Early lunar differentiation: 4.42-AE-old plagioclase clasts in Apollo 16 breccia 67435

We report on a⁴⁰Ar-³⁹Ar study of the Apollo 16 breccia 67435 and present ages of five samples representing matrix, lithic clasts and plagioclase clasts. While the matrix age spectrum does not have a well-defined plateau, the two lithic clasts gave plateau ages of 3.96 and 4.04 AE. Since all samples had apparent ages of ～1 AE in the fractions ≤600°C extraction temperature, the breccia might have been assembled in a rather mild process at about that time or even more recently out of material with different metamorphic ages. The two plagioclase samples, of which one was a single 9-mg mineral clast and the other a 15-mg composite of several clasts, also have ages of ～1 AE in the low-temperature release fractions, but are apparently undisturbed by any ～4-AE events since they both have well-defined plateaux at 4.42 AE. The age of these strongly calcic plagioclase clasts, believed to be remnants of the anorthositic lunar crust, establishes a lower age limit to the end of the early lunar differentiation and thus places a strong constraint to the lunar evolution.     

Mineralogy and stable isotope geochemistry of hydrothermally altered oceanic rocks

Data for the diffusion of cations in pyroxenes are relevant to a variety of sub-solidus processes including order-disorder and exsolution. Similar data must also be available if the reliability of geobarometers and geothermometers involving pyroxenes is to be assessed. Two types of diffusion experiment have been performed to determine cation diffusion rates in pyroxenes: (1) interdiffusion between single crystals of diopside and polycrystalline sinters enriched in Al and Fe, and (2) interdiffusion between single crystals of diopside and a glass of the same composition which was isotopically enriched in²⁶Mg and⁴³Ca. Following high-temperature annealing for periods up to several hundred hours, analysis of the diffusion couples, using an electron microprobe and an ion microprobe respectively, failed to show any measurable diffusion profiles. From these “null result” experiments the diffusion coefficients (D) for Al and Fe in diopside are estimated to be less than4×10^?14cm²s^?1 at 1200°C, and values ofD for Ca and Mg in diopside are estimated to be less than7 × 10^?14cm²s^?1 at 1250°C. These rates are significantly slower than published tracer-type diffusion data for Ca and Al.A review of studies of order-disorder, microstructural coarsening, and diffusion in pyroxenes suggest that activation energies for cation exchange are typically in excess of 60 kcal mol^?1. Transport rates will be assisted, and activation energies lowered by sample non-stoichiometry, inhomogeneities, high dislocation densities and the presence of water.The collective data for Al, Mg and Ca diffusion in diopside indicate diffusion coefficients? 10^?15cm²s^?1 at 1200°C. A comparison with data for diffusion in garnet, olivine and spinel suggests that pyroxenes may have the highest blocking temperatures.     

Cooling history of pyroxene chondrules in the Yamato-74191 chondrite (L3)—an electron microscopic study

Fine textures of clinopyroxene in an excentroradial pyroxene chondrule (EPC) and a comb-like pyroxene chondrule (CPC) in the Yamato-74191 chondrite (L3) have been studied by analytical electron microscopy. Both pyroxenes consist of three regions different in composition and texture; core, mantle and marginal regions, though the pyroxenes of the CPC are more Fe-rich than those of the EPC. The core region is the most Mg-rich with no Ca component and commonly shows polysynthetic (100) twins. The mantle region is slightly calcic, and the marginal region shows a rapid increase of Ca outward.The polysynthetic twins, cracks and subgrain boundaries in the core in the EPC and CPC must have formed during the transition from proto-type to clino-type pyroxenes. The exsolution textures in the mantle and marginal regions indicate initial crystallization of pigeonite-C followed by decomposition into pigeonite-P and augite. The decomposition must have taken place by nucleation growth in the mantle region and by spinodal decomposition in the marginal region. The periodicity of 15–20 nm in the spinodal decomposition textures indicates that the cooling rate of the pyroxenes, when passing through about 1000°C, was of the order of a few tens to several degrees centigrade per hour. The cooling history of the chondrules has been explained by a monotonous cooling controlled by the cooling rate of the surrounding medium.     

