Scott: Accretion and Collisional Evolution of Asteroids 697

Meteorite Evidence for the Accretion and

Collisional Evolution of Asteroids
Edward R. D. Scott
University of Hawai‘i

Meteorites contain a record of impacts during all stages of asteroid origin and evolution:
the formation and accretion of chondritic particles; the alteration, metamorphism and melting
of asteroids; and the erosion and disruption of asteroids by hypervelocity impacts. A review of
meteorite classification shows that numerous meteorites are not readily classified because they
do not fit simple models for asteroid formation and evolution that assume impacts were only
important during the final stage of asteroid evolution and because of inadequate understanding
of asteroidal impacts. Chronological, textural, and thermal constraints allow us to identify mete-
orite impact breccias that formed during accretion (e.g., Kaidun), when asteroids were partly
molten (e.g., mesosiderites), and during the subsequent disruption of asteroids (e.g., L chon-
drites). Studies of chondrites including Kaidun suggest that chondrules accreted with similar-
sized fragments of preexisting bodies that formed at greater heliocentric distances. In the inner
solar system, chondrules appear to have been crucial for initiating accretion. Without chon-
drules and rock fragments, dry dust failed to accrete to the nebular midplane because of nebular
turbulence and spiraled into the protosun. The tiny mass of asteroids may be partly due to
inefficient chondrule formation beyond 2 AU or less-efficient delivery of chondrules from near
the protosun.

1. INTRODUCTION the asteroids had cooled, they were eroded by impacts for
4.5 G.y. or disrupted. However, this simple model is inad-
Meteorites and returned asteroid samples are keys to equate as detailed petrologic studies and radiometric ages
understanding the chemical, mineralogical, and physical suggest that the four stages overlapped in time. Chondrules,
properties of near-Earth and main-belt asteroids. We know for example, appear to have accreted with fragments of pre-
a lot about the chemical, mineralogical, and physical prop- existing asteroids (e.g., Scott et al., 1996). Some chondrites
erties of meteorites but are still very ignorant about the and eucrites are breccias of materials excavated by impacts
corresponding properties of their parent asteroids and pos- from different depths that were metamorphosed after assem-
sible matches between meteorite classes and asteroid types bly (e.g., Yamaguchi et al., 1997). In addition, few chon-
(Burbine et al., 2002). To make progress in understanding drites and differentiated meteorites experienced simple cool-
the formation and evolution of asteroids, the origin of me- ing histories after peak temperatures were reached, probably
teorites, and the formation of chondritic components we ur- because of disruption by impacts (Scott and Keil,1999).
gently need to know more about the composition and struc- Here I review the meteorite evidence for impact process-
ture of asteroids. Our poor understanding of how impacts ing, both during and after accretion, in the light of studies of
modified asteroids and meteorites is a major barrier in terrestrial, lunar, and laboratory impacts and impact mod-
achieving these goals. Impacts controlled the formation of eling. I first review the variety of meteorites, emphasizing
asteroids and their destruction and were a major factor in those features and rocks that do not fit simple models for
their geological evolution (e.g., Scott et al., 1989). Under- the accretion and geological evolution of asteroids.
standing the role of impacts is critical if we wish to learn
how diverse types of chondritic materials accreted into as- 2. METEORITE TYPES AND
teroids, how and when 99.9% of the solid material in the PARENT BODIES
asteroid belt was lost, and how impacts changed the chemi-
cal, mineralogical, and physical properties of chondritic and Meteorites are divided into two classes: chondrites and
differentiated asteroids. differentiated or nonchondritic meteorites (Brearley and
A simple model for the geological evolution of chon- Jones, 1998; Mittlefehldt et al., 1998). Chondrites are aggre-
dritic and differentiated asteroids has four stages. Chon- gates of silicate and metallic particles and are remarkably
drules, metal grains, and other components were formed, similar in composition to the condensable material in the
probably in the solar nebula. These accreted together to form Sun. Except for H, He, C, N, O, and the inert gases, elemen-
parent asteroids. Some of these bodies were subsequently tal ratios are very largely within a factor of 2 of those in the
melted, forming metallic cores and olivine-rich mantles; solar photosphere (Wasson and Kallemeyn, 1988). Differ-
others were metamorphosed or aqueously altered. After entiated or nonchondritic meteorites are igneous rocks and

697
698 Asteroids III

differ considerably from the Sun in composition. Three types

are distinguished: irons, achondrites, and stony irons, most
of which are thought to be derived respectively from the
core, mantle and crust, and core-mantle boundary of melted
asteroids. Differentiated meteorites have elemental ratios
that may be up to 10–1000× different from solar values.
Largely as a result of meteorite discoveries in Antarctica
and various deserts around the world, the number of known
meteorites has increased in the past 20 years from 3000 to
around 20,000, allowing for pairing of fragments (Bischoff,
2001). Detailed studies of these meteorites suggest they
come from 100–150 different parent asteroids (Burbine et
al., 2002). However, a surprising number cannot be readily
classified as chondritic or differentiated. This suggests that
our ideas about the formation of meteorites and the evolu-
tion of their parent asteroids may require modification.

2.1. Chondrites

2.1.1. Components. Chondrites are mixtures of diverse Fig. 1. Polished slab of the Leoville CV3 chondrite showing
types of materials that formed in different parts of the solar numerous chondrules, one large and several small, white Ca-Al-
nebula under very different thermal conditions. Chondrules rich inclusions (CAIs), and three dark inclusions, which are
are rounded particles, typically millimeter-sized, that formed marked by arrows. Chondrules and CAIs formed in the solar
from molten silicate droplets (Rubin, 2000). Various kinds nebula but dark inclusions are probably fragments of preexisting
bodies that accreted with chondrules and CAIs. (Photo courtesy
of chondrules, associated metal grains, and the rarer Ca-Al-
of Alfred Kracher and Klaus Keil.)
rich inclusions (CAIs) formed at high temperatures (1200°–
2000°C) in the solar nebula (Ireland and Fegley, 2000).
Before these components accreted they were commonly
coated or mixed with volatile-rich, fine-grained matrix ma- few authors have argued that chondrules formed by impacts
terial, which contains micrometer- and submicrometer-sized of molten planetesimals (e.g., Sanders, 1996; Lugmair and
grains of circumstellar grains and clearly experienced less Shukolyukov, 2001) or impacts into asteroidal regoliths (e.g.,
thermal processing (Ott, 2001). Symes et al., 1998). However, formation of chondrules by
Chondrites also contain angular rock fragments. Many impacts between solid or molten asteroids appears incon-
are clearly pieces of chondrules and CAIs, some are for- sistent with many properties of chondrules and chondrites
eign fragments of known meteorite types (Lipschutz et al., (Taylor et al., 1983; Taylor and Scott, 2001).
1989), but most are matrix-rich chondritic fragments called It is commonly thought that CAIs and chondrules formed
dark inclusions (see Fig. 1) (Kracher et al., 1985; Johnson in the asteroid belt and were quickly accreted into asteroids
et al., 1990; Endress et al., 1994). Most dark inclusions are (e.g., Alexander et al., 2001). However, new models sug-
heavily altered, have sizes comparable to those of chon- gest that they may have formed close to the protosun and
drules and CAIs, may contain smaller chondrules than their were then distributed across the solar nebula by a bipolar
host chondrites, and some are rimmed by matrix material jet flow (Liffman and Brown, 1996; Shu et al., 2001; Ire-
(e.g., Fig. 1 of Scott et al., 1996). Their I-Xe ages suggest land and Fegley, 2000; Krot et al., 2001). If this model is
they were altered before their host chondrites (Hohenberg correct, the conclusion derived from inferred 26Al concen-
et al., 2001). Thus some dark inclusions appear to be frag- trations that CAIs predate chondrules by a million years or
ments of chondritic bodies that may have spiraled toward more (Wadhwa and Russell, 2000) may not be valid, as 26Al
the protosun and then accreted with chondrules to form new may have been produced by intense particle irradiation near
kinds of chondrites. the young Sun.
2.1.2. Formation of chondrules and CAIs: How and 2.1.3. Chondrite classes, groups, and origins. Three
where? Most authors infer that chondrules and CAIs main classes of chondrites, enstatite, ordinary, and carbon-
formed in the solar nebula by some kind of localized heat- aceous, are subdivided into 12 groups. One group, R chon-
ing mechanism such as nebular lightning, magnetic recon- drites, and several ungrouped chondrites do not fit into these
nection flares, or gas dynamic shock wave (Rubin, 2000; three classes (Fig. 2). Within each chondrite group, with the
Jones et al., 2000). Some authors invoke prior formation of possible exception of CR and CH chondrites, the propor-
chondrule-free planetesimals. Weidenschilling et al. (1998), tions of chondrules and other chondritic components and
for example, suggest that planetesimals in resonance with the bulk composition of each chondrite are surprisingly
Jupiter attained high eccentricities, generating bow shocks uniform. As for all meteorite groups, it is plausible that the
in the nebula gas that melted dust, forming chondrules. A members of each group come from a single body, but multi-
 Scott: Accretion and Collisional Evolution of Asteroids 699

