This article was originally published in Treatise on Geochemistry, Second Edition published by Elsevier, and the attached copy is provided by Elsevier for the author's benefit and for the benefit of the author's institution, for non- commercial research and educational use including without limitation use in instruction at your institution, sending it to specific colleagues who you know, and providing a copy to your institution’s administrator.
All other uses, reproduction and distribution, including without limitation commercial reprints, selling or licensing copies or access, or posting on open internet sites, your personal or institution’s website or repository, are prohibited. For exceptions, permission may be sought for such use through Elsevier's permissions site at:
http://www.elsevier.com/locate/permissionusematerial
Warren P.H., and Taylor G.J. (2014) The Moon. In: Holland H.D. and Turekian K.K. (eds.) Treatise on Geochemistry, Second Edition, vol. 2, pp. 213-250. Oxford: Elsevier.
© 2014 Elsevier Ltd. All rights reserved. Author's personal copy
2.9 The Moon
PH Warren, University of California, Los Angeles, CA, USA GJ Taylor, University of Hawai‘i, Honolulu, HI, USA
ã 2014 Elsevier Ltd. All rights reserved.
This article is a revision of the previous edition article by P. H. Warren, volume 1, pp. 559–599, © 2003, Elsevier Ltd.
2.9.1 Introduction: The Lunar Context 213 2.9.2 The Lunar Geochemical Database 214
2.9.2.1 Artificially Acquired Samples 214 2.9.2.2 Lunar Meteorites 214 2.9.2.3 Remote-Sensing Data 215 2.9.3 Mare Volcanism 216 2.9.3.1 Classification of Mare Rocks 216
2.9.3.2 Chronology and Styles of Mare Volcanism 218 2.9.3.3 Mare Basalt Trace Element and Isotopic Trends 224 2.9.4 The Highland Crust: Impact Bombardment and Early Differentiation 227 2.9.4.1 Polymict Breccias and the KREEP Component 227 2.9.4.2 Bombardment History of the Moon 229
2.9.4.3 Impactor Residues: Siderophile and Fragmental 230 2.9.4.4 Pristine Highland Rocks: Distinctiveness of the Ferroan Anorthositic Suite 231 2.9.4.5 The Magma Ocean Hypothesis 235 2.9.4.6 Alternative Models 237 2.9.5 Water in the Moon 238
2.9.5.1 Traditional View of a Dry Moon 238
2.9.5.2 Water in Pyroclastic Glasses 239 2.9.5.3 Water in Apatite in Mare Basalts and KREEP-Related Samples 239 2.9.5.4 Water in the Lunar Mantle 240 2.9.5.5 Implications of Water in the Lunar Interior 240 2.9.6 The Bulk Composition and Origin of the Moon 241
Acknowledgments 242 References 242
2.9.1 Introduction: The Lunar Context Longhi, 1992, 2003; Warren and Wasson, 1979b), and in this sense, the Moon more resembles a planet than an asteroid. Stable isotopic data suggest a remarkably similar pedigree for Another direct consequence of the Moon’s comparatively the constituent matter of the Moon and Earth, given the great small size was early, rapid decay of its internal heat engine. But isotopic diversity among sampled components of the solar the Moon’s thermal disadvantage has resulted in one great system (Warren, 2011; Zhang et al, 2012). Yet, lunar materials advantage for planetology. Lunar surface terrains, and many are obviously different. The Moon has no hydrosphere and of the rock samples acquired from them, retain for the most virtually no atmosphere, and lunar materials show strong part characteristics acquired during the first few hundred Ma of depletions of even mildly volatile constituents, not just N2, solar system existence. The Moon can thus provide crucial O , and H O (e.g., Wolf and Anders, 1980). Oxygen fugacity insight into the early development of the Earth, whereas the 2 2 is uniformly very low (BVSP, 1981; Nicholis and Rutherford, terrestrial record of early evolution was largely destroyed by 2009). These idiosyncrasies have direct and far-reaching impli- billions of years of geological activity and the first 500 Ma of cations for mineralogy and geochemical processes. Basically, Earth history are missing altogether. Lunar samples show that they imply that mineralogical (and thus process) diversity is the vast majority of the craters that pervade the Moon’s surface subdued, a factor that to some extent offsets the comparative are at least 3.9 Ga old (Norman et al., 2006). Impact cratering dearth of available data for lunar geochemistry. has been a key influence on the geochemical evolution of the
The Moon’s gross physical characteristics play an important Moon, and especially the shallow Moon. The popular giant role. Although exceptionally large (radius¼1738 km) in rela- impact model holds that the Moon originated as a form of tion to its parent planet, the Moon is only 0.012 times as impact spall after a collision between the proto-Earth and a massive as the Earth. By terrestrial standards, pressures inside the doomed Mars-sized (or larger) body (e.g., Cameron and Ward, Moon are feeble: The upper mantle gradient is 0.005 GPa km 1 1976; Canup, 2004). (vs. 0.033 GPa km 1 in Earth), and the central pressure is The uppermost few meters of the lunar crust, from which slightly less than 5 GPa. However, lunar interior pressures are all lunar samples derive, are a layer of loose, highly porous, sufficient to significantly influence igneous processes (e.g., fine impact-generated debris – regolith or lunar ‘soil.’ Processes
Treatise on Geochemistry 2nd Edition http://dx.doi.org/10.1016/B978-0-08-095975-7.00124-8 213 Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
214 The Moon
peculiar to the surface of an atmosphereless body, that is, effects such as a nearby impact melt (Cushing et al., 1999). But lunar of exposure to solar wind, cosmic rays, and micrometeorite granulitic breccias are almost invariably fine grained, and they bombardment, plus spheroidal glasses formed by in-flight tend to be ‘contaminated’ with meteoritic siderophile elements quenching of pyroclastic or impact-generated melt splashes, all (e.g., Cushing et al., 1999; Lindstrom and Lindstrom, 1986; are evident in any reasonably large sample of lunar soil (Eugster Warren et al., 1991b; however, cf. Treiman et al., 2010), so the et al., 2000; Keller and McKay, 1997). The lunar regolith is precursor rocks were probably mostly shallow impact breccias conventionally envisaged as having a well-defined lower bound- (brecciation and siderophile contamination being concen- ary, typically about 5–10 m below the surface (McKay et al., trated near the surface), and the heat source was probably 1991a); below the regolith is either (basically) intact rock or most often a proximal mass of impact melt. else a somewhat vaguely defined ‘megaregolith’ of loose but not Besides impactites, which are predominant near the bom- so finely ground-up material. Ancient highland terrains tend to barded surface, virtually all other lunar crustal rocks are igne- have roughly two to three times thicker developments of rego- ous or annealed-igneous. The superarid Moon has never lith than maria (Taylor, 1982). produced (by any conventional definition) sedimentary rock All lunar samples are derived through the regolith, so the and most assuredly has never hosted life. Even metamorphism detailed provenance of any individual lunar sample is rarely is of reduced scope, with scant potential for fluid-driven meta- obvious, and for ancient highland samples, never obvious. The somatism. Sampled lunar metamorphism is virtually confined closest approach to in-place sampling of bedrock came on the to impact-shock and thermal effects. Although regional burial
Apollo 15 mission, when many tens of clearly comagmatic metamorphism may occur (Stewart, 1975), deeply buried basalts were acquired within meters of their 3.3 Ga ‘young’ materials seldom find their way into the surface regolith, (and thus nearly intact) lava flow, so that their collective prov- whence all samples come. Annealing, among lunar samples, is enance is certain (Ryder and Cox, 1996). Lunar meteorites show more likely a product of simple postigneous slow cooling that impacts occasionally eject rocks clear off the Moon. How- (at significant original depth), or dry baking in proximity to ever, in a statistical way, most lunar rocks, even ancient highland an intrusion, or baking within a zone of impact heating. rocks, are found within a few hundred kilometers of their orig- The Moon’s repertoire of geochemical processes may seem inal locations. This conclusion stems from theoretical modeling limited and weird, but the Moon represents a key link between of cratered landscapes (Melosh, 1989; Shoemaker et al., 1970), the sampled asteroids and the terrestrial planets. Four billion plus observational evidence, such as the sharpness of geochem- years ago, at a time when all but microscopic bits of the ical boundaries between lava-flooded maria and adjacent Earth’s dynamic crust were fated for destruction, most of the highlands (e.g., Li and Mustard, 2000). Moon’s crust had already achieved its final configuration. Besides breaking up rock into loose debris, impacts create The Moon thus represents a unique window into the early important proportions of melt. A trace of melt along grain thermal and geochemical state of a moderately large object boundaries may suffice to produce new rock out of formerly in the inner solar system and into the cratering history of loose debris; the resultant rock would be classified as either near-Earth space. regolith breccia or fragmental breccia, depending upon whether surface fines were important in the precursor matter 2.9.2 The Lunar Geochemical Database (Sto¨ffler et al., 1980). Features diagnostic of a surface compo- nent include a smattering of glass spherules (of order 0.1 mm 2.9.2.1 Artificially Acquired Samples in diameter; typically a mix of endogenous mare-pyroclastic Six Apollo missions acquired a total of 382 kg of rocks and soil. glasses and impact-splash glasses) or abundant solar wind- Sampling was mostly either by simple scooping of bulk implanted noble gases (e.g., Eugster et al., 2000). regolith or by collection of large individual samples, mostly Elsewhere, impact melt may constitute a major fraction of much bigger than 0.1 kg (four of the six missions collected the volume of the material that becomes new rock, especially in individual rocks >8 kg). As a result, the particle size distribu- the largest events in which a planet’s gravitational strength limits tion of the overall sample is strongly bimodal. Of the total displacement and the kinetic energy of impact is mainly parti- Apollo collection, rocks big enough to not pass a 10 mm sieve tioned into heat (Melosh, 1989). Rocks formed in this manner comprise 70 wt%, yet the fraction between 1 and 10 mm adds are classified as impact-melt breccia and subclassified based on only 2–3 wt%, and the remaining 27–28 wt% of the material is whether they are clast-poor or clast-rich and whether their <1 mm fines (including core fines) (Vaniman et al., 1991). matrix is crystalline or glassy (Sto¨ffler et al., 1980). Obvious Three Russian unmanned Luna missions added a total of lithic and mineral clasts are very common in impact-melt brec- 0.20 kg of bulk regolith. The Luna samples are more valuable, cias, although the full initial proportion of clasts may not be gram-for-gram, than the Apollo samples because they represent evident in the final breccia. Some of the clasts may be so pul- three distinct sites. However, all nine of the lunar sample- verized that they become lost by digestion into commingled return sites are tightly clustered within the central-eastern superheated impact melt (Simonds et al., 1976). By some defi- region of the Moon’s nearside hemisphere. The nine sites can nitions, the term impact-melt breccia may be applied to prod- be encompassed within a polygon covering just 4.4% of the ucts of melt plus clast mixtures with initial melt proportion as Moon’s surface; if limited to rock-sampling (Apollo) sites, the low as 10 wt% (Papike et al., 1998; Simonds et al., 1976). polygon’s coverage is merely 2.7%. A few impactites feature a recrystallized texture, that is, they consist dominantly of a mosaic of grains meeting at 120 2.9.2.2 Lunar Meteorites triple junctions. These metamorphic rocks, termed granulitic breccias, may form from various precursor igneous or impac- R. L. Korotev (e.g., Korotev, 2005; 2012) maintains a fre- tite rocks, and the heat source may be regional (burial) or local, quently updated Internet list of lunar meteorites (or lunaites).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 215
As of mid-2012, the number of separate, unpaired finds was (Japan, 2007); Chang’e 1 (China, 2007); Chandrayaan 1 (India, rapidly approaching 80. Fortunately, in contrast to the Apollo 2008, with a spectrometry component, M3,fromNASA);Lunar and Luna sites, the distribution of source craters for lunar Reconnaissance Orbiter, LRO (NASA, 2009); Chang’e 2 (China, meteorites is almost random. This conclusion is based on the 2010); and Gravity Recovery and Interior Laboratory, GRAIL randomness of overall cratering on the present-day Moon (NASA, 2011). Among the pre-1994 remote-sensing geochemi- (Gallant et al., 2007; Ito and Malhotra, 2010; Le Feuvre and cal databases, the most notable (only recently superseded) is the Wieczorek, 2011; Morota et al., 2005) (cratering still shows a x-ray spectrometry data obtained for about 10% of the lunar slight bias toward the equator and 90 west longitude and surface on the Apollo 15 and 16 missions. But the key results, against both poles and 90 east) plus the finding (Gault, data for Mg/Si and Mg/Al, suffer from poor precision. For exam- 1983; Gladman et al., 1996) that the vast majority of Moon– ple, the average 1s error in Mg/Si reported for 22 large regions Earth journeys are not direct but involve a phase of geocentric by Adler et al. (1973) is 26%. or even heliocentric orbit. All of the lunar meteorites are finds, Clementine used four different cameras to map the global in some cases significantly weathered (Korotev, 2012). Lunar surface reflectance of the Moon at 11 different wavelengths, provenance is proven for these meteorites by a variety of evi- from the near ultraviolet (415 nm) to the near infrared dence, but the single most useful constraint comes from oxygen (2800 nm), in roughly one million images. Pixel resolution isotopes (Clayton and Mayeda, 1996; Hallis et al., 2010). generally varied from 100 to 300 m, but for selected areas, In addition to the usual potential for pairing among mete- Clementine’s highest resolution camera took 600000 images, oritic finds, lunar meteorites have an important potential for for four different wavelengths (415–750 nm), at typical pixel launch (source-crater) pairing (e.g., Zeigler et al., 2005). resolution of 7–20 m. Clementine multispectral images were Launch pairing can often be ruled out based on cosmic ray used to map surface mineralogy and, for a few elements, chem- exposure (CRE) constraints (e.g., Nishiizumi et al., 2002; Sokol ical composition, and even regolith maturity (i.e., average et al., 2008; Thalmann et al., 1996). The launch age is the sum extent of exposure to surface-regolithic processing) (e.g., of the terrestrial age (usually brief, of order 1–10 ka) plus the Lucey et al., 1998, 2000a,b; Pieters et al., 2002; Staid and p ‘4 ’ (all-directional exposure, Moon-to-Earth) CRE age. Unfor- Pieters, 2001; Wo¨hler et al., 2011). Translation from spectral tunately, the most unambiguous 4p CRE constraints, based on data into concentration is most straightforward for iron, using radioisotopic methods, become increasingly imprecise for CRE primarily data from 900 to 1000 nm, the region of a major þ ages much lower than 1 Ma and hopelessly imprecise for CRE Fe2 absorption band for pyroxene (Lucey et al., 2000a). The ages much less than 0.1 Ma, while physical modeling shows technique for titanium is more indirect, relying on the spectrum that, over time, most lunar meteorites have Moon-to-Earth slope as determined by the ratio between the 415 nm and journeys that take less than 0.1 Ma (Gault, 1983; Gladman longer wavelength (especially 750 nm) reflectances. Clementine et al., 1996). Another way of constraining launch pairing is data provided revelations about cryptomaria (Antonenko et al., based on the expectation that materials from a single source 1995), the petrology of moderately large impact craters as crater will generally show a degree of overall geochemical constraints on lateral and vertical heterogeneities within the similarity. This approach is most useful if one or both of the lunar crust (Cahill et al., 2009; Tompkins and Pieters, 1999), samples are a type of material (e.g., regolith breccia) that tends and compositional variations within the unfortunately remote to show region-specific composition. In this connection, it is (in relation to all Apollo and Luna sampling sites) South Pole– important to realize that the scale of the launch zone is far Aitken basin (Pieters et al., 2001). smaller than the full diameter of a crater, because ejection Lunar Prospector’s two most important geochemical map- velocity is a strong function of proximity to ground zero ping sensors were designed for gamma-ray spectrometry
(Warren, 1994). (GRS) and neutron spectrometry. These techniques measure In principle, a thorough sampling of the lunar surface the upper few decimeters of the regolith, whereas reflectance might yield pieces of Earth impact transported to the Moon and x-ray fluorescence (XRF) spectrometry measure no dee- (Armstrong et al., 2002), a large proportion of which would per than a few micrometers. The difference is not very im- presumably date from the era of heavy bombardment, 3.9 Ga portant, however, because the lunar regolith is extensively and before (see succeeding text). However, Earth’s atmosphere impact gardened. Prospector’s neutron spectrometer was is a major impediment to high-velocity rock ejection mainly designed to map the global distribution of regolith (Gladman and Chan, 2012), and rocks of terrestrial prove- hydrogen. Locally, very high H concentrations found near the nance have not been discovered among the meteorites hitting poles (Feldman et al., 2001) were interpreted as confirming Earth in modern times; only glasses (tektites) are known to Arnold’s (1979) suggestion that water liberated in impacts have reentered from space, and those are believed to have between comets and the Moon might have accumulated in reentered in the immediate aftermath of a cratering event. In cold traps within permanently shadowed regions near the any case, to date, no ‘terrene meteorite’ has been found among lunar poles. the lunar samples. Prospector’s GRS detector was of bismuth germanate type, with similar resolution to the NaI(Tl) detectors flown on the
Apollo 15 and 16 missions. The Prospector GRS database is superior because of vastly longer detector acquisition times 2.9.2.3 Remote-Sensing Data (i.e., better counting statistics) and because Prospector’s cover- The past two decades have been a golden age for lunar remote age was global, whereas the Apollo data covered no more than sensing, thanks to eight major missions: Clementine (BMDO/ 20% of the surface in two near-equatorial bands. The two ele- NASA, 1994); Lunar Prospector (NASA, 1997); SMART-1 ments most amenable to orbital GRS, iron and thorium, were (ESA, 2003); Selenological and Engineering Explorer, SELENE mapped to a spatial resolution of 45 km (Lawrence et al., 2007).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
216 The Moon
Locally, application of sophisticated deconvolution methods to heterogeneities within the lunar crust (Cahill et al., 2009; conspicuous terrain features allows an effective spatial resolu- Klima et al., 2011). Combined GRS and reflectance data have tion improvement by a factor of 1.5–2. Prospector’s GRS been applied to study the great South Pole–Aitken basin achieved spatial resolution of 60 km for potassium and tita- (Hagerty et al., 2011; cf. Yamamoto et al., 2012) and also the nium (Prettyman et al., 2006), and also for the rare earth ele- Moon’s volumetrically minor but nonetheless important silicic ment (REE) samarium, based on a complex technique involving hills (Besse et al., 2011; Glotch et al., 2010, 2011; Hagerty comparison with data from Prospector’s neutron spectrometer et al., 2006a; Lawrence et al., 2005); of which the most intrigu-
(Elphic et al., 2000). The GRS also yielded maps at spatial ing, because of its unusual (farside) location, is Compton– resolution of 150 km for oxygen, magnesium, aluminum, Belkovich (Jolliff et al., 2011). LRO’s Diviner instrument was silicon, calcium, and thorium (Prettyman et al., 2006). These primarily flown to make detailed diurnal temperature maps of data, particularly for thorium (an exemplary incompatible trace the lunar surface, but it has also been used to map various element; on the Moon, such elements are strongly concentrated aspects of surface composition (e.g., Allen et al., 2012; into KREEP), confirmed early hints from the Apollo GRS that Greenhagen et al., 2010; Kusuma et al., 2012). the Moon’s crust shows a remarkable degree of global geochem- Early results from the XRF spectrometers on SMART-1 and ical asymmetry. For example, average surface concentration of Chandrayaan 1 have been reported by Swinyard et al. (2009), Th is 3.5 times higher on the hemisphere centered over Oceanus Narendranath et al. (2011), and Weider et al. (2012). Both the Procellarum compared to the hemisphere antipodal to Procel- XRF of Chang’e 1 (Wu, 2012) and SELENE’s reflectance spec- larum (although it should be noted that at the low end of trometer (Ohtake et al., 2012) have detected small regions of
þ the concentration range calibration of the Prospector thorium the farside with exceptionally high mg ( MgO/[MgO FeO]). data was problematical; see Gillis et al., 2004;andWarren, The best defined of these are situated (1) a few kilometers 2005). Similar GRS results from SELENE and Chang’e 1 have south of Lacus Luxuriae, at the approximate center of the been reported by Yamashita et al. (2010, 2012) and Zhu et al. Freundlich–Sharonov basin, and (2) in the northern and west- (2010). The large region of the western nearside in which tho- ern parts of the Orientale basin. rium is conspicuously abundant roughly coincides with the northern Procellarum basin and has been dubbed the ‘Procellarum KREEP Terrane’ or ‘PKT’ (Gillis et al., 2004; Jolliff 2.9.3 Mare Volcanism et al., 2000). Similar results from GRS on SELENE have been 2.9.3.1 Classification of Mare Rocks reported by Yamashita et al. (2010).
Applying SELENE reflectance spectrometry to the high- The dark basalts that erupted to veneer the Moon’s flat, low- lands, Ohtake et al. (2009) found scattered areas of remark- lying ‘seas’ (maria) during the waning of lunar magmatism are ably near-pure plagioclase anorthosite. Areas with magnesian distinctive, even compared to other lunar samples. Mare spinel as the dominant mafic component were found using basalts have high (nearly always >16 wt%) FeO, low to mod- Chandrayaan 1 (M3) reflectance data by Pieters et al. (2011) erate mg, and usually also high TiO and low Al O (Figure 1; 2 2 3 and Lal et al. (2012). Other results from reflectance spectrom- Table 1). Calling these ‘basalt’ is potentially misleading. Most þ etry by Chandrayaan 1 have been reported by Isaacson et al. mare basalts are far more melanocratic (MgO FeO-rich, (2011) and Bhatt et al. (2012). Application of reflectance data feldspar-poor) than terrestrial basalt and arguably more to central-peak uplifts is a good way to map lateral and vertical analogs to a relatively leucocratic and low-mg komatiite.