Cation distribution in low-calcium pyroxenes: Dependence on temperature and calcium content and the thermal history of lunar and terrestrial pigeonites

Four pyroxenes with compositions En₄₈Fs₄₈Wo₄, En_47·5Fs_47·5Wo₅, En₄₅Fs₄₅Wo₁₀ and En₄₀Fs₄₀Wo₂₀, synthesized at 1200°C at atmospheric pressure, were heat-treated at 500, 600, 700, and 800°C for various lengths of time. These pyroxenes are variously ordered with respect to Fe²⁺ and Mg²⁺ without unmixing. The Fe²⁺-Mg²⁺ distribution over the two nonequivalent sites M1 and M2, determined through Mössbauer spectroscopy, is found to be a function of both temperature and concentration of Ca²⁺ at the M2 site. The preference of Fe²⁺ for the M2 site increases with decreasing temperature and increasing Ca²⁺. These data can be used to determine cation equilibration temperatures of lunar and terrestrial pigeonites. The lunar pigeonites usually indicate equilibration temperatures of 700–860°C, except the pigeonite from rock 14053, which may have been subjected to shock heating due to meteoritic impact.     

Petrology of the Green Knobs diatreme and implications for the upper mantle below the Colorado Plateau

Peridotite inclusions, crystal fragments, and kimberlite breccia at Green Knobs, New Mexico, have been studied to evaluate compositions and processes in the upper mantle below the Colorado Plateau. Most peridotite inclusions are spinel lherzolites and harzburgites, or their partly hydrated equivalents, in the Cr-diopside group. Orthopyroxene-rich websterites and olivine websterites comprise 3% of the peridotites and formed as cumulates. Typical anhydrous or slightly hydrated peridotites contain aluminous, calcic diopside (5–7% Al₂O₃), aluminous orthopyroxene (3–6% Al₂O₃), spinel, and olivine (near Fa₉). Geothermometers based on different mineral pairs yield temperatures from above 1100°C to below 700°C in single rocks. High values, derived from pyroxenes with included exsolution lamellae, may approximate temperatures of primary crystallization. Low values, based on olivine-spinel and olivine-clinopyroxene pairs, approach upper mantle temperatures before eruption. In rare samples, some spinel grains are rimmed by garnet while others are not rimmed; garnet formation was controlled by nucleation kinetics. About one-third of the peridotites were deformed shortly before eruption, with effects ranging from mild cataclasis to the production of ultramylonites.Discrete crystals of garnet, olivine (near Fa₈), and Cr-diopside represent garnet peridotite. Eclogites were not found. The garnet peridotite is more depleted than overlying spinel peridotite, and it is not a likely source for the minettes associated with the kimberlites.The mantle below Green Knobs consists of spinel peridotite from 45 to perhaps 60 km depth immediately underlain by more-depleted garnet peridotite. The position of the spinel-garnet transition may be fixed by kinetics. The kimberlite may have been produced when heat from ascending minette magma released volatiles from otherwise depleted garnet peridotite. Resulting gas-solid mixtures erupted along zones of deformation associated with Colorado Plateau monoclines. Sheared lherzolites formed during renewed movement along these zones.     

Nitrogen contents and isotopic ratios of clasts from the enstatite chondrite Abee

Nitrogen contents and isotopic ratios have been determined for three clasts from the enstatite chondrite Abee by stepwise heating. The clasts possess a wide range in nitrogen content, ranging from 254 to 850 ppm, whereas the nitrogen isotopic ratios are nearly identical atδ¹⁵N= ?29.2±0.6‰. A refractory inorganic nitrogen-bearing phase contains about 90% of the nitrogen which is released at temperatures of 1000°C and above. The stepwise heating experiments suggest the possible existence of two other distinct nitrogen components, released at low (770°C) and high (1500°C) temperatures.     