lihood that our samples come from ~24 or more chondritic

Chondrites
asteroids, it seems unwise to conclude that the interiors of
Enstatite Ordinary Carbonaceous chondritic asteroids are devoid of chondrules. In the absence
EH EL H L LL R CH CO CR CV CK CM CI
of returned samples from asteroids, we will assume that we
Matrix have sampled the interiors of chondritic asteroids.
(vol%) <2–15 10 –15 35 5 30 40 40 75 70 99
Possible links between chondrite groups and asteroid
types are discussed by Burbine et al. (2002). Enstatite chon-
Fig. 2. Classification and matrix concentrations in 13 chondrite drites, which may be derived from two or more M asteroids,
groups, 12 of which belong to the enstatite, ordinary, and carbon- probably formed closer to the Sun than other chondrites
aceous classes. R chondrites and ungrouped chondrites, which are (Wasson, 1988; Rubin and Wasson, 1995; Lugmair and Shu-
not shown, do not fit into these three classes. Chondrite groups kolyukov, 1998, 1999). The E-type asteroids, which appear to
are arranged in approximate order of increasing matrix concen- be the source of the closely related aubrites (enstatite achon-
tration (data from Scott et al., 1996). Properties of meteorites and
drites), are also found at the inner edge of the belt (Bell et
asteroids suggest that this sequence corresponds very approxi-
mately to increasing formation distance from the Sun (Rubin and
al., 1989). Ordinary chondrites are probably derived from
Wasson, 1995). Almost all the other mineralogical, chemical, and three or more S(IV)-type asteroids, fragments of which may
isotopic properties of these groups do not correlate with the ma- be classed as Q-type asteroids among near-Earth asteroids.
trix concentration, suggesting that chondritic asteroids did not CI, CM, CR, and CK chondrites are probably derived from
form simply from a disk of solid materials with compositions that C-type asteroids, which are concentrated further out in the
varied smoothly with heliocentric distance. asteroid belt (2.5–3.5 AU). CV and CO chondrites may
come from K-type asteroids, which are largely from the Eos
family at 3.0 AU (Bell et al., 1989). The most porous and
possibly very primitive chondrite, Tagish Lake, is a CM-
ple bodies cannot be excluded. The ungrouped chondrites like chondrite that is very rich in C (~5 wt%), H2O, and
probably come from another 13 or more bodies (Meibom presolar grains (Brown et al., 2000). It contains fewer chon-
and Clark, 1999). The rate at which new types of carbon- drules and CAIs than CM chondrites, and is thought to be
aceous chondrites are being discovered and the strange derived from one of the D-type asteroids (Hiroi et al.,
properties of recently discovered groups suggest that the 2001), which lie beyond 2.9 AU and probably dominate the
known chondrites may represent only a minor fraction of asteroid population beyond 4 AU (Jones et al., 1990).
the chondritic asteroids and the precursor material of dif- Properties of chondrites and asteroids suggest there were
ferentiated asteroids. some systematic variations in the properties of accreting
Oxygen-isotopic analyses show that chondrules in each planetesimals that survive today in the asteroid belt despite
class are quite distinct and must have been derived from at some mixing due to subsequent orbital evolution of the
least three separate reservoirs (Clayton, 1993). Chemical asteroids and their impact debris. At the inner edge of the
and mineralogical studies confirm that chondrules in H, L, asteroid belt, chondritic asteroids tend to be chondrule-rich
and LL chondrites are not fundamentally different and might and matrix-poor, like E and O chondrites (Fig. 2), and vola-
be derived from a single reservoir by size sorting. A similar tile concentrations in the matrix material are relatively low.
relationship holds for EH and EL chondrites, but the prop- In the outer parts of the belt, chondritic asteroids tend to
erties of carbonaceous chondrites are much more diverse have much higher proportions of matrix to chondrules and
and appear to require at least four separate sources. (Carbo- the matrix material is richer in water and other volatiles.
naceous chondrites have little in common except high abun- The degree of alteration and metamorphism experienced
dances of refractory elements.) Thus the parent asteroids by the parent bodies of chondrites also varied across the
of the chondrites probably formed from at least six sepa- belt (Keil, 2000). At the inner edge, peak temperatures were
rate reservoirs of chondrules with remarkably little mixing generally higher. Chondrites in the EH, EL, H, L, and LL
between the reservoirs. groups were all heated in the temperature range ~400°–
2.1.4. Chondritic asteroids. Sears (1998) and Akridge 950°C. The most deeply buried probably experienced the
et al. (1998) argued that chondrites are biased samples highest temperatures (type 6), whereas the least-metamor-
derived from the outer few kilometers of asteroids where phosed (type 3) were closer to the surface. Enstatite chon-
chondrules formed by impact and that we have not sampled drites experienced little if any aqueous alteration prior to
the asteroids’ chondrule-free dusty interiors. However, con- metamorphism, and ordinary chondrites experienced only
centrations of chondrules and matrix material are rather mild alteration in the most-fine-grained portions. The ex-
uniform in each group (Fig. 2) and we certainly have mete- ceptional CH chondrites resemble ordinary chondrites in
orite samples from the mantles and cores of differentiated that they contain very little water and matrix material but
asteroids (see below). Since ordinary chondrite parent bod- experienced lower peak temperatures (<300°C). They may
ies were baked at high temperatures, any loosely consoli- have formed closer to the Sun than other chondrites.
dated dirt should have been converted to rock capable of Further out in the asteroid belt, the carbonaceous parent
surviving the journey to Earth. In view of other difficulties bodies experienced aqueous alteration at temperatures of
with impact origins for chondrules (see above) and the like- ~0°–300°C (CM, CR, and CI bodies). In bodies that were
700 Asteroids III

heated to 300°–500°C, volatiles were lost and aqueous alter- they could be assigned to one of these two classes, and
ation merged into fluid-assisted metamorphism (CO and others that still have controversial classifications. Most of
CV). The most metamorphosed carbonaceous chondrites are the difficulties have arisen because of disagreements about
found in CK chondrites (types 4–6), which are almost com- the role of impacts in the formation of these meteorites.
pletely dry. 2.3.1. Impact melts. Meteorites identified as chondritic
impact melts are mixtures of unmelted chondritic material
2.2. Differentiated Meteorites and melt veins with chondritic composition and widths of
less than a few centimeters; e.g., Rose City and Chico
Differentiated meteorites formed in asteroids that melted (Bogard et al., 1995). However, two rocks that appear to be
shortly after they accreted (Wadhwa and Russell, 2000; melted chondrites are quite different. Patuxent Range 91501
Taylor et al., 1993), and are commonly divided into achon- is an 8.5-kg igneous-textured rock containing a few volume
drites, irons, and two kinds of stony irons, pallasites and meso- percent of metal-troilite nodules (Fig. 3). Without detailed
siderites (Shearer et al., 1998). Conventional views suggest bulk chemical and O-isotopic analyses and mineral com-
that differentiated meteorites should consist of easily char- positions, we would not know that it is probably an impact
acterized rocks from the core, mantle, or crust of asteroids. melt from the L-chondrite body (Mittlefehldt and Lindstrom,
The largest class of achondrites are the HED meteorites 2001). The second, Abee, was for many years the best-stud-
(~80% of achondrite falls): the howardites, eucrites, and ied EH4 chondrite, but subsequent studies showed that it
diogenites, which are very probably derived from Vesta is actually an impact melt and that its chondrule-shaped
(Keil, 2002). Another Vesta-like body is required for the ba- features are merely ghosts from the original structure (Rubin
saltic meteorite, Northwest Africa 011 (Yamaguchi et al., and Scott, 1997). Its origin is still poorly understood.
2001), conceivably the parent asteroid of 1459 Magnya (see 2.3.2. Metal-chondrite breccias. The best-studied me-
Burbine et al., 2002). Studies of the HED meteorites and teorite breccias that consist of metal veins or regions and
remote sensing indicate that Vesta differentiated to form a chondritic fragments are Netschaevo (classed as a IIE iron)
core, mantle, and crust. It experienced a high degree of and Portales Valley [classified as an H6 chondrite (Grady,
melting, possibly complete melting, probably due to 26Al 2000)]. Both must have been deeply buried when they
decay. Thus Vesta can be considered as the smallest terres- cooled as their metal portions developed Widmanstätten
trial planet (Keil, 2002). However, it may be incorrect to patterns (Fig. 4). Impacts created the chondritic fragments
consider Vesta as the archetypal differentiated asteroid as but it is not known how molten metal was formed and how
most may have lost their basaltic magma because of explo-
sive volcanism (Keil, 2000).
Excluding meteorites from Mars and the Moon, there are
four major groups of achondrites: ureilites, aubrites, brach-
inites, and angrites (Mittlefehldt et al., 1998). These me-
teorites appear to be derived from four parent asteroids
with very different igneous and impact histories from Vesta.
Ureilites are largely unbrecciated, coarse-grained, igneous
rocks made of olivine and pyroxene. Aubrites are brecci-
ated rocks largely composed of coarse-grained enstatite that
formed from a body distinct from the parent bodies of the
EH and EL chondrites (Keil, 1989). Brachinites are unbrec-
ciated, igneous rocks containing 80–95 vol% olivine and 0–
10 vol% plagioclase, and angrites are unbrecciated, plagio-
clase-poor, basaltlike igneous rocks. None of these groups
of achondrites is a close match for the products predicted
by simple models of asteroid differentiation, suggesting that
igneous processes in asteroids were much more complex
than commonly appreciated (Mittlefehldt et al., 1998).
Brachinites may be derived from A-type asteroids, au- Fig. 3. Polished face of the PAT 91501 meteorite showing a uni-
brites from E-type asteroids, and ureilites and angrites from form fine-grained rock with black vesicles under 3 mm in size and
S asteroids (Burbine et al., 2002). rounded or elongated, gray nodules of metal and troilite, which
are up to a centimeter in size. Chemical and O-isotopic analysis
of the silicates suggest that it is an impact melt from the L-chon-
2.3. Other Types of Meteorites
drite parent body (Mittlefehldt and Lindstrom, 2001). However,
unlike nearly all other impact-melted meteorites, it contains very
Textbooks imply that chondrites and differentiated me- few unmelted, chondritic grains, and none of these grains are
teorites are so different that even a simple-minded robot shocked. A cometary projectile may have been responsible for the
should be able to distinguish these two basic types of aster- high degree of melting and vaporization. (Vertical lines are saw
oidal material! However, our museums contain many strange marks. Photo courtesy of D. Mittlefehldt and NASA Johnson
meteorites that required detailed laboratory studies before Space Center.)
 Scott: Accretion and Collisional Evolution of Asteroids 701

