Early formation of evolved asteroidal crust

Abstract

Mechanisms for the formation of crust on planetary bodies remain poorly understood. It is generally accepted that Earth's andesitic continental crust is the product of plate tectonics, whereas the Moon acquired its feldspar-rich crust by way of plagioclase flotation in a magma ocean. Basaltic meteorites provide evidence that, like the terrestrial planets, some asteroids generated crust and underwent large-scale differentiation processes. Until now, however, no evolved felsic asteroidal crust has been sampled or observed. Here we report age and compositional data for the newly discovered, paired and differentiated meteorites Graves Nunatak (GRA) 06128 and GRA 06129. These meteorites are feldspar-rich, with andesite bulk compositions. Their age of 4.52 +/- 0.06 Gyr demonstrates formation early in Solar System history. The isotopic and elemental compositions, degree of metamorphic re-equilibration and sulphide-rich nature of the meteorites are most consistent with an origin as partial melts from a volatile-rich, oxidized asteroid. GRA 06128 and 06129 are the result of a newly recognized style of evolved crust formation, bearing witness to incomplete differentiation of their parent asteroid and to previously unrecognized diversity of early-formed materials in the Solar System.

Full text

LETTERS
Early formation of evolved asteroidal crust
James M. D. Day
1
, Richard D. Ash
1
, Yang Liu
2
, Jeremy J. Bellucci
1
, Douglas Rumble III
3
, William F. McDonough
1
,
Richard J. Walker
1
& Lawrence A. Taylor
2
Mechanisms for the formation of crust on planetary bodies remain
poorly understood
1
. It is generally accepted that Earth’s andesitic
continental crust is the product of plate tectonics
1,2
, whereas the
Moon acquired its feldspar-rich crust by way of plagioclase flo-
tation in a magma ocean
3,4
. Basaltic meteorites provide evidence
that, like the terrestrial planets, some asteroids generated crust
and underwent large-scale differentiation processes
5
. Until now,
however, no evolved felsic asteroidal crust has been sampled or
observed. Here we report age and compositional data for the newly
discovered, paired and differentiated meteorites Graves Nunatak
(GRA) 06128 and GRA 06129. These meteorites are feldspar-rich,
with andesite bulk compositions. Their age of 4.52 60.06 Gyr
demonstrates formation early in Solar System history. The iso-
topic and elemental compositions, degree of metamorphic re-
equilibration and sulphide-rich nature of the meteorites are most
consistent with an origin as partial melts from a volatile-rich,
oxidized asteroid. GRA 06128 and 06129 are the result of a newly
recognized style of evolved crust formation, bearing witness to
incomplete differentiation of their parent asteroid and to pre-
viously unrecognized diversity of early-formed materials in the
Solar System.
Formation of crust, the outermost solid shell of a planet, is a
fundamental process and its chemical nature is a reflection of the
formation, differentiation and cooling history of its parent body.
Thus, documenting causes of lithological diversity in crustal materi-
als is critical for understanding early Solar System processes and
planetary evolution. Knowledge regarding initial formation of plane-
tary crust is largely based on data gleaned from the oldest preserved
crustal rock and mineral remnants from Earth (,4.4 Gyr; refs 6, 7),
the Moon (,4.4 Gyr; ref. 8), Mars (.4.0 Gyr; ref. 9) and achondrite
meteorites (#4.56 Gyr; ref. 10). These materials record crustal
formation processes that were demonstrably diverse among these
bodies.
GRA 06128 and 06129 (hereafter referred to as GRA 06128/9) are
paired achondritic meteorites recovered from the same Antarctic ice-
field. The meteorites consist largely of sodium-rich plagioclase
(.75%), with olivine, two pyroxenes, phosphates and sulphides.
They have andesite to trachy-andesite bulk compositions (Fig. 1a; see
Methods and Supplementary Information). Major- and trace-element
compositions of silicate minerals are uniform within and between the
two meteorites. Minerals are compositionally unzoned and major sil-
icate phases have variable ranges of grain size (diameter ,0.1 to
.0.5 mm). Co-existing augite and orthopyroxene yield equilibration
temperatures of ,800uC. Oligoclase crystals have large positive Eu
anomalies, and merrillite and chlorapatite have elevated rare-earth
element (REE) abundances, rel ative to other minerals in the meteorites.
Consequently, the estimated bulk REE composition of GRA 06128/9,
determined by modal recombination, is dominated by feldspar for Eu
and phosphates for the other REEs (Fig. 1b). The meteorites are
enriched to only moderately depleted in volatile elements (for example
K, Na, S, Rb, Cl, Pb) relative to chondrites, and were formed at an
oxygen fugacity close to the iron–wu
¨stite 12 buffer (ref. 11).
