments on silicate melt compositions related to the lunar magma ocean (LMO) and its evolution with cooling. Our approach aims at constraining the primordial internal differentiation of the Moon into mantle and crust. We used graphite capsules in piston cylinder (1.35–0.80 GPa) and internally-heated pressure vessels (<0.50 GPa), over 1580–1020 °C, and produced melt compositions using a stepwise approach that reproduces fractional crystallization. Using our new experimental dataset, we define phase equilibria and equations predicting the saturation of liquidus phases, magma temperature, and crystal/melt partitioning for major elements relevant for the crystallization of the LMO. These empirical expressions are then used in a forward model that predicts the liquid line of descent and crystallization products of a 600 km-thick magma ocean. Our results show that the effects of changes in the bulk composition on the sequence of crystallization are minor. Our experiments also show the crystallization of a silica phase at ca. 1080 °C and we suggest that this phase might have contributed to the building of the lower anorthositic crust. Calculation of crustal thickness clearly shows that a thin crust similar to that revealed by GRAIL cannot have been generated through solidification of whole Moon magma ocean. We discuss the role of magma ocean depth, trapped liquid fraction (with implication for the alumina budget in the mantle and the crust), and the efficiency of plagioclase flotation in producing the thin crust. We also constrain the potential range of pyroxene compositions that could be incorporated into the crust and show that delayed crustal building during ca. 4% LMO crystallization on the nearside of the Moon may explain the dichotomy for Mg-number. Finally, we show that the LMO can produce magnesian anorthosites during the first stages of plagioclase crystallization.

Fractional crystallization of the lunar magma ocean: Updating the dominant paradigm

AbstractWe report results of systematic experimental simulation of fractional crystallization of a lunar magma ocean (LMO) with the Lunar Primitive Upper Mantle bulk composition. These results complement prior work that simulated equilibrium crystallization. In contrast to previous numerical models for investigating magma ocean solidification processes and implications, our combined program simulates these processes directly using petrologic experimentation. Our experiments mimic LMO crystallization that is fractional throughout the process, rather than switching from initially equilibrium to fractional crystallization partway through. To do this, we adopted an iterative approach in which the starting material for each run is synthesized using the composition of the melt phase from the prior run. We compare our results to those from long‐standing numerical models of LMO crystallization and show that although some features of those models are broadly reproduced, there are key differences in liquid lines of descent and the cumulate lithologies generated. Our results can be used to estimate the possible thickness of a primordial lunar crust formed from flotation of plagioclase during magma ocean solidification. Our estimate is greater than that from the recent Gravity Recovery and Interior Laboratory (GRAIL) mission, but consistent with the criteria on which the starting bulk composition was originally calculated. It assumes perfectly efficient separation of all plagioclase formed from the crystallizing magma ocean, which is likely not the case. We also demonstrate that a non‐chondritic bulk composition, with respect to trace elements, is not required in order to generate a KREEP (potassium, rare earth elements, and phosphorus) signature from magma ocean crystallization.

Anorthosites from an Archean continental arc in the Dharwar Craton, southern India: Implications for terrane assembly and cratonization

Anorthositic rocks have attracted attention in terms of their possible role as the primordial crust of our planet, and also as markers of Archean-Paleoproterozoic plate tectonic regimes. Here we investigate the Konkanhundi gabbro-anorthosite suite from Peninsular India, adjacent to the major collisional suture between the Western and Eastern Dharwar cratonic blocks. The anorthosites display high Al2O3 contents with negative correlation against MgO, high CaO and Sr contents, and positive Eu anomalies indicating plagioclase flotation in the magma chamber. The pyroxenite and gabbroic rocks of the suite show high Ni, Cr, Co contents and geochemical features typical of pyroxene cumulation. LA-ICPMS U-Pb analyses of magmatic zircon grains yield weighted mean 207Pb/206Pb ages of 2601 ± 12 Ma for pyroxenite, 2616 ± 12 Ma for gabbro, 2615 ± 13 Ma and 2627 ± 14 Ma for anorthositic gabbro, 2594 ± 16 Ma for anorthosite, and 2605 ± 27 Ma for microgabbro. The age data from the different rock types in the anorthosite suite are broadly consistent, marking the emplacement time as ca. 2.6 Ga, and providing insights on one of the oldest anorthosite complexes in Peninsular India. Zircon grains in the surrounding TTG gneisses into which the gabbro-anorthosite suite was emplaced show weighted mean 207Pb/206Pb age of 3321 ± 11 Ma suggesting Mesoarchean basement. The zircon grains display high Th/U values and REE patterns typical of magmatic crystallization with enriched HREE, positive Ce and Sm anomalies, and negative Pr, Nd, Eu anomalies. Zircon Lu-Hf analysis yield negative εHf(t) values of −4.9–−0.7 with TDMC (two stage model ages) of 3123–3376 Ma for pyroxenite, −4.5–−1.7 and 3192–3366 Ma for gabbro, −5.3–0.9 and 3034–3410 Ma for anorthositic gabbro, and −3.9–-2.4 and 3221–3311 Ma respectively, suggesting the incorporation of Mesoarchean components within the Neoarchean magmatic suite. However, zircon grains in the TTG gneiss possess more ‘juvenile’ εHf(t) values in the range of −0.2–1.3 with TDMC in the range of 3554–3646 Ma. The zircon Hf isotopes and trace element data, together with the whole rock geochemical features suggest that the parent magma of the Konkanhundi gabbro-anorthosite suite was derived from subduction-related depleted mantle source that also incorporated continental crustal components. The time of emplacement of the gabbro-anorthosite complex broadly correlates with widespread arc magmatism and crust production as well recycling in other domains of the Dharwar Craton. We envisage multiple convergence of microblocks during craton assembly at the end of Archean in Dharwar, with the greenstone belts marking zones of paleo-ocean closure.