tures and are depleted in plagioclase and troilite, which are

both concentrated in early silicate and metallic melts. Most
authors envisage that internal heating was responsible, but
impact heating has also been invoked (Rubin et al., 2001c).
McCoy et al. (2000) infer that the parent asteroid of these
rocks would have internal chemical and mineralogical het-
erogeneities on scales of >200 m. The asteroid Eros lacks
heterogeneities on this scale but minor igneous differentia-
tion cannot be excluded as impact mixing may have homo-
genized its surface (McCoy et al., 2001).
2.3.5. Metal-rich achondrites. Two meteorites, Mt.
Egerton and Itqiy, appear to be igneous rocks made of
~20 vol% of metallic Fe,Ni intergrown with coarse-grained
enstatite crystals (Mittlefehldt et al., 1998; Patzer et al.,
Fig. 4. Slice of the Portales Valley meteorite showing dark angu- 2001). Since molten silicate and metallic Fe,Ni should
lar clasts of what appears to be an H6 chondrite separated by light- readily separate in internally heated asteroids, they may
colored veins of metallic Fe,Ni, which have been etched to reveal have formed by impact. However, neither contains relict
a Widmanstätten pattern of oriented kamacite plates. An impact chondrules or clasts and their bulk composition is far from
created the angular clasts but it is not known how the molten metal chondritic. Five other metal-bearing enstatite meteorites
was formed, mixed with chondritic clasts, and then cooled slowly
have been identified as impact melts of enstatite chondrite
to form the Widmanstätten pattern. (Photo courtesy of Marvin
parentage by McCoy et al. (1995) and Burbine et al. (2000).
Killgore.)
Other workers favor an origin from indigenous melts from
the aubrite parent body.
2.3.6. Achondritic breccias containing chondritic mate-
it was mixed with chondritic material. Both impact heat- rial. Most meteorite breccias contain only a tiny fraction
ing and internal heating have been proposed (see Rubin et of material that could have come from a separate parent
al., 2001a). body (e.g., Lipschutz et al., 1989). However, the aubrite,
2.3.3. Metal-rich chondrites. Most chondrites contain Cumberland Falls, contains abundant chondritic clasts that
<15 vol% metallic Fe,Ni, but two newly discovered types were probably derived from a single projectile.
contain much more metal. CH chondrites contain 20 vol% 2.3.7. Chondritic breccias containing achondritic clasts.
metallic Fe,Ni, while two CB chondrites, QUE 94411 and Some chondrites contain achondritic clasts that are obvi-
HaH 237, have ~70 vol% metallic Fe,Ni; one of these was ously impact melted material. But a few clasts in type 3 ordi-
initially classified as an iron meteorite (Weisberg et al., nary chondrites appear to be fragments of differentiated as-
2001). Metal particles in these two chondrites are chon- teroids and may have been acquired during accretion (see
drule-sized and appear to have similar condensation histo- Mittlefehldt et al., 1998, pp. 168–171).
ries to the chondrules (Krot et al., 2001). The chondrites 2.3.8. Other oddball meteorites. A few meteorites,
also contain small amounts of hydrated matrix (Greshake e.g., Enon, Bocaiuva, and Northwest Africa 176, are com-
et al., 2002) and might be derived from hydrated M-class posed of ~50 vol% chondritic silicate inclusions in metal-
asteroids (Rivken et al., 1995). Because they are so differ- lic Fe,Ni (Liu et al., 2001). In these meteorites, the silicate
ent from other chondrites in their composition and texture, inclusions do not resemble rock fragments and the meteor-
some authors have suggested that metal-rich chondrites ites might be strongly metamorphosed metal-rich chondrites
might be impact products, not true chondrites (Wasson and or some kind of impact melt product.
Kallemeyn, 1990). Their bulk Fe/Mg ratios are 1.7–9× The abundance of meteorites that lie outside the enve-
above the solar ratio and their volatile elements are depleted lope of standard chondrites and differentiated meteorites
by factors of up to 30×, placing them well outside the en- suggests that our schemes for classifying meteorites are too
velope of chemical compositions of other chondrites. simple-minded and that geological processes on asteroids,
Three related meteorites, Bencubbin, Weatherford, and especially impacts, were more complex than we commonly
Gujba, all contain ~60 vol% metallic Fe,Ni and may also appreciate. In addition, these meteorites emphasize the limi-
be metal-rich chondrites. However, they lack CAIs and their tations of remote analyses of asteroids and the need for care-
classification is less certain (Rubin et al., 2001b). (Bencubbin ful characterization of returned samples.
was once classified as a mesosiderite.)
2.3.4. Partly differentiated meteorites. Acapulcoites 3. METEORITE EVIDENCE FOR
and lodranites appear to represent the transition between ASTEROIDAL IMPACTS
chondrites and achondrites. Acapulcoites contain relict chon-
drules and have chondritic compositions but were heated Information about asteroidal impacts has been culled
up to 1000°C so that millimeter- to centimeter-sized metal- from three types of meteoritic evidence: mineral and rock
lic veins formed. They were not hot enough to lose much textures, radiometric ages, and other constraints on the ther-
metal or basalt. Lodranites were heated to higher tempera- mal histories of minerals. The degree of shock in minerals
702 Asteroids III