Plagioclase lamellae in hypersthene,Tikkoatokhakh Bay,Labrador

Abundant lamellae of plagioclase are present in the (100) planes of hypersthene megacrysts in andesine anorthosite along Tikkoatokhakh Bay, northwest of Nain, Labrador. Spongy intergrowths of plagioclase in hypersthene also occur. Plagioclase lamellae have mean compositions ranging from An₄₃ to An₉₂, with extreme compositions from An₃₉ to An₉₇; the calcic compositions are the more abundant. Such lamellae are always accompanied in the hypersthene by grains or lamellar segments of magnetite, and rarely by lamellae of olivine, augite, magnetite, or ilmenite. Some calcic plagioclase lamellae contain antiperthitic spindles of orthoclase. The host rocks of the hypersthene megacrysts are layered leuconorites and anorthosites with mean plagioclase compositions ranging from An₄₁ to An₅₅. The plagioclase lamellae in hypersthene are characteristically much more calcic than the host-rock plagioclase. There is little doubt that the lamellae exsolved from a pyroxene host, dominantly by a coupled redox reaction which generated magnetite, thereby releasing silica to combine with the Ca-Tschermak and jadeite components of the precursor pyroxene. Rapid growth of megacrysts may account for their aluminous nature.     

Isotopic and REE studies of lunar basalt 12038: implications for petrogenesis of aluminous mare basalts

We report Sr, Nd, and Sm isotopic studies of lunar basalt 12038, one of the so-called aluminous mare basalts. A precise internal Rb-Sr isochron yields a crystallization age of 3.35±0.09 AE and initial⁸⁷Sr/⁸⁶Sr=0.69922?2 (2σ error limits, 1AE=10⁹ years, λ(⁸⁷Rb)=0.0139AE^?1). An internal Sm-Nd isochron yields an age of 3.28±0.23AE and initial¹⁴³Nd/¹⁴⁴Nd=0.50764?28. Present-day¹⁴³Nd/¹⁴⁴Nd is less than the “chondritic” value, i.e. ?(Nd, 0)=?2.3±0.4 where ?(Nd) is the deviation of¹⁴³Nd/¹⁴⁴Nd from chondritic evolution, expressed as parts in 10⁴. At the time of crystallization ?(Nd, 3.2AE)=1.5±0.6.We have successfully modeled the evolution of the Sr and Nd isotopic compositions and the REE abundances within the framework of our earlier model for Apollo 12 olivine-pigeonite and ilmenite basalts. The isotopic and trace element features of 12038 can be modeled as produced by partial melting of a cumulate mantle source which crystallized from a lunar magma ocean with a chondrite-normalized REE pattern of constant negative slope. Chondrite-normalized La/Yb=2.2 for this hypothetical magma ocean pattern. A plot of I(Sr) versus ?(Nd) for the Apollo 12 basalts clearly shows the influence of varying proportions of olivine, clinopyroxene, orthopyroxene, and plagioclase in the basalt source regions. A small percentage of plagioclase (～5%) in the 12038 source apparently is responsible for low I(Sr) and ?(Nd) in this basalt. Aluminous mare basalts from Mare Crisium (Luna 24) and by inference Mare Fecunditatis (Luna 16) occupy locations on the I(Sr)-?(Nd) plot similar to that of 12038, implying that some basalts from three widely separated lunar regions came from plagioclase-bearing source regions. A summary of model calculations for mare basalts shows a record of lunar mantle solidification during the period when REE abundances in the lunar magma ocean increased from ～20× chondritic to >100× chondritic. Although there is a general trend from olivine to clinopyroxene-dominated source regions with progressive magma ocean evolution, significant mineralogical heterogeneities in mantle composition apparently formed at any given stage of evolution, as evidenced in particular by the three Apollo 12 magma types.     