Although GRA 06128/9 had an igneous origin, they have also been
thermally metamorphosed and partially brecciated. Both meteorites
possess granoblastic textures, 120utriple junctions between coexist-
ing silicates, polysynthetic twinning in plagioclase and pentlandite-
troilite exsolution from a monosulphide solid solution. These fea-
tures are consistent with slow cooling and partial re-equilibration.
Using pyroxene exsolution lamellae, it has been estimated that GRA
06128/9 formed close to the surface (at depths of 15–20 m) of their
1
Department of Geology, University of Maryland, College Park, Maryland 20742, USA.
2
Department of Earth and Planetary Sciences, Planetary Geosciences Institute, University of
Tennessee, Knoxville, Tennessee 37996, USA.
3
Geophysical Laboratory, Carnegie Institution for Science, Washington DC 20015, USA.
SiO2 (wt%)
40 50 60 70
Phonolite
TP
PT
Basanite/
tephrite
Trachyte
Rhyolite
Dacite
Andesite
Basalt
and.
Basalt
Pic.
bas.
Foidite
Trachy-
andesite
BT A
TB
Measured and calculated
composition of
GRA 06128/9
Total alkalis (wt%)
Average terrestrial andesite
continental crust
Brachinites
GRA 06128/9
a
b
La Ce Pr Nd Sm Eu Gd Tb D
y
Ho Er Tm Yb
0
2
4
6
8
10
12
14
16
0.1
1
10
Sample/CI-chondrite
0.01
100
Figure 1
|
Bulk composition of the GRA 06128/9 achondrite meteorites.
a, Plot of total alkalis (Na
2
O and K
2
O) versus silica, showing multiple
measurements (filled circles) and calculated compositions (grey area) of the
GRA 06128/9 achondrite meteorites. Calculated compositions are based on
the variability in modal mineralogy and mineral major element
compositions (Supplementary Information). Abbreviations: TP, trachy-
phonolite; PT, phono-tephrite; TB, trachy-basalt; BTA, basaltic trachy-
andesite; Pic. bas., picro-basalt; Basalt and., basaltic andesite. b, Measured
REE patterns for the GRA 06128/9 meteorites. Shown are data for the
average terrestrial continental crust
1
and for the brachinites, Brachina, ALH
84025 and EET 99402/407 (ref. 26).
Vol 457
|
8 January 2009
|
doi:10.1038/nature07651
179
Macmillan Publishers Limited. All rights reserved
©2009

parent body
12
. This conclusion is consistent with an origin as evolved
crustal material.
Oxygen isotopes (as D
17
O values) provide a means to genetically
link Solar System materials (ref. 13; D
17
O notation is defined in
Methods). The D
17
O values for multiple pieces of GRA 06128/9
average 20.195 60.012%(Fig. 2; Table 1). This isotopic composi-
tion is different from most known differentiated bodies, including
the Earth, Moon and Mars.
The mean
207
Pb–
206
Pb age determined for chlorapatite in GRA
06128/9 is 4.517 60.060 Gyr (2s; Supplementary Methods). The
mean age for merrillite crystals is identical within the greater uncer-
tainty. Since
207
Pb–
206
Pb ages reflect the time of cessation of Pb
diffusion (ref. 14), the measured phosphate ages probably reflect
closure temperatures subsequent to the metamorphic event recorded
from pyroxene thermometry (#800 uC). Assuming that diffusion
characteristics are similar to those of terrestrial phosphates, the clos-
ure temperature of Pb in merrillite and chlorapatite in GRA 06128/9
is ,500 uC (ref. 14). Thus, the phosphate age of GRA 06128/9
demonstrates that crystallization, thermal metamorphism and cool-
ing below 500 uC occurred within ,100 Myr of the formation of the
Solar System at ,4.567 Gyr ago
15
. These ages can be used to argue
against an origin on any major planetary body, including Venus or
Mercury. The average age of the crust on Venus is estimated to be
,1 Gyr (ref. 16), and present knowledge of Mercury suggests it is
highly reduced, with a crust that is younger than 4.4 Gyr (ref. 17).
Taken with the oxygen isotope evidence, we conclude that GRA
06128/9 originated on an asteroid.