and clasts and the extent and nature of brecciation in mete- Models for accretion in the asteroid belt generally en-
orites provide direct clues to the impact history of meteor- visage micrometer- or submicrometer-sized dust grains
ites and their constituents (e.g., Scott et al., 1989; Stöffler sticking together and settling to the midplane of the nebula
et al., 1991; Bischoff and Stöffler, 1992). However, shock to form a dense layer of dust, which because of gravita-
and impact deformation effects are commonly heterogene- tional instabilities or some other mechanism was converted
ous and notoriously difficult to interpret. For example, sev- into a disk of kilometer-sized planetesimals. Several argu-
eral studies of the martian meteorite ALH 84001 have failed ments suggest that asteroids may not have accreted this way.
to agree on whether it suffered one major impact that formed First, as discussed earlier, many chondritic asteroids ac-
a large complex crater (Scott et al., 1998) or numerous im- creted largely from chondrules, not dust. In many cases, the
pacts and one thermal metamorphic event (Treiman, 1998). fine-grained dust accreted as rims on chondrules or other
Radiometric ages of meteorites combined with other large particles. In addition, chondrules appear to have ac-
kinds of chronological constraints such as cosmic-ray-expo- creted with fragments of preexisting bodies that were large
sure ages can provide invaluable constraints on the nature of enough to undergo alteration and possibly also melting.
impact events (Bogard, 1995). However, radiometric chro- Second, dry, solid silicate grains would probably not ad-
nometers are reset in impacts by postshock heating, which here to each other (Wood, 1996), and nebular turbulence
depends on shock pressure and the crater setting. Thus any may have prevented loose grain aggregates from settling to
proposed relationship between a radiometric age and a spe- the midplane to form a gravitationally unstable layer (Cuzzi
cific impact may be controversial. For example, Turner et et al., 1996, 2001). Third, as discussed below, the extraor-
al. (1997) argued that their 3.9-G.y. 40Ar-39Ar age for ALH dinarily diverse menagerie of chondrite groups appear to
84001 was a metamorphic age and that the plagioclase glass require more complex accretion models.
and other shock metamorphic features formed during an 3.1.1. Chondrite compositions. Although the matrix
impact 13 m.y. ago that removed the rock from Mars and concentration in chondrite groups appears to be correlated
exposed it to cosmic rays without significant reheating. with formation location (Fig. 2), it is not correlated with the
However, constraints from the magnetization record sug- concentration of refractory or moderately volatile elements
gest that the rock was not shocked or reheated significantly or the bulk O-isotopic composition. Thus chondritic plane-
when it left Mars, unlike other martian meteorites, and that tesimals did not simply form from a disk of solids with com-
the 3.9-Ga event was the major impact that damaged the positions that varied smoothly with radial distance. The bulk
rock (Weiss et al., 2000). composition of a chondrite group or chondritic asteroid
The ~3.9-G.y. 40Ar-39Ar ages and thermal histories of must have depended on two or more nebular parameters,
mesosiderites, which are essentially unshocked metal-sili- e.g., the time and place of accretion (e.g., Clayton, 1993).
cate breccias, were initially interpreted to result from a near- Since chondrules accreted from at least six separate
catastrophic impact that reheated the breccias at the time reservoirs, the accretion process must have ensured that
of heavy lunar bombardment (Bogard et al., 1990). How- each asteroid acquired chondrules from a single reservoir.
ever, additional studies of the thermal histories of metallic Whether chondrules formed near the protosun or in the as-
and silicate minerals suggested that mesosiderites were teroid belt, the time for an asteroid to accrete from a single
probably cooling slowly deep inside a large parent body at reservoir of chondrules must have been much less than the
that time. The major impact that created the metal-silicate time required to generate a new reservoir or chondrules at
breccias occurred much earlier, probably at ~4.45 Ga (see that location. In addition, early-formed asteroids had to be
Scott et al., 2001). isolated from late-formed reservoirs of accreting chon-
These studies show that understanding the impact his- drules. Thus chondrule formation and accretion may have
tories of meteorites may require several independent ther- been sporadic processes, ensuring that the composition of
mal constraints in addition to radiometric ages. Below I accreting material did not vary smoothly with time or ra-
summarize how meteorite evidence can illuminate our un- dial distance.
derstanding of various impact phenomena during the accre- 3.1.2. Role of chondrules. Chondrites with the lowest
tion and evolution of asteroids. The final section describes concentrations of chondrules, CI, CM, and Tagish Lake, all
in more detail how impact studies have been used to infer have heavily hydrated matrix material. Not one of the un-
the impact history of specific meteorite parent bodies. differentiated asteroids that we appear to have sampled
delivers chondrites that are composed largely of dry ma-
3.1. Accretion of Asteroids trix material with few chondrules. Even without chondrules,
such materials should be recognizable as chondritic from
For accretion to occur, impact velocities must be less their near-solar bulk compositions. It is therefore probable
than about twice the escape velocity, which for a 100-km- that planetesimals in the inner part of the asteroid belt and
radius asteroid is ~140 m/s. At these velocities, silicates are in the terrestrial planet region did not form without chon-
broken but not shocked or heated sufficiently to reset radio- drules and similar-sized particles. Either chondrule forma-
metric clocks. Thus accretional impacts can only be dated tion was remarkably efficient at these locations or fine dust
indirectly. failed to accrete without coating chondrules.
 Scott: Accretion and Collisional Evolution of Asteroids 703

Cuzzi et al. (1996, 2001) infer that chondrules played a tified material (Zolensky et al., 1996; Zolensky and Ivanov,
key role in triggering the accretion because chondrules and 2001). Some Kaidun clasts resemble the dark clasts in CR
particles with similar aerodynamic stopping distances were chondrites (Zolensky et al., 1996), which accreted with CR
concentrated at transient locations between turbulent eddies. chondrules. In addition, many clasts are unshocked and
Some of these accumulations settled to the midplane, drifted probably formed very early as the carbonates in all litholo-
inward, and accreted into planetesimals. Dust and fluffy grain gies appear to have formed <106 yr after CAIs (Hutcheon
aggregates were carried along by the gas into the protosun. et al., 1999). It seems probable that the clasts impacted at
Episodic addition of chondrules to the nebula and planetesi- relatively low velocities and that the parent body of Kaidun
mal accretion may have allowed chondrules from different res- accreted from fragmental debris from preexisting asteroids.
ervoirs to accrete without significant cross contamination. The Kaidun breccia might usefully be considered as a
The apparent tendency for the matrix concentration to chondrule-poor chondrite that is exceptionally rich in chon-
increase with increasing inferred formation distance sug- dritic clasts. If there was a pause in chondrule formation or
gests that in the cooler parts of the solar nebula, fine-grained deposition at some nebular location, the abundance of tiny
solids accreted without the assistance of chondrules. Ice asteroidal fragments suspended in turbulent gas may have
crystals and sticky organics may have allowed grain aggre- built up until rock chips, rather than chondrules, triggered
gates to grow large enough and dense enough to decouple planetesimal accretion.
from the turbulent nebular gas and descend to the midplane Asteroidal fragments may also have played a role in
to accrete. In the inner solar system, temperatures were transporting water across the asteroid belt during accretion.
probably too high for ice and organics to coat grains so that Experiments and theory suggest that slow kinetics prevented
only chondrules or particles with comparable aerodynamic the formation of hydrous silicates in the nebula and that
stopping distances acccreted. nebular temperatures in the asteroid belt probably exceeded
3.1.3. Kaidun chondritic breccia. The Kaidun chon- the condensation temperature of water ice [~180 K (Fegley,
drite (Fig. 5) is unique as it consists almost entirely of mil- 2000)]. Water may have accreted into asteroids in the form
limeter- and submillimeter-sized fragments of various types of particles of ice that drifted inward from beyond the nebu-
of chondritic material: CI, CM, CR, EH, and EL chondrites, lar snowline (Cyr et al., 1998), or fragments of outer plane-
material similar to the Tagish Lake chondrite, and uniden- tesimals that contained hydrous silicates.
Many details of asteroid accretion remain obscure but
the turbulent accretion model of Cuzzi et al. (2001) seems
to offer a good framework for understanding an early stage
in the accretion of components in chondrites. Further stud-
ies of chondrule and fragment sizes in Kaidun and other
chondrites are needed to test the ideas discussed above.

3.2. Mass Loss from the Asteroid Belt

Models for the solar nebula based on the existing bod-

ies suggest that ~99.9% of the solid material in the aster-
oid belt, 1 M or more, failed to accrete into asteroids or
was removed soon after. At Mars’ location, ~95% of the
solids may have been lost. If Jupiter formed rapidly in
~100 yr from gravitational instabilities, it is possible that
Jupiter prevented planetary embryos from accreting in the
belt while allowing the terrestrial planetary embryos to de-
velop (Kortenkamp and Wetherill, 2000). But if Jupiter
formed in 107 yr from a solid core that subsequently ac-
creted nebular gas, planetary embryos may have formed in
the asteroid belt. In this case, the mass of the asteroid belt
Fig. 5. Backscattered electron image of a thin section of the may have been removed largely by gravitational perturba-
unique Kaidun chondrite, which is mostly composed of millimeter- tions or collisional fragmentation followed by nongravita-
and submillimeter-sized fragments from many other chondritic bod-
tional removal of asteroid chips by, for example, gas drag,
ies. Some identified fragments include impact melt (a), enstatite
chondrite (j), CI chondrite (c), CM chondrite (d), and Tagish Lake-
or by some combination of all three processes (Wetherill,
type material (e). Kaidun, which has the highest proportion of rock 1989). The survival of Vesta’s crust argues against colli-
fragments to chondrules of any chondrite, may be an accretion- sional fragmentation as the major process. Most authors
ary breccia that formed by turbulent accretion of fragments in the have therefore investigated gravitational perturbations by
nebula when few chondrules were available. (Image supplied by the embryos themselves and the giant planets to clear the
Michael Zolensky.) asteroid belt (Chambers and Wetherill, 2001; Petit et al.,
704 Asteroids III