15 Apollo 11 Apollo 12 Apollo 14 12 Apollo 15 Apollo 16 Apollo 17 Luna 16 Luna 24
(wt%)
9 2
Ti-rich 6 Medium-Ti
Bulk-rock TiO
3
Medium-Ti Ti-poor (VLT)
0 20 30 40 50 60 70 80 Bulk-rock Mg/(Mg+Fe) (mol%)
Figure 1 Mare basalt major elements: TiO2 versus mg. Caveat: the apparent bimodality in TiO2 may be misleading (see text). Data are mainly from the compilation of Haskin and Warren (1991).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
Table 1 Summary of literature compositional data for lunar mare meteoritesa and other major varieties of mare basalt (see text)
Sample Na 2O MgO Al2O3 SiO2 K2O CaO TiO2 FeO sum Sc V Cr Co Ni Ga Rb Sr Zr Ba La Sm Eu Tb Lu Hf Ir Th wt% wt% wt% wt% wt% wt% wt% wt% wt% mgg 1 mgg 1 mg g 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 ng g 1 mgg 1
Dhofar 287 avg. 0.52 12.3 8.1 43.9 0.11 8.4 2.86 22.2 98.4 35 4.2 42 20.0 530 60 200 12.9 6.3 1.18 1.22 0.51 2.6 0.90 Kalahari 009 avg. 0.53 8.5 12.8 46.2 0.19 10.5 0.45 15.9 95.1 55 116 2.69 27 <12 2.2 7.0 110 14.5 66 0.80 0.56 0.40 0.17 0.20 0.4 0.09 LAP 02204 avg. 0.37 6.7 9.8 45.3 0.08 11.1 3.14 22.2 98.7 59 114 2.10 37 27.6 4.1 1.8 131 183 143 12.7 7.5 1.21 1.75 0.91 5.4 2.04 MIL 05035 avg. 0.24 7.6 9.1 47.7 0.02 12.0 1.17 21.4 99.1 101 106 2.16 26 9.6 3.0 0.5 109 34 27 1.71 1.8 0.77 0.54 0.40 1.12 0.28 NEA 003 avg. 0.29 13.6 8.0 44.7 0.08 9.2 1.34 20.6 97.8 51 3.8 51 84.0 117 252 3.0 1.7 0.60 0.46 0.28 1.10 0.43 NWA 032 avg. 0.35 7.9 9.1 44.7 0.11 10.9 3.10 22.0 98.2 59 132 2.67 41 49 4.3 1.8 137 193 326 11.5 6.7 1.12 1.62 0.86 5.0 1.97 NWA 773 avg. 0.14 26.7 4.4 43.0 0.09 5.6 0.32 19.4 99.6 21 2.17 89 216 45 159 134 9.4 4.1 0.32 0.83 0.39 3.7 1.37 NWA 4898 avg. 0.31 8.3 12.0 46.2 11.4 2.39 17.3 97.8 65 120 2.98 25 <180 145 4.7 4.6 1.00 1.06 0.57 4.5 0.44 b YA meteorite avg. 0.32 6.0 11.0 45.5 0.05 12.0 2.1 22.3 99.2 93 81 1.59 24 11 2.8 1.1 146 92 66 4.0 3.5 1.14 0.93 0.59 2.6 0.05 0.53 All Apollo 11 avg. 0.48 7.6 9.6 40.5 0.16 11.1 10.5 19.5 100.0 83 78 2.1 21 5.4 5.8 2.8 177 433 208 18 16 2.1 3.7 2.0 12.6 0.027 2.1 mare basalts SD 0.11 1.0 1.5 1.3 0.11 0.9 1.3 1.2 8 19 0.5 7 2.2 2.3 32 199 107 9 5 0.4 1.1 0.5 4.1 1.2 N 97 89 89 29 100 89 90 89 74 62 86 78 6 18 22 17 35 74 89 85 85 82 80 73 7 65
All Apollo 12 avg. 0.27 10.8 9.3 45.1 0.06 10.0 3.4 20.4 100.‘0 50 158 3.6 46 47 3.6 1.0 120 133 70 6.8 5.0 1.13 1.35 0.65 4.0 0.036 0.78 mare basalts SD 0.10 3.5 1.6 1.6 0.01 1.6 0.8 1.2 7 30 1.0 13 1.2 0.3 30 62 23 2.3 1.2 0.38 0.39 0.18 1.0 0.23 N9074787496798384683091792333446052825887708278562530 All Apollo 14 avg. 0.49 9.5 12.7 48.2 0.30 10.7 2.5 16.2 101.2 56 114 3.1 32 22 3.7 19 99 314 273 17 9 1.29 1.98 1.0 7.0 0.021 2.3 mare basalts SD 0.11 1.5 1.3 1.4 0.37 0.8 0.9 1.9 8 18 0.8 6 0.4 12 39 162 354 11 5 0.90 1.02 0.5 4.0 2.2 N94919024879491939359939322 24394681959394939393390
All Apollo 15 avg. 0.27 10.4 9.1 46.3 0.05 9.8 2.1 21.2 100.1 40 202 3.9 51 58 3.8 1.0 116 90 72 5.7 3.4 0.85 0.75 0.34 2.34 0.043 0.57 mare basalts SD 0.05 2.6 1.3 1.7 0.01 1.1 0.5 1.5 5 29 1.2 12 1.0 0.9 21 34 25 2.6 0.8 0.18 0.19 0.08 0.84 0.18 N 129 133 132 88 95 133 135 132 99 45 119 99 23 33 33 49 38 58 98 93 93 93 92 85 29 43 All Luna 16 avg. 0.52 6.4 13.4 43.8 0.18 11.7 5.0 18.7 100.2 64 75 1.56 17.5 79 3.7 1.9 404 300 335 18 13 3.27 2.53 1.13 0.60 1.9 mare basalts SD 0.02 0.5 0.2 0.04 0.3 0.2 0.4 6 10 0.12 0.9 80 2 2 0.32 0.16 0.10 0.3 N44414444 444 4111114444441 4
All Luna 24 avg. 0.27 9.4 11.2 45.7 0.03 11.6 0.8 20.1 99.7 46 163 2.1 40 43 1.8 103 50 43 2.0 1.6 0.66 0.36 0.23 1.4 0.19 mare basalts SD 0.07 4.3 1.9 1.2 0.01 2.1 0.3 1.6 8 13 0.8 7 8 5 1.1 0.7 0.08 0.16 0.11 1.4 0.04
N 19 17 17 9 12 17 17 19 13 11 19 13 5 1 4 1 3 13 13 11 12 13 12 5
Large (high-Ti) avg. 0.39 8.4 8.9 38.9 0.06 10.5 12.1 18.9 99.0 80 105 2.95 20.3 2.0 5.1 0.70 173 216 106 5.8 8.3 1.75 2.19 1.16 7.5 0.05 0.26 Apollo 17 SD 0.04 1.2 0.7 1.1 0.02 0.8 1.3 0.9 5 29 0.66 4.7 1.8 0.47 33 63 80 1.3 2.5 0.44 0.66 0.31 2.2 0.12 mare basalts N 170 169 169 50 170 169 170 169 165 115 170 165 6 18 45 72 50 79 159 158 158 158 154 143 5 38
Apollo 17 VLT avg. 0.15 12.5 10.4 48.2 0.02 9.4 0.7 18.3 100.7 47 221 5.3 46 1.7 8.8 1.81 2.24 1.23 8.0 0.40 mare basalts SD 0.04 3.0 1.2 0.7 0.01 1.1 0.2 1.7 11 30 1.0 15 1.3 2.0 0.42 0.56 0.28 1.9
N 10 10 10 4 10 10 10 10 6 6 10 6 6 9 9 9 9 8 1
All Apollo/Luna avg. 0.35 9.4 10.6 44.6 0.11 10.6 4.7 19.1 100.1 58 140 3.1 34 36 3.9 4 170 219 158 9.5 8.2 1.6 1.9 0.96 5.4 0.036 1.1 mare basaltsc SD 0.12 1.8 1.6 3.1 0.09 0.8 1.4 15 52 1.1 13 1.2
Data mainly from the compilation of Haskin and Warren (1991). A large set of additional trace element data have been reported by Neal (2001), but for a few elements (Mo and Sb), his data appear spuriously high. Abbreviations: avg, average; SD, standard deviation; and N, number of data averaged. aMass-weighted means of literature data; only pristine basalts are shown (mare-dominated polymict breccias are excluded). bMass-weighted mean of literature data for Y-793169 and Asuka-881757 probably paired mare-basaltic meteorites. cAverage of the 8 major Apollo/Luna varieties shown earlier.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
218 The Moon
The contrast with the rest of the Moon’s crust is stark. The for the Ti-rich samples from Apollo 17 (Neal et al., 1990). typical composition of highland crust, as represented by These site-specific classifications are intended to link samples lunar-meteoritic regolith breccias and regolith averages for with specific magma types or even specific lava flows. How- the Apollo 14, Apollo 16, and Luna 20 sites (Table 2), features ever, considering the great diversity of mare materials as a 5 wt% FeO, 0.4 wt% TiO2, and 27 wt% Al2O3. The distinc- whole, while it is natural to ask whether differences between tion between mare and highland materials is seldom contro- the Apollo 12 olivine and Apollo 12 pigeonite basalts indicate versial, but for quick-and-dirty mare versus highland heterogeneity within a single lava (Rhodes et al., 1977)or classification based exclusively on bulk-analysis data, Wood require separate multiple lavas (Neal et al., 1994), the big- (1975) proposed using a combination of weight percent TiO2 picture observation is that Apollo 12 pigeonite and Apollo 12 and Ca/Al ratio (Figure 2). olivine basalts are only subtly different from one another. The various subclassifications for mare basalt about to be To date, 9–11 lunar meteorite mare basalts have been described may seem arbitrary, but as will be discussed in the sufficiently characterized (this tally is cited as a range to reflect next section, mare basalt compositional diversity is in general the uncertain but probable launch pairing among Asuka not systematic but haphazard. On Earth, volcanic diversity is 881757, Yamato 793169, and MIL 05035; Arai et al., 2010; largely a function of systematic global tectonics, such as Liu et al., 2009) to show that they are exclusively Ti-poor and upwelling at mid-ocean ridges to generate MORB. But plate medium-Ti, with bulk TiO2 no higher than 3.1 wt% (both LAP tectonics presumably played no such role in mare volcanism. 02205 and NWA 032; TiO2 tally includes the monomict brec-
Mare basalts are classified primarily on the basis of their cia Kalahari 009, plus NWA 4734, which has been studied highly diverse bulk TiO2 contents. The original and long- petrologically; Wang et al., 2012). Their number is still too customary nomenclature and dividing criteria (e.g., Neal and small to have great statistical significance, but among the Taylor, 1992) arose from historical accident: Basalts with well-characterized ‘mingled’ lunar meteorites that consist of
<1.5 wt% TiO2 are ‘VLT’ (very-low titanium), those with subequal proportions and mare and highland debris, in most 1.5
Kramer, 2006). The Luna 24 Ti-poor basalts and three lunar mare basins, the deposits are typically irregular-oval in shape,
meteorites (the two ‘YA’ basalts and NWA 4898; Greshake not arcuate. Clementine data enabled identification of 90 et al., 2008) are mildly aluminous (Table 1), and the Luna much smaller pyroclastic deposits that show similar spectral 16 basalts are clearly aluminous. However, for the Luna 16 diversity but are much more widely distributed across the basalts, which were sampled strictly as tiny regolith particles, a Moon (Gaddis et al., 2000;cf.Gustafson et al., 2012). The caveat is in order. Most of the ‘bulk’ analyses in the literature mare pyroclastic glasses are classified by the same Ti-, Al-, and are uncorrected defocused-beam microprobe analyses, which K-based criteria as the crystalline mare basalts and also on the tend to give spuriously high Al2O3 (e.g., Kurat et al., 1976). expedient basis of color of the glass. Ti-rich glasses tend to be Most known mare basalt types are not ‘high-Al’ and far from orange or red, Ti-poor glasses green, and medium-Ti glasses saturated, as melt compositions, with respect to plagioclase tan or brown; devitrification of the glass can make rapidly (Longhi, 1992, 2006). accumulated spherules black (which may account for the
Aside from this overall classification, mare basalts are also ‘dark’ mantle deposits). classified based on more subtle distinctions among samples from a given landing site. For example, among the Apollo 12 2.9.3.2 Chronology and Styles of Mare Volcanism medium-Ti mare basalts, three major groupings are distin- guished, named for minerals that are especially abundant in The chronology of mare volcanism is constrained primarily by each type: olivine, pigeonite, and ilmenite. Among Apollo 15 isotopic ages for samples (Table 3; Figure 3) and secondarily, medium-Ti basalts, only two main types are recognized: but with broader application, by photogeologic (crater- Olivine-normative and quartz-normative, which probably cor- counting) methods, calibrated by extrapolation from the few respond to two distinct flows, the olivine-normative atop the isotopically dated surfaces. The oldest isotopically dated mare- quartz-normative (Ryder and Cox, 1996). The most elaborate type Apollo samples are high-Al, medium-Ti (formerly ‘low- subclassification of a suite of basically similar mare basalts is Ti’) cumulate clasts found in Apollo 14 highland breccias,
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
Table 2 Compositional diversity among lunar regolith samples, including some meteorite regolith breccias
Sample Sample type Na2O MgO Al2O3 SiO2 CaO TiO2 FeO mg K Sc V Cr Mn Co Ni Zn
1 m 1 m 1 1 1 m 1 m 1 m 1 Physical Geochemical wt% wt% wt% wt% wt% wt% wt% mol% mg g gg gg mg g mg g gg gg gg
Ap 11 average Soils (averaged) Mare 0.47 7.9 12.6 42.0 11.7 7.9 16.4 46 1.37 62 70 1.99 1.66 31 199 25 Ap 12 average Soils (averaged) Mare 0.41 10.4 12.1 46.2 9.9 2.6 17.2 52 2.1 37 114 2.47 1.60 41 260 6 Ap 15, Mare Soils (averaged) Mare 0.40 11.0 14.3 47.2 10.5 1.46 15.0 57 1.30 27 110 2.53 1.45 45 216 14 Ap 17, Mare Soils (averaged) Mare 0.39 9.8 12.1 40.8 11.1 8.5 16.6 51 0.66 60 76 3.1 1.78 34 170 35 Luna 16 Core soil Mare 0.38 8.8 15.6 44.4 11.7 3.30 16.4 49 0.95 52 80 2.20 1.64 32 167 26 Luna 24 Core soil Mare 0.29 9.9 11.9 44.6 11.4 1.04 19.2 48 0.27 40 140 2.91 1.97 47 129 13 NWA 773 Regolith breccia Mare 0.23 13.2 10.6 46.2 10.8 0.78 17.3 58 0.83 nd nd 2.74 2.01 nd nd nd QUE 94281 Regolith breccia Mainly mare 0.37 8.5 15.8 47.5 12.5 0.71 14.1 52 0.50 33 116 1.81 1.55 45 208 5 Yamato-793274 Regolith breccia Mainly mare 0.38 9.1 15.3 47.9 12.2 0.61 14.2 53 0.67 33 99 2.08 1.62 42 101 7 Ap 14 average Soils (averaged) Mainly highland 0.69 9.2 17.7 48.2 11.0 1.72 10.4 61 4.3 23 51 1.41 1.08 36 350 25 Ap 15, Apennine Front Core soil Mainly highland 0.47 10.3 20.0 46.2 11.4 1.24 10.0 65 nd 18.5 65 2.01 1.11 30 162 31 Ap 17, South Massif Soil 73141 Mainly highland 0.42 9.7 21.3 45.2 13.0 1.22 8.0 68 1.17 16.2 37 1.48 0.84 27 239 16 Luna 20 Core soil Mainly highland 0.38 9.2 23.0 45.0 14.4 0.46 7.3 69 0.62 16.4 39 1.27 0.85 30 246 20 Calcalong Creek Regolith breccia Mainly highland 0.49 nd 20.9 nd 16.1 0.80 10.9 nd 2.0 23 57 1.30 1.12 24 360 nd ALH 81005 Regolith breccia Highland 0.30 8.2 25.6 45.7 15.0 0.25 5.4 73 0.19 9.1 25 0.89 0.58 21 202 9 Ap 14, 14076 Regolith breccia Highland 0.44 3.3 30.5 44.1 16.8 0.33 3.8 61 0.73 7.8 17 0.50 0.45 15.8 231 7 Ap 14, 14315 Regolith breccia Highland 0.57 7.9 22.1 47.1 13.0 0.85 7.4 65 2.6 15.6 nd 1.24 0.78 33 430 nd Ap 16 average Soils (averaged) Highland 0.48 5.8 27.1 44.7 15.5 0.58 5.2 67 1.07 10.4 20 0.71 0.52 26 360 18 Dar al Gani 262 Regolith breccia Highland 0.36 4.7 27.7 44.6 16.1 0.20 4.3 66 0.42 7.6 26 0.61 0.50 18 241 < 40 Dhofar 025 Regolith breccia Highland 0.31 6.5 26.8 44.5 15.8 0.29 4.8 71 0.48 10.0 <30 0.77 0.58 15.7 141 < 30 MAC 88104/5 Regolith breccia Highland 0.33 4.0 28.1 45.0 16.6 0.24 4.2 63 0.23 8.5 18 0.63 0.48 14.6 148 7 QUE 93069 Regolith breccia Highland 0.34 4.6 28.5 44.9 16.3 0.27 4.3 66 0.30 7.5 22 0.57 0.47 22 295 15 Yamato-791197 Regolith breccia Highland 0.33 6.3 26.0 43.7 15.5 0.34 6.5 63 0.24 13.1 32 0.90 0.67 19.4 178 21 Average highland regolith Conceptual 0.39 5.7 26.9 44.9 15.6 0.37 5.1 67 0.70 10.0 23 0.76 0.56 20.6 247 13 Sample Ga Sr Zr Cs Ba La Ce Nd Sm Eu Tb Dy Yb Lu Hf Ta Os Ir Au Th U m gg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 mgg 1 Ap 11 average 5.0 163 330 130 220 18.0 55 54 13.8 1.71 3.2 17 10.8 1.59 11.6 1.59 7.8 8.6 2.9 2.34 0.52 Ap 12 average 4.7 138 560 310 360 31 87 67 15.1 1.89 3.6 20 10.7 1.54 12.8 1.35 5.2 5.6 2.4 5.2 1.68 Ap 15, Mare 4.0 138 320 170 200 21 50 30 9 1.27 2.1 13 6.6 0.98 7.1 0.9 6.3 6.4 3.0 3.0 0.8 Ap 17, Mare 5.0 168 229 nd 98 8.0 24 21.8 8.1 1.72 1.99 nd 7.2 0.99 6.8 1.18 nd 4.8 3.6 0.88 0.31 Luna 16 5.1 275 253 63 185 11.2 34 24.6 8.0 3.2 1.43 9.7 5.6 0.77 7.0 0.65 nd 10.3 2.4 1.12 0.30 Luna 24 1.2 91 78 52 41 3.2 8.3 6.4 2.02 0.66 0.49 3.0 1.74 0.27 1.69 0.25 7.7 5.7 6.9 0.45 0.10 NWA 773 nd nd nd nd nd 16 41 24 7 0.7 1.5 nd 5 0.7 nd nd nd nd nd nd nd QUE 94281 4.5 106 105 89 72 6.4 15.7 9.6 3.1 0.83 0.64 4.2 2.42 0.35 2.36 0.29 6.2 6.5 2.2 0.93 0.23
Yamato-793274 4.7 110 92 46 78 5.8 15.2 9.6 2.73 0.82 0.58 4.0 2.29 0.32 2.26 0.26 4.5 4.5 3.0 0.81 0.21 Ap 14 average 5.9 189 810 680 840 67 181 106 29.8 2.7 6.3 38 23 3.2 22 2.8 nd 17.5 5.8 13.8 3.3 Ap 15, Apennine Front5.1 145 363 260 256 99 59 34 10.8 1.36 2.02 nd 7.7 1.04 8.7 1.00 nd 4.4 1.9 3.8 1.24 Ap 17, South Massif2.6 137 219 236 154 15.5 38 24.9 7.0 1.2 1.50 9.3 5.4 0.80 5.2 0.75 nd 12 6 2.4 0.70 Luna 20 3.7 165 192 76 104 6.2 16.8 11.3 3.2 0.90 0.59 4.2 2.22 0.36 2.57 0.31 nd 9.5 3.6 1.06 0.37 Calcalong Creek nd nd nd nd nd 20.1 52 28.8 8.7 1.20 2.16 14.2 6.4 0.98 6.8 1.12 nd nd nd 4.2 1.13 ALH 81005 2.7 135 27 24 28 1.98 5.2 3.2 0.95 0.69 0.21 1.33 0.84 0.12 0.72 0.09 8.4 6.8 2.3 0.29 0.10 Ap 14, 14076 10.8 175 108 146 90 8.0 18.2 11.9 3.2 1.09 0.67 4.2 2.5 0.34 2.2 0.29 nd 7.9 7.4 1.34 0.34 Ap 14, 14315 6.6 140 520 420 380 36 91 53 14.3 1.56 3.1 19.9 10.4 1.55 11.3 1.32 nd 18.8 12.7 5.6 1.38 Ap 16 average 4.2 162 162 155 127 11.6 30 20.8 5.5 1.22 1.12 7.2 3.8 0.58 4.1 0.52 nd 12.2 8.0 1.97 0.57 Dar al Gani 262 4.0 210a 33 90 200a 2.2 6.1 3.4 1.03 0.75 0.22 1.59 0.84 0.12 0.77 0.10 nd 10.1 3.9 0.37 0.16a a a a Dhofar 025 3.1 1700 52 nd 120 3.3 8.0 5.0 1.43 1.02 0.30 1.57 1.11 0.18 1.08 0.13 nd 5.4 4.3 0.57 0.22 MAC 88104/5 3.5 151 34 38 32 2.52 6.3 4.0 1.15 0.79 0.24 1.49 0.97 0.14 0.86 0.11 7.3 7.2 2.7 0.39 0.10 QUE 93069 3.2 149 46 41 42 3.4 8.5 5.0 1.56 0.82 0.33 1.99 1.21 0.17 1.14 0.15 21 15.6 4.5 0.53 0.13 Yamato-791197 5.5 135 32 67 32 2.13 5.5 3.5 1.06 0.77 0.25 1.56 1.01 0.15 0.88 0.10 8.2 6.7 1.4 0.33 0.11 Average highland regolith4.8 149a 113 123 100a 7.87 19.9 12.2 3.34 0.97 0.71 4.54 2.51 0.37 2.56 0.31 11.3 10.1 5.2 1.27 0.4a aHot-desert meteorites, which tend to be grossly contaminated with Ba, U, and especially Sr (Korotev, 2012), are excluded from the ‘average highland regolith’ calculation. Data sources are too numerous to enumerate in full. Among the most noteworthy are previous compilations by Haskin and Warren (1991); Jerde et al. (1990) for Apollo 14 samples; Korotev (1987a) for Apennine Front core soil 15007; Warren and Kallemeyn (1986) for ALH81005, Luna 20 and South Massif soil (most Al-rich from Apollo 17) 73141; Korotev and Kremser (1992) for an average of Apollo 17 mare soils from the LM/ALSEP site near the center of the Taurus–Littrow Valley; Jolliff et al. (2003) for NWA773; and Warren et al. (2005) for other lunar meteorites.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
220 The Moon
Apollo 11 1.6 Apollo 12 Apollo 14 Apollo 15
1.4 Apollo 16 Apollo 17 Luna 16 Luna 24 1.2 Nonmare
1.0
0.8 K009
Bulk-rock Ca/Al, molar ratio Wood (1975) 0.6 Simpler, better? criterion Highland
0.4
0.1 1 10 100
Bulk-rock TiO 2 (wt%)
Figure 2 On a plot of molar Ca/Al versus weight percent TiO2, nearly all mare basalts plot above and to the right of a criterion proposed by Wood (1975) for distinguishing between mare and highland rocks, and well apart from the vast majority of highland rocks, which plot at Ca/Al¼0.5–0.6 and TiO2 <2 wt% (0.5 is the stoichiometric anorthite ratio; fields are based on Wood’s (1975) Figure 2). Plotted as individual filled green circles are five extraordinarily gabbronoritic pristine nonmare samples that straddle Wood’s line: sodic ferrogabbro clast 67915c (Marti et al., 1983), 14434 (Arai and Warren, 1997), 61224,11 (Marvin and Warren, 1980), 67667 (Warren and Wasson, 1978), and 73255c27 (James and McGee,
1979). Wood’s (1975) criterion is Ca/Al¼0.786–0.229 log10(TiO2), for Ca/Al as a molar ratio and TiO2 in wt%. Also plotted is a simpler variant proposed here: Ca/Al¼0.74–0.26 log10(TiO2) that seems better at including Ti-poor mare basalts as mare samples, without much changing the situation for the nonmare samples. ‘K009’ is the ancient monomict-brecciated Ti-poor basalt Kalahari 009, which is generally interpreted as a mare material, despite its position on this diagram (Sokol et al., 2008).
which extend up to 4.23 Ga (Dasch et al., 1987; Shih et al., By extrapolation from rigorously dated surfaces, Hiesinger 1986; Taylor et al., 1983). The age of the Kalahari 009 et al. (2003, 2010) derive an age of roughly 1.2 Ga. In general, meteorite’s Ti-poor mare component is similarly 4.3 Ga the youngest mare basalts in Procellarum tend to be Ti-rich (Shih et al., 2008; Sokol et al., 2008; Terada et al., 2007). But (Staid and Pieters, 2001). At the opposite extreme, the crypto- in general, mare clasts within highland breccias are rare. The maria presumably tend to be uncommonly old, by mare stan- overall scarcity of mare basalt among the thousands of diverse dards, and remote-sensing data indicate that mare basalt in clasts studied from highland breccias, the scarcity of crypto- cryptomaria tends to be Ti-poor, or at most medium-Ti maria (regions where mare surfaces have been covered, but (Giguere et al., 2003). In short, age seems a very poor predictor also sufficiently gardened to become detectible, by more recent of mare lava composition. impact cratering; Antonenko et al., 1995), and the sharpness of The majority of mare rock samples are noncumulate most mare-highland boundaries based on various remote- basalts, formed by melts of generally very low melt viscosity, sensing data (e.g., Li and Mustard, 2000), all indicate that of order 1 poise (calculated by method of Persikov et al., 1987) mare volcanism occurred mainly after the period of heavy in flows whose thicknesses were mostly less than 14 m bombardment that ended at 3.9 Ga (see below). (Robinson et al., 2012a), although by one estimation Most of the individual age data for mare samples are from (Hiesinger et al., 2002), they averaged 30–60 m and ranged Apollos 11, 12, 15, and 17, and these data led to an early up to 220 m. The most widely cited method for estimating misconception (e.g., Snyder et al., 2000) that among mare total accumulated thickness of mare lavas, based on crater basalts Ti concentration correlates with antiquity. If anything, submersion statistics (small, shallow craters are submerged the overall database, including photogeologic inferences, sug- sooner than big, deep craters), indicates the maria are generally gests an anticorrelation between Ti and antiquity. The youngest <0.5 km thick and average roughly 1 km (DeHon, 1979). isotopically dated mare rocks are the Ti-poor cumulate mete- Ho¨rz (1978) argued these estimates ignore prevolcanic erosion orite NWA 773 and the medium-Ti basaltic meteorite NWA of the crater depth/diameter ratio, and thus are too high by a 032, both 2.8 Ga (Borg et al., 2009; Fernandes et al., 2009a). factor of 4. However, Thomson et al. (2009) used an ejecta- Other mare meteorites have yielded intermediate ages (e.g., based approach to infer thicknesses a factor of 2 higher than Anand et al., 2006; Haloda et al., 2009). The very last gasp of the estimates of DeHon (1979). The abundance of highland mare volcanism, the least cratered mare lava, is believed to be a component within regolith at mare-interior locales (where the flow in Oceanus Procellarum (Hiesinger et al., 2003, 2010). highland component is probably mainly added by vertical
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 221
Table 3 Summary of age constraints for mare basalts and pyroclastic glasses
Sample TiO2 plotted Ar age Sr age Nd age Pb age
wt% age, Ga Ga Ga Ga Ga References
Lunar meteorites Dhofar 287 2.86 3.4 3.46 0.03 3.35 Shih et al. (2002), Terada et al. (2008) Kalahari 009 0.45 4.3 4.30 0.05 4.35 Sokol et al. (2008) and sources they cite LAP 02204 3.14 3.0 2.99 0.02 3.00 0.02 3.15 0.04 2.93 Fernandes et al. (2009a) and sources they cite MIL 05035 1.17 3.85 3.85 0.01 3.00 0.04 3.80 0.05 Fernandes et al. (2009a) and sources they cite NEA 003 1.34 3.09 3.09 0.06 Haloda et al. (2009) NWA 032 3.10 2.8 2.76 0.02 2.85 0.07 2.69 0.16 Fagan et al. (2002), Fernandes et al. (2009a) NWA 773 0.32 2.89 2.91 2.87 0.03 Fernandes et al. (2003), Borg et al.
(2004) NWA 4704 (low) 2.74 2.74 0.02 Fernandes et al. (2009a) NWA 4898 2.39 3.56 3.54 0.02 3.58 0.04 Gaffney et al. (2008), Fernandes et al. (2009b) YA meteorite 2.08 3.90 3.80 0.01 3.89 0.03 3.87 0.06 3.94 Misawa et al. (1993), Fernandes et al. (2009a) Apollo 11 10017 12.2 3.51 3.51 0.05 BVSP (1981) 10024 12.6 3.53 3.53 0.07 BVSP (1981) 10029 10.9 3.83 3.83 0.03 BVSP (1981) 10062 10.5 3.86 3.79 0.04 3.92 0.11 3.88 0.06 BVSP (1981) Apollo 12
12022 4.9 3.08 3.08 0.06 BVSP (1981) 12020 2.83 3.09 3.09 0.06 BVSP (1981) 12002 2.62 3.25 3.21 0.05 3.29 0.10 BVSP (1981) 12021 3.52 3.26 3.26 0.03 BVSP (1981) Apollo 14 14053 2.72 3.93 3.93 0.04 Ryder and Spudis (1980) 14072 2.57 4.04 4.04 0.05 4.05 0.08 Ryder and Spudis (1980) 14168 1.70 3.85 3.85 0.02 3.82 0.12 3.95 0.17 BVSP (1981), Shih et al. (1986) 14304,113c 3a 3.95 3.85 0.04 3.95 0.04 Shih et al. (1987) 14304,108c 3a 4.00 3.99 0.02 4.04 0.11 Shih et al. (1987) 14305,122-92c 3.60 4.23 4.23 0.05 Taylor et al. (1983) 14305,304c 2.40 3.85 3.85 0.05 3.83 0.08 3.91 0.16 Shih et al. (1986)
14321,184c 2.41 3.95 3.95 0.05 3.95 0.04 Ryder and Spudis (1980)
14321,223c 2.41 4.1 4.1 Dasch et al. (1987) Apollo 15 15668 2.49 3.10 3.10 0.05 BVSP (1981) 15065 1.70 3.21 3.21 0.04 BVSP (1981) Ap15 green glass 0.36 3.34 3.3 3.38 Snyder et al. (2000), Shih et al. (2001) 15385 2.19 3.35 3.35 0.05 BVSP (1981) 15682 2.27 3.37 3.37 0.07 BVSP (1981) 15388 3.10 3.38 3.36 0.04 3.42 0.07 Dasch et al. (1989) 15475 1.81 3.39 3.43 0.15 3.37 0.05 Snyder et al. (2000) Apollo 17 high-Ti 71055 13.9 3.56 3.56 0.09 BVSP (1981)
70017 13.4 3.59 3.59 0.18 BVSP (1981)
Ap17 orange glass 8.8 3.56 3.6 3.48 Snyder et al. (2000), Shih et al. (2001)
70215 13.1 3.79 3.79 0.04 BVSP (1981) 70255 11.4 3.84 3.84 0.02 BVSP (1981) 79001,2144 9.9 3.90 3.49 0.04 3.89 0.04 3.92 0.04 Shearer et al. (2001) Other Samples Apollo 17 VLT 0.71 4.01 ? – 4.01 Taylor et al. (1991) (p. 209) Luna 16 5.0 range 3.5–3.04 Fernandes and Burgess (2005) Luna 24 0.84 range 3.6–3.0 Fernandes and Burgess (2005)
Except as noted, TiO2 data are from compilation of Haskin and Warren (1991) or Table 1. For the sake of brevity, only the two oldest and two youngest members of each major sample type are shown. a TiO2 for the Shih et al. (1987) 14304 clasts estimated based on sum ‘difference’ and analogy to other Apollo 14 high-Al mare basalts.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
222 The Moon
15 Ap17 VLT (age lower limit) Luna 16 (age range) Luna 24 (age range) Apollo 11 12 Apollo 12 Apollo 14 Apollo 15 Apollo 15 green glass Apollo 17 high-Ti
9 Apollo 17 orange glass (wt%)
2 Lunar meteorites
6
Bulk-rock TiO
3
0 2.5 3.0 3.5 4.0 4.5
Age (Ga)
Figure 3 The TiO2 concentrations in mare basalts and pyroclastic glasses show no correlation with age. Age data are from compilations by Ryder and Spudis (1980), BVSP (1981), Fernandes and Burgess (2005), and (for pyroclastic glasses) Shih et al. (2001), plus a lower limit cited for Apollo 17
Ti-poor basalts by Taylor et al. (1991). The TiO data are averaged from the compilation of Haskin and Warren (1991). The five major Apollo basalt 2 types are shown with small symbols because each point represents one of many available samples from the given locale, whereas each of the lunar meteorites represents (probably) our only sample from its locale.