Owing to their tendency to strongly partition into metal relative to
silicate, the highly siderophile elements (HSE; including Re, Os, Ir,
Ru, Pt and Pd) are important recorders of primary planetary differ-
entiation. Concentrations of the HSE in bulk samples of GRA 06128/9
are elevated, in some cases within a factor of two or three of chondritic
abundances (Table 2; Fig. 3). These results demonstrate that a metallic
core had not segregated before generation of the GRA 06128/9 parental
melts. Some of the HSE are fractionated relative to one another in GRA
06128/9. These fractionated HSE compositions could not have been
incorporated into the rock via the impact of any known chondritic or
metal-rich meteorite material. Nor can they be explained through
terrestrial weathering processes. Although the meteorites are weath-
ered (Supplementary Information), laser ablation inductively coupled
plasma-mass spectrometry analysis of unaltered sulphides and FeNi
metal in the meteorites reveals that the HSE are hosted almost entirely
within these primary magmatic phases (Fig. 3). Furthermore, the
calculated initial
187
Os/
188
Os ratio for the GRA 06128/9 meteorites
(0.096 60.001) is within error of the initial Solar System value,
inconsistent with disturbance of the
187
Re–
187
Os system and consist-
ent with formation via early partial melting of an undifferentiated
parent body.
The formation mechanism for GRA 06128/9 was evidently unique
among known achondrites. Plagioclase-rich lunar ferroan anortho-
sites are considered to represent flotation cumulates from a large-
scale magma ocean melting event
3,4
. The elevated abundances of the
HSE, however, argue against this mode of origin for the GRA 06128/9
meteorites, as .50% melting of the silicate fraction of an asteroid
would lead to formation of a gravitationally separated metal–liquid
core (ref. 18). Consequently, large-scale differentiation leads to
marked depletion of HSE in the silicate portion of the body. For
example, lunar and terrestrial crustal rocks respectively contain less
than ,0.02 ng g
21
and ,0.04 ng g
21
Os (refs 1, 19), as compared
with ,200 ng g
21
Os for GRA 06128/9. Furthermore, HED meteor-
ites, which are considered to derive from the asteroid 4 Vesta (mean
diameter ,530 km) and are probably the result of a large-scale
magma ocean melting event
5
, also have very low abundances of the
HSE
20
—compositions at odds with those of GRA 06128/9.
It has been demonstrated that quartz-normative andesite compo-
sitions like that of GRA 06128/9 can be generated by partial melting
of volatile-rich chondritic precursors (see, for example, ref. 21).
These studies have shown that partial melting of an olivine-rich
lithology in the forsterite-anorthite-quartz system occurs at the peri-
tectic point, resulting in a melt that contains .50% plagioclase
22
.
Furthermore, Na-rich compositions lower liquidus temperatures
by up to 100 uC and shift the initial melt compositions to even higher
abundances of feldspar
23,24
. These experimental constraints, together
with the major-, trace-element and mineralogical data for GRA
06128/9, indicate that they most probably crystallized from magma
generated by partial melting of a primitive chondritic source.
The formation of GRA 06128/9 by partial melting and melt
segregation of a largely undifferentiated body implies the possibility
of a complementary ultramafic residue or cumulate. There are
numerous mafic and ultramafic achondrite meteorites in the terrest-
rial collection, of which the brachinites possess the most comple-
mentary characteristics to the GRA 06128/9 meteorites, including
overlapping D
17
O values (Fig. 2; ref. 25). The significant O isotope
variability present in brachinites (D
17
O520.15 to 20.31%) has
been attributed to partial or incomplete melting of their primitive
234567
Terrestrial fractionation line
Moon
Angrites
IIIAB
MGP
HED
Mesosiderites
δ18O (‰)
∆17O (‰)
–0.4
–0.3
–0.2
–0.1
0.0
Brachinites
GRA 06128/9
Figure 2
|
d
18
O
–
D
17
O plot for GRA 06128/9 versus achondrite meteorites
and lunar and terrestrial materials. MGP, main group pallasites (open
diamonds); IIIAB, IIIAB iron meteorites (filled grey squares); HED,
howardite-eucrite-diogenite meteorites (filled triangles). The GRA 06128/9
meteorites have oxygen isotope compositions most similar to brachinites.
Published data are from refs 5, 32 and references therein. Error bars for data
are smaller than symbols.
Table 1
|
Three oxygen isotope data for GRA 06128 and GRA 06129
Sample, specific number d
17
O(%)d
18
O(%)D
17
O(%)
GRA 06128,22 2.457 5.052 20.200
GRA 06128,22 2.330 4.822 20.207
GRA 06128 average 2.394 60.090 4.937 60.162 20.204 60.005
GRA 06129,92.421 4.968 20.192
GRA 06129,92.334 4.777 20.179
GRA 06129 average 2.378 60.062 4.873 60.135 20.186 60.009
Table 2
|
Whole-rock highly siderophile element data with initial Os isotopic compositions at 4.52 Gyr
Sample Specific number Mass (g) Os (p.p.b.) Ir (p.p.b.) Ru (p.p.b.) Pt (p.p.b.) Pd (p.p.b.) Re (p.p.b.)