Surface alteration of basalt due to cation-migration

Basaltic lava from Kilauea, Hawaii may have a red-brown surface, indicative of Fe-(hydr)oxides. This surface is not found where exposed to weathering, but at the interface between lava lobes, or in the interior of lava channels. We use several analytical techniques to determine how these Fe-(hydr)oxide surfaces may have developed. WDS-elemental distribution line profiles from the lava surface towards the lava′s interior detect an Fe-rich film of less than 5 μm thickness. Heat treatment of quenched, fresh lava samples of the same chemical composition between 600–1,090°C helps to replicate temperatures under which such an Fe-rich film might have formed. These experiments suggest that Fe-enrichment occurs above 1,020°C, whereas at lower temperatures Ca is enriched relative to Fe. One sample was treated below the glass transition temperature, at 600°C for 164 h. A depth profile with secondary neutral mass spectrometry shows an enrichment of Mg at the outer 50 nm of the glass surface. The formation of films requires cation migration, which is driven by an oxygen chemical potential between air and the reduced basalt (Fe²⁺/Fe³⁺ ratio of 13.3). The change of surface alteration from Mg to Ca film at lower temperatures, to predominantly Fe at high temperatures, is determined by a change of cation availability, largely controlled by crystallization that already occurs below 850°C, and volume crystallization that occurs above 925°C.     

Pyroxenes in Serra de Magé: cooling history in comparison with Moama and Moore County

Single-crystal X-ray, optical, and microprobe study of pyroxenes in the Serra de Magéfeldspar cumulate eucrite indicate complex exsolution features from a slow cooling history. Two pyroxenes now exist: “low” orthohypersthene ( P2₁ca) as host ( ～82 vol.%) and augite ( C2/c) in four distinct habits. This pyroxene pair yields an apparent “equilibration” temperature of ～900°. These relations are typical for orthopyroxene of both the Stillwater and Kintoki-San types, indicating an original pigeonite pyroxene with a bulk composition En₅₁Fs₃₉Wo₁₀. Variations in augite-hypersthene textural relationships suggest variable initial compositions from about Wo₈ to Wo₁₁. The bulk composition is intermediate to those of initial pigeonites in Moama and Moore County but the augite-hypersthene tie line is longer suggesting a slower cooling history. Our examinations of all three meteorites show that Serra de Magéaugite lamellae are as thick or thicker than those in the other meteorites, contrary to the measurement of Miyamoto and Takeda. The compositional data, textural relations, and existence of P2₁ca hypersthene suggest at least a comparable if not slower cooling history for Serra de Magé.     

The crystal structure and thermal history of orthopyroxene from lunar anorthosite 15415

A single crystal of untwinned orthopyroxene from lunar anorthosite sample 15415, with composition (Mg_1.14Fe_0.80Mn_0.02Ca_0.04)(Si_1.97Al_0.03)O₆, has a unit cell in space groupPbca witha = 18.310(15)Å,b = 8.904(10)Å,c = 5.214(7)Å, containing 2 formula units. A set of 742 counter-measured intensity data made with MoK_α radiation has been used to refine the crystal structure in isotropic thermal mode toR = 0.116. Anisotropic refinement led toR = 0.092, but thermal parameters are distorted by non-random errors resulting from poor crystal texture. The resulting structure is in close agreement with that obtained by Ghose [9] for a hypersthene from Greenland. A parameterq, which gives (Mg_qFe_1?q) for cation siteM(1) and (Mg_1.14?qFe_q?0.18Ca_0.04) for siteM(2), was included in the least-squares analysis, yieldingq = 0.90(1).This orthopyroxene has the high degree of cation order expected of pyroxenes subjected to Apollonian metamorphism at lower than 500–600°C. No evidence exists for a subsequent thermal event of sufficient intensity to disorder the pyroxene. On the basis of previous laboratory studies of argon-release patterns of lunar plagioclase and order-disorder kinetics of terrestrial pyroxenes, we attribute the reported isotopic age (3.9–4.1 AE) to cessation of metamorphism, perhaps caused by impact excavation.     

