Geological Society of America
Special Paper 356
2002
Comparison of the osmium and chromium isotopic methods
for the detection of meteoritic components in impactites:
Examples from the Morokweng and Vredefort
impact structures, South Africa
Christian Koeberl*
Institute of Geochemistry, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria
Bernhard Peucker-Ehrenbrink*
Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution,
Woods Hole, Massachusetts 02543-1541, USA
Wolf Uwe Reimold*
Impact Cratering Research Group, Department of Geology, University of the Witwatersrand,
Johannesburg 2050, South Africa
Alex Shukolyukov*
Scripps Institute of Oceanography, University of California, San Diego, California 92093-0212, USA
Günter W. Lugmair*
Scripps Institute of Oceanography, University of California, San Diego, California 92093-0212, USA,
and Max-Planck-Institute for Chemistry, P.O. Box 3060, 55020 Mainz, Germany
ABSTRACT
Breccias and melt rocks found at possible meteorite impact structures on Earth
may contain a minor extraterrestrial component. In the absence of evidence of shockmetamorphic effects in such rocks, the unambiguous detection of an extraterrestrial
component can be of diagnostic value regarding the impact origin of a geologic structure. Previously the concentrations of siderophile elements, mainly the platinumgroup elements (PGEs), or osmium isotopic studies have been used to detect such a
meteoritic component. The Cr isotopic method has shown great potential for identifying the type of impactor. Here we use a combination of trace element (PGE) and
Os and Cr isotopic data for rocks from the two large impact structures, Vredefort
and Morokweng, both in South Africa, to compare the two isotopic methods directly
for the first time. The percentage of a meteoritic component in Vredefort granophyre
was too low to yield a Cr isotopic signal, but there was a distinct Os isotope effect.
For Morokweng, both methods yielded positive results, the Cr isotope data suggesting
that the projectile had the composition of an ordinary chondrite, probably an L chondrite. Comparison of the two methods thus confirms that the Os isotopic method is a
*E-mails: Koeberl, [email protected]; Peucker-Ehrenbrink,
[email protected]; Reimold, [email protected]; Shukolyukov,
[email protected]; Lugmair, [email protected]
Koeberl, C., Peucker-Ehrenbrink, B., Reimold, W.U., Shukolyukov, A., and Lugmair, G.W., 2002, Comparison of the osmium and chromium isotopic methods
for the detection of meteoritic components in impactites: Examples from the Morokweng and Vredefort impact structures, South Africa, in Koeberl, C., and
MacLeod, K.G., eds., Catastrophic Events and Mass Extinctions: Impacts and Beyond: Boulder, Colorado, Geological Society of America Special Paper 356, p.
607–617.
607
608
C. Koeberl et al.
more sensitive tool for detecting an extraterrestrial component, whereas the Cr isotopic method, where applicable, can provide additional information regarding the
nature of the impactor.
INTRODUCTION
The verification of an extraterrestrial component in impactderived melt rocks or breccias can provide evidence confirming
an impact origin of a geologic structure. Similar approaches are
of great value in the investigation of distal ejecta layers; perhaps
the best example is that of the Cretaceous-Tertiary (K-T)
boundary, where the discovery of an extraterrestrial component
(Alvarez et al., 1980) opened the door to a reassessment of the
importance of impact events in the geological history of the
Earth. During impact, a small amount of the finely dispersed
meteoritic melt or vapor is mixed with a much larger quantity
of target rock vapor and melt, and this mixture later forms
impact-melt rocks, melt breccias, or impact glass. In most cases,
the contribution of meteoritic matter to these impactite lithologies is very small (K1%), leading to only slight chemical
changes in the resulting impactites.
The detection of such small amounts of meteoritic matter
within the normal upper crustal compositional signature of the
target rocks is extremely difficult. Only elements that have high
abundances in meteorites, but low abundances in terrestrial
crustal rocks (e.g., the siderophile elements) are useful. Another
complication is the existence of a variety of meteorite groups
and types (the three main groups are stony meteorites, iron meteorites, and stony-iron meteorites, in order of decreasing abundance; see, e.g., Wasson, 1985; Papike, 1998, for reviews on
meteorites), which have widely varying siderophile element
compositions. Distinctly higher siderophile element contents in
impact melts, compared to target rock abundances, can be indicative of the presence of either a chondritic or an iron meteoritic component. Achondritic projectiles (stony meteorites that
underwent magmatic differentiation) are much more difficult to
discern, because they have significantly lower abundances of
the key siderophile elements. Furthermore, in order to reliably
determine the target rock contribution of such elements, i.e., the
so-called indigenous component, absolute certainty must be attained that all contributing terrestrial target rocks have been
identified and their relative contributions to the melt mixture
are reasonably well known.