187
Os/
188
Os
m
2s.e.
187
Re/
188
Os 2s.e.
187
Os/
188
Os
i
2s.e.
GRA 06128 22 0.21 175.178.58 301.4 125.953.55 16.06 0.13100 0.00009 0.442 0.007 0.0964 0.0015
GRA 06129 9 0.26 265.095.56 378.1 143.054.93 24.30 0.13117 0.00005 0.442 0.007 0.0966 0.0015
Brachina USNM 535L0.06 156.0142.6228.9 110.667.45 10.12 0.12041 0.00014 0.312 0.005 0.0960 0.0011
m, measured; i, initial; s.e., standard error.
LETTERS NATURE
|
Vol 457
|
8 January 2009
180
Macmillan Publishers Limited. All rights reserved
©2009

parent body
26,27
, consistent with the projected parent body for GRA
06128/9. Further, Brachina (the ‘type’ for brachinites) has fractio-
nated HSE, sub-chondritic measured
187
Os/
188
Os (0.1204; versus
0.1311 for GRA 06128/9) and a depleted REE pattern, features that
are complementary if brachinites represent melt residues to the GRA
06128/9 meteorites (Figs 1 and 3). These characteristics are consistent
with an origin as rocks that did not experience metal–silicate equi-
libration, but underwent partial melting processes broadly comple-
mentary to those required for GRA 06128/9. All of these
characteristics provide possible evidence that GRA 06128/9 are either
genetically related to the brachinites or derive from a parent body
with a similar melting history.
Fractionation of the HSE in GRA 06128/9 can be potentially
explained by sulphide segregation. The HSE can be fractionated via
removal of a sulphide melt from a crystalline monosulphide solid
solution in the terrestrial mantle
28
. Similar styles of fractionation of
the HSE between GRA 06128/9 and terrestrial ores may implicate this
type of process as acting during generation of the GRA 06128/9
parental melts (see Supplementary Discussion).
It has been previously demonstrated that melting and metamorph-
ism of asteroid parent bodies were unlikely to have occurred through
impact-related heating, which has been shown to be a highly inef-
ficient process
29
. This conclusion is supported by the lack of evidence
for significant impactor contributions to the HSE inventories of the
GRA 06128/9 meteorites; energy released from initial accretion of the
parent body, or through decay of short-lived radionuclides (for
example,
26
Al), represent more viable heat sources. Constraints on
the size of the parent body of the GRA 06128/9 meteorites can only be
loosely applied. On asteroids ,100 km in radius, it is likely that
volatile-rich, low-density melts such as GRA 06128/9 would exceed
escape velocities through explosive pyroclastic volcanism and be lost
to space
30
. Thus, GRA 06128/9 could have originated either via extru-
sion from a large (.100 km radius) asteroid or emplaced intrusively
on a body of undetermined size.
Remote sensing of asteroids shows that, where detected, the pre-
ponderance of crust is basaltic. This is also true for the terrestrial
planets and the asteroid 4 Vesta. Feldspar-rich crust is not uncom-
mon, with the Moon’s crust and Earth’s continental crust being
feldspar-rich. Feldspar does not have a strong spectral wavelength
absorption in the near-infrared, but its high albedo is a diagnostic
feature, and a number of E-type asteroids have been detected in the
asteroid belt with this characteristic
31
. These asteroids may also have a
significant sulphide component
31
. The presence of E-type asteroids
implies that evolved crust may be extensive on some of these bodies.
So far, however, only one planet has been found to have a major
andesite crust component, namely Earth. It has been argued that,
to generate andesite crust, a significant volatile component is
required in the mantle of the parent body
2
. On Earth, this is achieved
through recycling of water into subduction zones
1
. The GRA 06128/9
meteorites require early partial melting of primitive, volatile-rich
source regions in an asteroidal body that did not suffer extensive
planetary differentiation, and thus point to an entirely new mode
of generation of andesite crust compositions.
METHODS SUMMARY
For mineralogical characterization we used an SX50 electron microprobe and
New Wave Research UP213 (213 nm) laser-ablation (LA) system coupled to a
ThermoFinnigan Element 2 inductively coupled plasma-mass spectrometer
(ICP-MS).