mixing) favors relatively thin maria (e.g., Ferrand, 1988). trace-element compositions reflect relatively inefficient segre- Based on DeHon’s calibration, the total volume of mare basalt gation of cumulus matter apart from ‘trapped’ mare melt. is roughly 7 106 km3, or 0.03% of the lunar volume. How- Many distinct mare melt types have been sampled via pyro- ever, conceivably during the era of mare volcanism, a larger clastic glasses. A compilation by Taylor et al. (1991) included volume of intrusive mare-like gabbros formed, and yet could 25 glass composition types. The process that produced the be unsampled, because a rare post-3.9 Ga large crater would be pyroclastic glasses (and dark mantle deposits) probably has a required to excavate them to the surface regolith. close analog on present-day Io. The Moon and Io are nearly Phenocrysts are common, especially in the Apollo 15 mare identical in size and density and in their lack of atmosphere. basalts, but they appear to have grown strictly during the main Lunar pyroclastic spherules have volatile-rich coatings (Butler stage of cooling, that is, during posteruptive flow across the and Meyer, 1976; Chou et al., 1975; Elkins-Tanton et al., surface, not in any deep magma chamber (Lofgren et al., 1975; 2003a; Fogel and Rutherford, 1995), and crystalline mare Walker et al., 1977). The phenocrysts are invariably mafic, basalts can be highly vesicular (Ryder, 1985). On Io, ongoing never (thus far) plagioclase. The absence of plagioclase in the volcanism produces steady, long-lived plumes that probably phenocryst assemblage and, moreover, its scarcity as a liquidus consist of tiny, diffuse droplets, akin to the lunar pyroclastic phase (Longhi, 1992, 2006) are powerful indications that glass spherules; these plumes have been observed to rise as mare magmas tended to transit directly from the mantle to high as 400 km (briefly) and spread as far as 700 km from the surface and seldom paused in crustal magma chambers their volcanic vents (McEwen et al., 1998). The force that drives (cf. Wilson and Head, 2003). The aluminous mare basalts are these plumes is the explosive expansion of gas bubbles upon plagioclase saturated, and among the best-documented suite of eruption into the vacuous ionian (or ancient lunar) atmo- this type, the aluminous mare basalts from Apollo 14, many sphere (Head and Wilson, 1992; McEwen et al., 1998). The appear to have assimilated a crustal (or crust–mantle area) explosive gas phase on Io is probably SO2; the composition of material, namely, KREEP (Dickinson et al., 1985). But as the predominant gas phase during mare volcanism can only be noted earlier, aluminous mare basalts are rare. Mare cumulates reconstructed by inference. It was probably dominated by CO managed to develop, as exemplified by rocks such as 12005 and CO2 (Fogel and Rutherford, 1995), but Elkins-Tanton et al. (Dungan and Brown, 1977; Rhodes et al., 1977), 15385 and (2003a) suggest that sulfur and/or halogens may have been
15387 (Ryder, 1985), 71597 (R. Warner et al., 1977), the significant, and recent evidence for relatively abundant lunar Apollo 14 clast that Taylor et al. (1983) found to be exception- water (see Section 2.9.5)promptedWetzel et al. (2012) to ally ancient, and the NWA773 meteorite (Jolliff et al., 2003). suggest that a carbon–hydrogen compound, methane (CH4), However, based on limited exsolution within their pigeonites may have been significant as well. The likely main precursor (cf. Arai and Warren, 1999), these cumulates formed for the gases, graphite, tends to undergo oxidation if entrained within rapidly cooling surface flows or lava ponds, and their in a hot magma that depressurizes during the late stages of its
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 223
ascent to the surface (Nicholis and Rutherford, 2009). Nonethe- the old nomenclature) basalts, with at gap at 6–9 wt%. But less, the smaller scale of the lunar dark mantle deposits com- Giguere et al. (2000) inferred from Clementine remotely deter- pared to their Io analogs confirms a tendency toward relatively mined TiO2 data that the bimodality is not globally significant. low proportions of volatile fuel in the lunar magmas. Also, when all known samples, including lunar meteoritic Within each pyroclastic glass type, there tend to be system- samples, are plotted on a log scale (Warren, 2003), the only atic, relatively simple differentiation trends, encompassing in bimodality appears between the Ti-poor and medium-Ti (plus some cases wide ranges. The Apollo 14 orange glass array has Ti-rich) types; that is, the portion of the range with conspicu- MgO ranging from 14.5 down to 9.5 wt% (Delano, 1986), and ously few samples is 1–2 wt%. Arai and Warren (1999) argued within an Apollo 14 green (Ti-poor) glass suite REE concentra- that a gap at 1–2 wt% TiO2 is more likely, from partial melting tions range over a factor of 2 (Shearer et al., 1990). But the in which absence versus presence of cumulus ilmenite plays a compositional vectors of these trends do not conform to expec- crucial role, than one at 6–9 wt% TiO2. Genesis of Ti-rich mare tations from any simple ‘batch’ partial melting or fractional basalts was conceivably complicated by the great density (i.e., crystallization model. Instead, it appears that complex processes negative buoyancy) that such compositions imply for melts, if during the formation and migration of melt in the mantle, sufficiently deep (within the innermost 2 vol%) of the lunar involving KREEP and ilmenite cumulates, are required (Elkins mantle (van Kan Parker et al., 2012). et al., 2000; Elkins-Tanton et al., 2003b; Hagerty et al., 2006b; Shearer et al., 1996). KREEP is an important material mainly associated with highland rocks, discussed in a later section. 18
Initial post-Apollo views of lunar evolution envisioned the mantle as processed by crystallization of a primordial magma 15 ocean into a ‘layer cake’ stratigraphic sequence of composition- ally distinct cumulate layers: deep, Ti-poor, high-mg cumulates, grading upward to FeO-rich, Ti-rich, and low-mg cumulates. 12
The Ti-poor mare basalts were assumed to come from the deep
cumulates, and the Ti-rich basalts from the shallow, late-stage (wt%) 9 cumulates (Taylor and Jakes, 1974). This view now appears 2 hopelessly oversimplified. There is little correlation between TiO 6 the Ti abundances of mare magmas and the experimentally estimated (by multiple-saturation pressure) depths of their source regions (reviewed by Taylor et al., 1991; but also see 3 Elkins et al., 2000). Ringwood and Kesson (1976) first articu- lated what has become a generally accepted feature of models 0 for mare petrogenesis: Ubiquitous modification of the initial (a) 4 6 8 10 12 14 16 18 20 deep mantle cumulates by varying degrees of hybridization and assimilation interactions with formerly shallow material, 14 swept down by convective stirring of the mantle. Aside from regular convective motions, many authors (e.g., Hess and Parmentier, 1995; Zhong et al., 2000) have noted that 12 the aftermath of magma ocean crystallization may have been a gravitationally unstable configuration (dense, FeO- 10 and ilmenite-rich residual mush atop relatively FeO- and
ilmenite-poor cumulates), which would imply an enhanced (wt%) 3 convective potential, and possibly a catastrophic overturn of O 2 8 the entire mantle. Moreover, in a body as large as the Moon, Al
‘polybaric’ partial melting probably occurs over a range of depths, by a gradual increase in melt fraction within rising 6 diapirs (Elkins et al., 2000; Longhi, 1992; Shearer and Papike, Basalts
2005). The depth of initial melting is difficult to constrain. Glasses Remarkably, Beard et al. (1998) inferred from 176Lu–176Hf 4 isotopic systematics (cf. Neal, 2001) that high-pressure garnet 4 6 8 10 12 14 16 18 20 played a role in the residua of some mare basalts. (b) MgO (wt%) The picture that emerges from the Ringwood and Kesson (1976) type modeling is an almost chaotic diversity of potential Figure 4 Mare pyroclastic glasses tend to have far higher MgO/Al2O3, yet a similar range in TiO , compared to crystalline mare basalts. For variations in the mixing between (already diverse) cumulates 2 basalts, plotted data are averages of many literature analyses for basalt and ‘juicy’ ingredients of the lunar mantle, such as downswept types (Table 1), except for lunar meteorites, which are individual rocks KREEP, during formation of precursors to mare volcanism. (shown with smaller filled squares). Other data sources are as cited by Nonetheless, it is often suggested or implied that the distribu- Arai and Warren (1999), that is, primarily the compilation of Taylor et al. tion of TiO2 among mare basalts is systematic, that is, bimodal. (1991); updated from Arai and Warren (1999) by adding lunar meteorites Usually, the notion is that one mode is Ti-rich, and the other (but not the cumulate NWA773, 26 wt% MgO, excluded because its mode is the combination of Ti-poor plus medium-Ti (‘low-Ti’ in composition is presumably unrepresentative of its magma type).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
224 The Moon
The pyroclastic glasses tend to be far more ‘primitive’ in alkalis (sodium, potassium, rubidium, and cesium) and by trace terms of MgO content (Figure 4). The two types of material metals, such as zinc, indium, bismuth, and cadmium (Figure 5). form a rough anticorrelation between MgO and Al2O3. But even The same trend of enormous depletions in comparison to the crystalline basalts are mostly picritic (far from plagioclase chondritic matter and even terrestrial basalt is also shown by saturation), and detailed consideration of the major element highland samples, but the evidence is clearest and most defini- trends suggests few, if any, of the sampled crystalline basalts tive from the mare basalts. Gross overall volatile depletion is are potentially related to spatially associated mare pyroclastic also evidenced by the near absence of hydrated minerals among glasses (Longhi, 1987). Arai and Warren (1999) suggested that lunar rocks. Various Apollo rocks, most impressively Apollo 16 the glasses tend to be more MgO-rich because the main fuel for breccia 66095, contain surface-associated ‘rust’ patches (mainly explosive volcanism, graphite (density, 2.2 g cm 3), must have FeOOH,Cl), but these are suspected to be products of terrestrial come from a mantle component that remained (more or less) oxidation, not endogenous lunar volatile processing (Papike solid and intact through the extensive primordial (magma- et al., 1991;cf.Shearer et al., 2012). Even extraordinarily sphere) melting of the Moon and thus remained MgO-rich. ‘evolved’ types of nonmare rock, such as a few tiny samples that are compositionally granite (a term not meant to imply the existence of lunar batholiths), contain pyroxene, not amphi- 2.9.3.3 Mare Basalt Trace Element and Isotopic Trends bole or mica, as their mafic component (Jolliff et al., 1999; Mare basalts, being young by lunar standards, are as a rule far less Warren et al., 1983). battered by impact processes than highland rocks; they are also Mare basalts indicate that the Moon’s pattern of crust– generally closer in composition to their parent melts (i.e., the mantle siderophile element depletions is roughly similar to proportion of cumulates is far higher among highland rocks). that of the Earth (Figure 6). The most ‘noble’ of the siderophile Thus, mare basalts are our most suitable samples for many elements in Figure 6 are osmium and iridium. For noble side- purposes of Moon/Earth comparison (Figures 5 and 6). Ti-poor rophile elements, individual rocks show great scatter (e.g., and medium-Ti types are preferred for this purpose because the Figure 7), but the indicated factor of 5 disparity between
Ti-rich types are unlike any common variety of terrestrial basalt. average mare and terrestrial basalt for osmium, iridium, ruthe- The most striking difference between mare basalt and terres- nium, platinum, and palladium appears significant. In lunar trial basalt is in their contents of volatile species. Besides H2O basalt, rhenium, gold, and germanium are more ‘nobly’ (Section 2.9.5), this disparity is manifested by data for volatile depleted than in the terrestrial regime, where these elements,
100 V O L A T I L I T Y >>>
10
1
0.1
avg medium-Ti mare basalt
Sample/CI chondrites (wt ratio) avg Ap-12 medium-Ti mare avg Ap-14 medium-Ti mare
0.01 avg Ap-15 medium-Ti mare
avg terrestrial basalt avg basaltic shergottite avg noncumulate eucrite 0.001 Th Sr V Si Cr Mn Cs Rb K Na GaBr Zn S Pb In Bi Cd
Figure 5 Volatile element concentrations in averaged medium-Ti mare basalts, normalized to CI chondrites (Wasson and Kallemeyn, 1988). Also shown for comparison are averages for terrestrial basalts, for basaltic shergottite (martian) meteorites, and for eucrite (asteroidal) meteorites. The elements are plotted in order of thermodynamically calculated solar nebula volatility (Wasson, 1985). The mare basalt data are mainly from the compilation of Haskin and Warren (1991); a noteworthy primary source is Wolf et al. (1979). The plotted overall average mare basalt composition is a 4:1:4 weighting of the Apollo 12, 14, and 15 types, respectively (the Apollo 14 type has been far less well studied and moreover may be idiosyncratically enriched in volatile-incompatible elements; Dickinson et al, 1989). The average for terrestrial basalt is based on Govindaraju’s (1994) compilation for USGS standards BCR-1, BHVO-1, BIR-1, and W-1, plus (given 1/5 weight) an average for MORB, based primarily on Hertogen et al. (1980). The average for eucrites is based on a large number of references, most notably Chou et al. (1976), Morgan et al. (1978), and Paul and Lipschutz (1990).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 225
10
avg medium-Ti mare basalt avg Ap-12 medium-Ti mare 1 avg Ap-14 medium-Ti mare avg Ap-15 medium-Ti mare avg Earth basalt
0.1
0.01
0.001
Sample/CI chondrites (wt ratio)
0.0001
V O L A T I L E Os IrRu Pt Pd Re Au Ni Ge Bi Sb Br Se Mo Co W
Figure 6 Siderophile element concentrations in averaged medium-Ti mare and terrestrial basalts, normalized to CI chondrites. The elements are plotted in order of CI-depletion factors in average medium-Ti mare basalt, but for some elements, volatility may account in large part for the depletion.
Data are from Day et al. (2007), Walker et al. (2004), and same sources as for Figure 5. To avoid an over-complex diagram, individual terrestrial
6 compositions are not plotted, but all show similar patterns at the scale of this diagram; the most noteworthy exceptions being low Ir (9 10 times CI) 3 and Ni (1 10 times CI) in BCR-1 and relatively low Os and Ir (virtually identical to mare basalts) and Sb (0.11 times CI) in MORB.
particularly rhenium, tend to exhibit incompatible lithophile Mare basalts Regolith samples Polymict breccias 100 tendencies (probably linked to Earth’s higher fO2). Also, some of the lunar depletion in elements such as germanium may reflect the Moon’s general depletion of volatile elements. The same trend of strong siderophile depletions in comparison to
10 chondritic matter, and even versus terrestrial igneous rocks, is also shown by highland samples (Day et al., 2010). But here, the evidence is far more complicated; a careful and, for side- rophile elements, possibly biased selection of ‘pristine’ rocks )
1 1 - (see succeeding text) is required to obtain compositions
representative of the endogenously igneous highland crust.
3 ϫ 10−4 Among the highly siderophile elements, iridium, osmium, times CI and nickel have been frequently and well determined, and all
Bulk Ir (ng g 0.1 conform with a general pattern among planetary igneous rocks (Walker, 2009; Warren et al., 1999) by showing a correlation
(albeit not linear) with MgO (Figure 8). These trends represent
0.01 14434 the strongest evidence that the mantle that produced the mare basalts was significantly siderophile-depleted in comparison to the terrestrial mantle. Figure 8 also includes martian meteorite data, but note that low MgO in the martian mantle may make 0.001 the iridium, osmium, and nickel levels of the martian trends 11 12 14 15 16 17 11 12 14 15 16 17 11 12 14 15 16 17 seem higher (vs. Earth and the Moon) than they truly are Apollo mission number (Warren et al., 1999). Figure 7 Despite great scatter in the data distribution, data for The role of sulfide-driven (chalcophile) fractionations in lunar a highly siderophile element such as iridium reveal a strong magmatism is difficult to constrain, but sulfides presumably play contrast between compositionally ‘pristine’ rocks, as exemplified by a lesser role on the Moon than on Earth, because the solubility of mare basalts, and lunar materials (regolith samples and polymict sulfide in mafic melt increases with decreasing fO (Elkins-Tanton breccias) contaminated by meteoritic debris. This contrast can be 2 exploited to help identify pristine nonmare rocks. A good example is et al., 2003a; Peach and Mathez, 1993). Compared to lithophile Apollo 14 diabase 14434, which has an ambiguous texture and elements of similar volatility, sulfur is exceptionally depleted in mineralogy (monomict-brecciated but not diagnostically coarse), yet is terrestrial and martian basalts, but not so depleted in lunar basalts judged very probably pristine based on its depletion in iridium (Arai and (Figure 5). Both medium-Ti and Ti-rich mare basalts are unsatu- Warren, 1997). rated with sulfide (Danckwerth et al., 1979).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
226 The Moon
1000
) 1 - NWA773 100 gg m
67667 Bulk Ni ( 10
ALH84
1 10
1
) 1
-
gg 0.1
ALH84 67667
Bulk Ir (n
0.01 Earth Mars Moon, mare basalt Moon, mare-pyroclastic Moon, nonmare 0.001
0 4 8 12 16 20 24 28 32 Bulk MgO (wt%)
Figure 8 (a) Ni and (b) Ir versus weight percent MgO, for igneous rocks from the Moon, Mars, and Earth. Lunar data are mainly from the compilation of Haskin and Warren (1991), the most noteworthy addition being lunar meteorite NWA 773. Data sources are too numerous to list but include Crocket and MacRae (1986), Bru¨gmann et al. (1987), Govindaraju (1994), Walker et al. (2004), Day et al. (2006, 2007, 2010), and Brandon et al. (2012).
Mare basalt incompatible element patterns are diverse could conceivably have formed without major plagioclase (Figure 9; a comparable variety of patterns are found for fractionation because pyroxene can impart a negative euro- mare pyroclastic glasses; Papike et al., 1998). Most have signif- pium anomaly. McKay et al. (1991b) reported relevant D icant negative europium anomalies, which have potentially data for lunar pigeonite. Brophy and Basu (1990) applied profound implications. In the reducing lunar environment, these and earlier D results to show that accounting for the negative europium anomalies are only to be expected in basalts mare europium anomalies without appeal to prior plagioclase that are plagioclase saturated, that is, left mantle residua with removal requires implausible assumptions about modal min- plagioclase, with which europium is thoroughly compatible eralogy and/or degree of melting in the source regions. Assim- (distribution coefficient D 1.1–1.2) when reduced to mostly ilation/mixing with KREEP swept down into mantle may have 2þ Eu by lunar fO2 (McKay et al., 1977). But most mare basalts affected many of the mare europium anomalies, but the mare (all those with Al2O3 less than about 12 wt%) are not saturated basalts with the largest negative europium anomalies tend not with plagioclase at any pressure (Longhi, 1992, 2006; Papike to have KREEP-like enriched La/Sm ratios (Figure 9). et al., 1998; Taylor et al., 1991). The inference, therefore, is that Data for isotopic tracer ratios, Ι , e , etc., in mare basalts Sr Nd long before the mare magmas formed, their source regions (i.e., have been reviewed by Snyder et al. (2000), Shearer et al. much of the lunar mantle) must have been predepleted in (2006), Borg et al. (2009), Taylor et al. (2009), and Nemchin plagioclase; a requirement that has long been viewed as a et al. (2011). Model ages based on strontium, neodymium, significant argument in favor of a primordial lunar magma and lead isotopes for the mantle source regions typically cluster ocean (Taylor and Jakes, 1974; Warren, 1985). Shearer and near 4.3–4.4 Ga, coinciding with a neodymium-based model Papike (1989) suggested that the negative europium anomalies age for urKREEP, 4.35 0.03 Ga (Lugmair and Carlson, 1978).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 227
1000
KREEP
100
10
1 Sample/CI chondrites (wt ratio) Ferroan anorthosite
La Ce Sm Eu Gd Tb Yb Lu 0.1
Figure 9 Chondrite-normalized REE concentrations in mare basalts, KREEP, and a representative ferroan anorthosite. Data for mare basalt types are from Table 1, with two deliberate omissions: the NWA773 cumulate, an individual rock presumably unrepresentative of its parent magma, and the Apollo 14 type, which is a suite with notoriously diverse REE abundances, although the patterns are generally parallel to KREEP (which is plotted). Individual mare basalt types are not labeled, but symbols on some of the patterns denote relatively Ti-rich varieties; the most Ti-rich basalts (largest symbols) tend to have the lowest (often subchondritic) La/Sm ratios. Data for high-K KREEP and ferroan anorthosite 15295c41 are from Table 4 and Warren et al. (1990), respectively.
e Most mare basalts have positive Nd; the exceptions are some The importance of pristine rocks as constraints on lunar eastern Procellarum (Apollo 12 and Apollo 14) samples whose crustal genesis will be discussed at length in the succeeding sources probably mixed to an unusual extent with KREEP, text, but polymict breccias, especially the arch-polymict rego- which is extremely abundant in the Procellarum region lith breccias, are useful as naturally produced blend-samples
(Lawrence et al., 2007). The Apollo 14 mare basalts also tend assembled from mostly local, but otherwise random, bits of Ι to have unusually high Sr, again consistent with KREEP assim- the lunar crust. The mixing diminishes the value of these rocks ilation. Medium-Ti mare basalts show remarkably diverse eHf as recorders of lunar igneous processes but makes them ideal (as high as 51) with again Apollo 14 and especially Apollo 12 for constraining variations in the regional bulk composition of samples accounting for much of the range (Unruh et al., 1984). the crust (Table 2). The most dramatic compositional varia-
Beard et al. (1998) inferred possible involvement of deep- tions among these samples involve their contents of the unique mantle garnet, a suggestion also supported by Neal (2001) lunar material, KREEP. Although named for its high contents based on high Zr/Y and Zr/Yb ratios in some mare pyroclastic of K, REE, and P, KREEP is actually rich in all of the incompat- glasses. In any event, the highly diverse isotopic ratios of the ible trace elements. The REE pattern of KREEP is shown in mare basalts demonstrate that volcanism arose from a remark- Figure 9. Among the many elements arbitrarily absent in the ably heterogeneous mantle, fundamentally different from the acronym are thorium and uranium, which are the most tectonically stirred, and thus homogenized, mantle of Earth. extremely enriched (vs. chondrites) in KREEP, and (with potas- sium) the main sources of long-lived radiogenic heating within planets. Because the many hundreds of Apollo polymict 2.9.4 The Highland Crust: Impact Bombardment impactites that are thorium- and REE-rich tend to contain all and Early Differentiation the incompatible elements in relatively constant (KREEP)
ratios, Warren and Wasson (1979a) suggested that nearly all 2.9.4.1 Polymict Breccias and the KREEP Component of the lunar crust’s complement of incompatible elements was In contrast with the systematically younger mare basalts, the vast derived from a common reservoir, possibly a magma ocean majority of rock samples from the ancient highland crust are residuum, which they dubbed urKREEP. Note, however, that impactites. The proportion of rocks that are completely unal- urKREEP is a strictly hypothetical material. It is expected/ tered by impact processing is so minor that in the highland inferred that none of this material remains in its original context, the term ‘pristine’ is used to denote samples that are in form, since almost as soon as it formed, it probably was mas- most cases monomict-brecciated impactites, but at least preserve sively involved in assimilative reactions between urKREEP and in (effectively) unaltered form the original chemical composi- magnesium-rich magmas (Shearer and Papike, 2005; Shervais tions of endogenously igneous lunar rocks (Papike et al., 1998; and McGee, 1999; Warren, 1988). There may even have been a Ryder et al., 1980; Warren and Wasson, 1977; Wolf et al., 1979). tendency for KREEP to differentiate (albeit temporarily and
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
228 The Moon
locally) by silicate liquid immiscibility (Jolliff et al., 1999; Neal Lunar meteorite highland regolith breccias, from widely and Taylor, 1991). The few KREEP-rich rocks that are pristine scattered random points, tend to be remarkably KREEP-poor are mostly basalts or, in rare cases, granitic (Papike et al., compared to the Apollo and Luna regolith samples (Table 2). 1998). Despite the ephemeral nature of the hypothetical The Lunar Prospector maps of the global distributions of urKREEP, it is useful to compare all other lunar sample com- thorium, uranium, potassium, and samarium (Elphic et al., positions to the average composition of the most concentrated 2000; Lawrence et al., 2007; Prettyman et al., 2006) revealed ‘high-K’ variety of polymict impactite KREEP, which occurs that the Apollo/Luna sampling region happens to be atypically among Apollo 12 and especially Apollo 14 samples (Table 4). KREEP-rich.
Table 4 Estimated composition of average ‘high-K’ KREEP, as derived by Warren (1989)
Warren and Wasson Warren (1989) KREEP Strength of correlation KREEP/CI chondrites Uncertainty classb a (1979) KREEP (avg. high-K) with KR wt ratio (see below)
m 1 þ Li gg 56 40 Moderate 25 III 1 Na mg g 6.4 7 Moderate þ II Mg mg g 1 64 50c Moderate II Al mg g 1 88 80 Weak 9.3 I Si mgg 1 224 235 Weakþ(?) I Pmgg 1 3.4 3.5 Moderate þ 3.4 III Kmgg 1 6.9 8 Weak þ 14 III Ca mg g 1 68 70 Weak 7.6 II Sc mgg 1 23 23 None 4.0 II Ti mg g 1 10 12 Very weak þ 29 III V mgg 1 43 40 Weak II
Cr mg g 1 1.3 1.2 Weak II 1 Mn mg g 1.08 1.05 None II 1 Fe mg g 82 80 Weak þ II 1 Co mgg 33 25 Weak IIII Ga mgg 1 7.5 9 Weak þ (III) Br mgg 1 Not estimated 120 Moderate þ IIII Rb mgg 1 22 22 Weak þ 9.9 III Sr mgg 1 200 200 Very weak þ 25 I Y mgg 1 300 400 Strong þ 278 (II) Zr mgg 1 1700 1400 Strong þ 368 III Nb mgg 1 80 100 Moderate þ 370 (III) Cs ng g 1 2000 1000 Weak þ 5.46 III
m 1 þ Ba gg 1200 1300 Strong 565 II m 1 þ La gg 110 110 Very strong 466 1 Ce mgg 270 280 Very strong þ 455 I 1 Pr mgg Not estimated 37 Very strong þ 398 I Nd mgg 1 180 178 Very strong þ 389 Sm mgg 1 49 48 Very strong þ 322 Eu mgg 1 3.0 3.3 Moderate þ 59 I Gd mgg 1 57 58 Very strong þ 294 Tb mgg 1 10 10.0 Very strong þ 282 Dy mgg 1 65 65 Very strong þ 265 Ho mgg 1 14 14 Strong þ 256 I Er mgg 1 39 40 Very strong þ 250
m 1 þ Tm gg Not estimated 5.7 Very strong 231 I m 1 þ Yb gg 36 36 Very strong 226 1 Lu mgg 5.0 5.0 Very strong þ 204 1 Hf mgg 37 38 Strong þ 317 I Ta mgg 1 4.0 5.0 Very strong þ 313 II W mgg 1 2.0 3.0 Strong þ 30 (III) Th mgg 1 18 22 Very strong þ 759 I U mgg 1 5 6.1 Strong þ 744 I Molar Mg/(MgþFe) 0.64 0.59c Moderate II aWarren’s (1989) parameter KR is the average of sample/KREEP ratios for a large set of incompatible trace elements. bEstimated uncertainties for the average high-K KREEP composition, expressed as maximum expected percentages of deviation between ‘true’ average and estimates: blank¼5%, I¼10%, II¼20%, III¼30%, and IIII¼40%. Parentheses denote elements for which extrapolation to high KR is required. cMg concentrations appear to be systematically higher in Apollo14 KREEP versus KREEP from other locales.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 229
2.9.4.2 Bombardment History of the Moon 1.0 and 0.3 Ga for the Copernicus and Tycho craters, and hugely uncertain crater count data for terrains resurfaced dur- The polymict impactites are extremely important as primary ing the formation of these craters. However, 40Ar–39Ar ages constraints on the impact bombardment of the Moon. This obtained by Culler et al. (2000) and Levine et al. (2005) for bombardment history can be extrapolated to Earth, where pre- hundreds of Apollo 12 and Apollo 14 impact glass spherules 3.8 Ga crust has been virtually eliminated by geodynamism. and a reanalysis of the Culler et al. data by Levine et al. suggest The polymict impactite samples have yielded ages that are that the cratering rate continued to decline by roughly a factor remarkably clustered near 3.9 Ga, especially for impact-melt of 2 between 3.1 and 2.0 Ga and then remained roughly con- breccias (Figure 10). This curiously unimodal age spectrum, stant until at 0.4 Ga, when it may have increased by a factor which was first noted in the initial Apollo 16 sample investi- of 4. Levine et al. (2005) stated that the data are consistent gations (e.g., Kirsten et al., 1973; Schaeffer and Husain, 1973), with, but do not require an increase in cratering rate in the last represents one of the most profound discoveries of planetary 400 Ma of lunar history. Zellner et al. (2009), using a similar sample research. It revealed that the rate of cratering (i.e., approach, found evidence for an increase at 0.8 Ga. These collisions between the Moon and asteroids and comets) was results should be viewed with caution. Impact-splash glasses vastly higher 3.9 Ga ago than it has been over the last 85% of are too small to churn the regolith as they land; they collect in solar system history. A few additional data in the 3.1–3.7 Ga the upper decimeter. The preponderance of young ages among time interval, derived from crater-abundance counts on mare such materials may say more about the history of local regolith lava terrains dated through Apollo and Luna mare basalt sam- gardening (by craters large enough and near enough to churn ples (BVSP, 1981; Hartmann et al., 2007; Hiesinger et al., 2003, effectively) than about the global history of cratering. 2010), indicate that the cratering rate declined rapidly, by a Within regolith breccia 14076, Warren (2008b) found a factor of 10, between 3.9 and 3.1 Ga. suite of impact spheroids, including clasts of accreted spher- For many years, the consensus view was that the cratering oids, with unusual SiO - and/or FeO-rich compositions that rate probably remained approximately constant since 3.1 Ga 2 indicate origin as impact-vapor condensates. (BVSP, 1981; Guinness and Arvidson, 1977; cf. Bogard et al., The first reported ages for impact-melt clasts from lunar 1994), based on an ‘Occam’s Razor’ preference for simplicity, meteorites (Cohen et al., 2000) showed a comparatively dif- but comparatively scant evidence: Indirectly inferred ages of fuse cluster, spread rather evenly from 2.6 to 4.0 Ga (Figure 10; diamond symbols). Most of the Cohen et al. (2000) lunar meteorite data are relatively imprecise (they were obtained
from tiny clasts), but these data are important because they presumably represent several very widely separated regions of Granulitic the Moon. Cohen et al. (2005) and Hudgins et al. (2011) breccias added additional data showing similarly diverse ages. The bombardment history before 3.9 Ga has been most > controversial. The relative scarcity of ages 3.9 Ga led many (originally Tera et al., 1974; most emphatically Ryder, 1990)to infer a spike in the global lunar cratering rate at 3.9 Ga; in other words, that the rate was considerably lower before 3.9 Ga. This cratering spike concept is known as the lunar
‘cataclysm’ hypothesis (Tera et al., 1974) but also somewhat
confusingly as the ‘late heavy bombardment’ hypothesis. Actu- ally, irrespective of the spike issue, a ‘heavy’ cratering rate at the ‘late’ date of 3.9 Ga is beyond dispute. The controversy con- cerns the degree to which the clustered 3.9 Ga ages reflect a
large factor and global spike, as opposed to a bump or inflec-
Impact melt breccias tion on a basically monotonic decline in the late-accretionary impact rate. From a celestial–mechanical standpoint, models exist (Morbidelli et al., 2001), including the popular ‘Nice model’ (Gomes et al., 2005; however, see Raymond et al., 2009), that predict continued intense cratering as late as