2002). However, it is not clear whether lunar-sized bodies unbrecciated samples. Regolith breccias contain only a per-
could have been removed from the asteroid belt without cent or so of foreign clasts, mostly carbonaceous chondrite
leaving a more traumatic collisional record in the meteor- material as most projectile material is lost during impacts
ites and surviving asteroids. on asteroids.
If chondrules were as critical for asteroid formation as Achondritic and many chondritic regolith breccias ap-
we infer, the efficiency of chondrule formation and accre- pear to have been lithified by localized shock melting. We
tion may have been important factors affecting the mass of lack iron and stony-iron meteorite breccias with solar-wind
the asteroids. Chondrule production, or chondrule deposi- gases, possibly because regolith on their parent bodies could
tion from bipolar flows, may have decreased with increas- not be converted to coherent rocks by localized shock melt-
ing distance from the protosun. If a large mass deficiency ing of minerals. Carbonaceous chondrite regolith breccias
existed in the asteroid belt prior to the accretion of plan- tend to have the lowest concentrations of track-rich grains
etesimals, few asteroids would have been able to accrete, and solar-wind gases and may have been lithified by alter-
according to Wetherill (1989). However, accreting solids ation. Shocked carbonaceous chondrites containing hydrous
may have been heterogeneously distributed in the belt. For minerals are not known, perhaps because of devolatilization
example, turbulent accretion of chondrules and particles on shock release.
with similar aerodynamic stopping distances may have
focused solids to specific locations where planetesimals 3.5. Fragmentation and Reaccretion
formed.
For seven groups of meteorites, there is textural and cool-
3.3. Impact Heating ing rate evidence suggesting that their parent bodies were
fragmented and then largely reaccreted so that material from
The effectiveness of impact heating of asteroids has been diverse depths of the target asteroid were mixed together.
a controversial issue in meteorite studies. Various authors These seven bodies are the parent bodies of the Shallowater
have argued that impacts caused extensive metamorphism enstatite achondrite, mesosiderites, H chondrites, L chon-
or melting on the parent bodies of almost every kind of drites, ureilites, IVA irons, and the parent body of IAB irons
meteorite (e.g., Takeda, 1993; Rubin, 1995). However, stud- and winonaites (Keil et al., 1994; Benedix et al., 2000; Scott
ies of terrestrial craters, shock experiments, and theoreti- et al., 2001). For all bodies except the H- and L-chondrite
cal considerations suggest that impacts were probably not bodies, the targets appear to have been hot enough prior to
a significant source of heat for metamorphism and melting impact to contain at least minor amounts of melt. This does
for asteroids under a few hundred kilometers in diameter not necessarily imply that few cold meteorite parent bodies
(Keil et al., 1997). The current mean impact velocity of as- were fragmented and reaccreted as it is easier to recognize
teroids, ~5 km/s (Bottke et al., 1994), is scarcely above the mixtures of materials from different depths when the targets
threshold impact velocity required to generate shock pres- contain some melt.
sures high enough to completely melt materials. Much of Low-resolution, smoothed-particle hydrodynamics com-
the impact melt that is generated on asteroids is ejected with puter simulations for 50–300-km-diameter targets suggest
velocities that exceed the asteroid’s escape velocity. The that for undifferentiated and fully differentiated target aster-
volume of melt may increase for very porous asteroids, but oids, near-catastrophic impacts can mix significant propor-
this does not alter the basic conclusion that mean tempera- tions of materials from all depths (Love and Ahrens, 1996;
tures of asteroids increase by <100°C even in the largest im- Scott et al., 2001). However, these studies have done little
pacts (Keil et al., 1997). more than prove the concept and more detailed modeling of
near-catastrophic impacts is needed.
3.4. Regolith Formation Existing studies suggest that impacts do not gradually
strip mantles from the cores of differentiated asteroids. Cores
For nearly all groups of chondrites and achondrites, we and mantles of differentiated asteroids were probably mixed
have samples that contain solar-wind gases and solar-flare together before disruption by catastrophic impacts. Thus the
tracks due to exposure in the top millimeter of their parent- rarity of olivine-rich, metal-free asteroids may reflect im-
body regoliths (Bunch and Rajan, 1988; McKay et al., pact scrambling rather than near-total fragmentation of dif-
1989). These regolith breccias provide our only tangible ferentiated asteroids smaller than Vesta (Burbine et al.,
samples of grains that have been on the top surfaces of as- 1996). Some S asteroids may be differentiated asteroids that
teroids. However, the proportion of grains exposed to so- were disguised by impact scrambling (Scott et al., 2001).
lar flares, which is lower than in typical lunar regolith brec-
cias, is too low to allow definitive inferences to be drawn 3.6. Case Histories
about the optical properties of asteroidal surfaces. Possibly
because of efficient mixing of irradiated and unirradiated 3.6.1. Ordinary chondrites. Despite many constraints
grains, there are only very minor chemical and mineralogi- from mineral cooling rates, 40Ar-39Ar ages, and studies of
cal differences between regolith breccias and unirradiated regolith and fragmental breccias, there are disagreements
fragmental breccias or, in the case of chondritic bodies, about the nature of the parent bodies of ordinary chondrites.
 Scott: Accretion and Collisional Evolution of Asteroids 705

The parent bodies of the H and L chondrites probably suf-

fered one or more major impacts that may have mixed
materials from all depths (Keil et al., 1994). Some ordinary
chondrites are breccias consisting of the most metamor-
phosed and presumably most deeply buried type 6 mate-
rial mixed with the least-metamorphosed type 3 material.
A few such breccias experienced temperatures above 500°C
after assembly suggesting that at least one major impact
occurred >4.4 G.y. ago when the parent asteroid was cool-
ing but still hot. The wide range of cooling rates in metal
grains in some regolith breccias combined with impact and
thermal modeling suggest that the H and L parent bodies
were broken up and reassembled between 4.4 G.y. and
exposure to cosmic rays <100 m.y. ago (Taylor et al., 1987).
However, the thermal modeling depends critically on as-
sumptions about the porosity and thermal conductivity of
the ordinary chondrite bodies (see McSween et al., 2002).
Layers of regolith <1 km thick can drastically reduce the Fig. 6. Polished slice of the Barea mesosiderite showing a mixture
rate of cooling of 100-km-radius asteroids and enhance the of centimeter-sized spherules of metallic Fe,Ni (light) and angu-
thermal gradients near the surface of the asteroid (Haack lar fragments of basaltic rock (dark) embedded in an intimate
et al., 1990). Thus Akridge et al. (1998) argue that H chon- metal-silicate matrix. This breccia formed when molten metal from
drites are derived from depths of <10 km from a 100-km the core of an asteroid was mixed with mantle and crustal rocks.
The impact was probably a near-catastrophic event that scrambled
radius asteroid and that most of the interior of the parent body
metal and silicate from a single asteroid that broke up and re-
has not been sampled, contrary to Bennett and McSween accreted (Scott et al., 2001). The alternative explanation, that the
(1996). Further thermal modeling to fit the constraints im- molten metal and rock were derived from separate asteroids that
posed by 40Ar-39Ar ages would be useful. collided at low velocity during accretion (Rubin and Mittlefehldt,
One candidate parent body of H chondrites, 6 Hebe, is 1993), appears less plausible as radiometric ages suggest that
an atypical S(IV) type as it shows rotational spectral varia- mixing occurred 100–150 m.y. after the solar system formed, long
tions. Gaffey and Gilbert (1998) suggest that these are due after asteroids accreted.
to impact-formed melt sheets on Hebe or residues of metal-
rich projectiles, and that these supply meteorites such as
Portales Valley (Fig. 4) and the IIE iron, Netschaevo. How- fragments are pieces of basalts, gabbros, and pyroxenites
ever, the link between IIE and H chondrites is not well es- with minor amounts of dunite. Thus their constituents could
tablished (Bogard et al., 2000), Portales Valley probably have been derived from all levels of an igneously stratified
formed at depth as it cooled slowly, and known asteroids asteroid that was similar but not identical to the parent body
are too small to generate impact-formed melt sheets (Keil of the howardites, eucrites, and diogenites, presumably
et al., 1997). Studies of complementary metal-poor samples Vesta (Keil, 2002). Geochemical evidence suggests that the
would help us to understand the formation of metal-chon- metal was molten prior to the impact that mixed metal and
drite breccias and possible links with H chondrites but such silicate. The major puzzle for understanding mesosiderites
samples have not yet been identified. has been the small amount of olivine, which should have
The high proportion of heavily shocked L chondrites been abundant in differentiated asteroids. One model sug-
with 40Ar-39Ar ages around 500 m.y. shows that the L-chon- gests that the olivine is confined to metal-poor regions that
drite parent body suffered a major impact at this time. Haack are poorly sampled and that the ingredients in mesosiderites
et al. (1996) argue that the L chondrites are derived from a are derived from a single body that was scrambled by a
kilometer-sized rubble pile or fragment ejected from near large impact (Scott et al., 2001). An alternative idea is that
the impact point of a catastrophic dispersion event. This the mesosiderite body formed when molten metal from the
event may have been responsible for the deposition of nu- core of a projectile was mixed with crust from the meso-
merous ordinary chondrites in 450-m.y.-old limestone in siderite parent body in a 1-km/s impact (see Rubin and
Sweden when the meteorite flux on Earth may have been 10– Mittlefehldt, 1993). However, impacts at this velocity would
100× higher for ~30 m.y. (Schmitz et al., 1997). Nesvorný probably have been very rare at the time of metal-silicate
et al. (2002) suggest that the parent body of L chondrites mixing, which has been inferred to be 100–150 m.y. after
and the Flora asteroid family may have been the same body. the solar system formed. At the current mean impact speed
3.6.2. Mesosiderites. These differentiated meteorites of 5 km/s, projectile material is almost entirely lost (Love
are all breccias composed of roughly equal proportions of and Ahrens, 1996).
metallic Fe,Ni and silicate. The silicate portion consists of The large M-type asteroid, 16 Psyche, has been sug-
fragments of rocks and silicate minerals in a fine-grained gested as a scrambled parent body for mesosiderites as it
fragmental or igneous matrix (Fig. 6). The rock and mineral lacks a dynamical family that might have formed if Psyche
706 Asteroids III