207
Pb–
206
Pb ages were obtained via LA-ICP-MS and were corrected
for mass fractionation using an exponential fractionation law by means of brack-
eting the phosphate analyses with standard reference materials (SRMs: NIST 610,
NIST 612, BCR-2g). Ratios of
207
Pb/
206
Pb for each SRM were used to calculate
the fractionation factor (a). Differences in abetween the three SRMs had a
negligible effect on calculated ages. Os isotopic and platinum-group elemental
abundance measurements were made using isotope dilution and solvent extrac-
tion/anion exchange purification methodologies. Os isotopes and concentra-
tions were measured via thermal ionization mass spectrometry, and Ir, Ru, Pt,
Pd and Re abundances were measured using solution ICP-MS. Oxygen isotope
analysis was performed via laser fluorination of pre-leached powder whole-rock
aliquots stripped of magnetic minerals. Standardization of delta values was
achieved by comparison with the Gore Mountain garnet standard, USNM
107144, analysed during every analytical session.
Full Methods and any associated references are available in the online version of
the paper at www.nature.com/nature.
Received 26 May; accepted 17 November 2008.
1. Rudnick, R. L. & Gao, S. in The Crust (ed. Rudnick, R. L.) Vol. 3, Treatise on
Geochemistry (eds Holland, H. D. & Turekian, K. K.) 1
–
64 (Elsevier-Pergamon,
2003).
2. Campbell, I. H. & Taylor, S. R. No water, no granites — no oceans, no continents.
Geophys. Res. Lett. 10, 1061
–
1064 (1985).
3. Wood, J. A., Dickey, J. S., Marvin, U. B. & Powell, B. N. Lunar anorthosites and a
geophysical model of the Moon. Proc. Apollo 11 Lunar Sci. Conf. 965
–
988 (1970).
4. Smith, J. A. et al. Petrologic history of the Moon inferred from petrography,
mineralogy, and petrogenesis of Apollo 11 rocks. Proc. Apollo 11 Lunar Sci. Conf.
1149
–
1162 (1970).
Re Os Ir Ru Pt Pd
Sample/CI-chondrite
187Re/188Os
0.30 0.35 0.40 0.45 0.50
187Os/188Os
GRA 06128/9
Brachina
Carbonaceous
Enstatite
Ordinary
ab
4.56 Gyr isochron
Calculated range of whole-rock
assuming 100% of HSE are
retained in sulphide and FeNi metal
0.01
0.1
1
10
0.120
0.122
0.124
0.126
0.128
0.130
0.132
Figure 3
|
Highly siderophile element and Re
–
Os isotope systematics of
GRA 06128/9 and Brachina. a, Measured HSE patterns for GRA 06128/9
and Brachina showing significant fractionation of Ir, Pt and Pd from Re, Os
and Ru in the meteorites relative to CI chondrite (Orgueil). Also shown is the
calculated whole-rock compositional field (shaded) for GRA 06128/9 using
measured modal abundances and HSE compositions of pentlandite/FeNi
metal and FeS in the meteorites (Supplementary Information). These
calculated estimates are in broad agreement with measured whole-rock
values, indicating that HSE are dominantly hosted within sulphide and FeNi
metal. b,
187
Re/
188
Os–
187
Os/
188
Os diagram for GRA 06128/9 and Brachina
versus chondritic meteorites. GRA 06128/9 plot to elevated values relative to
all chondrite groups in
187
Re/
188
Os–
187
Os/
188
Os space. Brachina (open
circle) has present-day sub-chondritic
187
Os/
188
Os. GRA 06128/9 (filled
circles) and Brachina plot along a ,4.56 Gyr isochron, with an
187
Os/
188
Os
value close to the Solar System initial value, which passes through the field of
chondritic meteorite data. Normalization and chondrite data are from ref.
33. Error bars for GRA 06128/9 and Brachina data are smaller than symbols.
NATURE
|
Vol 457
|
8 January 2009 LETTERS
181
Macmillan Publishers Limited. All rights reserved
©2009

5. Greenwood, R. C., Franchi, I. A., Jambon, A. & Buchanan, P. C. Widespread magma
oceans on asteroidal bodies in the early Solar System. Nature 435, 916
–
918
(2005).
6. Bowring, S. A. & Williams, I. S. Priscoan (4.00
–
4.03 Ga) orthogneiss from
northwestern Canada. Contrib. Mineral. Petrol. 134, 3
–
16 (1999).