The mafic component of the lunar crust: Constraints on the crustal abundance of mantle and intrusive rock, and the mineralogy of lunar anorthosites

Most models of early lunar evolution predict that the anorthositic highlands crust is the result of plagioclase flotation on a magma ocean. However, the lunar highlands crust typically contains 4 wt% FeO and so is more mafic than the strict definition of the anorthosites thought to comprise it. We used new Clementine-based mineral maps of the Moon as inputs to a series of mixing models that calculate the abundance and distribution of major highland rock types and shed light on three possible sources of excess mafic material in the lunar highlands: mafic (15 vol% mafic minerals) anorthosites, post-magma ocean igneous activity, and mafic basin ejecta. Mixing models that feature pure anorthosites like the purest anorthosite (PAN) described by Ohtake et al. (2009) and Pieters et al. (2009) are most compatible with the data. They allow us to place an upper limit of 10–20 vol% mantle material that could be mixed with the primary highlands crust. The upper limit on mantle material indicated by the mixing models is significantly lower than the 30–40 vol% mantle material expected from simple geometric calculations of the major lunar basins’ excavation cavities based on an excavation cavity depth/diameter ratio of 1/10; this discrepancy allows us to conclude that the excavation cavities of the three largest lunar basins may have been significantly shallower than those of the smaller basins. Our results are consistent with excavation cavity depth/diameter ratios for these largest basins in the range of 0.035 to 0.06, which agrees with previous gravity measurements by Wieczorek and Phillips (1999).

Estimation of trace element concentrations in the lunar magma ocean using mineral‐ and metal‐silicate melt partition coefficients

AbstractThis study uses experimentally determined plagioclase‐melt D values to estimate the trace element concentrations of Sr, Hf, Ga, W, Mo, Ru, Pd, Au, Ni, and Co in a crystallizing lunar magma ocean at the point of plagioclase flotation. Similarly, experimentally determined metal‐silicate partition experiments combined with a composition model for the Moon are used to constrain the concentrations of W, Mo, Ru, Pd, Au, Ni, and Co in the lunar magma ocean at the time of core formation. The metal‐silicate derived lunar mantle estimates are generally consistent with previous estimates for the concentration of these elements in the lunar mantle. Plagioclase‐melt derived concentrations for Sr, Ga, Ru, Pd, Au, Ni, and Co are also consistent with prior estimates. Estimates for Hf, W, and Mo, however, are higher. These elements may be concentrated in the residual liquid during fractional crystallization due to their incompatibility. Alternatively, the apparent enrichment could reflect the inappropriate use of bulk anorthosite data, rather than data for plagioclase separates.

Sm–Nd systematics of lunar ferroan anorthositic suite rocks: Constraints on lunar crust formation