were an exposed core from a 500-km-diameter asteroid Acknowledgments. I thank D. W. Mittlefehldt, M. Killgore,
(Davis et al., 1999). However, an estimate of the density M. Zolensky, and A. Kracher for supplying meteorite photos; G. J.
of Psyche by Viateau (2000) is too low to be consistent with Taylor, A. N. Krot, B. A. Cohen, C. A. Goodrich, K. Keil, and
this model. Field work would be useful to test these ideas other colleagues for valuable discussions and comments; and H. Y.
McSween and W. Bottke for helpful reviews. This work was sup-
on mesosiderite origins and to develop better models for
ported in part by NASA Grant NAG 5-4212 (K. Keil, P.I.). This
large impacts.
is Hawai‘i Institute of Geophysics and Planetology Publication
3.6.3. Ureilite achondrites. Ureilites, which probably No. 1231 and School of Ocean and Earth Science and Technology
formed at depth in a partly melted asteroid, consist largely Publication No. 6025.
of coarse-grained olivine and pyroxene grains with minor
interstitial metal and graphite. If these phases were in equi- REFERENCES
librium, ureilites must have formed at depth in a body that
was at least 100 km in radius (Mittlefehldt et al., 1998). One Akridge G., Benoit P. H., and Sears D. W. G. (1998) Regolith and
of the most remarkable features of ureilites is that they all megaregolith formation of H-chondrites: Thermal constraints
cooled rapidly from 1250° to <650°C at ~10°C/h (see Good- on the parent body. Icarus, 132, 185–195.
rich et al., 2001, 2002). To cool in a few days, the hot rock Alexander C. M. O’D., Boss A. P., and Carlson R. W. (2001) The
must have been broken into meter-sized pieces. Fragmenta- early evolution of the inner solar system: A meteoritic perspec-
tive. Science, 293, 64–68.
tion of impact debris into such tiny pieces may have been
Bell J. F., Davis D. R., Hartmann W. K., and Gaffey M. J. (1989)
aided by formation of CO-CO2 gas in grain boundaries on Asteroids: The big picture. In Asteroids II (R. P. Binzel et al.,
pressure release (Keil et al., 1994). eds.), pp. 921–945. Univ. of Arizona, Tucson.
A large number of the ureilites contain shock-formed Benedix G. K., McCoy T. J., Keil K., and Love S. G. (2000) A
diamonds and other shock metamorphic features. The high petrologic study of the IAB iron meteorites: Constraints on the
proportion of shocked ureilites and their unusual thermal formation of the IAB-winonaite parent body. Meteoritics &
history suggest that they are derived from near the point of Planet. Sci., 35, 1127–1141.
impact. Since ureilite regolith breccias are made from simi- Bennett M. E. and McSween H. Y. (1996) Revised model calcula-
lar material, it seems likely that ureilites are derived from tions for the thermal histories of ordinary chondrite parent
one member of a family of asteroids that accreted a few bodies. Meteoritics & Planet. Sci., 31, 783–792.
days or more after the catastrophic breakup of the first Bischoff A. (2001) Meteorite classification and the definition of
new chondrite classes as a result of successful meteorite search
ureilite body. Since the catastrophic impact probably oc-
in hot and cold deserts. Planet. Space Sci., 49, 769–776.
curred 4.5 G.y. ago, the asteroid family may no longer be Bischoff A. and Stöffler D. (1992) Shock metamorphism as a fun-
recognizable. damental process in the evolution of planetary bodies: Infor-
mation from meteorites. Eur J. Mineral., 4, 707–755.
4. IMPLICATIONS Bogard D. D. (1995) Impact ages of meteorites: A synthesis.
Meteoritics & Planet. Sci., 30, 244–268.
The meteorite record contains a wealth of information Bogard D. D., Garrison D. H., Jordan J. L., and Mittlefehldt D. W.
about the origin and evolution of asteroids and the criti- (1990) 39Ar-40Ar ages and thermal history of mesosiderites:
cal role of impacts at all stages. Meteorite breccias and fea- Evidence for major parent body disruption <4 Ga ago. Geo-
tures can be identified that reflect impacts during accretion; chim. Cosmochim. Acta, 54, 2549–2564.
during melting, metamorphism, and alteration of asteroids; Bogard D. D., Garrison D. H., Norman M., Scott E. R. D., and
Keil K. (1995) 39Ar-40Ar age and petrology of Chico: Large-
and during the erosion and disruption of cold asteroids.
scale impact melting on the L chondrite parent body. Geochim.
However, many meteorites do not fit comfortably into our Cosmochim. Acta, 59, 1383–1399.
simple models for understanding asteroid accretion and evo- Bogard D. D., Garrison D. H., and McCoy T. J. (2000) Chronol-
lution. This probably results from the complexity of geologi- ogy and petrology of silicates from IIE iron meteorites: Evi-
cal processes on asteroids and our inadequate understand- dence of a complex parent body evolution. Geochim. Cosmo-
ing of what impacts do to hot and cold asteroids during and chim. Acta, 64, 2133–2154.
after accretion. Bottke W. F., Nolan M. C., Greenberg R., and Kolvoord R. A.
New meteorites and better constraints on the chronology, (1994) Velocity distribution among colliding asteroids. Icarus,
impact, and thermal history of chondrites and differentiated 107, 255–268.
meteorites have greatly helped to advance our understand- Brearley A. J. and Jones R. H. (1998) Chondritic meteorites. In
ing of asteroid formation and evolution since the review by Reviews in Mineralogy, Vol. 36, Planetary Materials (J. J.
Papike, ed.), pp. 3-1 to 3-398. Mineralogical Society of Amer-
Scott et al. (1989). But they have also emphasized the ur-
ica, Washington.
gent need for fieldwork on main-belt and near-Earth aster- Brown P. G. and 21 colleagues (2000) The fall, recovery, orbit
oids and laboratory study of returned samples. Without and composition of the Tagish Lake meteorite: A new type
these we cannot hope to understand fully how meteorites, of carbonaceous chondrite. Science, 290, 320–325.
asteroids, and planets formed; how asteroids and planets Bunch T. E. and Rajan R. S. (1988) Meteorite regolithic breccias.
interact; and how the harmful effects of such interactions In Meteorites and the Early Solar System (J. F. Kerridge and
can be minimized. M. S. Matthews, eds), pp. 144–164. Univ. of Arizona, Tucson.
 Scott: Accretion and Collisional Evolution of Asteroids 707