2.2 2.6 3.0 3.4 3.8 4.2 4.6 3.9 Ga. But the abrupt solar system reconfiguration and resul- Age (Ga) tant major cratering spike implied by the ‘Nice model’ could happen at practically any time; 600 Ma after solar system origin Figure 10 Ages of lunar polymict breccias, especially impact-melt is hardly an obvious outcome. breccias, show strong clustering near 3.9 Ga. 40Ar–39Ar ages are shown by Skeptics of the cratering spike hypothesis have argued that filled symbols and 87Rb–87Sr ages by open symbols. Squares: Apollo the clustering of ages near 3.9 Ga may be a ‘stonewall’ effect samples, data compiled by Papike et al. (1998), except excluding (in order to avoid complicating the plot) five much younger ages (0.4–2.3 Ga) for (Chapman et al., 2007; Hartmann, 1975; Wetherill, 1981). The impact-melt glass samples and the granulite ages of Hudgins et al. (2008). stonewall model invokes a saturation of the pre-3.9 Ga crust Diamonds: clasts in four different lunar meteorites (Bogard et al, 2000; with so many impacts that isotopic clocks were constantly Cohen et al, 2000). Note: error bars are generally 2s (or unstated), but only being reset within a heated megaregolith, until the rocks 1s for the Cohen et al. (2000) meteorite clast data. that eventually became samples were excavated to the surface
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
230 The Moon
regolith (sensu stricto) by a few major impacts. This might seem impact-melt breccias from Apollo 16. Also, basin ejecta thick- a plausible explanation for the paucity of pre-3.9 ages, if the ness modeling (Haskin et al., 2002) suggests that ejecta from 40 40 3.8–3.9 Ga clustering was limited to the easily reset K– Ar distant Imbrium and Serenitatis outweigh ejecta from the 87 87 age system. However, Papike et al. (1998) cite many Rb– Sr- nearby but relatively small and older Nectaris, in the upper based ages clustering near 3.8 Ga. Chapman et al. (2007) also (sampled) Apollo 16 megaregolith. Considering all of these critique the stonewall model. constraints, an age of 4.2 Ga for Nectaris seems entirely The relevance of the impact-melt age spectrum as portrayed possible, if not probable. by Ryder (1990) and Dalrymple and Ryder (1993, 1996) can The age of a fourth basin, Crisium, can in principle be still be questioned, however, on the grounds that these authors constrained using Luna 20 samples. Cohen et al. (2001) always assumed that only samples with classic (near- obtained 40Ar–39Ar ages for six Luna 20 rocklets and reviewed subophitic) impact-melt breccia textures are appropriate for literature data (mainly from Swindle et al., 1991) for 12 others. dating impact events. J. Warner et al. (1977) suggested that Swindle et al. (1991) found a loose clustering of ages at 3.75– on the early (pre-3.9 Ga) Moon, where the ambient crustal 3.90 Ga. Swindle et al. (1991) suggested (in a ‘tentative’ way) temperature was still relatively close to the solidus, granulitic that the oldest sample in this cluster, 3.895 0.017 Ga, might breccias tended to form in major impacts, in analogs propor- date the Crisium impact. However, it is not even clear that the tion and position to where on the post-3.9 Ga Moon impact- sample in question is an impact-melt breccia (no thin section melt breccias would form. This model, from a team of leading was made). The only two certain impact-melt breccias dated by authorities on impactite genesis, has never been effectively Swindle et al. (1991) yielded ages of 0.52 0.01 Ga and
refuted. As was already a point of emphasis to Warren and 4.09 0.02 Ga. Among the samples dated by Cohen et al. Wasson (1977), granulitic breccias are frequently older than (2001), most are in our opinion probably either impact-melt 3.9 Ga (Figure 10). breccias or annealed impact-melt breccias (genuine, pristine Skeptics of the cratering spike hypothesis also note (e.g., ‘gabbros’ seldom have grain sizes of <200 mm like rocklet Chapman et al., 2007) that the Apollo 3.9 Ga impact-melt 2004D; pristine troctolites seldom have grain sizes of < m samples come exclusively from sites clustered in the central 100 m like rocklet 2004C). Considering all of the 13 or so 40 39 nearside and thus are likely from preponderantly just two or Luna 20 rocklets that have yielded Ar– Ar ages (reviewed by three impacts: Imbrium, Serenitatis (the connection between Cohen et al., 2001) and are likely impact-melt breccias, the Apollo 17 and Serenitatis has always been considered obvious, data (excluding the 0.52 Ga outlier) show an almost even but see Spudis et al., 2011), and, with greatest uncertainty, distribution across the range 3.75–4.19 Ga. In other words,
Nectaris. Cratering theory (Grieve and Cintala, 1992; Melosh, the age of the Crisium impact is not yet constrained beyond 1989) indicates that the proportional yield of impact melt, as being probably within the range 3.8–4.2 Ga. opposed to solid ejecta, is vastly higher in large basin-forming By extrapolation to other solar system bodies, the Moon’s events than in smaller scale impacts. Warren (1996) applied bombardment history represents a crucial series of chronologic this principle to the particular case of the Moon. Liu et al. benchmarks for planetology. This extrapolation involves com-
(2012) have proposed a slight upward correction for the age plex correction for the flux and prevailing velocity of the of Imbrium, to 3.91 Ga, based on U–Pb ages for Apollo 12 impactors (asteroids and comets) that are functions of, mainly, zircons. heliocentric distance (e.g., Chyba, 1991). Bogard (1995) found Most photogeologic–stratigraphic interpretations hold that that ages of numerous impactite meteorites from the HED Nectaris is older than Imbrium and Serenitatis, and two-thirds asteroid (Vesta?), and many more from the (probably) sepa- of the Moon’s still recognizable basins appear older than Nec- rate mesosiderite asteroid, show a similar clustering near taris (Wilhelms, 1987). Impact-melt breccias of Nectaris origin 3.9 Ga. The sole martian meteorite older than 1.3 Ga has also are presumably present among the Apollo 16 samples, yielded a 40Ar–39Ar age of 3.9 Ga (Ash et al., 1996). Even if the acquired 550 km from the center of Nectaris. 40Ar–39Ar ages cratering spike (cataclysm) hypothesis is incorrect, vastly for Apollo 16 impact-melt breccias mostly cluster from 3.87 to higher inner solar system cratering rates at 3.9 Ga, as dem-
3.92 Ga (Dalrymple et al., 2001; Norman et al., 2006, 2010; onstrated primarily and still most impressively from lunar
Papike et al., 1998). Still, as discussed by Norman et al. (2010), samples, are a well-established fact. Implications for evolution the absolute age of Nectaris remains unclear because the sam- of Earth, including the biosphere, are profound (e.g., Kring and ple ages may mostly record the Imbrium event. A large group of Cohen, 2002). Apollo 16 ‘light matrix breccias’ (classified as impact-melt breccias by Sto¨ffler et al., 1980) yielded 40Ar–39Ar ages in the 2.9.4.3 Impactor Residues: Siderophile and Fragmental range 4.12–4.26 Ga (Mauer et al., 1978; Schaeffer and Husain, 1973). Mauer et al. (1978) assumed these ‘Group 1’ samples The siderophile elements in polymict lunar materials come cannot be basin-related because they tend to be distinctly almost entirely from meteoritic contamination added to KREEP-poor compared to the other Apollo 16 impact melts; the outer Moon in impacts. Siderophile elements thus may the assumption was that all impacts big enough to form basins give clues to the nature of the materials that have bombarded would plumb into a KREEP layer in the lower crust. However, the Moon. Unfortunately, however, the limited expanse of the in light of what Prospector data (Lawrence et al., 2007) Apollo/Luna sampling region, with samples dominated by just revealed about the extreme concentration of KREEP into the three to four major basins, again severely restricts the general Procellarum (central-eastern nearside) region, a Nectaris prov- applicability of this approach. enance seems more plausible for the ‘Group 1’ samples than The degree to which the largest basins dominate the mega- for the more typical, younger but unsuitably KREEP-rich, regolith siderophile budget (as they clearly dominate the
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 231
impact-melt inventory; Warren, 1996) has been controversial. the IAB type) impacted at both Serenitatis and the Apollo 16 The Edward Anders group at the University of Chicago region. Morgan et al. (2001) argued that the high-Au/Ir signa- reported a wealth of excellent lunar siderophile analyses in a ture of late lunar accretion matches the composition they infer series of many papers (e.g., Hertogen et al., 1977) and inter- for the latest major accretion on Earth. Based on osmium preted some rather diffuse clustering in impactite siderophile isotopic evidence, Puchtel et al. (2008) found that the domi- composition as signatures of several different basins. Wasson nant impactor residues in Apollo 14 (Imbrium?) and Apollo et al. (1975) and Korotev (1987a, 1994), and Korotev et al. 17 (Serenitatis?) impactites may represent a type of primitive
(2006) have argued that basin signatures tend to be obscured material not currently delivered to Earth as meteorites. by major siderophile inputs from smaller craters. Both views Joy et al. (2012) carefully scrutinized ancient (>3.4 Ga) may be partly correct. The kinetic energy of impact is the main Apollo regolith samples (regolith-assembly ages inferred by control over basin (cavity, ejecta, and even melt) volume, and Joy et al., 2011) to find a suite of intact (albeit minute) impac- unfortunately for the aim of a simple, general answer to the tor fragments. These authors argued that because most of their basins versus craters controversy, the ratio of impact energy to ancient impactor remnants appear to be of asteroidal (not projectile mass is sensitive to a wide-ranging and unrecover- cometary) derivation, the early bombardment of the Moon able parameter: collision velocity. Velocity of impact with the must have been dominantly of asteroidal provenance. Asteroi- Moon is typically about three times faster for comets than for dal domination is also supported by crater size-frequency sta- asteroids (e.g., Chyba, 1991). Apart from impact velocity, the tistics (Strom et al., 2005), but the interpretation of Joy et al. problem is partially constrained by the size-frequency spec- (2012) seems oversimplified. Comets consist mostly of ice,
> trum of projectiles striking the Moon. This size distribution hardly likely to survive for 3.4 Ga at the near-equatorial conforms to a power law with number of particles propor- Apollo sampling locations. The far higher impact velocities tional to L p, where L is the projectile diameter and p is a and icy compositions of comets both work against more than constant. Numerous studies, reviewed by Melosh (1989), indi- tiny proportions of cometary impactor debris sticking to the cate that the slope p ranges from 1.8 for large L to 3.5 for small Moon in the first place (Ong et al., 2010). L, with a rather sharp inflection at L 10 m. Thus (given that for large projectiles p<<3, and assuming velocity, density, and 2.9.4.4 Pristine Highland Rocks: Distinctiveness of the siderophile concentrations, all remain roughly constant across Ferroan Anorthositic Suite the size spectrum), most siderophile contamination, over time, should come from a comparative few large collisions. How- The distinction between pristine rocks and polymict breccias is ever, it may be that in most regions, no single basin dominates a crucial first step for any attempt to accurately gauge the the megaregolith siderophile budget. original diversity and thus the igneous processes involved in In any event, the steeper (p>3) slope for impactor L<10 m genesis of the nonmare lunar crust (the terms nonmare and implies that for the regolith (sensu stricto), a different regime highland are essentially synonymous; arguably, ‘highland’ is probably prevails, with micrometeorites important, if not an inapt term for the bulk of the Moon’s crust, most of which is dominant, in determining the siderophile budget. This infer- lower than the ‘lowland’ mare basalt veneers). The vast major- ence is confirmed by regolith sample data for iridium, the best ity of nonmare rocks are polymict breccias, and unfortunately, studied of the ‘noble’ siderophile elements. Mare regoliths, identifying the pristine exceptions is seldom as easy as we which form by relatively small-scale impact gardening atop would like it to be. In rare cases, vestiges of a coarse igneous intact, siderophile-poor bedrock, show a strong correlation cumulate texture, more consistent with an endogenous igne- between Ir concentration and regolith maturity (Wasson ous origin than with an impact-melt origin, are preserved (e.g., et al., 1975). Highland regoliths show little correlation with Dymek et al., 1975; Warren and Wasson, 1980b). But the most maturity, probably because their siderophile enrichments are straightforward and generally applicable criterion is based on largely inherited from impact gardening at much larger scale, siderophile elements. Nonpristine rocks, that is, products of that is, from megaregolith. meteorite impact-induced mixing, generally contain enough
The Hertogen et al. (1977) interpretation of basin- meteoritic debris to impart high concentrations of siderophile dominated siderophile signatures has been refined in many elements such as Ir, in comparison to the levels characteristic of subsequent works, usually emphasizing impact-melt breccias, pristine rocks, as exemplified by mare basalts (Figure 7). for example, Morgan et al. (2001) and Norman et al. (2002). Warren and Wasson (1977) suggested 3 10 4 times the CI The polymict highland impactites from the Apollo 14, Apollo chondritic concentration as the ‘cutoff’ level. 15, and Apollo 16 sites, where the siderophile components It should be emphasized that this siderophile cutoff should should be dominated by some combination of Imbrium, never be construed as an upper limit, sine qua non, and nor Serenitatis, and (possibly, for Apollo 16 only) Nectaris ejecta, should data below the cutoff be taken as complete proof of tend to feature remarkably nonchondritic siderophile patterns: pristine composition. In evaluating suspected pristine samples, Au/Ir is typically three times, and Ni/Ir two times, the CI- all relevant traits, such as texture and mineralogy (silicates in chondritic ratios. The polymict impactites from Apollo 17, gross disequilibrium, or more obviously FeNi metal with typ- presumably dominated by ejecta from adjacent Serenitatis, ical meteoritic kamacite composition, can be tip-offs of feature similar but less extremely high Au/Ir and Ni/Ir. impact-mixing), other aspects of bulk composition (absence Among chondrites (Wasson and Kallemeyn, 1988), only the of KREEP contamination can be a mildly favorable indicator), EH enstatite type has Au/Ir approaching the typical Apollo 14, and isotopic data (an extremely old age can be mildly Apollo 15, and Apollo 16 ratio. Korotev (1987a; 1998) has favorable), should be assessed, if possible (cf. Ryder et al., conjectured that high-Au/Ir and Ni/Ir iron meteorites (akin to 1980; Warren, 1993). A fine-grained texture, disequilibrium
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
232 The Moon
(zoned) silicates, and KREEP contamination, all are inevitable evidence for infiltration metasomatism from the mineralogy of (but fortunately, high siderophile concentrations are not), if symplectites in Mg-suite troctolite 76525. the pristine rock happens to be KREEP basalt. When pristine rocks alone are considered, a remarkable The rate of discovery of pristine nonmare rocks has been geochemical bimodality is manifested within the nonmare limited by the scarcity of large lunar rock samples. Identifica- crust. During the 1970s, petrologists gradually noticed that tion of a coarse cumulate or otherwise ‘plutonic’ texture is the most anorthositic pristine rocks tend to feature distinctively difficult without a thin section at least several millimeter low mg in comparison to otherwise comparable nonmare across, and determination of trace siderophile elements rocks. Dowty et al. (1974) coined the term ‘ferroan becomes increasingly difficult (and prone to the ‘nugget’ effect) anorthosite,’ and Warner et al. (1976) were the first to postu- if available sample mass falls below 0.1 g. Lunar geochemists late a separate ‘Mg-rich plutonic’ suite to account for nearly all must resist a temptation to overestimate the likelihood that other pristine nonmare rocks. In more recent usage, the origi- new samples may be pristine. As a rather severe example, nal terms may be replaced with ‘ferroan anorthositic suite’ (or
Snyder et al. (1995) classified 12 out of 18 small rock clasts simply ‘ferroan suite’) and ‘Mg suite,’ but their basic meanings from Apollo 14 as ‘probably pristine,’ even though ‘(n)one of have not changed. The distinctiveness of the ferroan anortho- the samples ... are texturally pristine,’ and their siderophile sitic (FA) suite was first noticed on the basis of mineral com- data included no concentration (or limit) lower than 6 10 4 position data, namely, plots of plagioclase molar Ca/(CaþNa) times CI (and even that level was approached for only one of versus mafic silicate mg. Essentially, the same pattern is found the 12 samples in question). Rocklet 14286 (Figure 11), which by simply plotting bulk-rock mg versus molar Na/(NaþCa) arguably might pass for an uncommonly fine-grained pristine (Figure 12(a)). The Mg-suite rocks distribute along a normal norite, except it contains a huge mass of meteoritic kamacite fractionation path, with Na/(NaþCa) increasing as mg (Albrecht et al., 1995), indicates that potential pristine rocks decreases, to form a diagonal trend on Figure 12(a). The FA should be evaluated with great caution. suite forms a distinct cluster to the low Na/(NaþCa), low mg Results from recent studies related to pristine nonmare rocks side of the Mg-suite trend, and there is a noticeable gap (or at include a tiny (0.35 0.15 mm) spinel-rich clast in the ALH least, a sparsely populated region) between the two groups. 81005 meteorite that Gross and Treiman (2011) interpret as Warren and Kallemeyn (1984) found the same bimodality pristine and possibly related to the spinel detected from orbit using other plagiophile element ratios, such as Ga/Al or Eu/ by M3 (Pieters et al., 2011). Takeda et al. (2006;cf.Yamaguchi Al, in place of Na/(NaþCa). Most impressive is Eu/Al et al., 2010) interpreted a 3.0 1.3 mm clast in the Dhofar 489 (Figure 12(b)), which results in a wide gap between the two meteorite as ‘magnesian anorthosite’ (actually, the texture is too pristine rock groups. fine-grained to justify regarding this material as pristine, and Besides having distinctive combinations of mg and plagio- moreover at the scale of this sample, the modal distinction phile ratios, FA suite rocks tend to be more anorthositic than between anorthosite and troctolite cannot possibly be signifi- other types of pristine rocks (Figure 13). Not all of the ferroan cant). Ohtake et al. (2012) assumed that this Dhofar ‘magnesian pristine rocks are anorthosites, sensu stricto (>90 vol% plagio- anorthosite’ has relevance to the magnesian and anorthositic clase: Sto¨ffler et al., 1980), but prevalence of low mg among the compositions (which may be impact-mixtures) that they have most anorthositic components of the crust is borne out by observed, mainly within the churned interiors of basins, using major element variations among regolith samples SELENE data. Working on more plausibly pristine rocks, (Figure 14). The obvious yet far-reaching implication McCallum et al. (2006) studied pyroxene exsolution structures (Warren and Wasson, 1980a) is that the FA suite, and no to constrain cooling histories, and Elardo et al. (2012) found other adequately sampled type of pristine lunar rock (note
that Figure 13 includes no samples smaller than 1 g), would have been buoyant over its parental magma. If the magma ocean hypothesis (see Section 2.9.4.5) has any validity, the FA suite represents the only rock type plausibly formed as
magma ocean flotation crust, and the manifestly sunken cumu-
lates of the Mg suite must come from intrusions that were emplaced into an older, FA-dominated crust. In the few FA suite rocks that (as sampled) contain more than 1–2% mafic silicates, the mafics are typically a mix of 1mm olivine and coarsely exsolved low-calcium pyroxene (igneous
pigeonite) (Papike et al., 1991, 1998; Taylor et al., 1991). Another apparent geochemical discontinuity between the FA and Mg suites is manifested by nickel–cobalt systematics in olivines (Shearer and Papike, 2005). Early models designed to account for all pristine nonmare
rocks as magma ocean flotation cumulates (e.g., Longhi and Figure 11 Transmitted light image of rocklet 14286,5 (Albrecht et al, 1995), which consists mainly of Ni-rich meteoritic metal (kamacite, black), Boudreau, 1979; Wood, 1975) assumed that the mafic silicates yet has peripheral embayments that host a noritic silicate-impact-melt formed from locally high proportions of trapped melt. Namur lithology in which the grains are relatively coarse and equant, almost in et al. (2011) recently attempted to revive this hypothesis. the manner of pristine rocks. Such a texture should give pause to anyone The Stillwater Complex is often cited as an analog (Raedeke who contemplates relaxed criteria for identifying pristine rocks. and McCallum, 1980), but the Stillwater is of complex,
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 233
1 situ measurements (Floss et al., 1998), have only strengthened that evidence. Namur et al. (2011) also argued that the poiki-
litic textures of some FA suite rocks imply abundant trapped melt. But Barnes and Hill (1995) showed conclusively that 0.3 poikilitic texture in cumulate rocks is not an indication of abundant trapped melt. Also, some ureilites, feldspar-free meteorites with only trace amounts of trapped melt, are clear-
0.1 ly poikilitic (Goodrich et al., 2004). 72275c FA suite rocks have distinctive but not entirely uniform com- positions, and James et al. (1989) suggested a subclassification of the suite into four ‘subgroups’ based on subtly different mg
Bulk molar Na/(Na + Ca) 0.03 ratios of mafic silicates and alkali contents of plagioclase.
Obviously, such terms can be useful for descriptive purposes, but our sampling of the FA suite (mainly from one site, Apollo (a) 16) is probably inadequate for assessing whether the proposed subgroups reflect compositional discontinuities within the suite 100 or arbitrary divisions of a compositional continuum.
Most of the other pristine nonmare rocks are Mg-suite
rocks: troctolitic, noritic, and gabbronoritic rocks that, based 30 on evidence from a few texturally pristine samples (e.g., Dymek 1000 72275c et al., 1975; Marvin and Warren, 1980), probably all formed as ϫ cumulates. The highest mg cumulates include several with ultramafic modes, but only one of these, dunite 72415 10 (Dymek et al., 1975), is over 1 g (Figure 13), so many of them may actually be grossly unrepresentative samples of troc- tolites. High-calcium pyroxene apparently came relatively late in the crystallization sequence of most Mg-suite magmas; gab- 3 Bulk Eu/Al wt ratio FA suite bronorites are relatively rare, and they tend to have lower mg Mg suite þ Gabbronorites and higher Na/(Na Ca) than norites (Figure 12(a)). They Evolved rock types were first recognized as a distinctive subdivision of the Mg Mare basalts 1 suite by James and Flohr (1983). As reviewed by Papike et al. 0 102030405060708090100 (1998; cf. Shervais and McGee (1999)), the most evolved (b) Bulk Mg/(Mg + Fe), mol% pristine rock types, alkali (high Na/Ca) suite rocks and rare
Figure 12 Examples of the geochemical bimodality of pristine nonmare granites (and similar quartz monzodiorites; also fine-grained rocks: (a) Na/(NaþCa) molar ratio and (b) Eu/Al weight ratio versus bulk- felsites), may be extreme differentiates related to the Mg suite rock mg. The category ‘evolved rocks’ here includes such rock types as and/or KREEP. Some of the best-sampled granitic rocks show granite, felsite, quartz monzodiorite, and KREEP basalt. The database clear millimeter(þ)-scale effects of silicate liquid immiscibility used for these diagrams comprises all known nonmare rocks with (Jolliff et al., 1999; Warren et al., 1987).
‘pristinity confidence index’ 6 as compiled by Warren (1993); more Isotopic studies have constrained the genesis chronology of recent data sources include Arai and Warren (1997), Jolliff et al. (1999), the various types of pristine rock (Figure 15). Dating the Zeigler et al. (2000) (gabbronorite 62283,7-15 only), Neal and Kramer always REE- and rubidium-poor FA suite poses a severe analyt- (2003), plus a few unpublished analyses by the author. On the Eu/Al ical challenge (the 40K–40Ar system seems far more prone than diagram (b), the single gray diamond represents quartz monzodiorite 147Sm–143Nd and 87Rb–87Sr to undergo resetting during the 14161; its actual Eu/Al ( 1000) is 126 (Jolliff et al, 1999), but it is plotted slow cooling and multiple intense shocks that ancient pristine at a slightly lower value (90) to avoid compressing the y-axis of the diagram. For gabbronorite 62283,7-15, the mg ratio is inferred to be rocks have typically endured). Considering the unique suitabil- 40 mol% based on published mineralogy (Zeigler et al., 2000). ity of the FA suite to represent flotation crust from the putative magma ocean, 147Sm–143Nd isochron ages obtained for FA multimagma origin, and Haskin and Salpas (1992) showed samples have been surprisingly diverse, ranging from from trace-element data that the Stillwater’s anorthosites (the 4.56 0.07 (Alibert et al., 1994) down to 4.29 Ga (for
supposed FA suite analogs) contain very little trapped melt: 62236; Borg et al., 1999). Borg et al. (2011) determined a “We might have expected ... that the proportion of trapped Sm–Nd age of 4.37 Ga for 60025, and also, using Pb isotopes, intercumulus liquid would not exceed about 50%; it never 4.36 Ga for the same sample. These authors, cognizant of the reaches 9%” and averages 4.5%. Likewise, as shown by Warren extreme technical difficulty in making these measurements, and Wasson (1980a; Figure 3), the more mafic members of the argue that their 4.36 Ga Pb result is currently the most trust-
FA suite, such as 62236 (Borg et al., 1999)and62237(Warren worthy age for FA suite origin. Unfortunately, this age may and Wasson, 1978), have only slightly higher REE concentra- not reflect original igneous crystallization because 60025 is tions than the purest anorthosites, indicating that the mafics probably not fully pristine. A detailed petrological study are almost purely cumulus (or geochemically equivalent (James et al., 1991) indicated that 60025 is “polymict, consist- ‘adcumulus’) material; that is, in FA rocks, the trapped melt ing of a mixture of ferroan-anorthosite lithologies that differ components are uniformly low. Newer analyses, including in significantly.”
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
234 The Moon
100
90 Float ??? 80 Sink
70
60
50
40
Modal feldspar (vol%) 30
20
10 FA suite Other rock types 0 1 10 100 1000 10000 Sample mass (g)
Figure 13 Ferroan suite rocks tend to be more anorthositic than any other type of pristine lunar rock. Most of the ‘other’ large pristine rocks are troctolitic and noritic cumulates of the Mg suite. Data are from the compilation of Warren (1993).
90 Another factor that must be considered in interpreting ancient lunar ages is the strong possibility that such ages typ- ically postdate igneous crystallization by many tens of Ma due 80 S161 to subsolidus cooling within or below a vacuous-porous, insu- A81 lating megaregolith (Warren et al., 1991a) that may have been 70 L20 subjected to intense impact-shocks. The importance of the D081 megaregolith annealing effect has probably been underesti- 60 Ap14 14076 mated in the past. GRAIL data suggest that the porosity of the upper crust is ‘substantially’ greater than previously believed 50 (Wieczorek et al., 2013). In a most important study, Gaffney
et al. (2011) have shown that in preheated samples shock can
Mg/(Mg+Fe), mol% 40 KREEP/mare-poor regolith have major effects on isotopic systems, even 147Sm–143Nd. Other KREEP-poor lunaites Linear (Mg/Fe) regression Norman et al. (2003) inferred that the plagioclase in the 30 Mare-highland mixtures FA suite rocks may have been compositionally modified, most Pure mare basalt meteorites likely during a large impact event many hundreds of Ma Komatiitic Earth rocks 20 after igneous crystallization. When Norman et al. (2003)
0 5 10 15 20 25 30 35 constructed a 147Sm–143Nd isochron using mafic silicates Al O , wt% 2 3 only (from four different FA samples, including 62236), they found an age of 4.456 0.040 Ga, and they suggest this may Figure 14 mg versus Al2O3 for lunar meteorites and averaged highland regolith samples from Apollo and Luna sites. The data set be closer to the true igneous crystallization age than any of used for KREEP-poor highland regoliths comprises lunar meteorite the individual FA rock isochrons. breccias ALH81005, DaG 262, Dhofar 025, Dhofar 1428, MAC88105, As reviewed by Shearer et al. (2006) and Snyder et al. NWA 2200, QUE93069, Shisr 160 and Shisr 161; Apollo 14 sample (2000), Mg-suite rock ages distribute fairly evenly across a 14076,1 (Jerde et al, 1990); and average Apollo 16 station 11 regolith range from 4.5 Ga (very close to the age of the Moon itself) (Korotev, 1981). The other KREEP-poor lunar meteorites are polymict down to 4.1 Ga. However, for one of the oldest samples, breccias DaG 400, Dhofar 026, Dhofar 081, NWA 482, and Y-82192. Edmunson et al. (2009) recently determined a slightly younger The mare-highland mixture lunar meteorites shown are EET87521, age of 4.33 0.04 Ga. The oldest 147Sm–143Nd-based age is QUE94281, and Y-793274. The pure mare basalts are Asuka-881757, NWA 032, and Y-793169. Data for komatiitic Earth rocks are from 4.46 0.07 Ga, from a clast in breccia 15445 (Shih et al., BVSP (1981). The curve represents an extrapolated linear regression 1993). In a very general way, the sequence of typical or median for all of the plotted KREEP-poor highland samples. Also shown is the age for the various major pristine rock types appears to be FA position inferred for the pure highland component of the Luna 20 suite, Mg suite, gabbronorite (subset of the Mg suite), alkali regolith sample. suite, granites and felsites, KREEP basalts, and mare basalts.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 235
Origin of the Solar System Origin of Earth Origin of the Moon
Heavy impact ? Bombardment ?
Mare basalt source ? FA FA suite ? Suite
Mg KREEP source Relative Suite volume of Mg nonmare Suite crust Alkali Suite Granites & Felsites KREEP Basalts
Mare Volcanism
2.6 3.0 3.4 3.8 4.2 4.6
Age range, Ga
Figure 15 Summary of the chronology of lunar crustal genesis, based mainly on data previously reviewed by Papike et al. (1998). Other noteworthy sources include Carlson and Lugmair (1979) and Shih et al. (1992) for KREEP and mare basalt ages, including source model ages, and Borg et al. (1999) for the lowest of arguably accurate/relevant ages for a ferroan anorthositic (FA) suite rock ( 4.29 Ga, for 62236). The inset cartoon indicates (roughly) the relative volume of crust produced at the various stages; that is, the main volume of the crust probably formed as FA suite flotation cumulate, whereas evolved rocks like granites and KREEP basalts are probably volumetrically minor.
Caveat: Before leaving the topic of pristine rocks, it should studies of large terrestrial impact sites have found evidence be noted that whole classes of samples classified as pristine for a limited form of fractional crystallization, that is, produc- might conceivably be products of exceptionally large-scale tion of rocks that to some extent resemble pristine-igneous impact melting. The Mg-suite rocks are probably the most materials (Darling and Moser, 2012; O’Connell-Cooper et al., likely candidates for such an impact-melt genesis (Delano 2012). However, the layering in these impact melts may be and Ringwood, 1978; Hess, 1994; A. E. Ringwood, pers. largely inherited from never-homogenized compositional comm., 1993). The diverse Mg-suite ages (Figure 15) imply structure within the target terrain (Zieg and Marsh, 2005). any impact-melt model is inconsistent with the bombardment For example, Darling et al. (2010) inferred from lead isotopic cataclysm hypothesis, at least in its extreme form as advocated data that for the Sudbury Complex, “it seems unlikely that the by Ryder (1990). The finite number of old basins is a major impact melts were fully homogenised at any stage.” stumbling block for the Mg-suite-as-impact-melt-products hypothesis because the hypothesis requires derivation of each 2.9.4.5 The Magma Ocean Hypothesis Mg-suite sample through a combination of two major impacts.