We have measured Sm–Nd systematics, including the short-lived 146Sm–142Nd chronometer, in lunar ferroan anorthositic suite (FAS) whole rocks (15415, 62236, 62255, 65315, 60025). At least some members of the suite are thought to be primary crystallization products formed by plagioclase flotation during crystallization of the lunar magma ocean (LMO). Most of these samples, except 62236, have not been exposed to galactic cosmic rays for a long period and thus require minimal correction to their 142Nd isotope composition. These samples all have measured deficits in 142Nd relative to the JNdi-1 terrestrial standard in the range −45 to −21ppm. The range is −45 to −15ppm once the 62236 142Nd/144Nd ratio is corrected from neutron-capture effects. Analyzed FAS samples do not define a single isochron in either 146Sm–142Nd or 147Sm–143Nd systematics, suggesting that they either do not have the same crystallization age, come from different sources, or have suffered isotopic disturbance. Because the age is not known for some samples, we explore the implications of their initial isotopic compositions for crystallization ages in the first 400Ma of solar system history, a timing interval that covers all the ages determined for the ferroan anorthositic suite whole rocks as well as different estimates for the crystallization of the LMO. 62255 has the largest deficit in initial 142Nd and does not appear to have followed the same differentiation path as the other FAS samples. The large deficit in 142Nd of FAN 62255 may suggest a crystallization age around 60–125Ma after the beginning of solar system accretion. This result provides essential information about the age of the giant impact forming the Moon. The initial Nd isotopic compositions of FAS samples can be matched either with a bulk-Moon with chondritic Sm/Nd ratio but enstatite-chondrite-like initial 142Nd/144Nd (e.g. 10ppm below modern terrestrial), or a bulk-Moon with superchondritic Sm/Nd ratio and initial 142Nd/144Nd similar to ordinary chondrites.

Implications for the origins of pure anorthosites found in the feldspathic lunar meteorites, Dhofar 489 group

Remote observation by the reflectance spectrometers onboard the Japanese lunar explorer Kaguya (SELENE) showed the purest anorthosite (PAN) spots (>98% plagioclase) at some large craters. Mineralogical and petrologic investigations on the feldspathic lunar meteorites, Dhofar 489 and Dhofar 911, revealed the presence of several pure anorthosite clasts. A comparison with Apollo nearside samples of ferroan anorthosite (FAN) indicated that of the FAN samples returned by the Apollo missions, sample 60015 is the largest anorthosite with the highest plagioclase abundance and homogeneous mafic mineral compositions. These pure anorthosites (>98% plagioclase) have large chemical variations in Mg number (Mg# = molar 100 × Mg/(Mg + Fe)) of each coexisting mafic mineral. The variations imply that these pure anorthosites underwent complex formation processes and were not formed by simple flotation of plagioclase. The lunar highland samples with pure anorthosite and the PAN observed by Kaguya suggest that pure anorthosite is widely distributed as lunar crust lithology over the entire Moon.

Magma emplacement at anomalous spreading ridge: Constraints due to plagioclase crystals from basalts of Marsili seamount (Southern Tyrrhenian back-arc)

Plagioclase crystals from four basalt samples dredged from different portions of the Marsili seamount (Southern Tyrrhenian, Italy) provide a detailed record of the changes occurring in the surrounding magma during its ascent in the crust. The textural features and chemical zoning (i.e., An content and Fe, Mg, Sr and Ba concentrations) of the plagioclases from each sample show that aggregation of various generations of crystals occurred in the host melt prior to eruption. Plagioclases resulting from crystallisation in small sill intrusions, where local flotation of plagioclase may have occurred, are ubiquitous. Instead, large (up to 1mm) homogeneous An-rich plagioclases are rare and have only been found in two of the studied basalts, recovered from the northern and southern portions of the volcano, respectively. These crystals record crystallisation events occurring long before eruption, and probably derive from deep-seated crystalline material. At the highest An values, the Sr and Ba concentrations of the studied plagioclases also indicate the existence of two distinct mafic magmas within the lower crustal plumbing system of the Marsili volcano, one with lower Sr and Ba contents than the other. Three of the four studied basalts came from magma with low Sr and Ba, whereas the fourth, which erupted from the southern portion of the volcano, sampled magma richer in Sr and Ba. The ascent of magma caused resorption of the previously formed plagioclase crystals and regrowth of normally or inversely zoned portions, depending on the water-saturated or water-undersaturated conditions of the surrounding magma, respectively. Small shallow magmatic reservoirs may be located under the southern part of the Marsili volcano, as testified by the resorbed sodic cores (An52–57) found in the two basalt samples from this portion.