Burbine T. H., Meibom A., and Binzel R. P. (1996) Mantle mate- Science, 293, 2234–2236.
rial in the main belt: Battered to bits? Meteoritics & Planet. Hohenburg C. M., Meshik A. P., Pravdivtseva O. V., and Krot
Sci., 31, 607–620. A. N. (2001) I-Xe dating: Dark inclusions from Allende CV3
Burbine T. H., McCoy T. J., and Dickinson T. L. (2000) Origin of (abstract). Meteoritics & Planet. Sci., 36, A83.
plagioclase-“enriched”, igneous, enstatite meteorites (abstract). Hutcheon I. D., Weisberg M. K., Phinney D. L., Zolensky M. E.,
Meteoritics & Planet. Sci., 35, A36. Prinz M., and Ivanov A. V. (1999) Radiogenic 53Cr in Kaidun
Burbine T. H., McCoy T. J., Meibom A., Gladman B., and Keil K. carbonates: Evidence for very early aqueous activity (abstract).
(2002) Meteoritic parent bodies: Their number and identifica- In Lunar and Planetary Science XXX, Abstract #1722. Lunar
tion. In Asteroids III (W. F. Bottke Jr. et al., eds.), this volume. and Planetary Institute, Houston (CD-ROM).
Univ. of Arizona, Tucson. Ireland T. R. and Fegley B. (2000) The solar system’s earliest
Chambers J. E. and Wetherill G. W. (2001) Planets in the asteroid chemistry: Systematics of refractory inclusions. Intl. Geol.
belt. Meteoritics & Planet. Sci., 36, 381–399. Rev., 42, 865–894.
Clayton R. N. (1993) Oxygen isotopes in meteorites. Annu. Rev. Johnson C. A., Prinz M., Weisberg M. K., Clayton R. N., and
Earth Planet. Sci., 21, 115–149. Mayeda T. K. (1990) Dark inclusions in Allende, Leoville, and
Cuzzi J. N., Dobrovolskis A. R., and Hogan R. C. (1996) Turbu- Vigarano: Evidence for nebular oxidation of CV3 constituents.
lence, chondrules, and planetesimals. In Chondrules and the Geochim. Cosmoshim. Acta, 54, 819–830.
Protoplanetary Disk (R. H. Hewins et al., eds.), pp. 35–43. Jones R. H., Lee T., Connolly H. C., Love S. G., and Shang H.
Cambridge Univ., New York. (2000) Formation of chondrules and CAIs: Theory and obser-
Cuzzi J. N., Hogan R. C., Paque J. M., and Dobrovolskis A. R. vation. In Protostars and Planets IV (V. Mannings et al., eds.),
(2001) Size selective concentration of chondrules and other pp. 927–962. Univ. of Arizona, Tucson.
small particles in protoplanetary nebula turbulence. Astrophys. Jones T. D., Lebovsky L. A., Lewis J. S., and Marley M. S. (1990)
J., 546, 496–508. The composition and origin of the C, P, and D asteroids: Water
Cyr K. E., Sears W. D., and Lunine J. I. (1998) Distribution and as a tracer of thermal evolution in the outer belt. Icarus, 88,
evolution of water ice in the solar nebula: Implications for 172–192.
solar system body formation. Icarus, 135, 537–548. Keil K. (1989) Enstatite meteorites and their parent bodies. Mete-
Davis D. R., Farinella P. and Marzari F. (1999) The missing oritics, 24, 195–208.
Psyche family: Collisionally eroded or never formed? Icarus, Keil K. (2000) Thermal alteration of asteroids: Evidence from
137, 140–151. meteorites. Planet. Space Sci., 48, 887–903.
Endress M., Keil K., Bischoff A., Spettel B., Clayton R. N., and Keil K. (2002) Geological history of asteroid 4 Vesta: The “small-
Mayeda T. K. (1994) Origin of dark clasts in the Acfer 059/El est terrestrial planet”. In Asteroids III (W. F. Bottke Jr. et al.,
Djouf 001 CR2 chondrite. Meteoritics, 29, 26–40. eds.), this volume. Univ. of Arizona, Tucson.
Fegley B. Jr. (2000) Kinetics of gas-grain reactions in the solar Keil K., Haack H., and Scott E. R. D. (1994) Catastrophic frag-
nebula. Space Sci. Rev., 92, 177–200. mentation of asteroids: Evidence from meteorites. Planet.
Gaffey M. J. and Gilbert S. L. (1998) Asteroid 6 Hebe: The prob- Space Sci., 42, 1109–1122.
able parent body of the H-type ordinary chondrites and the IIE Keil K., Stöffler D., Love S. G., and Scott E. R. D. (1997) Con-
irons. Meteoritics & Planet. Sci., 33, 1281–1295. straints on the role of impact heating and melting in asteroids.
Goodrich C. A., Fioretti A. M., Tribaudino M., and Molin G. Meteoritics & Planet. Sci., 32, 349–363.
(2001) Primary trapped melt inclusions in olivine in the oliv- Kortenkamp S. J. and Wetherill G. W. (2000) Formation of the
ine-augite-orthopyroxene ureilite Hughes 009. Geochim. Cosmo- asteroid belt (abstract). In Lunar and Planetary Science XXXI,
chim. Acta, 65, 621–652. Abstract #1813. Lunar and Planetary Institute, Houston (CD-
Goodrich C. A., Krot A. N., Scott E. R. D., Taylor G. J., Fioretti ROM).
A. M., and Keil K. (2002) Formation and evolution of the ureil- Kracher A., Keil K., Kallemeyn G. W., Wasson J. T., Clayton
ite parent body and its offspring (abstract). In Lunar and Plan- R. N., and Huss G. I. (1985) The Leoville accretionary breccia.
etary Science XXXIII, Abstract #1379. Lunar and Planetary Proc. Lunar Planet. Sci. Conf. 16th, in J. Geophys. Res., 90,
Institute, Houston (CD-ROM). D123–D135.
Grady M. M. (2000) Catalogue of Meteorites. Cambridge Univ., Krot A. N., Meibom A., Russell S. S., Alexander C. M. O’D.,
Cambridge. 689 pp. Jeffries T. E., and Keil K. (2001) A new astrophysical setting
Greshake A., Krot A. N., Meibom A., Weisberg M. K., Zolensky for chondrule formation. Science, 291, 1776–1779.
M. E., and Keil K. (2002) Heavily hydrated lithic clasts in CH Liffman K. and Brown M. J. I. (1996) The protostellar jet model
chondrites and the related metal-rich chondrites Queen Alex- of chondrule formation. In Chondrules and the Protoplanetary
andra Range 94411 and Hammadah al Hamra 237. Mete- Disk (R. H. Hewins et al., eds), pp. 285–302. Cambridge Univ.,
oritics & Planet. Sci., 37, 281–293. New York.
Haack H., Rasmussen K. L., and Warren P. H. (1990) Effects of Lipschutz M. E., Gaffey M. J., and Pellas P. (1989) Meteoritic
regolith/megaregolith on the cooling histories of differentiated parent bodies: Nature, number, size and relationship to present-
asteroids. J. Geophys. Res., 95, 5111–5124. day asteroids. In Asteroids II (R. P. Binzel et al., eds.), pp. 740–
Haack H., Farinella P., Scott E. R. D., and Keil K. (1996) Mete- 777. Univ. of Arizona, Tucson.
oritic, asteroidal, and theoretical constraints on the 500 Ma Liu M., Scott E. R. D., Keil K., Wasson J. T., Clayton R. N.,
disruption of the L chondrite parent body. Icarus, 119, 182– Mayeda T., Eugster O., Crozaz G., and Floss C. (2001) North-
191. west Africa 176: A unique iron meteorite with silicate inclu-
Hiroi T., Zolensky M. E., and Pieters C. M. (2001) The Tagish sions related to Bocaiuva (abstract). In Lunar and Planetary
Lake meteorite: A possible sample from a D-type asteroid. Science XXXII, Abstract #2152. Lunar and Planetary Institute,
708 Asteroids III

Houston (CD-ROM). chondritic asteroids. Icarus, 113, 156–167.