Impact 1 is required to engender the exceptionally large pond The original lunar magma ocean hypothesis (Wood et al., of impact melt that forms the Mg-suite cumulate. Impact 2 is 1970) was tied to a discovery that among 1700 tiny (nearly required to excavate the cumulate and add it to the surface all <2 mm) rocklets from the Apollo 11 Mare Tranquillitatis regolith. Two overlapping basins are probably required for soil, a distinctive minority (5%) of exotic (i.e., highland) origin such a scenario. Also, as noted by Warren et al. (1996),inan had anorthositic or ‘anorthositic gabbro’ mineralogy or com- event as gigantic as the oft-conjectured Procellarum impact, the position. Based on these few tiny samples, Wood et al. (1970) distinction between pristine and nonpristine becomes blurred inferred that the Moon’s main, highland crust formed as a and potentially even misleading, because most of the energy in globe-wide ‘surface layer of floating anorthite crystals’ over a the impact melt is inherited from the very warm (and deep) zone of “gabbroic liquid, of composition near the eutectic target region, and the mantle beneath ground zero probably value of the Moon as a whole.” Wood et al. (1970) claimed does not respond in a strictly passive way to the aftermath of to discern cumulate textures but admitted their samples were such a colossal near-surface heating event. Unfortunately, the “much too fine grained to have formed in such a grand event.” great South Pole–Aitken basin, located approximately antipo- In any event, Wood et al. (1970) saw that “the cumulate dal to the Apollo–Luna sampling region, still can only be interpretation must (also) be based on the specialized compo- studied from orbit (Garrick-Bethell and Zuber, 2005, 2009; sitions of anorthositic rocks.” Hagerty et al., 2011; Pieters et al., 2001; Yamamoto et al., The most impressive argument for a lunar magma ocean is 2012). In addition to the evidence of 14286 (Figure 11), still that same straightforward observation: the bulk
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
236 The Moon
composition of the globe-wide highland crust is far too anor- Time steps thositic to have formed by piecemeal aggregation of basaltic (1) Initial melt (or, for LMO, peak melt) partial melts ascended from the deeper interior. The normative ® 3 (2) First (major) plag crystallization plagioclase contents of internally generated partial melts are (3) Late; most of (2) melt has crystallized Moderate-degree melting OR a Al-Ca-rich Moon constrained by phase equilibria (e.g., Longhi and Pan, 1988)to be less than about 55 wt%, whereas the highly anorthositic ( 75 wt% normative plagioclase) composition of the global lunar crust has been amply confirmed in by lunar meteorites (or … liquidus crystals) 3 (Table 2) and remote sensing (Prettyman et al., 2006). meltingExtraordinarily and an Earthlike extensive Moon As discussed in the previous section, the premise for º Warren’s (1985) second argument favoring the magma ocean 2 1 2 hypothesis, the extreme antiquity of many pristine nonmare 1 rocks, has been called into question by Borg et al. (2011) but Eu/Al of melt strengthened by other work (Norman et al., 2003). ® A slightly newer argument, which at this juncture seems mg of melt (or, scaled, mg of its liquidus crystals) more impressive, is based on the geochemical bimodality of Figure 16 Schematic equivalent of Figure 12(b), illustrating the model the pristine nonmare rocks (Figure 12). The singularly low-mg of Warren (1986) for producing the geochemical bimodality among the composition of the FA suite must reflect a special process of pristine nonmare rocks. In the case of crystallization of a magma ocean, origin. Warren (1986) argued that a deep, originally ultramafic provided initial (or peak) melting was extensive enough and the bulk- magma ocean is the only plausible scenario for engendering a Moon material was not too rich in the plagioclase-limiting oxides Al2O3 global component of low-mg cumulate anorthosite while bury- and CaO, for much of the crystallization sequence, going from (1) to (2), ing the comparable (or larger) volume of low-mg mafic removal of mafic silicates (olivine and pyroxene) from the ultramafic magma ocean drives down its mg ratio while leaving plagiophile ratios (sunken) cumulates that must have formed simultaneously such as Eu/Al virtually unchanged. Only after plagioclase (i.e., the ferroan with the anorthosite so deep that they are absent among sam- anorthositic suite) begins to form, at (2), does the Eu/Al begin to pled lunar rocks. This scenario accounts for the peculiarly low fractionate. In the contrasting case of Mg-suite crystallization (or, in mg of the flotation crust, provided the initial magma ocean was principle, magma ocean crystallization, if Al2O3 and CaO are initially ultramafic, so that onset of (copious) plagioclase crystalliza- high), plagioclase is on the liquidus of the moderately mafic (basaltic) tion was preceded by extensive fractional crystallization of melt from the beginning (1 2), and the mg ratio never undergoes
fractionation without accompanying fractionation of plagiophile ratios mafic silicates, driving down the mg, yet not fractionating plagiophile ratios such as Na/(NaþCa) and Eu/Al, in the such as Eu/Al. In principle, assuming a more Al2O3- and CaO-rich Moon residual melt (Figure 16). A similarly low mg yet low Na/ and/or less extensive primordial melting, one of the intermediate trends (NaþCa) composition was probably never reproduced by would have developed, with the ferroan anorthositic suite less distinct from the Mg-suite. subsequent, post-magma ocean magmatism because these magmas originated by moderate (more or less conventional) partial melting of the mantle and thus were already at or near A fourth argument in support of the magma ocean hypoth- plagioclase saturation even as they ascended into the crust. esis, formulated by various authors in the mid-1970s (most Note that this magma ocean aspect of the model requires that notably Taylor and Jakes, 1974), holds that the prevalence of the bulk-Moon composition (and thus the initial or peak (–) europium anomalies among plagioclase-poor mare basalts product of very extensive magma ocean melting) is earthlike implies predepletion of massive proportions of plagioclase,
þ in the sense of not having great enrichments in the plagioclase- with ( ) europium anomaly, from mare basalt source regions. limiting oxides, Al2O3 and CaO, or in FeO/MgO ratio. Also, the As discussed earlier, the assumption that the mare basalt Mg-suite aspect of the model works best if the initial magma europium anomalies can only reflect plagioclase fractionation ocean melting was not so extreme (near total-Moon) as to has been challenged, but it still appears sound (Brophy and eliminate substantial Al O - and CaO-rich materials in the Basu, 1990). 2 3 deep mantle because such materials are needed to yield initial Warren (1985) also cited the uniformity of the trace- Mg-suite magmas at or near plagioclase saturation. element pattern of KREEP, suggesting derivation from the Even if the parent Mg-suite magmas of some of the highest- residuum (urKREEP) of a single, global magma. This argument mg dunites and troctolites initially reached the crust undersat- has been weakened because remote sensing (Lawrence et al., urated with plagioclase, assimilation of FA country rock by 2007; cf. Warren, 2003) and lunar meteorites (Table 2) have superheated, adiabatically decompressed melt may have cur- revealed that KREEP is only abundant in one relatively small tailed the extent of the fractional crystallization of mafic sili- area, that is, the Procellarum–Imbrium (PKT) region. How- cates prior to plagioclase saturation (Ryder, 1991; Warren, ever, the same revelation about KREEP distribution justifies a 1996). Hess (1994) noted that such assimilation could never sixth argument: the need to account for the extremely hetero- be very extensive, but Warren (1986) assumed that real-world geneous distribution of KREEP, in stark contrast to the almost assimilation would seldom involve total melting. The assimi- uniform (anorthositic) major element composition of the lation could be mostly physical; that is, plagioclase grains that global crust. Both the Procellarum–Imbrium region and the became loosened by disaggregation of anorthositic wall rock giant, nearly antipodal South Pole–Aitken basin have excep- (at say 30% melting), and after a while settled at the base of an tionally mafic (noritic) compositions, yet strong KREEP enrich- Mg-suite intrusion, might be impossible to distinguish from ment is found only at Procellarum–Imbrium (Lawrence et al., plagioclase grown entirely from the parent melt. 2007; Pieters et al., 2001); so the PKT anomaly is not simply a
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 237
byproduct of the large-scale cratering of that region. The The Mg-suite plutons would inevitably interact extensively uniqueness of the PKT region suggests that a globe-wide, with urKREEP stewing at the base of the FA crust, and this extremely KREEP-concentrated magmatic plumbing system, seems a major advantage of the model because pristine such as a declining magma ocean, channeled and concentrated KREEP basalts have surprisingly high mg ratios (Warren, incompatible elements into this one region. The failure of 1988), and Mg-suite cumulates appear to have formed from KREEP to concentrate as well into the South Pole–Aitken remarkably KREEP-rich parent melts (e.g., Shearer and Papike, basin may be a consequence of its origin at a later stage, after 2005; Shervais and McGee, 1999). the magma ocean had already (virtually) solidified. Garrick- One obvious reason for skepticism regarding the magma Bethell et al. (2010) argued that interaction between tidal force ocean hypothesis is the colossal energy input that is required to and the magma ocean was responsible for the greater thickness engender a fully molten layer encompassing a large portion of of the farside crust (cf. Loper and Werner, 2002). the Moon (e.g., Wetherill, 1981). That concern has been A seventh argument in support of the magma ocean stems allayed by the giant impact hypothesis of lunar origin from the work of Longhi et al. (2010), who noted that a variant (Canup, 2004; Pahlevan et al., 2011). Today, skepticism of that hypothesis represents the most likely explanation for stems mainly from isotopic studies of FA suite rocks. Part of the paradox that Mg-suite rocks tend to have lower Ni and Co the problem lies in the young apparent ages (younger than concentrations in olivine in comparison to mare basalts some Mg-suite rocks of manifest non-magma ocean origin) (Shearer and Papike, 2005). Longhi et al. (2010) noted that for samples such as 62236, 4.29 Ga (Borg et al., 1999) the partitioning of nickel and cobalt are such that among and 60025, 4.37 Ga (Borg et al., 2011). However, as dis- cumulates from a deep magma ocean the late, FeO-rich cumu- cussed earlier, 60025 appears polymict, Norman et al. (2003) lates (i.e., future mare basalt sources) would have been mark- found that neodymium isotopic data from four FA suite rocks edly enriched compared to the earliest, MgO-rich cumulates define a ‘robust’ isochron, aged 4.456 0.040 Ga, and ancient (future Mg-suite sources). ages are probably sensitive to the siting of the material relative Many authors have constructed detailed models for the to the Moon’s porous megaregolith. e crystallization of a lunar magma ocean (e.g., Brophy and A decade ago (Warren, 2003), the positive Nd typically Basu, 1990; Elardo et al., 2011; Elkins-Tanton et al., 2011; found for FA suite rocks was widely perceived as a major cause Longhi et al., 2010; Loper and Werner, 2002; Snyder et al., for skepticism regarding magma ocean genesis. Norman et al.
1992; Suckale et al., 2012; Warren, 1990; cf. the early review of (2003) found an initial eNd of þ0.8 0.5 for FA suite sample Warren, 1985). One key constraint is that silicate adiabats have 67215c. Other positive initial e for FA suite rocks were Nd small dP/dT relative to melting curves, so crystallization occurs reported by Alibert et al. (1994), Borg et al. (1999), and Nyquist primarily along the base of the magma ocean (Thomson, et al. (2006, 2010). As discussed by Warren (2003), these 1864). This is important because pressures of the order 0.1– positive initial eNd ratios arguably imply derivation of the FA 4 GPa enhance stability of pyroxene at the expense of olivine. suite from material that had spent the prior few tens of Ma Geochemists commonly assume that the lunar magma ocean (between origin of the solar system and igneous crystallization) produced mainly olivine until late in its crystallization with a remarkably fractionated, light REE-depleted Sm/Nd sequence. More realistically, even the early cumulates consisted ratio. No conventional magmasphere model predicts light largely of pyroxene (Warren and Wasson, 1979b), with impor- REE depletion at any point in the magma ocean’s evolution, tant compositional implications for the initial magma ocean, and least of all in its later stages. On the contrary, magma ocean for its early and deep cumulates, and for the bulk-Moon com- crystallization models (e.g., Snyder et al., 1992) indicate that position. All models indicate that the late magma ocean crystallization of pyroxene tends to enrich light REE over heavy becomes (thanks to its low fO2) highly enriched in FeO, REE as the magma ocean evolves to approach plagioclase which gives the FA suite parent melt a high density, which in saturation (i.e., FA suite flotation). However, the Borg et al. turn allowed a considerable proportion of mafic silicate to be (2011) 147Sm–143Nd isochron for 60025 (with its Pb–Pb- rafted within the FA flotation crust, as modeled quantitatively supported age of 4.37 Ga) gives a precise eNd that is negative, by Warren (1990). –0.24 0.09. Also, Norman et al. (2003) and Nyquist et al. e As reviewed by Shearer and Papike (2005), origin of the (2006) noted that the value of Nd is sensitive to the choice of e other major class of ancient pristine rocks, the Mg suite, is normalizing ratio. Thus, Nyquist et al. (2006) regarded an Nd widely ascribed to more conventional layered intrusions, of þ0.19 0.13 as consistent with crystallization “from a emplaced in the FA crust during and shortly after the magma feldspar-rich lunar magma ocean evolved from ‘chondritic’ ocean waned into its final dregs of urKREEP (e.g., Ryder, 1991; relative REE abundances.” Yet another complication is that
Shearer and Papike, 2005; Warren and Wasson, 1980a). In a recent evidence (discussed in the final section of this chapter) variant of this model, Hess (1994) suggested the layered intru- suggests that the bulk Moon may have a depleted Sm/Nd ratio. sions may have formed as the impact melts of a few of the largest lunar accretionary events (cf. Section 2.9.4.4). Mg-suite 2.9.4.6 Alternative Models plutonism may have been triggered by the enhanced convec- tive potential that many authors (e.g., Hess, 1994; Hess and Models have been proposed that invoke no magma ocean, Parmentier, 1995; Ryder, 1991; Warren and Wasson, 1980a; although not in any detail for many years. Walker (1983) Zhong et al., 2000) have noted would likely ensue if reviewed the evidence for petrogenetic diversity among the the magma ocean crystallized into a gravitationally unstable various types of pristine rocks and suggested that the lunar configuration: dense, FeO- and ilmenite-rich residual mush crust formed by ‘serial magmatism’ as a series of flows and atop relatively FeO- and ilmenite-poor early cumulates. plutons. However, Walker (1983) offered little justification for
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
238 The Moon
his implicit assumption that the entire lunar magma ocean depletion of aluminum in the source region of Ti-poor mare hypothesis should be rejected simply because important (yet pyroclastic glasses. subordinate to the FA suite) components of the crust formed Shirley (1983) proposed a model that is in some respects by processes other than magma ocean cumulate flotation. intermediate between ‘serial’ magmatic models and the magma Walker (1983) conceded that a magma ocean is implied “if ocean hypothesis. In this model, only a relatively thin zone the bulk crust ...(is) significantly more feldspathic than inter- near the top of the system is fully molten; the rest of the nally generated partial melt.” As discussed earlier, it is now differentiating system, or at least a large portion immediately clear that the lunar crust (or at least its upper half) contains below the magma ocean, is a convective, mostly crystalline upward of 70 wt% plagioclase. mush that Shirley (1983) called ‘magmifer.’ In a quasi-steady Wetherill (1981) argued that accretional heating could not state, the magmifer continually bleeds partial melt into the melt a large fraction of the Moon, and yet, he suggested, it magma ocean, which replenishes the mush layer by precipitat- might have produced the Moon’s anorthositic crust. In this ing mafic crystals (while simultaneously floating buoyant model, the Moon is zone-refined by production of many anorthosite to form a FA suite crust). Warren (1985) suggested þ regional magma chambers as the central impact melts of the the term magmasphere for the combined (magmifer thin biggest accretionary events. As the Moon grows, these impact- magma ocean) partially molten system. This model has an melt plutons differentiate so that incompatible elements and obvious thermal advantage: most of the Moon can be differen- low-density plagioclase concentrate upward. Melting is discon- tiated by the magmasphere, without fully melting the differen- tinuous in space and time, but the ultimate, cumulative effect is tiated volume. Realistically, any ‘magma ocean’ must have a global surface layer of anorthosite. Similar models were some magmifer/magmasphere characteristics, but it is unclear earlier (and briefly) mentioned by Wetherill himself, by whether a lunar magmifer could sustain itself against upward Alfve´n and Arrhenius (1976), and by Hess et al. (1977). percolation of buoyant melt (McKenzie, 2000) over a thick The most detailed variant of this type of model was pro- enough depth range, with a large enough fraction of melt, to posed by Longhi and Ashwal (1985). In their model, the have any appreciable effect. A major role for equilibration upward concentration of anorthosite is enhanced by diapiric between the magma ocean and a thick magmifer would have detachment and ascent of feldspathic portions of the layered a major disadvantage: It would dampen, if not cancel, the impact-melt plutons of the Wetherill (1981) model. Partial potential for early fractional crystallization of ultramafic, melting, partly due to pressure release, would lubricate the plagioclase-undersaturated magma (see above) to engender diapirs and enhance their potential for poking through the the peculiarly low-mg ratio of the FA suite pristine rocks. It cool, mostly solid near-surface layer (and would be consistent would also tend to reduce the tendency for heterogeneity to with an argument (Haskin et al., 1981) that ferroan anortho- develop in the mantle (the magmifer is well-stirred, like a melt- sites may be residual solids from episodic partial melting). lubricated variant of Earth’s present-day asthenosphere), and Through the years (e.g., Borg et al., 1999), this anorthosite thus come into conflict with evidence (e.g., Figure 1; hafnium diapirism model has been widely cited as the strongest com- isotopic data of Beard et al., 1998; Unruh et al., 1984) for petition for the magma ocean hypothesis. However, the diapir- extreme heterogeneity in the source mantle for mare basalts. ism model has some serious drawbacks. It implies that the anorthositic early crust should consist mainly of annealed (and possibly sheared and deformed) restite. But among the 2.9.5 Water in the Moon few pristine anorthositic (FA suite) samples that are little- 2.9.5.1 Traditional View of a Dry Moon brecciated, only one (15415: Taylor et al., 1991) shows an annealed texture consistent with a restite origin. The majority Beginning with the first studies of Apollo 11 samples, which (e.g., 62237, Dymek et al., 1975; 66035c, Warren and Wasson, contained abundant fresh basalts containing no hydrous min- 1980b; 64435c, James et al., 1989) have igneous textures. Also erals or signs of alteration in spite of being billions of years old, (in common with all other published models for early lunar available evidence indicated that the Moon was essentially petrogenesis without a magma ocean), the diapirism model anhydrous. Even intrusive rocks, which would have lost less seems ill-suited to account for the geochemical bimodality of water than erupted lavas, were not shown to contain water- pristine rocks, that is, the observation that nearly all of the bearing minerals. In addition, many samples contain small pristine rocks with the distinctive ‘ferroan’ geochemical signa- grains of metallic iron that show no signs of oxidation or ture (Figure 12) are anorthosites (Figure 13), and vice versa. reaction with hydrogen species. Bulk chemical analyses of Why should such a relationship have developed, and in such lunar basalts indicate <0.02 wt% H O(Haskin and Warren, 2 consistent way, if the crust formed by aggregation of diapiri- 1991), an unknown amount of which might be terrestrial cally mobilized components from many separate and isolated contamination. In contrast, fresh terrestrial basalts erupted on intrusions? Highly inconsistent results seem inevitable, consid- land typically contain a few tenths of percent H2O(BVSP, ering that the original plutons are supposed to form and crys- 1981), although much of the H2O is added by aqueous alter- tallize at variety of depths (i.e., pressures), and would be prone ation. Seismic data (e.g., Tittmann et al., 1972) showed that
` to diverse fates, vis-a-vis regional remelting events during seismic waves were not attenuated in the Moon as much as the prolonged but spotty accretional heating of the Moon. By they are on Earth, implying an absence of water-bearing rocks the same token, the diapirism model seems ill-suited to and wet sediments in the Moon. These data were interpreted account for the characteristic uniformity of the KREEP compo- to mean that the Moon contains many orders of magnitude nent (Warren and Wasson, 1979a). Longhi (2003) advocated a less water than does Earth. However, improved analytical tech- deep primordial magma ocean, citing evidence for major niques, particularly secondary ion mass spectrometry (SIMS),
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 239
show that this long-held interpretation is incorrect. Water is Thus, using the extremes of measured apatite H2O, the mare present inside the Moon, and its abundance and isotopic com- magmas contained between 100 and 1500 mgg 1 H O when 2 position may provide useful tracers of lunar formation and apatite crystallized. Apatite crystallized late, after 95–99% of magmatic evolution. the lava had crystallized (Tollari et al., 2006), and because H2O behaves as an incompatible element (e.g., Bolfan-Casanova, 2005), it would have increased in concentration by a factor of 2.9.5.2 Water in Pyroclastic Glasses 20–100 over the value in preerupted magma. This implies that
Using SIMS techniques developed to analyze H2O and other the initial water content of the preeruptive magmas was 1 volatiles in terrestrial volcanic glasses and nominally anhy- between 1 and 75 mgg . However, the magma would have drous minerals (e.g., Hauri et al., 2006), Saal et al. (2008) lost a significant fraction of its water during lava crystallization. 1 reported H2O contents up to about 30 mgg in Apollo 15 The Saal et al. (2008) calculation for diffusive loss from pyro- green glass and up to 15 mgg 1 in Apollo 17 orange glass. clastic lava suggests water loss might be as high as 98%. Thus < m 1 Most analyses were above the detection limits of 6 gg water loss might be compensated by water increase in the H2O. (We express the H abundance as equivalent H2O, residual magma during crystallization, suggesting that, to first although it is probably present as OH– in the glasses and order, initial water concentrations are roughly the same as the þ apatite. A significant amount may be present as H .) Eruption water content of the residual magma in which apatite formed, 1 leads to significant loss of H2O because of the strong pressure between 100 and 1500 mgg . Note that this is similar to the dependence on H O solubility (e.g., Moore et al., 1998), so range in pyroclastic glasses. 2 these values of a few tens of microgram per gram were certainly These data clearly change our view of the Moon as an higher in the preeruptive magma. To estimate this, Saal et al. anhydrous body. However, Sharp et al. (2010) suggested that used a nanoSIMS to measure H2O, chlorine, fluorine, and we not throw out the old view entirely. They found that chlo- sulfur from center to edge of a homogeneous green (very low rine isotopic compositions in lunar soils, breccias, and basalts, Ti) glass bead. Using measured diffusion coefficients, they including the pyroclastic orange glass, show a large range in the 37 35 estimated the cooling time and initial H2O and volatile ele- Cl/ Cl ratio. In contrast, this ratio is strikingly uniform in ment contents required to produce the profiles. The surprising terrestrial samples (from both the crust and mantle) and in result was an estimated H2O concentration of between 260 meteorites. Sharp et al. (2010; cf. McCubbin et al., 2011) and 745 mgg 1, in the range of values reported for quenched suggested that the difference between Earth and Moon arises volcanic glass at mid-ocean ridges (e.g., Saal et al., 2002). because of the lack of available hydrogen in lunar magmas. On 35 To avoid the complication of loss during eruption and Earth, the lighter Cl isotope is lost preferentially to a vapor 37 associated errors in calculating the initial H2O concentrations, phase, while the heavier Cl isotope is incorporated preferen- Hauri et al. (2011) searched for, found, and measured H2Oin tially into HCl gas. These processes balance one another, blunt- trapped melt inclusions in olivine microphenocrysts in Apollo ing the fractionation of the chlorine isotopes. If no water 17 orange glass beads (sample 74220). Concentrations ranged (hydrogen) is present to provide the hydrogen to make HCl from 615 to 1410 mgg 1, confirming the diffusion calculations gas, the lighter isotope is preferentially lost and the rocks end 37 and the similarity in H2O concentrations to terrestrial mid- up with higher Cl, as shown by the data. Using a distillation ocean ridge basalts. The clear implication is that some regions calculation, Sharp et al. (2010) estimated that the lunar 1 of the lunar interior have H2O contents like those in the interior’s H concentration is roughly 10 ng g , less by several terrestrial mantle. orders of magnitude than the concentration inferred from
SIMS analyses. However, given the unambiguous identification
of H O in pyroclastic glasses and in apatites in mare basalts, it 2.9.5.3 Water in Apatite in Mare Basalts and KREEP-Related 2 seems likely that the chlorine isotopic data record factors other Samples than low lunar H2O content, such as the effect of low oxygen Of course, pyroclastic materials, almost by definition, are prod- fugacity and hence higher H/OH in lunar compared to terres- ucts of extraordinarily volatile-rich eruptions. The mineral apa- trial magmas. tite is common as a late-crystallizing phase in crystalline mare Apatite grains in KREEP-containing samples contain signif- basalts and nonmare KREEP-related rocks, such as KREEP icantly less H2O than those in mare lavas (Greenwood et al., basalts, alkali suite rocks, and many norites and troctolites. It 2011; McCubbin et al., 2010; Robinson et al., 2012b,c), rang- contains a crystallographic site for monovalent negative ions, ing from below detection limit ( 25 mgg 1) up to only usually Cl–,F–,andOH–, although others can also be present. 190 mgg 1. Even the largest concentration reported so far is
OH concentrations (usually reported as H2O) in mare basalt 5–30 times lower than in mare basalts. The low H2O in KREEP- apatites have been reported by McCubbin et al. (2010), Boyce rich magmas is significant for two reasons. First, the samples et al. (2010), Greenwood et al. (2011),andBarnes et al. (2012). measured (alkali anorthosite, quartz monzodiorites, and fel- The analyzed apatite grains (in both medium-Ti and Ti-rich sites and ferrobasalts that probably formed by liquid immisci- mare basalts) contain between 1000 and 6000 mgg 1 H O. bility) are products of fractional crystallization in intrusions, 2 Estimating the H2O content of the preeruptive magmas is hence subject to considerably less H2O loss than mare basaltic fraught with uncertainties. It requires knowing the partition lava flows. Assuming no water loss during crystallization in coefficient of H2O between melt and apatite, how much crys- intrusions and that apatite crystallized after 95% crystallization tallization occurred before apatite crystallized, and how much of KREEP magma, these data imply that the initial water con- hydrogen was lost from the lava. McCubbin et al. (2010) centration of the magmas was between 30 and 90 mgg 1. estimated that D (apatite/melt) is between 0.1 and 0.25. Second, as noted by Elkins-Tanton and Grove (2011),H2Oas
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
240 The Moon
an incompatible element ought to be enriched in the crystalli- high H2O and high concentrations of highly volatile elements, zation products from the magma ocean, hence in KREEP rocks. such as bismuth and thallium; these are notoriously enriched
This bears on the question of water delivery to the Moon (see in orange glass but low in mare basalt, yet both magma types Section 2.9.5.5). and source regions have similar H2O contents. The heteroge- Greenwood et al. (2011) and Robinson et al. (2012b,c) neity in H2O concentration makes an estimate of bulk lunar measured the D/H ratio and total H2O in apatite in KREEP- H2O highly uncertain. rich samples (Figure 17). Besides being distinctly lower in
H O, alkali anorthosite, felsite, and quartz monzodiorite 2 2.9.5.5 Implications of Water in the Lunar Interior have lower dD (230–430%) than mare basalts (390– % 1010 ). The mare lavas must have lost considerable H2O The low (though measureable) H2O in KREEP magmas upon eruption, and their elevated dD probably reflects that. is surprising given its incompatible behavior. Crystallization
Loss of H2O and concomitant fractionation of D from H is of the magma ocean produced the trace element enriched more difficult to assess for the intrusive KREEPy samples, but urKREEP after 98% or 99% crystallization of the magma may be minimal, implying that at least some water in the lunar ocean, yet KREEP basalt mantle source regions contained at interior has higher D/H than the bulk Earth appears to have least ten times less water than other basaltic source regions. d % ( Dof 62.5 ; Robert et al., 2000). This indicates that the magma ocean had a low H2Oconcentra- tion (Elkins-Tanton and Grove, 2011). It further implies, assum-
ing that the Moon formed fully molten, that the newly accreted 2.9.5.4 Water in the Lunar Mantle < m 1 Moon contained 1 gg H2O in order for the last dregs of The vast difference in water content between mare basalts and the magma ocean to contain <10 mgg 1. This can be explained pyroclastic glasses on the one hand and KREEP-related rocks by loss in the protolunar disk, assuming a giant impact origin on the other implies a heterogeneous distribution of H Oin of the Moon. A preliminary assessment of water loss from the 2 the lunar mantle. We can estimate the H2O concentrations in disk has been presented by Desch and Taylor (2011a,b, 2012), the lunar interior by assuming that on average magma produc- who conclude tentatively that almost all ( 98%) initial H2Oin tion involved 10% partial melting of mantle source regions. the disk would not accrete to the Moon. It is conceivable that
Since H2O is incompatible, source regions contained 10% of the KREEP-related magmas record the primary lunar water and the concentrations in intrusive magmas and preeruptive volca- that the somewhat elevated dD compared to Earth is due to nic. Thus, mare basalt source regions had H O contents of preferential hydrogen loss from the disk (Desch and Taylor, 2 m 1 10 and 150 gg , pyroclastic magma source regions had 2011, 2012; Robinson et al., 2012c). 1 60–140 mgg , and KREEP basalt sources contained between If the Moon did not accrete with significant H2O, how did 3 and 9 mgg 1. There is clearly no trend of increasing water the mare basalt and pyroclastic mantle sources regions acquire concentration with increasing REE. Nor is there a trend of 100 mgg 1? It is plausible that lunar water is dominated by
Apatite in lunar rocks 1000
800
600
Mare basalt d D (per mil) 400 Alk. anorthosite
Felsite (12013)
200 Felsite (77538)
QMD (14161)
0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
H2O (wt%) Figure 17 Hydrogen isotopic compositions and water contents of apatite crystals in mare basalts (Greenwood et al, 2011) and KREEP-related intrusive rocks. KREEP-related samples include an alkali anorthosite and felsite clasts in 12013 (Greenwood et al, 2011), felsite and ferrobasaltic clasts in 77538 (Robinson et al., 2012b,c), and a quartz monzodiorite fragment in 14161 (Robinson et al., 2012b,c).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 241
water added after lunar formation. If so, the late addition must The canonical late twentieth-century interpretation of the have arrived after the magma ocean had crystallized to prevent Apollo seismic data was that the thickness of the crust in
KREEP from being rich in H2O. The volatile-rich impactors the central nearside (Apollo 12/Apollo 14) region is 60 km and must blast through the FAN crust and deposit volatiles (both the global average thickness about 73 km (e.g., Taylor, 1982). water and highly volatile elements) as deep as the 500 km However, more thorough processing of the Apollo lunar seismic source depths for some pyroclastic magmas. The dDof dataset suggests the central nearside crustal thickness is the late-accreting water is not known. Mare basalts have quite only 34–38 km (Khan and Mosegaard, 2002; Lognonne´, d high D(Figure 17), but loss of 95–98% of the preeruptive 2005;cf.Ishihara et al., 2009). Wieczorek (2003) calculated water (as H and D) by Rayleigh fractionation is sufficient from global mass distribution constraints that the thickness to raise a terrestrial dD (–62.5%) to the range observed in cannot be less than 33 km, but the newest results from GRAIL mare basalts. It is imperative to constrain the preeruptive dD indicate that the crust’s bulk density is lower than previously of lunar basalts. supposed (Wieczorek et al., 2013;cf.Huang and Wieczorek,
2012), which implies a commensurate reduction in the esti- mated mass of the crust. In any case, the required Al2O3 and CaO to account for the lunar crust now appears greatly reduced. 2.9.6 The Bulk Composition and Origin of the Moon Meanwhile, the global thorium map (Lawrence et al., 2007;cf. Warren, 2005) and lunar meteorites (Table 2) have revealed
A vast literature exists concerning attempts to divine the bulk that the Apollo (central nearside) sampling region is highly composition of the Moon and to assess the implications for unrepresentative for KREEPy refractory lithophile elements, models of the origin of the Moon and Earth. The most obvious again relaxing the need to invoke refractory enrichments. areas of compositional contrast versus Earth are the lunar vol- A higher upper-crustal porosity also has implications for atile depletions (Figure 5) and the gross depletion in Fe–Ni interpretation of the precious few data on lunar heat flow (i.e., core matter) implied by the Moon’s low bulk density. The (Hagermann and Tanaka, 2006) and their implications for
Fe–Ni depletion is a key advantage of the giant impact model, the bulk-Moon contents of thorium and uranium (and indirectly which discriminates against central-interior matter as it forms all other refractory lithophile elements including Al2O3 the Moon (Canup and Asphaug, 2001; Canup, 2004). Geo- and CaO). physical studies (Khan et al., 2006; Weber et al., 2011) indicate The anticorrelation between Al2O3 and mg shown by that the lunar core’s radius is roughly 330 km, implying a KREEP-poor highland regolith samples including new lunar relative volume only 0.04 times that of Earth’s core. meteorites (Figure 14) extrapolates to a high mg for the mafic More controversial have been claims (e.g., Taylor, 1982; component of the highland crust. As noted by Warren (2005), Taylor and Jakes, 1974) that the bulk Moon is enriched this and other evidence (e.g., the extremely high mg of some roughly twofold in cosmochemically refractory lithophile ele- Mg-suite rocks; Figure 12) is difficult to reconcile with a bulk ments (a class that includes the REE, the heat sources thorium Moon that is greatly enriched in FeO compared to Earth’s and uranium, and the major oxides Al2O3 and CaO) and that primitive mantle. Apart from extrapolations based on mare compared to Earth’s primitive mantle, the Moon’s silicate mg basalts (whose sources may have been atypical of the overall ratio is much lower, and its FeO concentration is much higher. mantle), arguments in favor of FeO enrichment include geo- Neither of these claims has been confirmed by subsequent physical modeling for constraints such as the Moon’s moment developments, which include the advent of global thorium of inertia factor (Khan et al., 2006; Taylor et al., 2006). and samarium maps (Lawrence et al., 2007; Prettyman et al., The saga of the estimated bulk-Moon refractory lithophile
2006), data from lunar meteorites, and some radically changed composition took a surprising twist in recent years, as evidence interpretations of the Apollo seismic database. mounted that the bulk Moon (Boyet and Carlson, 2007) as well Starting with Cameron and Ward (1976), the putative refrac- as Earth and Mars (Brandon et al., 2009; Caro et al., 2008) may tory enrichments have been claimed as an advantage of the high be depleted in light REE relative to chondrites. The depletion is temperatures implied by most variants of the giant impact model constrained to have happened very early, probably within of Earth–Moon origin. It has often been claimed, or implied, that 20 Ma of the origin of the solar system (Caro, 2011). A popular the well-documented lunar volatile depletions (Figure 5)areso conjecture for Earth is that light REE were largely sequestered by great, that they must be a unique signature of origin by giant a petrologic quirk of the primordial Earth into a deep ‘hidden’ impact. However, a similar volatile depletion pattern is found for reservoir (e.g., Tolstikhin et al., 2006). But such a model cannot the eucrite meteorites (Figure 5), which are products of an aster- possibly be valid for Mars or (given the time constraints oid (or asteroids) that presumably was never involved in an discussed by Caro, 2011) the Moon. More generic processes Earth–Mars scale collision (see Chapter 1.6). Another common that have been suggested as mechanisms for depletion range misconception holds that the bulk-Moon volatile depletions from impact erosion (Korenaga, 2009; O’Neill and Palme, imply, by mass balance, a significant degree of refractory enrich- 2008) to planetesimal explosive volcanism (Warren, 2008a). ment. In fact, total depletion of every constituent with cosmo- Another complication is the likelihood (Warren, 1992) chemical volatility (solar nebula condensation temperature: that geochemical stratification in the impactor would be Wasson, 1985) between those SiO2 and H2O, from even the inherited by the Moon (i.e., the orbiting spall from the giant most volatile-rich (CI) type of chondrite, would increase the impact, biased against deep-interior matter), which would concentrations of all elements more refractory than Si by a factor thus acquire a bulk composition with slight thorium and of only 1.22. For 10 other types of chondrites (Jarosewich, 1990), light-REE enrichments relative to the two parent bodies the increase would be merely a factor of 1.02–1.09 (average 1.05). (Earth and the impactor).