Anorthosite formation by plagioclase flotation in ferrobasalt and implications for the lunar crust

The Sept Iles layered intrusion (Quebec, Canada) is dominated by a basal Layered Series made up of troctolites and gabbros, and by anorthosites occurring (1) at the roof of the magma chamber (100–500 m-thick) and (2) as cm- to m-size blocks in gabbros of the Layered Series. Anorthosite rocks are made up of plagioclase, with minor clinopyroxene, olivine and Fe–Ti oxide minerals. Plagioclase displays a very restricted range of compositions for major elements (An 68–An 60), trace elements (Sr: 1023–1071 ppm; Ba: 132–172 ppm) and Sr isotopic ratios ( 87Sr/ 86Sr i : 0.70356–0.70379). This compositional range is identical to that observed in troctolites, the most primitive cumulates of the Layered Series, whereas plagioclase in layered gabbros is more evolved (An 60–An 38). The origin of Sept Iles anorthosites has been investigated by calculating the density of plagioclase and that of the evolving melts. The density of the FeO-rich tholeiitic basalt parent magma first increased from 2.70 to 2.75 g/cm 3 during early fractionation of troctolites and then decreased continuously to 2.16 g/cm 3 with fractionation of Fe–Ti oxide-bearing gabbros. Plagioclase (An 69–An 60) was initially positively buoyant and partly accumulated at the top of the magma chamber to form the roof anorthosite. With further differentiation, plagioclase (<An 60) became negatively buoyant and anorthosite stopped forming. Blocks of anorthosite (autoliths) even fell downward to the basal cumulate pile. The presence of positively buoyant plagioclase in basal troctolites is explained by the low efficiency of plagioclase flotation due to crystallization at the floor and/or minor plagioclase nucleation within the main magma body. Dense mafic minerals of the roof anorthosite are shown to have crystallized from the interstitial liquid. The processes related to floating and sinking of plagioclase in a large and shallow layered intrusion serve as a proxy to refine the crystallization model of the lunar magma ocean and explain the vertically stratified structure of the lunar crust, with (gabbro-)noritic rocks at the base and anorthositic rocks at the top. We propose that the lunar crust mainly crystallized bottom-up. This basal crystallization formed a mafic lower crust that might have a geochemical signature similar to the magnesian-suite without KREEP contamination, while flotation of some plagioclase grains produced ferroan anorthosites in the upper crust.

The lunar magma ocean: Reconciling the solidification process with lunar petrology and geochronology

The Moon is thought to have originated with a magma ocean that produced a plagioclase flotation crust as solidification proceeded. Ages of anorthositic crust range over at least 200 million years. The model for solidification presented here integrates chemical and physical constraints of lunar magma ocean solidification to determine (1) the final thickness of flotation crust generated by a fractionally solidifying magma ocean, (2) the timescale of crystallization before plagioclase is a stable phase, (3) the timescale of solidification after the formation of the plagioclase flotation crust, and (4) the post-overturn lunar mantle composition and structure. We find that magma oceans of as much as 1000 km depth are consistent with creating an anorthositic crust 40 to 50 km in thickness. Solidification of the magma ocean prior to formation of the flotation crust may occur on the order of 1000 years, and complete solidification would require additional ten to tens of millions of years. Reconciling these short model timescales with radiometric dates of crustal samples requires either a very late-forming Moon combined with finding older crustal ages to be incorrect, or calling on tidal heating of the crust by the early Earth to prolong solidification. Gravitationally driven overturn of cumulates during tidal heating provides a mechanism for creating the compositions and ages of the lunar Mg suite of crustal rocks. Further, we find that upon crystallization, the Moon likely began with an azimuthally heterogeneous, gravitationally stable mantle, after magma ocean cumulate overturn. This result may help explain the experimentally determined origin of picritic glasses at similar depths but from different source materials.

Origin of Anorthosite and Magnetitite Layers in the Bushveld Complex, Constrained by Major Element Compositions of Plagioclase

The Bushveld Complex, a layered mafic intrusion in South Africa, shows extreme vertical differentiation in terms of mineral compositions and modal proportions from dunite to ferrodiorite. In a continuous borehole core drilled through the uppermost 2� 8k m of the intrusion, typical rocks range upwards from troctolite, through gabbronorite and ferrogabbronorite to ferrodiorite, with extreme examples of anorthosite, magnetitite and feldspathic pyroxenite. The An content of plagioclase has previously been determined for 420 samples and decreases upward from An78 to An36, with six minor, slow reversals. Variations in modal proportions of plagioclase have been calculated based on 2200 density determinations on whole-rocks. Forty-five anorthosite layers have been identified, ranging from 1 to 23 m thick. None of these layers is associated with the abovementioned reversals in An content in plagioclase and nearly all have leucocratic rocks below and above, with more than the likely cotectic proportions of plagioclase. These observations argue against an origin for anorthosite related to magma addition or to supersaturation and oscillatory nucleation. Rhythmically pulsed crystallization, possibly associated with pressure changes, followed by crystal settling and sorting of minerals of different densities is a hypothesis consistent with all the observations. Twenty layers of magnetitite have been identified. There is a significant reversal in An content in the overlying plagioclase compared with the underlying sample across only one such layer. Again, this observation challenges hypotheses that such layers result from magma addition, but is consistent with a pressure-change hypothesis for triggering magnetite crystallization. The upper contacts of magnetitite layers that grade into anorthosite over many centimetres possibly also reflect settling and sorting. Rocks forming the uppermost 100 m of the intrusion contain the most sodic plagioclase compositions, demonstrating that there is no downward crystallizing roof facies. Furthermore, this uppermost 100 m section is depleted in plagioclase relative to its cotectic proportions. Hence, we find no evidence supporting flotation or prolonged suspension of plagioclase.