Oxygen isotope measurements of phosphate from fish teeth and bones

In situ measurements of lunar surface brightness temperatures made as a part of the Apollo Lunar Surface Experiments Package at the Apollo 15 Hadley Rille landing site are reported. Data derived from 5 thermocouples of the Heat Flow Experiment, which are lying on or just above the surface, are used to examine the thermal properties of the upper 15 cm of the lunar regolith using eclipse and nighttime cool-down temperatures. Application of finite-difference techniques in modeling the lunar soil shows the thermocouple data are best fit by a model consisting of a low-density and low-thermal conductivity surface layer approximately 2 cm thick overlying a region increasing in conductivity and density with depth. Conductivities on the order of 1 × 10^?5 W/cm-°K are postulated for the upper layer, with conductivity increasing to the order of 1 × 10^?4 W/cm-°K at depths exceeding 20 cm. An increase in mean temperature with depth indicates that the ratio of radiative to conductive transfer at 350°K is 2.7 for at least the upper few centimeters of lunar soil; this value is nearly twice that measured for returned lunar fines. The thermal properties model deduced from Apollo 15 surface temperatures is consistent with earth-based microwave observations if electrical properties measured on returned lunar fines are assumed.     

60025: relict of primitive lunar crust?

New isotopic analyses are presented for 3 plagioclase-rich fractions and one mafic fraction from ferroan anorthosite 60025. The observed²⁰⁶Pb/²⁰⁴Pb ratios vary between 52.5 and 60.5, all much higher than the ratio for terrestrial contamination. In a²⁰⁷Pb/²⁰⁶Pb²⁰⁴Pb/²⁰⁶Pb correlation diagram, the plagioclase data define a model PbPb age of 4.520 ± 0.007 AE using meteoritic primordial lead for the non-radiogenic component. In the concordia diagram the plagioclase data yield intersections at 4.503 ± 0.007 and 0.28 AE. The meaning of the lower intercept is obscure. The earlier 60025 analysis of Tera and Wasserburg [1], with an observed²⁰⁶Pb/²⁰⁴Pb of 23.0, agrees closely with the new plagioclase data in the isotope correlation and concordia diagrams. Since the apparent age does not correlate with the²⁰⁶Pb/²⁰⁴Pb ratios and U contents of the samples, it does not appear to be controlled by terrestrial lead contamination. The time-averaged μ values for the plagioclase leads are exceptionally low, 16–55, and agree within factors of 2 with the observed μ values in the samples. These are much lower than the values observed for mafic rocks or their sources, showing that the anorthosite lead has never been associated for a substantial length of time with any high μ source. In this way the 60025 data differ substantially from UPb data for two other lunar anorthosites, 15415 and 60015. The results suggest that the averaged model Pb ages of 4.51 ± 0.01 AE closely approximate the crystallization age for the plagioclase fraction of the anorthosite, and that it dates back to an early phase in lunar history. One sample from the mafic fraction of 60025 yields a younger model Pb age of 4.42 AE. The age may have been lowered by post-crystallization disturbances or perhaps this fraction is not coeval with the plagioclase fraction.     

Oxidation of titanomagnetites in mafic and felsic intrusive rocks

We report opaque mineralogical observations and magnetic properties of primary titanomagnetites in Tertiary submarine gabbros from DSDP, Legs 30 and 37 and in a late Archean, continental granitic pluton, the Shelley Lake granite. The titanomagnetites and silicates in all the submarine gabbros have been deuterically oxidized. There is no indication of subsequent low-temperature oxidation, although serpentization of olivines is pervasive in the deeper Leg 37 units. The Leg 30 samples, from a single thick sill, contain abundant coarse (≈100 μm) titanomagnetites with fully developed ilmenite exsolution lamellae. Curie temperatures are 515–550°C; there are no low Curie temperatures that would indicate surviving unoxidized titanomagnetite. The unserpentinized Leg 37 gabbros contain scarce opaques with pure magnetite Curie points that are barely resolvable microscopically; most occur as inclusions in pyroxene. In the Shelley Lake granite, on the other hand, many samples exhibit bimodal blocking-temperature spectra, with blocking temperature peaks at 250–300°C and 550–575°C. The low-blocking-temperature phase is unidentified. No pyrrhotite was seen in thin section. Optically homogeneous grains coexist with fully exsolved neighbours, but the electron microprobe indicates no titanium. The lamellae appear to be haematite, not ilmenite, and the primary composition of the opaques is pure magnetite. The oxidation state of the opaques is very inhomogeneous, even on a fine scale.     