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
242 The Moon
The timing of the giant impact is in principle constrained by Alibert C, Norman MD, and McCulloch MT (1994) An ancient age for a ferroan the 182Hf–182 W isotopic system, as tungsten is moderately anorthosite clast from lunar breccia 67016. Geochimica et Cosmochimica Acta 58: 2921–2926. siderophile (i.e., tungsten partitions into cores), whereas haf- Al-Kathiri A, Gnos E, and Hofmann BA (2007) The regolith portion of the lunar meteorite nium is lithophile. Interpretations of the tungsten isotopic data Sayh al Uhaymir 169. Meteoritics and Planetary Science 42: 2137–2152. lead to a range of lower limits for the time span between origin Allen CC, Greenhagen BT, Donaldson Hanna KL, and Paige DA (2012) Analysis of lunar of the solar system and the giant impact: 30 Ma (Yu and pyroclastic deposit FeO abundances by LRO Diviner. Journal of Geophysical Jacobsen, 2011), 42 Ma (Kleine et al., 2009), and 70 Ma Research, Planets 117: E00H28. Anand M, Taylor LA, Floss C, Neal CR, Terada K, and Tanikawa S (2006) Petrology and (Halliday, 2008). For further discussion, see Righter (2007) geochemistry of LaPaz Icefield 02205: A new unique low-Ti mare-basalt meteorite. and Nimmo et al. (2010). Geochimica et Cosmochimica Acta 70: 246–264. The precise match between Earth and the Moon in terms Antonenko I, Head JW, Mustard JF, and Hawke BR (1995) Criteria for the detection of of stable isotopes for a variety of elements, including oxygen, lunar cryptomaria. Earth Moon Planets 69: 141–172. Arai T, Hawke BR, Giguere TA, Misawa K, Miyamoto M, and Kojima H (2010) Antarctic silicon, titanium, and chromium (Armytage et al., 2012; lunar meteorites Yamato-793169, Asuka-881757, MIL 05035, and MET 01210 Liu et al., 2010; Spicuzza et al., 2007; Warren, 2011; Wiechert (YAMM): Launch pairing and possible cryptomare origin. Geochimica et et al., 2001; Zhang et al., 2012), although curiously not iron Cosmochimica Acta 75: 2231–2248. (Poitrasson et al., 2009; Weyer et al., 2005), represents a Arai T and Warren PH (1997) “Large” (1.7 g) compositionally pristine diabase 14434: powerful constraint. As noted by Hallis et al. (2010),itis A lithology transitional between Mg-gabbronorite and high-Al mare basalt. Meteoritics and Planetary Science 32: A7–A8. doubtful whether such an array of precise isotopic matches Arai T and Warren PH (1999) Lunar meteorite QUE94281: Glass compositions and other should be seen as a confirmation of the giant impact model. evidence for launch pairing with Yamato-793274. Meteoritics and Planetary Science In most variants of the model, the material that ultimately 34: 209–234. forms the Moon is derived mainly not from the proto-Earth Armstrong JC, Wells LE, and Gonzalez G (2002) Rummaging through Earth’s attic for remains of ancient life. Icarus 160: 183–196. but from the Mars-sized intruder. In the estimation of Armytage RMG, Georg RB, Williams HM, and Halliday AN (2012) Silicon isotopes in Canup and Asphaug (2001), the intruder’s contribution was lunar rocks: Implications for the Moon’s formation and the early history of the Earth. probably 80 wt%. Considering the easily measurable differ- Geochimica et Cosmochimica Acta 77: 504–514.
Arnold JR (1979) Ice in the lunar polar regions. Journal of Geophysical Research ences between the real Mars and Earth for most of these 84: 5659–5668. isotope systems (e.g., for chromium and titanium; Trinquier Ash RD, Knott SF, and Turner G (1996) A 4-Gyr shock age for a martian meteorite and et al., 2007, 2009), the giant impact model seems to require a implications for the cratering history of Mars. Nature 380: 57–59. remarkable degree of similarity between the putative giant Barnes JJ, Anand M, Franchi IA, et al. (2012) The hydroxyl content and hydrogen impactor and the final Earth. A possible explanation is effi- isotope composition of lunar apatites. Lunar and Planetary Science 43: #1797. cient homogenization by convective stirring between the Barnes SJ and Hill RET (1995) Poikilitic chromite in komatiitic cumulates. Mineralogy and Petrology 54: 85–92. protolunar impact-generated disk and the hot postimpact Beard BL, Taylor LA, Scherer EE, Johnson CM, and Snyder GA (1998) The source Earth (Pahlevan and Stevenson, 2007; Pahlevan et al., 2011). region and melting mineralogy of high-titanium and low-titanium lunar However, that model invokes a temperature for the protolunar basalts deduced from Lu-Hf isotope data. Geochimica et Cosmochimica Acta disk (4000 K) far higher than the temperatures typically found 62: 525–544. in the giant-impact modeling of Canup (2004). In all of Benoit PH, Sears DWG, and Symes SJK (1996) The thermal and radiation exposure history of lunar meteorites. Meteoritics and Planetary Science 31: 869–875. Canup’s 47 ‘favorable’ simulations, the initial protolunar disk Besse S, Sunshine JM, Staid MI, et al. (2011) Compositional variability of the Marius was found to consist mainly of liquid, with the proportion of Hills volcanic complex from the Moon Mineralogy Mapper (M3). Journal of vapor averaging 19% and maximally 30%. “The impactor Geophysical Research, Planets 116: E00G13. material that ends up in orbit is primarily that portion of the Bhatt M, Mall U, Bugiolacchi R, et al. (2012) Lunar iron abundance determination using the 2-mm absorption band parameters. Icarus 220: 51–64. object that was heated the least, having avoided direct collision Bogard DD (1995) Impact ages of meteorites, a synthesis. Meteoritics with the Earth.” Also, the Pahlevan et al. (2011) model can 30: 244–268. only account for the widely speculated bulk-lunar enrichment Bogard DD, Garrison DH, and Nyquist LE (2000) Argon-39-Argon-40 ages of lunar in FeO by invoking a melt-vapor differentiation process that highland rocks and meteorites. Lunar and Planetary Science 31: #1138. 39 40 should have generated bulk-lunar enrichment in heavy silicon Bogard DD, Garrison DH, Shih C-Y, and Nyquist LE (1994) Ar- Ar dating of two lunar granites – The age of Copernicus. Geochimica et Cosmochimica Acta isotopes, evidence for which has been sought (Armytage et al., 58: 3093–3100. 2012) but not found. Bolfan-Casanova N (2005) Water in Earth’s mantle. Mineralogical Magazine 69: 229–257. Borg LE, Connelly JN, Boyet M, and Carlson RW (2011) Chronological evidence that Acknowledgments the Moon is either young or did not have a global magma ocean. Nature 477: 70–73. Borg LE, Gaffney AM, Shearer CK, et al. (2009) Mechanisms for incompatible-element This research was supported by NASA grants NNX09AE31G enrichment on the Moon deduced from the lunar basaltic meteorite Northwest Africa (Cosmochemistry) and NNX11AE85G (LASER). 032. Geochimica et Cosmochimica Acta 73: 3963–3980. Borg L, Norman M, Nyquist L, et al. (1999) Isotopic studies of ferroan anorthosite 62236: A young lunar crustal rock from a light rare-earth-element-depleted source. Geochimica et Cosmochimica Acta 63: 2679–2691. Boyce JW, Liu Y, Rossman GR, et al. (2010) Lunar apatite with terrestrial volatile References abundances. Nature 466: 466–469.
Adler I, Trombka JI, Schmadebeck R, et al. (1973) Results of the Apollo 15 and 16 x-ray Boyet M and Carlson RW (2007) A highly depleted moon or a non-magma ocean origin experiment. Proceedings of the Fourth Lunar Science Conference, vol. 4, pp. for the lunar crust? Earth and Planetary Science Letters 262: 505–516. 2783–2792. Brandon AD, Lapen TJ, Debaille V, Beard BL, Rankenburg K, and Neal C (2009) Albrecht A, Herzog GF, Klein J, et al. (1995) Trace elements, 26Al and 10Be, and noble Re-evaluating 142Nd/144Nd in lunar mare basalts with implications for the early gases in lunar rock 14286. Lunar and Planetary Science 26: 13–14. evolution and bulk Sm/Nd of the Moon. Geochimica et Cosmochimica Acta Alfve´n H and Arrhenius G (1976) Evolution of the Solar System. Washington DC: NASA. 73: 6421–6445.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 243
Brandon AD, Puchtel IS, Walker RJ, Day JMD, Irving AJ, and Taylor LA (2012) Dasch EJ, Shih C-Y, Bansal BM, Wiesmann H, and Nyquist LE (1987) Isotopic Evolution of the martian mantle inferred from the 187Re–187Os isotope and highly analysis of basaltic fragments from lunar breccia 14321: Chronology and siderophile element abundance systematics of shergottite meteorites. Geochimica et petrogenesis of pre-Imbrium mare volcanism. Geochimica et Cosmochimica Acta Cosmochimica Acta 76: 206–235. 51: 3241–3254. Brophy JG and Basu A (1990) Europium anomalies in mare basalts as a consequence of Day JMD, Pearson DG, and Taylor LA (2007) Highly siderophile element mafic cumulate fractionation from an initial lunar magma. Proceedings of the Lunar constraints on accretion and differentiation of the Earth-Moon system. Science and Planetary Science Conference, vol. 20, pp. 25–30. 315: 217–219. Bru¨gmann GE, Arndt NT, Hofmann AW, and Tobschall HJ (1987) Noble metal Day JMD, Taylor LA, Floss C, Patchen AD, Schnare DW, and Pearson DG (2006) abundances in komatiite suites from Alexo, Ontario and Gorgona Island, Columbia. Comparative petrology, geochemistry, and petrogenesis of evolved, low-Ti lunar Geochimica et Cosmochimica Acta 51: 2159–2169. mare basalt meteorites from the LaPaz Icefield, Antarctica. Geochimica et Butler P Jr. and Meyer C Jr. (1976) Sulfur prevails in coatings on glass droplets: Apollo Cosmochimica Acta 70: 1581–1600. 15 green and brown glasses and Apollo 17 orange and black (devitrified) glasses. Day JMD, Walker RJ, James OB, and Puchtel IS (2010) Osmium isotope and highly Proceedings of the Lunar Science Conference, vol. 7, pp. 1561–1582. siderophile element systematics of the lunar crust. Earth and Planetary Science BVSP (1981) Basaltic Volcanism on the Terrestrial Planets. New York: Pergamon. Letters 289: 595–605. Cahill JTS, Lucey PG, and Wieczorek MA (2009) Compositional variations of the lunar DeHon RA (1979) Thickness of the western mare basalts. Proceedings of the Lunar and crust: Results from radiative transfer modeling of central peak spectra. Journal of Planetary Science Conference, vol. 10, pp. 2935–2955. Geophysical Research, Planets 114: E09001. Delano JW (1986) Pristine lunar glasses: Criteria, data, and implications. Proceedings Cameron AGW and Ward WR (1976) The origin of the moon. Lunar Science of the Lunar and Planetary Science Conference, vol. 16, pp. D201–D213. 7: 120–122. Delano JW and Ringwood AE (1978) Siderophile elements in the lunar highlands: Canup RM (2004) Dynamics of lunar formation. Annual Review of Astronomy and Nature of the indigenous component and implications for the origin of the Moon. Astrophysics 42: 441–475. Proceedings of the Lunar and Planetary Science Conference, vol. 9, pp. 111–159. Canup RM and Asphaug E (2001) Origin of the Moon in a giant impact near the end of Desch SJ and Taylor GJ (2011) A model of the Moon’s volatile depletion. Lunar and the Earth’s formation. Nature 412: 708–712. Planetary Science 42: #2005. Carlson RW and Lugmair GW (1979) Sm-Nd constraints on early lunar differentiation Desch SJ and Taylor GJ (2012) Volatile depletion from the protolunar disk. European and the evolution of KREEP. Earth and Planetary Science Letters 45: 123–132. Planetary Science Conference, abstract EPSC2012-271. Caro G (2011) Early silicate Earth differentiation. Annual Review of Earth and Planetary Dickinson T, Taylor GJ, Keil K, and Bild RW (1989) Germanium abundances in lunar Sciences 39: 31–58. basalts: Evidence of mantle metasomatism? Proceedings of the Lunar and Planetary
Caro G, Bourdon B, Halliday AN, and Quitte´ G (2008) Super-chondritic Sm/Nd ratios in Science Conference, vol. 19, pp. 189–198.
Mars, the Earth and the Moon. Nature 452: 336–339. Dickinson T, Taylor GJ, Keil K, Schmitt RA, Hughes SS, and Smith MR (1985) Apollo
Chapman CR, Cohen BA, and Grinspoon DH (2007) What are the real constraints on the 14 aluminous mare basalts and their possible relationship to KREEP. Proceedings
existence and magnitude of the late heavy bombardment? Icarus 189: 233–245. of the Lunar and Planetary Science Conference, vol. 15, pp. C365–C374. Chou C-L, Boynton WV, Bild RW, Kimberlin J, and Wasson JT (1976) Trace element Dowty E, Prinz M, and Keil K (1974) Ferroan anorthosite: A widespread and distinctive evidence regarding a chondritic component in howardite meteorites. Proceedings of lunar rock type. Earth and Planetary Science Letters 24: 15–25. the Lunar Science Conference, vol. 7, pp. 3501–3518. Dungan MA and Brown RW (1977) The petrology of the Apollo 12 ilmenite basalt suite. Chou C-L, Boynton WV, Sundberg LL, and Wasson JT (1975) Volatiles on the surface Proceedings of the Lunar Science Conference, vol. 8, pp. 1339–1381. of Apollo 15 green glass and trace-element distributions among Apollo 15 soils. Dymek RF, Albee AL, and Chodos AA (1975) Comparative petrology of lunar cumulate Proceedings of the Lunar Science Conference, vol. 6, pp. 1701–1727. rocks of possible primary origin: Dunite 72415, troctolite 76535, norite 78235, Chyba CF (1991) Terrestrial mantle siderophiles and the lunar impact record. Icarus and anorthosite 62237. Proceedings of the Lunar Science Conference,vol.6, 92: 217–233. pp. 301–341. Clayton RN and Mayeda TK (1996) Oxygen isotopic studies of achondrites. Geochimica Edmunson J, Borg LE, Nyquist LE, and Asmeron Y (2009) A combined Sm–Nd, Rb–Sr, et Cosmochimica Acta 60: 1999–2017. and U–Pb study of Mg–suite norite 78238: Further evidence for early differentiation Cohen BA, Snyder GA, Hall CM, Taylor LA, and Nazarov MA (2001) Argon-40-Argon- of the Moon. Geochimica et Cosmochimica Acta 73: 514–527. 39 chronology and petrogenesis along the eastern limb of the Moon from Luna 16, Elardo SM, Draper DS, and Shearer CK Jr. (2011) Lunar Magma Ocean crystallization 20 and 24 samples. Meteoritics and Planetary Science 36: 1345–1366. revisited: Bulk composition, early cumulate mineralogy, and the source regions of Cohen BA, Swindle TD, and Kring DA (2000) Support for the lunar cataclysm the highlands Mg-suite. Geochimica et Cosmochimica Acta 75: 3024–3045. hypothesis from lunar meteorite impact melt ages. Science 290: 1754–1756. Elardo SM, McCubbin FM, and Shearer CK Jr. (2012) Chromite symplectites in Mg- Cohen BA, Swindle TD, and Kring DA (2005) Geochemistry and 40Ar-39Ar suite troctolite 76535 as evidence for infiltration metasomatism of a lunar layered geochronology of impact-melt clasts in feldspathic lunar meteorites: Implications intrusion. Geochimica et Cosmochimica Acta 87: 154–177. for lunar bombardment history. Meteoritics and Planetary Science 40: 755–777. Elkins LT, Fernandes VA, Delano JW, and Grove TL (2000) Origin of lunar ultramafic Crocket JH and MacRae WE (1986) Platinum-group element distribution in komatiitic green glasses, constraints from phase equilibrium studies. Geochimica et and tholeiitic volcanic rocks from Munro Township, Ontario. Economic Geology Cosmochimica Acta 64: 2339–2350. 81: 1242–1251. Elkins-Tanton LT, Burgess S, and Yin Q-Z (2011) The lunar magma ocean: Reconciling Culler TS, Becker TA, Muller RA, and Renne PR (2000) Lunar impact history from the solidification process with lunar petrology and geochronology. Earth and 40Ar/39Ar dating of glass spherules. Science 287: 1785–1788. Planetary Science Letters 304: 326–336. Cushing JA, Taylor GJ, Norman MD, and Keil K (1999) The granulitic impactite suite – Elkins-Tanton LT, Chatterjee N, and Grove TL (2003a) Experimental and petrological impact melts and metamorphic breccias of the earliest lunar crust. Meteoritics and constraints on lunar differentiation from the Apollo 15 green picritic glasses. Planetary Science 34: 185–195. Meteoritics and Planetary Science 38: 515–527. 40 39 Dalrymple GB and Ryder G (1993) Ar/ Ar age spectra of Apollo 15 impact melt rocks Elkins-Tanton LT, Chatterjee N, and Grove TL (2003b) Magmatic processes that by laser step-heating and their bearing on the history of lunar basin formation. produced lunar fire fountains. Geophysical Research Letters 30: 1513 (4 pp). Journal of Geophysical Research, Planets 98: 13085–13095. Elkins-Tanton LT and Grove TL (2011) Water (hydrogen) in the lunar mantle: Results Dalrymple GB and Ryder G (1996) Argon-40/argon-39 age spectra of Apollo 17 from peteology and magma ocean modeling. Earth and Planetary Science Letters highlands breccia samples by laser step heating and the age of the Serenitatis basin. 307: 173–179. Journal of Geophysical Research, Planets 98: 13085–13095. Elphic RC, Lawrence DJ, Feldman WC, et al. (2000) Lunar rare earth element 40 39 Dalrymple GB, Ryder G, Duncan RA, and Huard JJ (2001) Ar- Ar ages of Apollo 16 distribution and ramifications for FeO and TiO2: Lunar Prospector neutron impact melt rocks by laser step heating. Lunar and Planetary Science 32: #1225. spectrometer observations. Journal of Geophysical Research, Planets Danckwerth PA, Hess PC, and Rutherford MJ (1979) The solubility of sulfur in high- 105: 20333–20345.
TiO mare basalts. Proceedings of the Lunar and Planetary Science Conference, vol. Eugster O, Polnau E, Salerno E, and Terribilino D (2000) Lunar surface exposure 2 10, pp. 517–530. models for meteorites Elephant Moraine 96008 and Dar al Gani 262 from the Moon. Darling JR, Hawkesworth CJ, Storey CD, and Lightfoot PC (2010) Shallow impact: Meteoritics and Planetary Science 35: 1177–1182. Isotopic insights into crustal contributions to the Sudbury impact melt sheet. Fagan TJ, Taylor GJ, Keil K, et al. (2002) Northwest Africa 032, product of lunar Geochimica et Cosmochimica Acta 74: 5680–5696. volcanism. Meteoritics and Planetary Science 37: 371–394. Darling JR and Moser DE (2012) Impact induced crustal differentiation: New insights Feldman WC, Maurice S, Lawrence DJ, et al. (2001) Evidence for water ice near the from the Sudbury structure. Lunar and Planetary Science 42: #2164. lunar poles. Journal of Geophysical Research, Planets 106: 23231–23252.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
244 The Moon
Fernandes VA and Burgess R (2005) Volcanism in Mare Fecunditatis and Mare Crisium: Grieve RAF and Cintala MJ (1992) An analysis of differential impact melt-crater scaling Ar-Ar age studies. Geochimica et Cosmochimica Acta 69: 4919–4934. and implications for the terrestrial record. Meteoritics 27: 526–538. Fernandes VA, Burgess R, and Morris A (2009a) 40Ar-39Ar age determinations of lunar Gross J and Treiman AH (2011) Unique spinel-rich lithology in lunar meteorite ALHA basalt meteorites Asuka 881757, Yamato 793169, Miller Range 05035, La Paz 81005: Origin and possible connection to M3 observations of the farside highlands. Icefield 02205, Northwest Africa 479, and basaltic breccia Elephant Moraine 96008. Journal of Geophysical Research, Planets 116: E10009. Meteoritics and Planetary Science 44: 805–821. Guinness EA and Arvidson RE (1977) On the constancy of the lunar cratering Fernandes VA, Korotev RL, and Renne PR (2009b) 40Ar-39Ar ages and chemical flux over the past 3.3 109 yr. Proceedings of the Lunar Science Conference, vol. 8, composition for lunar mare basalts: NWA 4734 and NWA 4898. Lunar and Planetary pp. 3475–3494. Science 40: #1045. Gustafson JO, Bell JF III, Gaddis LR, Hawke BR, and Giguere TA (2012) Characterization Fernandes VA, Burgess R, and Turner G (2003) 40Ar-39Ar chronology of lunar of previously unidentified lunar pyroclastic deposits using Lunar Reconnaissance meteorites Northwest Africa 032 and 773. Meteoritics and Planetary Science Orbiter Camera data. Journal of Geophysical Research, Planets 117: E00H25. 38: 555–564. Hagermann A and Tanaka S (2006) Ejecta deposit thickness, heat flow, and a critical Ferrand WH (1988) Highland contamination and minimum basalt thickness in northern ambiguity on the Moon. Geophysical Research Letters 38: L19203. Mare Fecunditatis. Proceedings of the Lunar and Planetary Science Conference, vol. Hagerty JJ, Lawrence DJ, and Hawke BR (2011) Thorium abundances of basalt ponds in 18, pp. 319–329. South Pole-Aitken basin: Insights into the composition and evolution of the far side Floss C, James OB, McGee JJ, and Crozaz G (1998) Lunar ferroan anorthosite lunar mantle. Journal of Geophysical Research, Planets 116: E06001. petrogenesis, clues from trace element distributions in FAN subgroups. Geochimica Hagerty JJ, Lawrence DJ, Hawke BR, Vaniman DT, Elphic RC, and Feldman WC (2006) et Cosmochimica Acta 62: 1255–1283. Refined thorium abundances for lunar red spots: Implications for evolved, nonmare Fogel RA and Rutherford MJ (1995) Magmatic volatiles in primitive lunar glasses: I. volcanism on the Moon. Journal of Geophysical Research 111: E06002. FTIR and EPMA analyses of Apollo 15 green and yellow glasses and revision of the Hagerty J, Shearer CK, and Papike JJ (2005) Trace element characteristics of the Apollo volatile-assisted fire-fountain theory. Geochimica et Cosmochimica Acta 14 high-alumina basalts: Implications for early magmatism on the Moon. 59: 201–215. Geochimica et Cosmochimica Acta 69: 5831–5845. Gaddis LR, Hawke BR, Robinson MS, and Coombs C (2000) Compositional analyses of Hagerty JJ, Shearer CK, and Vaniman DT (2006) Heat-producing elements in the lunar small lunar pyroclastic deposits using Clementine multispectral data. Journal of mantle: Insights from ion microprobe analyses of lunar pyroclastic glasses. Geophysical Research, Planets 105: 4245–4262. Geochimica et Cosmochimica Acta 70: 3457–3476. Gaffney AM, Borg LE, Asmerom Y, Shearer CK, and Burger PV (2011) Disturbance of Halliday AN (2008) A young Moon-forming giant impact at 70–110 million years isotope systematics during experimental shock and thermal metamorphism of a accompanied by late-stage mixing, core formation and degassing of the Earth.
lunar basalt with implications for martian meteorite chronology. Meteoritics and Philosophical Transactions of the Royal Society A366: 4163–4181.