7. Wilde, S. A., Valley, J. W., Peck, W. H. & Graham, C. M. Evidence from detrital
zircons for the existence of continental crust and oceans on the Earth 4.4 Gyr ago.
Nature 409, 175
–
178 (2001).
8. Carlson, R. W. & Lugmair, G. The age of ferroan anorthosite 60025: Oldest crust
on a young Moon? Earth Planet. Sci. Lett. 90, 119
–
130 (1988).
9. Ash, R. D., Knott, S. F. & Turner, G. A 4-Gyr shock age for a martian meteorite and
implications for the cratering history of Mars. Nature 380, 57
–
59 (1996).
10. Mittlefehldt, D. W., McCoy, T. J., Goodrich, C. A. & Kracher, A. in Planetary
Materials (ed. Papike, J. J.) Ch. 4 (Mineralogical Society of America, 1998).
11. Shearer, C. K. et al. GRA 06129: A meteorite from a new asteroidal geochemical
reservoir or Venus? Lunar Planet. Sci. Conf. Abstr. XXXIX, 1825 (2008).
12. Mikouchi, T. & Miyamoto, M. Mineralogy and pyroxene cooling rate of unique
achondrite meteorite GRA 06129. Lunar Planet. Sci. Conf. Abstr. XXXIX, 2297
(2008).
13. Clayton, R. N. Oxygen isotopes in meteorites. Annu. Rev. Earth Planet. Sci. 21,
115
–
149 (1993).
14. Cherniak, D. J., Landord, W. A. & Ryerson, F. J. Lead diffusion in apatite and zircon
using ion implantation and Rutherford back-scattering techniques. Geochim.
Cosmochim. Acta 55, 1663
–
1673 (1991).
15. Amelin, Y., Krot, A. N., Hutcheon, I. D. & Ulyanov, A. A. Lead isotopic ages of
chondrules and calcium-aluminium-rich inclusions. Science 297, 1678
–
1683
(2002).
16. Strom, R. G., Schaber, G. G. & Dawson, D. D. The global resurfacing of Venus. J.
Geophys. Res. 99 (E5), 10899
–
10926 (1994).
17. Taylor, G. J. & Scott, E. R. D. in Meteorites, Comets, and Planets (ed. Davis, A. M.)
Vol. 1, Treatise on Geochemistry (eds Holland, H. D. & Turekian, K. K.) 477
–
486
(Elsevier-Pergamon, 2003).
18. Taylor, G. J. Core formation in asteroids. J. Geophys. Res. 97, 717
–
726 (1992).
19. Day, J. M. D., Pearson, D. G. & Taylor, L. A. Highly siderophile element constraints
on accretion and differentiation of the Earth-Moon system. Science 315, 217
–
219
(2007).
20. Birck, J.-L. & Alle
`gre, C. J. Contrasting Re/Os magmatic fractionation in planetary
basalts. Earth Planet. Sci. Lett. 124, 139
–
148 (1994).
21. Jurewicz, A. J. G., Mittlefehldt, D. W. & Jones, J. H. Experimental partial melting of
the Allende (CV) and Murchison (CM) chondrites and the origin of asteroidal
basalt. Geochim. Cosmochim. Acta 57, 2123
–
2139 (1995).
22. Morse, S. A. Basalts and Phase Diagrams (Springer, 1980).
23. Tuttle, O. F. & Bowen, N. L. Origin of granite in the light of experimental studies in
the system NaAlSi
3
O
8
-KAlSi
3
O
8
-SiO
2
-H
2
O. Geol. Soc. Am. Mem. 74, 153p (1958).
24. Kushiro, I. On the nature of silicate melt and its significance in magma genesis;
regularities in the shift of the liquidus boundaries involving olivine, pyroxene, and
silica minerals. Am. J. Sci. 275, 411
–
431 (1975).
25. Ziegler, R. A. et al. Petrology, geochemistry and likely provenance of unique
achondrite Graves Nunatak 06128. Lunar Planet. Sci. Conf. Abstr. XXXIX, 2456
(2008).
26. Mittlefehldt, D. W., Bogard, D. D., Berkley, J. L. & Garrison, D. H. Brachinites:
Igneous rocks from a differentiated asteroid. Meteorit. Planet. Sci. 38, 1601
–
1625
(2003).
27. Rumble, D., Irving, A. J., Bunch, T. E., Wittke, J. H. & Kuehner, S. M. Oxygen
isotopic and petrological diversity among Brachinites NWA 4872, NWA 4874,
NWA 4882 and NWA 4969: How many ancient parent bodies? Lunar Planet. Sci.
Conf. Abstr. XXXIX, 1974 (2008).