40Ar-39Ar chronology of isolated phases from Bununu and Malvern howardites

High resolution⁴⁰Ar-³⁹Ar age spectra have been measured on plagioclase and glass from two howardites. Both the plagioclase and glass from the gas-rich Bununu howardite show well-defined age plateaux, yielding distinct ages of 4.42 ± 0.04 and 4.24 ± 0.05 AE, respectively. These age patterns are rather well behaved and are interpreted as representing the distinct times of formation of plagioclase from igneous processes and of glass fragments produced by impact on the meteorite body. The release pattern for the glass from the heavily shocked Malvern howardite is undulating at low and intermediate temperatures but does have a high-temperature plateau. Its age spectrum indicates little apparent diffusion loss, but rather an extensive redistribution of either⁴⁰Ar during the shock event or of³⁹Ar during the neutron irradiation or both. The total K-Ar age of Malvern glass is 3.64 ± 0.04 AE and the high-temperature plateau is 3.73 ± 0.05 AE. The age spectrum of the Malvern plagioclase has an intermediate temperature “plateau” at 3.80 AE that represents 20% of the total⁴⁰Ar content and increases towards a high-temperature plateau at 4.29 ± 0.04 AE containing 26% of the total gas release. It seems likely that the event which formed the Malvern glass also reset part of the plagioclase. The distinct histories observed for the different phases of these howardites are consistent with their formation from a regolith. The present results along with similar young ages for igneous clasts from Kapoeta clearly show that the regoliths were extant on the parent bodies of howardites and that they were subjected to violent impact events at least as recently as 3.7 AE ago.     

Josephinite: Crystal structures and phase relations of the metals

Josephinite, a complex, metal-bearing rock from the region of the Josephine Peridotite in southwest Oregon, contains FeNiCo metal alloy phases having exsolution textures. Scanning and transmission electron microscopy have revealed Widmanstaetten patterns in which the lamellae are polysynthetically twinned, ordered, face-centered-cubic FeNi₃ surrounded by untwinned, ordered, face-centered-cubic FeNi₃ and body-centered-cubic FeCo. These exsolution textures require a temperature in excess of 500°C for their formation. This is consistent with a mantle derivation of josephinite prior to obduction of the peridotite in one of the Klamath ophiolites.     

Thermal diffusivity of lunar rocks under atmospheric and vacuum conditions

Thermal diffusivity, k, of three lunar rocks (10049 and 10069; Type A, Apollo 11 and 14311; Apollo 14) and a terrestrial basalt (alkaline olivine basalt, Oki-do?go, Japan) was measured under one atmosphere and in vacuum conditions (10^?3 ～ 10^?5 mmHg) in the temperature range from 85 to 850°K. The semi-empirical curve of k =A + B/T +CT³ is fitted to the data in each condition. The porosity of rocks strongly affects the thermal diffusivity at low temperature ( T ? 500°K) in vacuum condition. At 150°K, thermal diffusivity of lunar rocks with porosity of 5.5% (10049) and 11% (10069) at one atmosphere is about 1.7 and 3.2 times of that in vacuum, respectively. The difference between the values at one atmosphere and those in vacuum decreases as the temperature increases. Measurements of k should be made at gas pressures at least lower than 10^?3 mmHg to estimate the value under lunar surface conditions.     