Love S. G. and Ahrens T. J. (1996) Catastrophic impacts on grav- Rubin A. E. (2000) Petrogic, geochemical and experimental con-
ity-dominated asteroids. Icarus, 124, 141–155. straints on models of chondrule formation. Earth Sci. Rev., 50,
Lugmair G. W. and Shukolyukov A. (1998) Early solar system 3–27.
timescales according to 53Mn-53Cr systematics. Geochim. Rubin A. E. and Mittlefehldt D. W. (1993) Evolutionary history
Cosmochim. Acta, 62, 2863–2886. of the mesosiderite asteroid: A chronologic and petrologic syn-
Lugmair G. W. and Shukolyukov A. (1999) The 53Mn-53Cr sys- thesis. Icarus, 101, 201–212.
tematics of the enstatite chondrites (abstract). In Lunar and Rubin A. E. and Scott E. R. D. (1997) Abee and related EH chon-
Planetary Science XXX, Abstract #1093. Lunar and Planetary drite impact-melt breccias. Geochim. Cosmochim. Acta, 61,
Instute, Houston (CD-ROM). 425–435.
Lugmair G. W. and Shukolyukov A. (2001) Early solar system Rubin A. E. and Wasson J. T. (1995) Variations of chondrite prop-
events and timescales. Meteoritics & Planet. Sci., 36, 1017– erties with heliocentric distance (abstract). Meteoritics &
1026. Planet. Sci., 30, 569.
McCoy T. J. and 9 colleagues (1995) Origin and history of im- Rubin A. E., Ulff-Møller F., Wasson J. T., and Carlson W. D.
pact-melt rocks of enstatite chondrite parentage. Geochim. (2001a) The Portales Valley meteorite breccia: Evidence for
Cosmochim. Acta, 59, 161–175. impact-induced melting and metamorphism of an ordinary
McCoy T. J., Nittler L. R., Burbine T. H., Trombka J. I., Clark chondrite. Geochim. Cosmochim. Acta, 65, 323–342.
P. E., and Murphy M. E. (2000) Anatomy of a partially differ- Rubin A. E., Kallemeyn G. W., Wasson J. T., Clayton R. N.,
entiated asteroid: A “NEAR”-sighted view of acapulcoites and Mayeda T. K., Grady M. M., and Verchovsky A. B. (2001b)
lodranites. Icarus, 148, 29–36. Gujba: A new Bencubbin-like meteorite fall from Nigeria (ab-
McCoy T. J., Burbine T. H., McFadden L., Starr R. D., Gaffey stract). In Lunar and Planetary Science XXXII, Abstract #1779.
M. J., Nittler L. R., Evans L. G., Izenberg N., Lucey P., Lunar and Planetary Institute, Houston (CD-ROM).
Trombka J. I., Bell J. F. III, Clark B. E., Clark P. E., Squyres Rubin A. E., Wasson J. T., Kallemeyn G. W., Clayton R. N., and
S. W., Chapman C. R., Boynton W. V. and Veverka J. (2001) Mayeda T. K. (2001c) Ungrouped iron NWA 468: An impact-
The composition of 433 Eros: A mineralogical-chemical syn- melt product with low-Ca clinopyroxene-bearing silicate inclu-
thesis. Meteoritics & Planet. Sci., 36, 1661–1672. sions (abstract). Meteoritics & Planet. Sci., 36, A178–179.
McKay D. S., Swindle T. D., and Greenberg R. (1989) Asteroi- Sanders I. (1996) A chondrule-forming scenario involving molten
dal regoliths: What we do not know. In Asteroids II (R. P. planetesimals. In Chondrules and the Protoplanetary Disk
Binzel et al., eds.), pp. 617–642. Univ. of Arizona, Tucson. (R. H. Hewins et al., eds.), pp. 327–334. Cambridge Univ.,
McSween H. Y. Jr., Ghosh A., Grimm R. E., Wilson L., and Young New York.
E. D. (2002) Thermal evolution models of asteroids. In Aster- Schmitz B., Peucker-Ehrenbrink B., Lindström M., and Tassinari
oids III (W. F. Bottke Jr. et al., eds.), this volume. Univ. of M. (1997) Accretion rates of meteorites and cosmic dust in the
Arizona, Tucson. early Ordovician. Science, 278, 88–90.
Meibom A. and Clark B. E. (1999) Evidence for the insignificance Scott E. R. D. and Keil K. (1999) Metamorphism of meteorites:
of ordinary chondrite material in the asteroid belt. Meteoritics Clues to the geology of asteroids (abstract). Meteoritics &
& Planet. Sci., 34, 7–24. Planet. Sci., 34, A103–A104.
Mittlefehldt D. W. and Lindstrom M. M. (2001) Petrology and Scott E. R. D., Taylor G. J., Newsom H., Herbert F., Zolensky
geochemistry of the Patuxtent Range 91501, a clast-poor im- M., and Kerridge J. F. (1989) Chemical, thermal and impact
pact melt from the L-chondrite parent body and Lewis Cliff processing of asteroids. In Asteroids II (R. P. Binzel et al., eds.),
88663, an L7 chondrite. Meteoritics & Planet. Sci., 36, 439– pp. 701–739. Univ. of Arizona, Tucson.
457. Scott E. R. D., Love S. G., and Krot A. N. (1996) Formation of
Mittlefehldt D. W., McCoy T. J., Goodrich C. A., and Kracher A. chondrules and chondrites in the protoplanetary nebula. In
(1998) Non-chondritic meteorites from asteroidal bodies. In Chondrules and the Protoplanetary Disk (R. H. Hewins et al.,
Reviews in Mineralogy, Vol. 36, Planetary Materials (J. J. eds.), pp. 87–96. Cambridge Univ., New York.
Papike, ed.), pp. 4-1 to 4-195. Mineralogical Society of Amer- Scott E. R. D., Krot A. N., and Yamaguchi A. (1998) Carbonates
ica, Washington. in fractures in martian meteorite Allan Hills 84001: Petrologic
Nesvorný D., Morbidelli A., Vokrouhlický, Bottke W. F., and Broz evidence for impact origin. Meteoritics & Planet. Sci., 33, 709–
M. (2002) The Flora family: A case of the dynamically dis- 719.
persed collisional swarm? Icarus, 157, 155–172. Scott E. R.D., Haack H., and Love S. G. (2001) Formation of
Ott U. (2001) Presolar grains in meteorites: An overview and mesosiderites by fragmentation and reaccretion of a large dif-
some implications. Planet. Space Sci., 49, 763–767. ferentiated asteroid. Meteoritics & Planet. Sci., 36, 869–881.
Patzer A., Hill D. H., and Boynton W. V. (2001) Itqiy: A metal- Sears D. W. G. (1998) The case for rarity of chondrules and cal-
rich enstatite meteorite with achondritic texture. Meteoritics & cium-aluminum-rich inclusions in the early solar system and
Planet. Sci., 36, 1495–1505. some implications for astrophysical models. Astrophys. J., 498,
Petit J.-M., Chambers J., Franklin F., and Nagasawa M. (2002) 773–778.
Primordial excitation and depletion of the main belt. In Aster- Shearer C. K., Papike J. J., and Rietmeijer F. J. M. (1998) The
oids III (W. F. Bottke Jr. et al., eds.), this volume. Univ. of Ari- planetary sample suite and environments of origin. In Reviews
zona, Tucson. in Mineralogy, Vol. 36, Planetary Materials (J. J. Papike, ed.),
Rivken A. S., Howell E. S., Britt D. T., Lebovsky L. A., Nolan pp. 1-1 to 1-28. Mineralogical Society of America, Washington.
M. C., and Branston D. D. (1995) 3-µm spectrophotometric Shu F. H., Shang H., Gounelle M., and Glassgold A. E. (2001)
survey of M- and E-class asteroids. Icarus, 117, 90–100. The origin of chondrules and refractory inclusions in chondritic
Rubin A. E. (1995) Petrologic evidence for collisional heating of meteorites. Astrophys. J., 548, 1029–1050.
 Scott: Accretion and Collisional Evolution of Asteroids 709

Stöffler D., Keil K., and Scott E. R. D. (1991) Shock metamor- Wasson J. T. (1988) The building stones of planets. In Mercury
phism of ordinary chondrites. Geochim. Cosmochim. Acta, 55, (F. Vilas et al., eds.), pp. 622–650. Univ. of Arizona, Tucson.
3845–3867. Wasson J. T. and Kallemeyn G. W. (1988) Compositions of chon-
Symes S. J. K., Sears D. W. G., Akridge D. G., Huang S., and drites. Phil. Trans. R. Soc. Lond., A325, 535–544.
Benoit P. (1998) The crystalline lunar spherules: Their forma- Wasson J. T. and Kallemeyn G. W. (1990) Allan Hills 85085: A
tion and implications for the origin of meteoritic chondrules. subchondritic meteorite of mixed nebular and regolithic heri-
Meteoritics & Planet. Sci., 33, 13–29. tage. Earth Planet. Sci. Lett., 101, 148–161.
Takeda H. (1993) Chemical differentiation during collision and Weidenschilling S. J., Marzari F., and Hood L. L. (1998) The ori-
accretion of meteorite parent bodies. In Primitive Solar Nebula gin of chondrules at Jovian resonances. Science, 279, 681–684.
and Origin of Planets (H. Oya, ed.), pp. 375–394. Terra Sci- Weisberg M. K., Prinz M., Clayton R. N., Mayeda T. K., Sugiura
entific, Tokyo. N., Zashu S., and Ebihara M. (2001) A new metal-rich chon-
Taylor G. J. and Scott E. R. D. (2001) Molten planetesimals in drite grouplet. Meteoritics & Planet. Sci., 36, 401–418.
the early solar system (abstract). Meteoritics & Planet. Sci., Weiss B. P., Kirschvink J. L., Baudenbacher F. J., Vali H., Peters
36, A204. N. T., Macdonald F. A., and Wikswo J. P. (2000) A low tem-
Taylor G. J., Scott E. R. D., and Keil K. (1983) Cosmic setting for perature transfer of ALH84001 from Mars to Earth. Science,
chondrule formation. In Chondrules and Their Origin (E. A. 290, 791–795.
King, ed.), pp. 262–278. Lunar and Planetary Institute, Houston. Wetherill G. W. (1989) Origin of the asteroid belt. In Asteroids II
Taylor G. J., Maggiore P., Scott E. R. D., Rubin A. E., and Keil K. (R. P. Binzel et al., eds.), pp. 661–680. Univ. of Arizona, Tucson.
(1987) Original structures and fragmentation and reassembly Wood J. A. (1996) Unresolved issues in the formation of chon-
histories of asteroids: Evidence from meteorites. Icarus, 69, drules and chondrites. In Chondrules and the Protoplanetary
1–13. Disk (R. H. Hewins et al., eds.), pp. 55–69. Cambridge Univ.,
Taylor G. J., Keil K., McCoy T., Haack H., and Scott E. R. D. New York.
(1993) Asteroid differentiation: Pyroclastic volcanism to Yamaguchi A., Taylor G. J., and Keil K. (1997) Metamorphic
magma oceans. Meteoritics, 28, 34–52. history of the eucritic crust of 4 Vesta. J. Geophys. Res., 102,
Treiman A. H. (1998) The history of Allan Hills 84001 revised: 13381–13386.
Multiple shock events. Meteoritics & Planet. Sci., 33, 753– Yamaguchi A., Misawa K., Haramura H., Kojima H., Clayton
764. R. N., Mayeda T. K., and Ebihara M. (2001) Northwest Af-
Turner G., Knott S. F., Ash R. D., and Gilmour J. D. (1997) Ar- rica 011, a new basaltic meteorite (abstract). Meteoritics &
Ar chronology of the Martian meteorite ALH84001: Evidence Planet. Sci., 36, A228.
for the timing of the early bombardment of Mars. Geochim. Zolensky M. E. and Ivanov A. V. (2001) Kaidun: A smorgasbord
Cosmochim. Acta, 61, 3835–3850. of new asteroid samples (abstract). Meteoritics & Planet. Sci.,
Viateau B. (2000) Mass and density of asteroids (16) Psyche and 36, A233.
(121) Hermione. Astron. Astrophys., 354, 725–731. Zolensky M. E., Ivanov A. V., Yang S. V., Mittlefehldt D. W., and
Wadhwa M. and Russell S. S. (2000) Timescales of accretion and Ohsumi K. (1996) The Kaidun meteorite: Mineralogy of an
differentiation in the early solar system. In Protostars and unusual CM1 lithology. Meteoritics & Planet. Sci., 31, 484–
Planets IV (V. Mannings et al., eds.), pp. 995–1018. Univ. of 493.
Arizona, Tucson.