28. Bockrath, C., Ballhaus, C. & Holzheid, A. Fractionation of the platinum-group
elements during mantle melting. Science 305, 1951
–
1953 (2004).
29. Keil, K., Sto
¨ffler, D., Love, S. G. & Scott, E. R. D. Constraints on the role of impact
heating and melting in asteroids. Meteorit. Planet. Sci. 32, 349
–
363 (1997).
30. Wilson, L. & Keil, K. Consequences of explosive eruptions on small solar system
bodies: The case of the missing basalts on the aubrite parent body. Earth Planet.
Sci. Lett. 140, 191
–
200 (1991).
31. Clark, B.-E. et al. E-type asteroid spectroscopy and compositional modelling. J.
Geophys. Res. 109, doi:10.1029/2003JE002200 (2004).
32. Spicuzza, M. J., Day, J. M. D., Taylor, L. A. & Valley, J. W. Oxygen isotope
constraints on the origin and differentiation of the Moon. Earth Planet. Sci. Lett.
253, 254
–
265 (2007).
33. Horan, M. F., Walker, R. J., Morgan, J. W., Grossman, J. N. & Rubin, A. E. Highly
siderophile elements in chondrites. Chem. Geol. 196, 27
–
42 (2003).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank the ANSMET 2006/2007 field team, the Meteorite
Working Group and the Smithsonian Institution of Washington for collection and
provision of the GRA 06128/9 and Brachina meteorites. D. Mittlefehldt and
R. Greenwood provided reviews that improved the quality of this paper. A. Patchen
and P. Piccoli provided assistance with electron microprobe analysis. Portions of
this study were supported by the NASA Cosmochemistry Program:
NNX07AM29G (R.J.W.), NNX08AH76G (W.F.M.), NNG05GG03G (L.A.T.).
Author Contributions All authors participated in data collection and interpretation
and commented on the manuscript. J.M.D.D led the project and wrote the paper.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. Correspondence and requests for materials should be
addressed to J.M.D.D. (jamesday@umd.edu).
LETTERS NATURE
|
Vol 457
|
8 January 2009
182
Macmillan Publishers Limited. All rights reserved
©2009

METHODS
Mineralogical investigations of polished thick and thin sections of GRA 06128
(sub-sections 42 and 51) and GRA 06129 (22 and 25) were performed using a
Cameca SX50 electron microprobe analyser
34
(EMPA; University of Tennessee).
Concentrations of minor and trace elements were determined in minerals using a
New Wave Research UP213 (213 nm) laser-ablation system coupled to a
ThermoFinnigan Element 2 ICP-MS (University of Maryland). Minerals were
analysed using individual spots with a 15–80 mm diameter, a laser repetition rate
of 7 Hz and a photon fluence of 2–2.5 J cm
22
. Th/ThO production was ,0.07%
for all analytical sessions. Backgrounds on the ICP-MS sample gas were collected
for ,20 s followed by ,40 s of laser ablation of the sample. Washout time
between analyses was .2 min. Data were collected in time-resolved mode so
that effects of inclusions, mineral zoning and laser beam penetration could be
evaluated. The NIST 610 glass standard was used for calibration of relative
element sensitivities. Replicate LA-ICP-MS analyses of the BIR-1g glass standard
run at intervals during analytical sessions yielded an external precision of better
than 3% (1srelative standard deviation) for all measured element compositions
of silicates and phosphates. Replicate LA-ICP-MS analyses of the University of
Toronto JB Sulphide standard run at intervals during analysis of sulphides
yielded an external precision of better than 1% (1srelative standard deviation)
for highly siderophile element abundances.
207
Pb–
206
Pb ages were obtained using the same laser and mass spectrometer
settings as those for minerals and glasses analysed by LA-ICP-MS. Chlorapatite
and merrillite were measured because the high concentrations of U (0.1–
3 p.p.m.) make them suitable for LA-ICP-MS
207
Pb–
206
Pb dating. All data reduc-
tion was made offline using Microsoft Excel. Background Pb signals were taken
on mass, and subtracted from each isotopic measurement during ablation. Each
ratio was determined using the background corrected Pb isotopic measure-
ments. The average and 2s
mean
of the background corrected ratios, after ratios
outside 3swere discarded, were used to determine the age and error for each
phosphate. An exponential fractionation law was used to correct for mass frac-
tionation by means of bracketing the phosphate analyses with standard reference
materials (SRM: NIST 610, NIST 612, BCR-2g). Ratios of
207
Pb/
206
Pb for each
SRM were used to calculate the fractionation factor (a; ref. 35). Differences in a
between the three SRMS had a negligible effect on calculated ages. The
207
Pb–
206
Pb ages were calculated using Isoplot/Ex (ref. 36).