Ferroan anorthosite: A widespread and distinctive lunar rock type

Eight of eleven Apollo 16 rake-sample anorthosites are very similar to each other, to hand-specimen Apollo 16 anorthosites, and to Apollo 15 anorthosites. They have feldspar An_96.6, both high- and low-Ca pyroxene with a restricted range of (low-magnesium) composition, minor olivine (～ Fo₆₀), traces of ilmenite and chromite, and originally coarse-grained, but now cataclastic texture. Such ferroan anorthosite is evidently a coherent, distinctive and widespread lunar rock type of cumulate origin which may not necessarily be very closely related genetically to other highland rock types.     

Contrasting origins of Cenozoic silicic volcanic rocks from the western Cordillera of the United States

Two fundamentally different types of silicic volcanic rocks formed during the Cenozoic of the western Cordillera of the United States. Large volumes of dacite and rhyolite, mostly ignimbrites, erupted in the Oligocene in what is now the Great Basin and contrast with rhyolites erupted along the Snake River Plain during the Late Cenozoic. The Great Basin dacites and rhyolites are generally calc-alkaline, magnesian, oxidized, wet, cool (<850°C), Sr-and Al-rich, and Fe-poor. These silicic rocks are interpreted to have been derived from mafic parent magmas generated by dehydration of oceanic lithosphere and melting in the mantle wedge above a subduction zone. Plagioclase fractionation was minimized by the high water fugacity and oxide precipitation was enhanced by high oxygen fugacity. This resulted in the formation of Si-, Al-, and Sr-rich differentiates with low Fe/Mg ratios, relatively low temperatures, and declining densities. Magma mixing, large proportions of crustal assimilation, and polybaric crystal fractionation were all important processes in generating this Oligocene suite. In contrast, most of the rhyolites of the Snake River Plain are alkaline to calc-alkaline, ferroan, reduced, dry, hot (830–1,050°C), Sr-and Al-poor, and Nb-and Fe-rich. They are part of a distinctly bimodal sequence with tholeiitic basalt. These characteristics were largely imposed by their derivation from parental basalt (with low fH₂O and low fO₂) which formed by partial melting in or above a mantle plume. The differences in intensive parameters caused early precipitation of plagioclase and retarded crystallization of Fe–Ti oxides. Fractionation led to higher density magmas and mid-crustal entrapment. Renewed intrusion of mafic magma caused partial melting of the intrusive complex. Varying degrees of partial melting, fractionation, and minor assimilation of older crust led to the array of rhyolite compositions. Only very small volumes of distinctive rhyolite were derived by fractional crystallization of Fe-rich intermediate magmas like those of the Craters of the Moon-Cedar Butte trend. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

SmNd constraints on early lunar differentiation and the evolution of KREEP

The Sm-Nd systematics of lunar KREEP basalt 15386 reflects two chronologically distinct events in the development of the incompatible element-rich materials of the moon. The measured Sm-Nd mineral isochron of 15386 indicates an age of 3.85 ± 0.08 AE which is consistent with the reported Rb-Sr and³⁹Ar-⁴⁰Ar ages of many other KREEP-rich samples. This age is interpreted as the time at which 15386 crystallized from a liquid on or near the lunar surface. The frequent occurrence of this age for KREEP-dominated samples, as well as the restricted location of KREEP near major lunar near-side impact basins, suggests that the eruption of these incompatible element-rich liquids was related to deep impact events during the postulated final bombardment phase of the surface of the moon. However, the lower than chrondritic initial¹⁴³Nd/¹⁴⁴Nd of 15386 and the essentially identical Sm-Nd evolution of other KREEP-rich samples require that the light REE enrichment which characterizes KREEP was established considerably before 3.85 AE. Within the limits imposed by model assumptions in the various radiometric systems, it is concluded that the extremely narrow spread of Sm-Nd model ages for these samples around 4.36 AE, and the compatibility of this age with that indicated by the U-Pb and Rb-Sr systems, indicate that the source of later KREEP volcanism was produced in the closing stages of an early global scale lunar differentiation episode.