Planetary Science 46: 35–52. Hallis LJ, Anand M, Greenwood RC, Miller MF, Franchi IA, and Russell SS (2010) The
Gallant J, Gladman B, and Cuk M (2007) Current bombardment of the Earth–Moon oxygen isotope composition, petrology and geochemistry of mare basalts: Evidence
system: Emphasis on cratering asymmetries. Icarus 202: 371–382. for large-scale compositional variation in the lunar mantle. Geochimica et Garrick-Bethell I, Nimmo F, and Wieczorek MA (2010) Structure and formation of the Cosmochimica Acta 74: 6885–6899. lunar farside highlands. Science 330: 949–951. Haloda J, Ty´cova´ P, Korotev RL, et al. (2009) Petrology, geochemistry, and age of low- Garrick-Bethell I and Zuber MT (2005) An indigenous origin for the South Pole-Aitken Ti mare-basalt meteorite Northeast Africa 003-A: A possible member of the Apollo basin thorium anomaly. Geophysical Research Letters 32: 13203. 15 mare basaltic suite. Geochimica et Cosmochimica Acta 73: 3450–3470. Garrick-Bethell I and Zuber MT (2009) Elliptical structure of the lunar South Pole-Aitken Hartmann WK (1975) “Lunar cataclysm”: A misconception? Icarus 24: 181–187. basin. Icarus 204: 399–408. Hartmann WK, Quantin C, and Mangold N (2007) Possible long-term decline Gault DE (1983) The terrestrial accretion of lunar material. Lunar and Planetary Science in impact rates. 2. Lunar impact-melt data regarding impact history. Icarus 14: 8. 186: 11–23. Giguere TA, Taylor GJ, Hawke BR, and Lucey PG (2000) The titanium contents of lunar Haskin LA, Korotev RL, Gillis JJ, and Jolliff BL (2002) Stratigraphies of Apollo and Luna mare basalts. Meteoritics and Planetary Science 35: 193–200. highland landing sites and provenances of materials from the perspective of basin Giguere TA, Hawke BR, Blewett DT, et al. (2003) Remote sensing studies of the impact ejecta modeling. Lunar and Planetary Science 33: #1364. Lomonosov-Fleming region of the Moon. Journal of Geophysical Research, Planets Haskin LA, Lindstrom MM, Salpas P, and Lindstrom DJ (1981) On compositional 108: E002069. variations among lunar anorthosites. Proceedings of the Lunar and Planetary Gillis JJ, Jolliff BL, and Korotev RL (2004) Lunar surface geochemistry: Global Science Conference, vol. 12, pp. 41–66. concentrations of Th, K, and FeO as derived from Lunar Prospector and Clementine Haskin LA and Salpas PA (1992) Genesis of compositional characteristics of Stillwater data. Geochimica et Cosmochimica Acta 68: 3791–3805. AN-I and AN-II thick anorthosite units. Geochimica et Cosmochimica Acta Gladman BJ, Burns JA, Duncan M, Lee P, and Levison HF (1996) The exchange of 56: 1187–1212. impact ejecta between terrestrial planets. Science 271: 1387–1392. Haskin LA and Warren PH (1991) Lunar chemistry. In: Heiken G, Vaniman D, and Gladman B and Chan CS (2012) Terrene meteorites: Effects of planetary atmospheres French B (eds.) Lunar Sourcebook: A User’s Guide to the Moon, pp. 357–474. on ejecta launch. Meteoritics and Planetary Science 47: #5147. New York: Cambridge University Press. Glotch TD, Hagerty JJ, Lucey PG, et al. (2011) The Mairan domes: Silicic volcanic Hauri EH, Gaetani GA, and Green TH (2006) Partitioning of water during melting of the constructs on the Moon. Geophysical Research Letters 38: L21204 (5 pp). Earth’s upper mantle at H2O-undersaturated conditions. Earth and Planetary Science Glotch TD, Lucey PG, Bandfield JL, et al. (2010) Highly silicic compositions on the Letters 248: 715–734. Moon. Science 329: 1510–1513. Hauri EH, Weinreich T, Saal AE, Rutherford MC, and Van Orman JA (2011) High Gnos E, Hofmann BA, Al-Kathiri A, et al. (2004) Pinpointing the source of a lunar pre-eruptive water contents preserved in lunar melt inclusions. Science meteorite: Implications for the evolution of the Moon. Science 305: 657–659. 333: 213–215. Gomes R, Levison HF, Tsiganis K, and Morbidelli A (2005) Origin of the Head JW and Wilson L (1992) Lunar mare volcanism: Stratigraphy, eruption conditions, cataclysmic late heavy bombardment period of the terrestrial planets. Nature and the evolution of secondary crusts. Geochimica et Cosmochimica Acta 435: 466–469. 56: 2155–2175. Goodrich CA, Scott ERD, and Fioretti AM (2004) Ureilitic breccias: Clues to the Hertogen J, Janssens M-J, and Palme H (1980) Trace elements in ocean ridge basalt petrologic structure and impact disruption of the ureilite parent asteroid. Chemie der glasses: Implications for fractionations during mantle evolution and petrogenesis. Erde 64: 283–327. Geochimica et Cosmochimica Acta 44: 2125–2143. Govindaraju K (1994) 1994 compilation of working values and sample description for Hertogen J, Janssens M-J, Takahashi H, Palme H, and Anders E (1977) Lunar basins 383 geostandards. Geostandards Newsletter 18: 1–158. and craters: Evidence for compositional changes of bombarding population. Greenhagen BT, Lucey PG, Wyatt MB, et al. (2010) Global silicate mineralogy of the Proceedings of the Lunar Science Conference, vol. 8, pp. 17–45.
Moon from the Diviner Lunar Radiometer. Science 329: 1507–1509. Hess PC (1994) Petrogenesis of lunar troctolites. Journal of Geophysical Research,
Greenwood JP, Itoh S, Sakamoto N, Warren P, Taylor LA, and Yurimoto H (2011) Planets 99: 19083–19093. Hydrogen isotope ratios in lunar rocks indicate delivery of cometary water to the Hess PC and Parmentier EM (1995) A model for the thermal and chemical evolution of Moon. Nature Geoscience 4: 79–82. the Moon’s interior: Implications for the onset of mare volcanism. Earth and Greshake A, Irving AJ, Kuehner SM, Korotev RL, Gellissen M, and Palme H (2008) Planetary Science Letters 134: 501–514. Northwest Africa 4898: A new high-alumina mare basalt from the Moon. Lunar and Hess PC, Rutherford MJ, and Campbell HW (1977) Origin and evolution of LKFM Planetary Science 39: #1631. basalt. Proceedings of the Lunar Science Conference, vol. 8, pp. 2357–2374.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 245
Hiesinger H, Head JW III, Wolf U, Jaumann R, and Neukum G (2002) Lunar mare basalt Kirsten T, Horn P, and Kiko J (1973) 39Ar-40Ar dating and rare gas analysis of flow units, thicknesses determined from crater size-frequency distributions. Apollo 16 rocks and soils. Proceedings of the Lunar Science Conference,vol.4, Geophysical Research Letters 29: 8891–8894. pp. 1757–1784. Hiesinger H, Head JW III, Wolf U, Jaumann R, and Neukum G (2003) Ages and Kleine T, Touboul M, Bourdon B, et al. (2009) Hf–W chronology of the accretion and stratigraphy of mare basalts in Oceanus Procellarum, Mare Nubium, Mare Cognitum, early evolution of asteroids and terrestrial planets. Geochimica et Cosmochimica and Mare Insularum. Journal of Geophysical Research, Planets 108: 5065 (27 pp). Acta 73: 5150–5188. Hiesinger H, Head JW, Wolf U, Jaumann R, and Neukum G (2010) Ages and Klima RL, Pieters CM, Boardman JW, et al. (2011) New insights into lunar petrology: stratigraphy of lunar mare basalts in Mare Frigoris and other nearside maria based Distribution and composition of prominent low-Ca pyroxene exposures as observed on crater size-frequency distribution measurements. Journal of Geophysical by the Moon Mineralogy Mapper (M3). Journal of Geophysical Research, Planets Research, Planets 115: E03003. 116: E00G06. Ho¨rz F (1978) How thick are lunar mare basalts? Proceedings of the Lunar and Planetary Korenaga J (2009) A method for estimating the composition of the bulk Silicate Earth in Science Conference, vol. 9, pp. 3311–3331. the presence of a hidden geochemical reservoir. Geochimica et Cosmochimica Acta Huang Q and Wieczorek MA (2012) Density and porosity of the lunar crust from gravity 73: 6952–6964. and topography. Journal of Geophysical Research, Planets 117: E05003. Korotev RL (1981) Compositional trends in Apollo 16 soils. Proceedings of the Lunar Hudgins JA, Kelley SP, Korotev RL, and Spray JG (2011) Mineralogy, geochemistry, and Planetary Science Conference, vol. 12, pp. 577–605. 40 39 and Ar- Ar geochronology of lunar granulitic breccia Northwest Africa 3163 and Korotev RL (1987a) Mixing levels, the Apennine Front soil component, and paired stones: Comparisons with Apollo samples. Geochimica et Cosmochimica compositional trends in the Apollo 15 soils. Proceedings of the Lunar and Planetary Acta 75: 2865–2881. Science Conference, vol. 17, pp. E411–E431. Hudgins JA, Spray JG, Kelley SP, Korotev RL, and Sherlock SC (2008) A laser probe Korotev RL (1987b) The meteoritic component of Apollo 16 noritic impact 40 39 Ar/ Ar and INAA investigation of four Apollo granulitic breccias. Geochimica et melt breccias. Proceedings of the Lunar and Planetary Science Conference, vol. 17, Cosmochimica Acta 72: 5781–5798. pp. E491–E512. Isaacson PJ, Pieters CM, Besse S, et al. (2011) Remote compositional analysis of lunar Korotev RL (1994) Compositional variation in Apollo 16 impact melt breccias and 3 olivine-rich lithologies with Moon Mineralogy Mapper (M ) spectra. Journal of inferences for the geology and bombardment history of the Central Highlands of the Geophysical Research, Planets 116: E00G11. Moon. Geochimica et Cosmochimica Acta 58: 3931–3969. Ishihara Y, Goossens S, Matsumoto K, et al. (2009) Crustal thickness of the Korotev LR (2005) Lunar geochemistry as told by lunar meteorites. Chemie der Erde Moon: Implications for farside basin structures. Geophysical Research Letters 65: 297–346. 36: L19202. Korotev RL (2012) Lunar meteorites from Oman. Meteoritics and Planetary Science
Ito T and Malhotra R (2010) Asymmetric impacts of near-Earth asteroids on the Moon. 47: 1365–1402.
Astronomy and Astrophysics 519: A63. Korotev RL, Haskin LA, and Jolliff BL (1998) Ir/Au ratios and the origin of Apollo
James OB and Flohr MK (1983) Subdivision of the Mg-suite noritic rocks into Mg- 17 and other Apollo impact-melt breccias. Lunar and Planetary Science
gabbronorites and Mg-norites. Proceedings of the Lunar and Planetary Science 29: #1231. Conference, vol. 13, pp. A603–A614. Korotev RL and Kremser DT (1992) Compositional variations in Apollo 17 soils and James OB, Lindstrom MM, and Flohr MK (1989) Ferroan anorthosite from lunar breccia their relationship to the geology of the Taurus-Littrow site. Proceedings of the Lunar 64435: Implications for the origin and history of lunar ferroan anorthosites. and Planetary Science Conference, vol. 22, pp. 275–301. Proceedings of the Lunar and Planetary Science Conference, vol. 19, pp. 219–243. Korotev RL, Zeigler RA, and Jolliff BL (2006) Feldspathic lunar meteorites Pecora James OB and McGee JJ (1979) Consortium breccia 73255: Genesis and history of two Escarpment 02007 and Dhofar 489: Contamination of the surface of the lunar coarse-grained “norite” clasts. Proceedings of the Lunar and Planetary Science highlands by post-basin impacts. Geochimica et Cosmochimica Acta Conference, vol. 10, pp. 713–743. 70: 5935–5956. James OB, McGee JJ, and Lindstrom MM (1991) Lunar ferroan anorthosite 60025: Korotev RL, Zeigler RA, Jolliff BL, Irving AJ, and Bunch TE (2009) Compositional and Petrology and chemistry of mafic lithologies. Proceedings of the Lunar and lithological diversity among brecciated lunar meteorites of intermediate iron Planetary Science Conference, vol. 21, pp. 63–87. concentration. Meteoritics and Planetary Science 44: 1287–1322. Jarosewich E (1990) Chemical analyses of meteorites: A compilation of stony and iron Kring DA and Cohen BA (2002) Cataclysmic bombardment throughout the inner solar meteorite analyses. Meteoritics 25: 323–337. system. Journal of Geophysical Research, Planets 107: 5009. Jerde EA, Morris RV, and Warren PH (1990) In quest of lunar regolith breccias of exotic Kurat G, Kracher A, Keil K, Warner R, and Prinz M (1976) Composition and origin of provenance – A uniquely anorthositic sample from the Fra Mauro (Apollo 14) Luna 16 aluminous mare basalts. Proceedings of the Lunar Science Conference, highlands. Earth and Planetary Science Letters 98: 90–108. vol. 7, pp. 1301–1322. Jolliff BL, Floss C, McCallum IS, and Schwartz JM (1999) Geochemistry, petrology, and Kusuma KN, Sebastian N, and Murty SVS (2012) Geochemical and mineralogical cooling history of 14161,7373, a plutonic lunar sample with textural evidence of analysis of Gruithuisen region on Moon using M3 and DIVINER images. Planetary granitic-fraction separation by silicate-liquid immiscibility. American Mineralogist and Space Science 67: 46–56. 84: 821–837. Lal D, Chauhan P, Shah RD, Bhattacharya S, Ajai, and Kumar ASK (2012) Detection of Jolliff BL, Gillis JJ, Haskin LA, and Korotev RL (2000) Major lunar crustal terranes: Mg spinel lithologies on central peak of crater Theophilus using Moon Mineralogy Surface expressions and crust-mantle origins. Journal of Geophysical Research Mapper (M3) data from Chandrayaan-1. Journal of Earth System Science 105: 4197–4216. 121: 847–853. Jolliff BL, Korotev RL, Zeigler RA, and Floss C (2003) Northwest Africa 773: Lunar mare Lawrence DJ, Hawke BR, Hagerty JJ, et al. (2005) Evidence for a high-Th, evolved breccia with a shallow-formed olivine-cumulate component, inferred very-low-Ti lithology on the Moon at Hansteen Alpha. Geophysical Research Letters (VLT) heritage, and a KREEP connection. Geochimica et Cosmochimica Acta 32: L07201. 67: 4857–4879. Lawrence DJ, Puetter RC, Elphic RC, et al. (2007) Global spatial deconvolution of Lunar Jolliff BL, Wiseman SA, Lawrence SJ, et al. (2011) Non-mare silicic volcanism on the Prospector Th abundances. Geophysical Research Letters 34: L03201. lunar farside at Compton-Belkovich. Nature Geoscience 4: 566–571. Le Bas MJ (2001) Report of the Working Party on the classification of the lunar igneous Joy KH, Kring DA, Bogard DD, McKay DS, and Zolensky ME (2011) Re-examination of rocks. Meteoritics and Planetary Science 36: 1183–1188. the formation ages of the Apollo 16 regolith breccias. Geochimica et Cosmochimica Le Feuvre M and Wieczorek MA (2011) Nonuniform cratering of the Moon and a revised Acta 75: 7208–7225. crater chronology of the inner Solar System. Icarus 214: 1–20. Joy KH, Zolensky ME, Nagashima K, et al. (2012) Direct detection of projectile relics Levine J, Becker TA, Muller RA, and Renne PR (2005) 40Ar/39Ar dating of Apollo 12 from the end of the lunar basin-forming epoch. Science 336: 1426–1429. impact spherules. Geophysical Research Letters 32: L15201. Keller LP and McKay DS (1997) The nature and origin of rims on lunar grains. Li L and Mustard JF (2000) Compositional gradients across mare-highland contacts: Geochimica et Cosmochimica Acta 61: 2331–2341. Importance and geological implication of lateral transport. Journal of Geophysical
Khan A, Maclennan J, Taylor SR, and Connolly JAD (2006) Are the Earth and the Moon Research, Planets 105: 20431–20450.
compositionally alike? Inferences on lunar composition and implications for lunar Lin Y, Shen W, Liu Y, et al. (2012) Very high-K KREEP-rich clasts in the impact melt origin and evolution from geophysical modeling. Journal of Geophysical Research, breccia of the lunar meteorite SaU 169: New constraints on the last residue of the Planets 111: E05005. Lunar Magma Ocean. Geochimica et Cosmochimica Acta 85: 19–40. Khan A and Mosegaard K (2002) An inquiry into the lunar interior: A nonlinear Lindstrom MM and Lindstrom DJ (1986) Lunar granulites and their precursor inversion of the Apollo lunar seismic data. Journal of Geophysical Research, Planets anorthositic norites of the early lunar crust. Proceedings of the Lunar and Planetary 107: 5036. Science Conference, vol. 16, pp. D263–D276.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
246 The Moon
Liu Y, Floss C, Day JMD, Hill E, and Taylor LA (2009) Petrogenesis of lunar mare McKenzie D (2000) Constraints on melt generation and transport from V-series activity basalt meteorite Miller Range 05035. Meteoritics and Planetary Science ratios. Chemical Geology 162: 81–94. 44: 261–284. Melosh HJ (1989) Impact Cratering: A Geologic Process. New York: Oxford University Liu D, Jolliff BL, Zeigler RA, et al. (2012) Comparative zircon U-Pb geochronology of Press. impact melt breccias from Apollo 12 and lunar meteorite SaU 169, and implications Misawa K, Tatsumoto M, Dalrymple GB, and Yanai K (1993) An extremely low U/Pb for the age of the Imbrium impact. Earth and Planetary Science Letters source in the Moon: U-Th-Pb, Sm-Nd, Rb-Sr, and 40Ar-39Ar systematics of 319: 277–286. lunar meteorite Asuka 881757. Geochimica et Cosmochimica Acta 57: Liu Y, Spicuzza MJ, Craddock PR, et al. (2010) Oxygen and iron isotope constraints on 4687–4702. near-surface fractionation effects and the composition of lunar mare basalt source Moore G, Vennemann T, and Carmichael ISE (1998) An empirical model for the regions. Geochimica et Cosmochimica Acta 74: 6249–6262. solubility of H2O in magmas to 3 kilobars. American Mineralogist 83: 36–42. Lofgren GE, Donaldson CH, and Usselman TM (1975) Geology, petrology, and Morbidelli A, Petit JM, Gladman B, and Chambers JE (2001) A plausible cause of the crystallization of Apollo 15 quartz-normative basalts. Proceedings of the Lunar late heavy bombardment. Meteoritics and Planetary Science 36: 371–380. Science Conference, vol. 6, pp. 79–100. Morgan JW, Higuchi H, Takahashi H, and Hertogen J (1978) A “chondritic” eucrite Lognonne´ P (2005) Planetary seismology. Annual Review of Earth and Planetary parent body: Inference from trace elements. Geochimica et Cosmochimica Acta Sciences 33: 571–604. 42: 27–38. Longhi J (1987) On the connection between mare basalts and picritic volcanic glasses. Morgan JW, Walker RJ, Brandon AD, and Horan MF (2001) Siderophile elements in Proceedings of the Lunar and Planetary Science Conference, vol. 17, pp. E349–E360. Earth’s upper mantle and lunar breccias, data synthesis suggests manifestations of Longhi J (1992) Experimental petrology and petrogenesis of mare volcanics. the same late influx. Meteoritics and Planetary Science 36: 1257–1275. Geochimica et Cosmochimica Acta 56: 2235–2251. Morota T, Ukai T, and Furumoto M (2005) Influence of the asymmetrical cratering rate Longhi J (2003) A new view of lunar ferroan anorthosites: Postmagma ocean on the lunar cratering chronology. Icarus 173: 322–324. petrogenesis. Journal of Geophysical Research, Planets 108: 5083. Namur O, Charlier B, Pirard C, Hermann J, Lie´geois J-P, and Auwera JV (2011) Longhi J (2006) Petrogenesis of picritic mare magmas: Constraints on the extent of Anorthosite formation by plagioclase flotation in ferrobasalt and implications for the early lunar differentiation. Geochimica et Cosmochimica Acta 70: 5919–5934. lunar crust. Geochimica et Cosmochimica Acta 75: 4998–5018. Longhi J and Ashwal LD (1985) Two-stage models for lunar and terrestrial anorthosites: Narendranath S, Athiray PS, Sreekumar P, et al. (2011) Lunar X-ray fluorescence Petrogenesis without a magma ocean. Proceedings of the Lunar and Planetary observations by the Chandrayaan-1 X-ray Spectrometer (C1XS): Results from the Science Conference, vol. 15, pp. C571–C584. nearside southern highlands. Icarus 214: 53–66. Longhi J and Boudreau AE (1979) Complex igneous processes and the formation of the Neal CR (2001) Interior of the Moon: The presence of garnet in the primitive deep lunar
primitive lunar crustal rocks. Proceedings of the Lunar and Planetary Science mantle. Journal of Geophysical Research, Planets 106: 27865–27885.
Conference, vol. 10, pp. 2085–2105. Neal CR, Hacker MD, Snyder GA, Taylor LA, Liu Y-G, and Schmitt RA (1994) Basalt
Longhi J, Durand SR, and Walker D (2010) The pattern of Ni and Co abundances in generation at the Apollo 12 site, Part 1: New data, classification, and re-evaluation.
lunar olivines. Geochimica et Cosmochimica Acta 74: 784–798. Meteoritics 29: 334–348. Longhi J and Pan V (1988) A reconnaissance study of phase boundaries in low-alkali Neal CR and Kramer G (2003) The composition of KREEP: A detailed study of KREEP basaltic liquids. Journal of Petrology 29: 115–147. basalt 15386. Lunar and Planetary Science 34: #2023. Loper DE and Werner CL (2002) On lunar asymmetries 1. Tilted convection and crustal Neal CR and Kramer GY (2006) The petrogenesis of the Apollo 14 high-Al mare basalts. asymmetry. Journal of Geophysical Research, Planets 107: 5046 (7 pp). American Mineralogist 91: 1521–1535.
Lucey PG, Blewett DT, and Hawke BR (1998) Mapping the FeO and TiO2 content of the Neal CR and Taylor LA (1991) Evidence for metasomatism of the lunar highlands and lunar surface with multispectral imagery. Journal of Geophysical Research, Planets the origin of whitlockite. Geochimica et Cosmochimica Acta 55: 2965–2980. 103: 3679–3699. Neal CR and Taylor LA (1992) Petrogenesis of mare basalts: A record of lunar Lucey PG, Blewett DT, and Jolliff BL (2000) Lunar iron and titanium abundance volcanism. Geochimica et Cosmochimica Acta 56: 2177–2211. algorithms based on final processing of Clementine ultraviolet-visible images. Neal CR, Taylor LA, Hughes SS, and Schmitt RA (1990) The significance of fractional Journal of Geophysical Research, Planets 105: 20297–20305. crystallization in the petrogenesis of Apollo 17 type A and B high-Ti basalts. Lucey PG, Blewett DT, Taylor GJ, and Hawke BR (2000) Imaging of lunar surface Geochimica et Cosmochimica Acta 54: 1817–1833. maturity. Journal of Geophysical Research, Planets 105: 20377–20386. Nemchin AA, Whitehouse MJ, Grange ML, and Muhling JR (2011) On the elusive Lugmair GW and Carlson RW (1978) The Sm-Nd history of KREEP. Proceedings of the isotopic composition of lunar Pb. Geochimica et Cosmochimica Acta Lunar Science Conference, vol. 9, pp. 689–704. 75: 2940–2964. Marti K, Aeschlimann U, Eberhardt P, et al. (1983) Pieces of the ancient lunar crust: Neukum G, Ivanov BA, and Hartmann WK (2001) Cratering records in the inner Ages and composition of clasts in consortium breccia 67915. Proceedings of the solar system in relation to the lunar reference system. Space Science Reviews Lunar and Planetary Science Conference, vol. 14, pp. B165–B175. 96: 55–86. Marvin UB and Warren PH (1980) A pristine eucrite-like gabbro from Descartes and its Nicholis MG and Rutherford MJ (2009) Graphite oxidation in the Apollo 17 orange exotic kindred. Proceedings of the Lunar Science Conference, vol. 11, pp. 507–521. glass magma: Implications for the generation of a lunar volcanic gas phase. Mauer PPE, Geiss J, Gro¨gler N, et al. (1978) Pre-Imbrian craters and basins: Ages, Geochimica et Cosmochimica Acta 73: 5905–5917. compositions, and excavation depths of Apollo 16 breccias. Geochimica et Nimmo F, O’Brien DP, and Kleine T (2010) Tungsten isotopic evolution during late- Cosmochimica Acta 42: 1687–1720. stage accretion: Constraints on Earth–Moon equilibration. Earth and Planetary McCallum IS, Domeneghetti MC, Schwartz JM, et al. (2006) Cooling history of lunar Science Letters 292: 363–370. Mg-suite gabbronorite 76255, troctolite 76535 and Stillwater pyroxenite SC-936: Nishiizumi K, Okazaki R, Park J, Nagao K, Masarik J, and Finkel RC (2002) Exposure The record in exsolution and ordering in pyroxenes. Geochimica et Cosmochimica and terrestrial histories of Dhofar 019 martian meteorite. Lunar and Planetary Acta 70: 6068–6078. Science 33: #1366. McCubbin FM, Jolliff BL, Nekvasil H, et al. (2011) Fluorine and chlorine abundances in Norman MD, Bennett VC, and Ryder G (2002) Targeting the impactors: Siderophile lunar apatite: Implications for heterogeneous distributions of magmatic volatiles in element signatures of lunar impact melts from Serenitatis. Earth and Planetary the lunar interior. Geochimica et Cosmochimica Acta 75: 5073–5093. Science Letters 202: 217–228. McCubbin FM, Steele A, Hauri EH, Nekvasil H, Yamashita S, and Hemley RJ (2010) Norman MD, Borg LE, Nyquist LE, and Bogard DD (2003) Chronology, geochemistry, Nominally hydrous magmatism on the Moon. Proceedings of the National Academy and petrology of a ferroan noritic anorthosite clast from Descartes breccia 67215: of Sciences 27: 11223–11228. Clues to the age, origin, structure, and impact history of the lunar crust. Meteoritics McEwen AS, Keszthelyi LP, Spencer JR, et al. (1998) High-temperature silicate and Planetary Science 38: 645–661. volcanism on Jupiter’s moon Io. Science 281: 87–90. Norman MD, Duncan RA, and Huard JJ (2006) Identifying impact events within the McKay DS, Heiken G, Basu A, et al. (1991a) The lunar regolith. In: Heiken G, lunar cataclysm from 40Ar-39Ar ages and compositions of Apollo 16 impact melt
Vaniman D, and French B (eds.) Lunar Sourcebook: A User’s Guide to the Moon, rocks. Geochimica et Cosmochimica Acta 70: 6032–6049.
pp. 285–356. New York: Cambridge University Press. Norman MD, Duncan RA, and Huard JJ (2010) Imbrium provenance for the Apollo 16 McKay GA and Weill DF (1977) KREEP petrogenesis revisited. Proceedings of the Lunar Descartes terrain: Argon ages and geochemistry of lunar breccias 67016 and 67455. Science Conference, vol. 8, pp. 2339–2355. Geochimica et Cosmochimica Acta 74: 763–783. McKay G, Le L, and Wagstaff J (1991b) Constraints on the origin of the mare basalt Nyquist L, Bogard D, Yamaguchi A, et al. (2006) Feldspathic clasts in Yamato-86032: europium anomaly: REE partition coefficients for pigeonite. Lunar and Planetary Remnants of the lunar crust with implications for its formation and impact history. Science 22: 883–884. Geochimica et Cosmochimica Acta 70: 5990–6015.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 247
Nyquist LE, Shih C-Y, Reese YD, et al. (2010) Lunar crustal history recorded in lunar Robinson MS, Ashley JW, Boyd AK, et al. (2012a) Confirmation of sublunarean anorthosites. Lunar and Planetary Science 41: #1383. voids and thin layering in mare deposits. Planetary and Space Science O’Connell-Cooper CD, Dickin AP, and Spray JG (2012) The Manicouagan impact melt 69: 18–27. sheet: Evidence for isotopic homogenization with limited assimilation. Earth and Robinson KL, Taylor GJ, Hellebrand E, and Nagashima K (2012b) Water in evolved Planetary Science Letters 335–336: 48–58. lunar rocks: Implications for water distribution in the lunar mantle. Lunar and O’Neill HStC and Palme H (2008) Collisional erosion and the non-chondritic Planetary Science 43: #1727. composition of the terrestrial planets. Philosophical Transactions of the Royal Robinson KL, Taylor GJ, Hellebrand E, and Nagashima K (2012c) KREEP, KREEP Society 366: 4205–4238. derivatives, and the lunar D enrichment. European Planetary Science Conference, O’Connell-Cooper CD, Dickin AP, and Spray JG (2012) The Manicouagan impact melt abstract EPSC2012–283. sheet: Evidence for isotopic homogenization with limited assimilation. Earth and Ryder G (1985) Catalog of Apollo 15 Rocks, Part 2. Houston: NASA Johnson Space Planetary Science Letters 335: 48–58. Center (Curatorial Branch Publication 72). Ohtake M, Matsunaga T, Haruyama J, et al. (2009) The global distribution of pure Ryder G (1990) Lunar samples, lunar accretion and the early bombardment of the anorthosite on the Moon. Nature 461: 236–241. Moon. Eos 71: 313–323. Ohtake M, Takeda H, Matsunaga T, et al. (2012) Asymmetric crustal growth on the Ryder G (1991) Lunar ferroan anorthosites and mare basalt sources: The mixed Moon indicated by primitive farside highland materials. Nature Geoscience connection. Geophysical Research Letters 18: 2065–2068. 5: 384–388. Ryder G and Cox BT (1996) An Apollo 15 mare basalt fragment and lunar mare Ong L, Asphaug EI, Korycansky D, and Coker RF (2010) Volatile retention from provinces. Meteoritics and Planetary Science 31: 50–59. cometary impacts on the Moon. Icarus 207: 578–589. Ryder G, Delano JW, Warren PH, Kallemeyn GW, and Dalrymple GB (1996) A glass Pahlevan K and Stevenson DJ (2007) Equilibration in the aftermath of the lunar-forming spherule of equivocal impact origin from the Apollo 15 landing site: Unique target giant impact. Earth and Planetary Science Letters 262: 438–449. mare basalt. Geochimica et Cosmochimica Acta 60: 693–710. Pahlevan K, Stevenson DJ, and Eiler JM (2011) Chemical fractionation in the Ryder G, Norman MD, and Score RA (1980) The distinction of pristine from meteorite- silicate vapor atmosphere of the Earth. Earth and Planetary Science Letters contaminated highlands rocks using metal compositions. Proceedings of the Lunar 301: 433–443. and Planetary Science Conference, vol. 11, pp. 471–479. Papike JJ, Ryder G, and Shearer CK (1998) Lunar samples. In: Papike JJ (ed.) Planetary Ryder G and Spudis PD (1980) Volcanic rocks in the lunar highlands. In: Papike JJ and Materials. Reviews in Mineralogy, vol. 36, pp. 5.1–5.234. Chantilly, VA: Merrill RB (eds.) Proceedings of the Conference on the Lunar Highlands Crust, Mineralogical Society of America. pp. 353–375. New York: Pergamon. Papike J, Taylor L, and Simon S (1991) Lunar minerals. In: Heiken G, Vaniman D, and Saal AE, Hauri EH, Cascio ML, Van Orman JA, Rutherford MC, and Cooper RF (2008)
French B (eds.) Lunar Sourcebook: A User’s Guide to the Moon, pp. 121–181. Volatile content of lunar volcanic glasses and the presence of water in the Moon’s
New York: Cambridge University Press. interior. Nature 454: 192–195.