Compositions of Mercury's earliest crust from magma ocean models

The size of the Mercurian core and the low ferrous iron bearing silicate content of its crust offer constraints on formation models for the planet. Here we consider a bulk composition that allows endogenous formation of the planet's large core, and by processing the mantle through a magma ocean, would produce a low-iron oxide crust consistent with observations. More Earth-like bulk compositions require silicate removal, perhaps by a giant impact, to create the planet's large core fraction. We find that the endogenous model can produce a large core with either a plagioclase flotation crust or a low-iron oxide magmatic crust. Because a magma ocean creates a gradient in iron oxide content in the resulting planetary mantle, the parts of the mantle removed by a putative giant impact could result in either a high-iron oxide mantle in contradiction to current crustal measurements, or a low-iron oxide mantle consistent with the current understanding of Mercury. If a giant impact cannot preferentially remove shallow mantle material then the proto-Mercury must have had a bulk low iron-oxide composition. Thus a specific bulk composition is required to make Mercury endogenously, and either a specific process or a specific composition is required to make it exogenously through giant impact. Measurements taken by the MESSENGER mission, when compared to predictions given here, may help resolve Mercury's formation process.

Paired melt lenses at the East Pacific Rise and the pattern of melt flow through the gabbroic layer at a fast-spreading ridge

From geophysics, two thin and narrow sill-like melt lenses occur directly beneath the ridge axis on the East Pacific Rise (EPR) near 9°N. One is at the base of sheeted dikes, and the other at the base of the crust in the mantle transition with largely crystalline material in between. Along-axis continuity indicates that the upper one is a steady-state feature along much of the EPR.What causes melt to pool at two levels in the crust, with little melt in between? Normative densities of gabbro sampled from Hess Deep, eastern Pacific, at solidus temperatures are greater than melt densities calculated from compositions of basalt glass from the EPR. Therefore, abyssal-tholeiite liquids of all compositions are buoyant in a hot gabbroic crystalline matrix with no more than a few percent melt. There is therefore no level in the gabbroic layer at which pure basaltic melt attains neutral buoyancy, and by this means produces sill-like rather than dike-like intrusives. Instead, the upper melt lens appears to collect at a combined compositional and permeability barrier, defined by the coarse-grained base of the dike sequence sampled on the northern rift margin at Hess Deep. About 200m of massive dike rocks with ophitic to subophitic coarse-grained texture occur just above gabbronorite and oxide–gabbro cumulates which are the crystallized relics of highly differentiated ferrobasaltic and ferroandesitic liquids in the melt lens. At temperatures of crystallization of cumulates at the level of the melt lens (1100–1000°C), the basal dike rocks of less differentiated basaltic composition just above this were 90–98% crystalline, and they annealed to the point where dike contacts cannot be seen on outcrop surfaces. These slightly molten rocks need to be fractured by rift dilatancy or overpressure from beneath for magmas to ascend beyond the melt lens. Neutral buoyancy occurs in fine-grained dikes above where densities of escaping magma finally match those of the fractured and porous formation and this is the level that can be laterally intruded.The base of the crust, on the other hand, presents an abrupt reduction in the density contrast between ascending melt and surrounding rock, and a higher concentration of melt pockets and channels. Neutral buoyancy can be attained where primitive gabbros are first encountered if the ascending magmas are picritic, laden with olivine phenocrysts. High olivine concentrations in the narrow bodies of moving magma can be attained by a combination of sinking, relative plagioclase flotation, and flow differentiation. Sill-like intrusions that crystallize to dunite and plagioclase–peridotite cumulates will result. Such rocks are found but their structural orientation is unknown at the base of the intra-rift ridge at Hess Deep; sill-like analogs are found in some ophiolites. Lateral leakage of deep melt lenses into transform faults might account for both primitive picritic tholeiites and strongly plagioclase-phyric basalts found, for example, at Siqueiros Fracture Zone at 8°30′N along the EPR.