Fused-bead major element concentrations were analysed using the CAMECA
SX-50 EMPA and protocols for glass analyses outlined in ref. 34. Minor- and
trace-element concentrations were measured on the same beads using LA-ICP-
MS protocols outlined above, with 150-mm raster paths and obtaining 20 s of
background and ,60 s of analysis. Os isotopic and platinum-group elemental
analyses were performed at the University of Maryland using protocols outlined
in Supplementary Information. Isotopic compositions of Os were measured in
negative ion mode by thermal ionization mass spectrometry. Re, Pd, Pt, Ru and
Ir were measured using an Aridus desolvating nebuliser coupled to an Element 2
ICP-MS in low-resolution mode. External precision for
187
Os/
188
Os, determined
via measurement of standards bracketed with the meteorite samples, was 2.5%
(2s). External reproducibility on PGE analyses using the Element 2 was better
than 0.5% for 0.1 p.p.b. solutions and 0.3% for 1p.p.b. solutions. Total proced-
ural blanks run with the samples had an average
187
Os/
188
Os isotope composi-
tion of 0.1448 60.0024, with average concentrations of 1.5 pg (Re), 37 pg (Pd),
20 pg (Pt), 5 pg (Ru), 2 pg (Ir) and ,1 pg (Os); blank corrections were negligible.
Oxygen isotope analyses were performed at the Geophysical Laboratory,
Carnegie Institution for Science and are reported in d
18
O, d
17
O(d
X
O
n
is the
per mil (%) deviation of
X
O/
16
Oinnfrom the international standard (std)
V-SMOW given by the relationship: d
X
O
n
51,000 3((
X
O/
16
O
n
)/
(
X
O/
16
O
std
)21), where Xis either 17 or 18 and nrepresents the unknown)
and D
17
O notation, which represents deviations from the terrestrial fractiona-
tion line (l50.526 (ref. 37); D
17
O51,000ln((d
17
O/1,000) 11) –
0.526 31,000ln((d
18
O/1,000) 11)) (after ref. 38). The value of 0.526 was
obtained by linear regression of linearized values for d
17
O and d
18
O of terrestrial
silicate minerals
37,39
. Samples were loaded in a Sharp reaction chamber (ref. 40).
Successive, repeated blanks with BrF
5
and vacuum pumping were carried out for
12 h until there was less than 150 mm of non-condensable gas pressure remaining
after a blank run. Quantitative release of oxygen by fluorination reaction was
performed by heating samples individually with a CO
2
laser in the presence of
BrF
5
. Standardization of delta values was achieved by comparison with the Gore
Mountain garnet standard, USNM 107144, analysed during every analytical
session.
34. Day, J. M. D. et al. Comparative petrology, geochemistry, and petrogenesis of
evolved, low-Ti mare basalt meteorites from the LaPaz Icefield, Antarctica.
Geochim. Cosmochim. Acta 70, 1581
–
1600 (2006).
35. Baker, J., Peate, D., Waight, T. & Meyzen, C. Pb isotopic analysis of standards and
samples using a
207
Pb-
204
Pb double spike and thallium to correct for mass bias
with a double-focusing MC-ICP-MS. Chem. Geol. 211, 275
–
303 (2004).
36. Ludwig, K. R. Isoplot. Program and documentation, version 2.95. (Revised edition
of US Open-File report, 91-445, 2003).
37. Rumble, D., Miller, M. F., Franchi, I. A. & Greenwood, R. C. Oxygen three-isotope
fractionation lines in terrestrial silicate minerals: An inter-laboratory comparison
of hydrothermal quartz and eclogitic garnet. Geochim. Cosmochim. Acta 71,
3592
–
3600 (2007).
38. Clayton, R. N. & Mayeda, T. K. Oxygen isotope studies of achondrites. Geochim.
Cosmochim. Acta 60, 1999
–
2017 (1996).
39. Miller, M. F. Isotopic fractionation and the quantification of
17
O anomalies in the
oxygen three-isotope system: an appraisal and geochemical significance.
Geochim. Cosmochim. Acta 66, 1881
–
1889 (2002).
40. Sharp, Z. D. A laser-based microanalytical method for the in situ determinat ion of
oxygen isotope ratios of silicates and oxides. Geochim. Cosmochim. Acta 54,
1353
–
1357 (1990).
doi:10.1038/nature07651
Macmillan Publishers Limited. All rights reserved
©2009