Paul RL and Lipschutz ME (1990) Chemical studies of differentiated meteorites: I. Labile Saal AE, Hauri EH, Langmuir CH, and Perfit MR (2002) Vapour undersaturation in
trace elements in Antarctic and non-Antarctic eucrites. Geochimica et primitive mid-ocean-ridge basalt and the volatile content of the Earth’s upper Cosmochimica Acta 54: 3185–3195. mantle. Nature 419: 451. Peach CL and Mathez EA (1993) Sulfide melt-silicate melt distribution coefficients for Schaeffer OA and Husain L (1973) Early lunar history: Ages of 2 to 4 mm soil nickel and iron and implications for the distribution of other chalcophile elements. fragments from the lunar highlands. Proceedings of the Lunar Science Conference, Geochimica et Cosmochimica Acta 57: 3013–3021. vol. 4, pp. 1847–1864. Persikov ES, Bukhtiyarov PG, and Kalinicheva TV (1987) Effects of composition, Sharp ZD, Shearer CK, McKeegan KD, Barnes JD, and Wang Y (2010) The chlorine temperature, and pressure on magma fluidity. Geokhimiya 33: 483–498. isotope composition of the Moon and implications for an anhydrous mantle. Pieters CM, Besse S, Boardman J, et al. (2011) Mg-spinel lithology: A new Science 329: 1050–1053. rock type on the lunar farside. Journal of Geophysical Research, Planets Shearer CK, Hess PC, Wieczorek Mark A, et al. (2006) Thermal and magmatic evolution 116: E00G08. of the Moon. Reviews in Mineralogy and Geochemistry 60: 365–518. Pieters CM, Head JW III, Gaddis L, Jolliff BL, and Duke M (2001) Rock types of South Shearer CK and Papike JJ (1989) Is plagioclase removal responsible for the negative Eu Pole-Aitken basin and extent of basaltic volcanism. Journal of Geophysical anomaly in the source regions of mare basalts. Geochimica et Cosmochimica Acta Research, Planets 106: 28001–28022. 53: 3331–3336. Pieters CM, Stankevich DG, Shkuratov YG, and Taylor LA (2002) Statistical analysis of Shearer CK and Papike JJ (2005) Early crustal building processes on the moon: Models the links between lunar mare soil mineralogy, chemistry and reflectance. Icarus for the petrogenesis of the magnesian suite. Geochimica et Cosmochimica Acta 155: 285–298. 69: 3445–3461. Poitrasson F, Roskosz M, and Corgne A (2009) No iron isotope fractionation between Shearer CK, Papike JJ, and Layne GD (1996) Deciphering basaltic magmatism on the molten alloys and silicate melt to 2000 C and 7.7 GPa: Experimental evidence and Moon from the compositional variations in Apollo 15 very low-Ti picritic magmas. implications for planetary differentiation and accretion. Earth and Planetary Geochimica et Cosmochimica Acta 60: 509–528. Science Letters 278: 376–385. Shearer CK, Papike JJ, Simon SB, Shimizu N, Yurimoto H, and Sueno S (1990) Ion Prettyman TH, Hagerty JJ, Elphic RC, et al. (2006) Elemental composition of the lunar microprobe studies of trace elements in Apollo 14 volcanic glass beads: surface: Analysis of gamma ray spectroscopy data from Lunar Prospector. Journal Comparisons to Apollo 14 mare basalts and petrogenesis of picritic magmas. of Geophysical Research, Planets 111: E12007. Geochimica et Cosmochimica Acta 54: 851–867. Puchtel IS, Walker RJ, James OB, and Kring DA (2008) Osmium isotope and highly Shearer CK, Sharp ZD, McCubbin FM, et al. (2012) Chlorine distribution and its isotope siderophile element systematics of lunar impact melt breccias: Implications for the composition, alteration mineralogy, and micro-textural analysis of “Rusty Rock” late accretion history of the Moon and Earth. Geochimica et Cosmochimica Acta 66095: Implications for the petrogenesis of “Rusty Rock”, origin of “rusty” 72: 3022–3042. alteration, and volatile element behavior on the Moon. Lunar and Planetary Science Raedeke LD and McCallum IS (1980) A comparison of fractionation trends in the lunar 43: #1416. crust and in the Stillwater Complex. In: Papike JJ and Merrill RB (eds.) Proceedings Shervais JW and McGee JJ (1999) KREEP cumulates in the western lunar highlands, of the Conference on the Lunar Highlands Crust, pp. 133–153. New York: ion and electron microprobe study of alkali-suite anorthosites and norites from Pergamon. Apollo 12 and 14. American Mineralogist 84: 806–820. Raymond SN, O’Brien DP, Morbidelli A, and Kaib NA (2009) Building the Shih C-Y, Nyquist LE, Bansal BM, and Wiesmann H (1992) Rb-Sr and Sm-Nd terrestrial planets: Constrained accretion in the inner Solar System. Icarus chronology of an Apollo 17 KREEP basalt. Earth and Planetary Science Letters 203: 644–662. 108: 203–215. Rhodes JM, Blanchard DP, Dungan MA, Brannon JC, and Rodgers KV (1977) Shih C-Y, Nyquist LE, Bogard DD, et al. (1986) Geochronology and petrogenesis of Chemistry of Apollo 12 mare basalts: Magma types and fractionation processes. Apollo 14 very high potassium basalts. Proceedings of the Lunar and Planetary
Proceedings of the Lunar Science Conference, vol. 8, pp. 1305–1338. Science Conference, vol. 16, pp. D214–D228.
Righter K (2007) Not so rare Earth? New developments in understanding the origin of Shih C-Y, Nyquist LE, Dasch EJ, Bogard DD, Bansal BM, and Wiesmann H (1993) Age the Earth and Moon. Chemie der Erde 67: 179–200. of pristine noritic clasts from lunar breccias 15445 and 15455. Geochim Ringwood AE and Kesson SE (1976) A dynamic model for mare basalt petrogenesis. Cosmochim Acta 57: 915–931. Proceedings of the Lunar Science Conference, vol. 7, pp. 1697–1722. Shih C-Y, Nyquist LE, Reese YD, and Bischoff A (2008) Sm-Nd and Rb-Sr isotopic Robert F, Gautier D, and Dubrulle B (2000) The solar system D/H ratio: Observations studies of meteorite Kalahari 009: An old VLT mare basalt. Lunar and Planetary and theories. Space Science Reviews 92: 201–224. Science 39: #2165.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
248 The Moon
Shih C-Y, Nyquist LE, Reese Y, Wiesmann H, Nazarov MA, and Taylor LA (2002) Terada K, Sasaki Y, Anand M, Sano Y, Taylor LA, and Horie K (2008) Uranium lead The chronology and petrogenesis of the mare basalt clast from lunar meteorite systematics of low-Ti basaltic meteorite Dhofar 287A: Affinity to Apollo 15 green Dhofar 287: Rb–Sr and Sm–Nd isotopic studies. Lunar and Planetary Science glasses. Earth and Planetary Science Letters 270: 119–124. 38: #1344. Thalmann C, Eugster O, Herzog GF, et al. (1996) History of lunar meteorites Queen Shih C-Y, Nyquist LE, Reese Y, Wiesmann H, and Schwandt C (2001) Rb-Sr and Sm- Alexandra Range 93069, Asuka 881757, and Yamato 791639 based on noble gas Nd isotopic constraints on the genesis of lunar green and orange glasses. Lunar and abundances, radionuclide concentrations, and chemical composition. Meteoritics Planetary Science 32: #1401. and Planetary Science 31: 857–868. Shirley DN (1983) A partially molten magma ocean model. Proceedings of the Lunar Thomson W (1864) On the secular cooling of the Earth. Transactions of the Royal and Planetary Science Conference, vol. 13, pp. A519–A527. Society of Edinburgh 23: 157–169. Shoemaker EM, Hait MH, Swann GA, et al. (1970) Origin of the lunar regolith at Thomson BJ, Grosfils EB, Bussey DBJ, and Spudis PD (2009) A new technique for Tranquillity base. Proceedings of the Lunar Science Conference, vol. 1, pp. estimating the thickness of mare basalts in Imbrium Basin. Geophysical Research 2399–2412. Letters 36: L12201. Simonds CH, Warner JL, Phinney WC, and McGee PE (1976) Thermal model for impact Tittmann BR, Abdel-Gawad M, and Housley RM (1972) Elastic velocity and Q factor breccia lithification: Manicouagan and the moon. Proceedings of the Lunar Science measurements on Apollo 12, 14, and 15 rocks. Proceedings of the Lunar Science Conference, vol. 7, pp. 2509–2528. Conference, vol. 3, pp. 2565–2575. Snyder GA, Borg LE, Nyquist LE, and Taylor LA (2000) Chronology and isotopic Tollari N, Tooplis MJ, and Barnes S-J (2006) Predicting phosphate saturation in silicate constraints on lunar evolution. In: Canup R and Righter K (eds.) Origin of the Earth magmas: An experimental study of the effects of melt composition and temperature. and Moon, pp. 361–395. Arizona: University of Arizona Press. Geochimica et Cosmochimica Acta 70: 1518–1536. Snyder GA, Neal CR, Taylor LA, and Halliday AN (1995) Processes involved in the Tolstikhin IN, Kramers JD, and Hofmann AW (2006) A chemical Earth model with whole formation of magnesian-suite plutonic rocks from the highlands of Earth’s Moon. mantle convection: The importance of a core–mantle boundary layer (D’’) and its Journal of Geophysical Research, Planets 100: 9365–9388. early formation. Chemical Geology 226: 79–99. Snyder GA, Taylor LA, and Neal CR (1992) The sources of mare basalts: A model Tompkins S and Pieters CM (1999) Mineralogy of the lunar crust: Results from involving lunar magma ocean crystallization, plagioclase flotation and trapped Clementine. Meteoritics and Planetary Science 34: 25–41. instantaneous residual liquid. Geochimica et Cosmochimica Acta 56: 3809–3823. Treiman AH, Maloy AK, Shearer CK, and Gross J (2010) Magnesian anorthositic Sokol AK, Fernandes VA, Schulz T, et al. (2008) Geochemistry, petrology and ages of granulites in lunar meteorites Allan Hills A81005 and Dhofar 309: Geochemistry and the lunar meteorites Kalahari 008 and 009: New constraints on early lunar evolution. global significance. Meteoritics and Planetary Science 45: 163–180. Geochimica et Cosmochimica Acta 72: 4845–4873. Trinquier A, Birck J-L, and Allegre CJ (2007) Widespread 54Cr heterogeneity in the
Spicuzza MJ, Day JMD, Taylor LA, and Valley JW (2007) Oxygen isotope constraints on inner solar system. Astrophysical Journal 655: 1179–1185.
the origin and differentiation of the Moon. Earth and Planetary Science Letters Trinquier A, Elliott T, Ulfbeck D, Coath C, Krot AN, and Bizzarro M (2009) Origin of
253: 254–265. nucleosynthetic isotope heterogeneity in the solar protoplanetary disk. Science
Spudis PD, Wilhelms DE, and Robinson MS (2011) The Sculptured Hills of the Taurus 324: 374–376. Highlands: Implications for the relative age of Serenitatis, basin chronologies and the Unruh DM, Stille P, Patchett PJ, and Tatsumoto M (1984) Lu-Hf and Sm-Nd evolution cratering history of the Moon. Journal of Geophysical Research, Planets 116: E00H03. in lunar mare basalts. Proceedings of the Lunar and Planetary Science Conference, Staid MI and Pieters CM (2001) Mineralogy of the last lunar basalts: Results from vol. 14, pp. B459–B477. Clementine. Journal of Geophysical Research, Planets 106: 27887–27900. van Kan Parker M, Sanloup C, Sator N, et al. (2012) Neutral buoyancy of titanium-rich Stewart DB (1975) Apollonian metamorphic rocks – The products of prolonged melts in the deep lunar interior. Nature Geoscience 5: 186–189. subsolidus equilibration. Lunar Science 6: 774–776. Vaniman D, Dietrich J, Taylor GJ, and Heiken G (1991) Exploration, samples, and recent Sto¨ffler D, Kno¨ll H-D, Marvin UB, Simonds CH, and Warren PH (1980) Recommended concepts of the Moon. In: Heiken G, Vaniman D, and French B (eds.) Lunar classification and nomenclature of lunar highland rocks – A committee report. Sourcebook, A User’s Guide to the Moon, pp. 5–26. Cambridge: Cambridge In: Papike JJ and Merrill RB (eds.) Proceedings of the Conference on the Lunar University Press. Highlands Crust, pp. 51–70. New York: Pergamon. Walker D (1983) Lunar and terrestrial crust formation. Proceedings of the Lunar Science Strom RG, Malhotra R, Ito T, Yoshida F, and Kring DA (2005) The origin of planetary Conference, vol. 14, pp. B17–B25. impactors in the inner solar system. Science 309: 1847–1850. Walker D, Longhi J, Lasaga AC, Stolper EM, Grove TL, and Hays JF (1977) Slowly Suckale J, Elkins-Tanton LT, and Sethian JA (2012) Crystals stirred up: 2. Numerical cooled microgabbros 15555 and 15065. Proceedings of the Lunar Science insights into the formation of the earliest crust on the Moon. Journal of Geophysical Conference, vol. 8, pp. 1521–1547. Research, Planets 117: E008005. Walker RJ (2009) Highly siderophile elements in the Earth, Moon and Mars: Update and Swindle TD, Spudis PD, Taylor GJ, Korotev RL, Nichols RH Jr., and Olinger CT (1991) implications for planetary accretion and differentiation. Chemie der Erde Searching for Crisium Basin ejecta: Chemistry and ages of Luna 20 impact melts. 69: 101–125. Proceedings of the Lunar and Planetary Science, vol. 21, pp. 167–181. Walker RJ, Horan MF, Shearer CK, and Papike JJ (2004) Low abundances of highly Swinyard BM, Joy KH, Kellett BJ, et al. (2009) X-ray fluorescence observations of the siderophile elements in the lunar mantle: Evidence for prolonged late accretion. moon by SMART-1/D-CIXS and the first detection of Ti Ka from the lunar surface. Earth and Planetary Science Letters 224: 399–413. Planetary and Space Science 57: 744–750. Wang Y, Hsu W, Guan Y, et al. (2012) Petrogenesis of the Northwest Takeda H, Yamaguchi A, Bogard DD, et al. (2006) Magnesian anorthosites and a deep Africa 4734 basaltic lunar meteorite. Geochimica et Cosmochimica Acta crustal rock from the farside crust of the Moon. Earth and Planetary Science Letters 92: 329–344. 247: 171–184. Warner JL, Phinney WC, Bickel CE, and Simonds CH (1977) Feldspathic granulitic Taylor SR (1982) Planetary Science: A Lunar Perspective. Houston: Lunar and Planetary impactites and pre-final bombardment lunar evolution. Proceedings of the Lunar Institute. Science Conference, vol. 8, pp. 2051–2066. Taylor SR and Jakes P (1974) The geochemical evolution of the Moon. Proceedings of Warner RD, Keil K, and Taylor GJ (1977) Coarse-grained basalt 71597: A product of the Lunar Science Conference, vol. 5, pp. 1287–1305. partial olivine accumulation. Proceedings of the Lunar Science Conference, vol. 8, Taylor DJ, McKeegan KD, and Harrison TM (2009) Lu–Hf zircon evidence for rapid pp. 1429–1442. lunar differentiation. Earth and Planetary Science Letters 279: 157–164. Warren PH (1985) The magma ocean concept and lunar evolution. Annual Review of Taylor LA, Shervais JW, Hunter RH, et al. (1983) Pre-4.2 AE mare basalt volcanism in Earth and Planetary Sciences 13: 201–240. the lunar highlands. Earth and Planetary Science Letters 66: 33–47. Warren PH (1986) Anorthosite assimilation and the origin of the Mg/Fe-related Taylor SR, Taylor GJ, and Taylor LA (2006) The Moon: A Taylor perspective. bimodality of pristine Moon rocks: Support for the magmasphere hypothesis. Geochimica et Cosmochimica Acta 70: 5904–5918. Proceedings of the Lunar and Planetary Science Conference, vol. 16, Taylor GJ, Warren P, Ryder G, Delano J, Pieters C, and Lofgren G (1991) Lunar rocks. pp. D331–D343.
In: Heiken G, Vaniman D, and French B (eds.) Lunar Sourcebook: A User’s Guide to Warren PH (1988) The origin of pristine KREEP: Effects of mixing between urKREEP and
the Moon, pp. 183–284. New York: Cambridge University Press. the magmas parental to the Mg-rich cumulates. Proceedings of the Lunar and Tera F, Papanastassiou DA, and Wasserburg GJ (1974) Isotopic evidence for a terminal Planetary Science Conference, vol. 18, pp. 233–241. lunar cataclysm. Earth and Planetary Science Letters 22: 1–21. Warren PH (1989) KREEP: Major-element diversity, trace-element uniformity (almost). Terada K, Anand M, Sokol AK, Bischoff A, and Sano Y (2007) Cryptomare In: Taylor GJ and Warren PH (eds.) Workshop on Moon in Transition: Apollo 14, magmatism 4.35 Gyr ago recorded in lunar meteorite Kalahari 009. Nature KREEP, and Evolved Lunar Rocks (Technical Report 89-03), pp. 149–153. Houston: 450: 849–852. Lunar and Planetary Institute.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
The Moon 249
Warren PH (1990) Lunar anorthosites and the magma ocean hypothesis: Importance of Warren PH and Wasson JT (1980b) Further foraging for pristine nonmare rocks: FeO enrichment in the parent magma. American Mineralogist 75: 46–58. Correlations between geochemistry and longitude. Proceedings of the Lunar Warren PH (1992) Inheritance of silicate differentiation during lunar origin by giant Science Conference, vol. 11, pp. 431–470. impact. Earth and Planetary Science Letters 112: 101–116. Wasson JT (1985) Meteorites: Their Record of Early Solar System History. New York: Warren PH (1993) A concise compilation of key petrologic information on possibly Freeman. pristine nonmare Moon rocks. American Mineralogist 78: 360–376. Wasson JT, Boynton WV, Chou C-L, and Baedecker PA (1975) Compositional evidence Warren PH (1994) Lunar and martian meteorite delivery services. Icarus regarding the influx of interplanetary materials onto the lunar surface. Moon 111: 338–363. 13: 121–141. Warren PH (1996) Global inventory of lunar impact melt as a function of parent crater Wasson JT and Kallemeyn GW (1988) Compositions of chondrites. Philosophical size. Lunar and Planetary Science 27: 1379–1380. Transactions of the Royal Society London A325: 535–544. Warren PH (2003) The Moon. In: Holland HD and Turekian KK (eds.) Treatise on Weber RC, Lin P-Y, Garnero EJ, Williams Q, and Lognonne´ P (2011) Seismic detection Geochemistry, vol. 1, pp. 559–599. Oxford: Elsevier. of the lunar core. Science 331: 309–312. Warren PH (2005) “New” lunar meteorites: Implications for composition of the global Weider SZ, Kellett BJ, Swinyard BM, et al. (2012) The Chandrayaan-1 X-ray lunar surface, of the lunar crust, and of the bulk Moon. Meteoritics and Planetary Spectrometer: First results. Planetary and Space Science 60: 217–228. Science 40: 477–506. Wetherill GW (1981) Nature and origin of basin-forming projectiles. In: Schultz PH and Warren PH (2008a) A depleted, not ideally chondritic bulk Earth: The explosive-volcanic Merrill RB (eds.) Multi-ring Basins, pp. 1–18. New York: Pergamon. basalt loss hypothesis. Geochimica et Cosmochimica Acta 72: 2217–2235. Wetzel DT, Jacobsen SD, Rutherford MJ, Hauri EH, and Saal AE (2012) The solubility Warren PH (2008b) Lunar rock-rain: Diverse silicate impact-vapor condensates and speciation of carbon in lunar picritic magmas. Lunar and Planetary Science in an Apollo-14 regolith breccia. Geochimica et Cosmochimica Acta 43: #1535. 72: 3562–3585. Weyer S, Anbar AD, Brey GP, Munker C, Mezger K, and Woodland AB (2005) Iron Warren PH (2011) Stable-isotopic anomalies and the accretionary assemblage of the isotope fractionation during planetary differentiation. Earth and Planetary Science Earth and Mars: A subordinate role for carbonaceous chondrites. Earth and Letters 240: 251–264. Planetary Science Letters 311: 93–100. Wiechert U, Halliday AN, Lee D-C, Snyder GA, Taylor LA, and Rumble D (2001) Oxygen Warren PH, Claeys P, and Cedillo-Pardo E (1996) Mega-impact melt petrology isotopes and the Moon-forming giant impact. Science 294: 345–348. (Chicxulub, Sudbury, and the Moon): Effects of scale and other factors on potential Wieczorek MA (2003) The thickness of the lunar crust: How low can you go? Lunar and for fractional crystallization and development of cumulates. In: Ryder G, Planetary Science 34: #1330. Fastovsky D, and Gartner S (eds.) The Cretaceous-Tertiary Event and Other Wieczorek MA, Neumann GA, Nimmo F, et al. (2013) The crust of the Moon as seen by
Catastrophes in Earth History, pp. 105–124. Geological Society of America Special GRAIL. Science (December 5, 2012), http://dx.doi.org/10.1126/science.1231530.
Paper 307. Available at: http://www.sciencemag.org/content/early/2012/12/04/science.
Warren PH, Haack H, and Rasmussen KL (1991) Megaregolith insulation and the 1231530.abstract.
duration of cooling to isotopic closure within differentiated asteroids and the Moon. Wilhelms DE (1987) The Geologic History of the Moon. Washington, DC: US Geological Journal of Geophysical Research, Planets 96: 5909–5923. Survey (USGS Professional Paper 1348). Warren PH, Jerde EA, and Kallemeyn GW (1987) Pristine Moon rocks: A “large” felsite Wilson L and Head JW III (2003) Depth generation of magmatic gas on the Moon and and a metal-rich ferroan anorthosite. Proceedings of the Lunar and Planetary implications for pyroclastic eruptions. Geophysical Research Letters 30: http://dx. Science Conference, vol. 17, pp. E303–E313. doi.org/10.1029/2002GL016082. Warren PH, Jerde EA, and Kallemeyn GW (1990) Pristine Moon rocks: An alkali Wo¨hler C, Berezhnoy A, and Evans R (2011) Estimation of elemental abundances of the anorthosite with coarse augite exsolution from plagioclase, a magnesian lunar regolith using Clementine UVVISþNIR data. Planetary and Space Science harzburgite, and other oddities. Proceedings of the Lunar and Planetary Science 59: 92–110. Conference, vol. 21, 31–59. Wolf R and Anders E (1980) Moon and Earth: Compositional differences inferred from Warren PH, Jerde EA, and Kallemeyn GW (1991) Pristine Moon rocks: Apollo 17 siderophiles, volatiles, and alkalis in basalts. Geochimica et Cosmochimica Acta anorthosites. Proceedings of the Lunar and Planetary Science Conference, vol. 21, 44: 2111–2124. pp. 51–61. Wolf R, Woodrow A, and Anders E (1979) Lunar basalts and pristine highland rocks: Warren PH and Kallemeyn GW (1984) Pristine rocks (8th foray): “Plagiophile” Comparison of siderophile and volatile elements. Proceedings of the Lunar and element ratios, crustal genesis, and the bulk composition of the Moon. Planetary Science Conference, vol. 10, pp. 2107–2130. Proceedings of the Lunar and Planetary Science Conference, vol. 15, Wood JA (1975) Lunar petrogenesis in a well-stirred magma ocean. Proceedings of the pp. C16–C24. Lunar Science Conference, vol. 6, pp. 1087–1102. Warren PH and Kallemeyn GW (1986) Geochemistry of lunar meteorite Yamato-791197: Wood JA, Dickey JS, Marvin UB, and Powell BN (1970) Lunar anorthosites and a Comparison with ALHA81005 and other lunar samples. Proceedings of the NIPR geophysical model of the Moon. Proceedings of the Apollo 11 Lunar Science Symposium on Antarctic Meteorites (Tokyo), vol. 10, pp. 3–16. Conference, pp. 965–988. Warren PH, Kallemeyn GW, and Kyte FT (1999) Origin of planetary cores: Evidence from Wu Y (2012) Major elements and Mg# of the Moon: Results from Chang’E-1 highly siderophile elements in martian meteorites. Geochimica et Cosmochimica Interference Imaging Spectrometer (IIM) data. Geochimica et Cosmochimica Acta Acta 63: 2105–2122. 93: 214–234. Warren PH, Taylor GJ, Keil K, Shirley DN, and Wasson JT (1983) Petrology and Yamaguchi A, Karouji Y, Takeda H, et al. (2010) The variety of lithologies in the Yamato- chemistry of two “large” granite clasts from the Moon. Earth and Planetary Science 86032 lunar meteorite: Implications for formation processes of the lunar crust. Letters 64: 175–185. Geochimica et Cosmochimica Acta 74: 4507–4530. Warren PH, Ulff-Møller F, and Kallemeyn GW (2005) “New” lunar meteorites: Impact Yamamoto S, Nakamura S, Matsunaga T, et al. (2012) Olivine-rich exposures in the melt and regolith breccias and large-scale heterogeneities of the upper lunar crust. South Pole-Aitken Basin. Icarus 218: 331–344. Meteoritics and Planetary Science 40: 989–1014. Yamashita N, Hasebe N, Reedy RC, et al. (2010) Uranium on the Moon: Global Warren PH and Wasson JT (1977) Pristine nonmare rocks and the nature of distribution and U/Th ratio. Geophysical Research Letters 37: L10201. the lunar crust. Proceedings of the Lunar Science Conference, vol. 8, Yu G and Jacobsen SB (2011) Fast accretion of the Earth with a late Moon-forming pp. 2215–2235. giant impact. Proceedings of the National Academy of Sciences Warren PH and Wasson JT (1978) Compositional-petrographic investigation of pristine 108: 17604–17609. nonmare rocks. Proceedings of the Lunar and Planetary Science Conference, vol. 9, Zeigler RA, Jolliff BL, Korotev RL, and Haskin LA (2000) Petrology, geochemistry, and 185–217. possible origin of monomict mafic lithologies of the Cayley Plains. Lunar and Warren PH and Wasson JT (1979a) The origin of KREEP. Reviews of Geophysics and Planetary Science 31: #1859. Space Physics 17: 73–88. Zeigler RA, Korotev RL, Jolliff BL, and Haskin LA (2005) Petrography and geochemistry
Warren PH and Wasson JT (1979b) Effects of pressure on the crystallization of of the LaPaz Icefield basaltic lunar meteorite and source crater pairing with
a “chondritic” magma ocean and implications for the bulk composition of the Moon. Northwest Africa 032. Meteoritics and Planetary Science 40: 1073–1101. Proceedings of the Lunar and Planetary Science Conference, vol. 10, Zellner NEB, Delano JW, Swindle TD, Barra F, Olsen E, and Whittet DCB (2009) pp. 2051–2083. Evidence from 40Ar/39Ar ages of lunar impact glasses for an increase in the impact Warren PH and Wasson JT (1980a) Early lunar petrogenesis, oceanic and extraoceanic. rate 800 Ma ago. Geochimica et Cosmochimica Acta 73: 4590–4597. In: Merrill RB and Papike JJ (eds.) Proceedings of the Conference on the Lunar Zhang J, Dauphas N, Davis AM, Leya I, and Fedkin A (2012) The proto-Earth as a Highlands Crust, pp. 81–99. New York: Pergamon. significant source of lunar material. Nature Geoscience 5: 251–255.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250
Author's personal copy
250 The Moon
Zhang A-C and Hsu W-B (2009) Petrography, mineralogy, and trace element Zhu M-H, Ma T, and Chang J (2010) Chang’E-1 gamma ray spectrometer and geochemistry of lunar meteorite Dhofar 1180. Meteoritics and Planetary Science preliminary radioactive results on the lunar surface. Planetary and Space Science 44: 1265–1286. 60: 217–228. Zhong S, Parmentier EM, and Zuber MT (2000) A dynamic origin for the global Zieg MJ and Marsh BD (2005) The Sudbury Igneous Complex: Viscous emulsion asymmetry of lunar mare basalts. Earth and Planetary Science Letters differentiation of a superheated impact melt sheet. Geological Society of America 177: 131–140. Bulletin 117: 1427–1450.
Treatise on Geochemistry, Second Edition, (2014), vol. 2, pp. 213-250