Geochemical methods have been used to determine the
presence of the traces of such an extraterrestrial component (see
review in Koeberl, 1998). In the absence of actual meteorite
fragments, it is necessary to chemically search for traces of
meteoritic material mixed in with the target rocks in breccias
and melt rocks. Meteoritic components have been identified for
⬃40 impact structures (see Koeberl, 1998, for a list), of the
more than 160 impact structures that have so far been identified
on Earth. This number reflects mostly the extent to which these
structures have been studied in detail, because only a few of
these impact structures were first identified by finding a meteoritic component (the majority were confirmed by the identification of shock-metamorphic effects). The identification of a
meteoritic component can be achieved by determining the concentrations and interelement ratios of siderophile elements, especially the platinum group elements (PGEs), which are several
orders of magnitude more abundant in meteorites than in terrestrial upper crustal rocks. This is illustrated for one of the
PGEs, Ir, in Figure 1. Iridium is most often determined as a
proxy for all PGEs, because it can be measured with the best
detection limit of all PGEs by neutron activation analysis
(which was, for a long time, the only more or less routine
method for Ir measurements at abundance levels less than parts
per billion in small samples).
The use of PGE abundances and ratios avoids some of the
ambiguities that result if only moderately siderophile elements
(e.g., Cr, Co, Ni) are used in an identification attempt. However,
problems may arise if the target rocks have high abundances of
siderophile elements or if the siderophile element concentrations in the impactites are very low. In such cases, the Os and
Cr isotopic systems can be used to establish the presence of a
meteoritic component in a number of impact-melt rocks and
breccias (e.g., Koeberl and Shirey, 1997; Shukolyukov and
Lugmair, 1998). In the past, PGE data were used to estimate
the type or class of meteorite for the impactor (e.g., Morgan,
1978; Palme, 1982; Palme et al., 1978, 1979, 1981), but these
attempts were not always successful. It is difficult to distinguish
among different chondrite types based on siderophile element
(or even PGE) abundances, which has led to conflicting conclusions regarding the nature of the impactor at a number of
structures (see Koeberl, 1998, for details). Clearly, the identification of a meteoritic component in impactites is not a trivial
problem. In this study we use a combination of trace element
(PGE) analyses and the results from both Os and Cr isotopic
studies to illustrate the pros and cons of each method with two
case studies, the Vredefort and Morokweng impact structures
of South Africa.
EXPERIMENTAL METHODS: BACKGROUND AND
ANALYTICAL PROCEDURES
Osmium isotopic method
The use of the Os isotopic system is based on the formation
Os (one of seven stable isotopes of Os) by b-decay of
187
Re (half-life of 42.3 Ⳳ 1.3 Ga). Meteorites (and the terrestrial mantle) have Os and other PGEs contents higher by factors
of
187
Comparison of the osmium and chromium isotopic methods
609
Figure 1. Approximate range of iridium
concentrations in variety of terrestrial
rocks, impactites, distal impact ejecta
(i.e., tektites and Cretaceous-Tertiary
[K-T] boundary clays), and meteorites
(for data sources, see Koeberl, 1998), in
comparison with data for target rocks
and impact-melt rocks at Vredefort and
Morokweng impact structures (data
from French et al., 1989; Koeberl et al.,
1997; Koeberl and Reimold, 2002).
of 102–105 than terrestrial crustal rocks. In addition, meteorites
have relatively low Re and high Os abundances, resulting in
Re/Os ratios ⱕ0.1, whereas the Re/Os ratio of terrestrial crustal
rocks is usually no less than 10. Similar to the Rb-Sr and SmNd isotopic methods, the abundance of radiogenic 187Os is normalized to the abundance of nonradiogenic 188Os. The use of
the method is based on the fact that the 187Os/188Os isotopic
ratios for meteorites and terrestrial crustal rocks are significantly different. As a result of the high Re and low Os concentrations in old crustal rocks, their 187Os/188Os ratio increases
rapidly with time (average upper crustal 187Os/188Os ⳱ 1–1.2).
In contrast, meteorites have low 187Os/188Os ratios of ⬃0.11–
0.18, and, because Os is much more abundant in meteorites
than Re, only small changes in the meteoritic 187Os/188Os ratio
occur with time. In the laboratory, the abundances of Os (in
some cases also of Re) and the 187Os/188Os isotopic ratio are
measured by sensitive negative thermal ionization mass spectrometry (NTIMS) (e.g., Creaser et al., 1991).
The first application of this isotope system to impactrelated materials was reported by Luck and Turekian (1983),
who measured the Os isotopic composition of K-T boundary
clays, which was feasible because of the relatively high Os
abundances (tens of parts per billion) in these rocks. However,
the first application for impact-crater studies, by Fehn et al.
(1986), was hampered by a lack in sensitivity, because target
rocks and typical impactites have very low (sub-parts per billion) Os abundances. This changed with the introduction of the
NTIMS technique (Völkening et al., 1991), which allowed the
measurement of Os isotopic ratios in samples of a few grams
mass containing sub-parts per billion amounts of Os. The first
successful application of this method dealt with the Bosumtwi
impact structure in Ghana (Koeberl and Shirey, 1993).
As a result of the relatively high Os abundances in meteorites, the addition of even a small amount of meteoritic matter
to the crustal target rocks leads to an almost complete change
of the Os isotopic signature of the resulting impact melt or
breccia. Contamination from an achondritic meteorite requires
much higher percentages of meteoritic addition due to the much
lower PGE abundances in achondrites compared to chondritic
and iron meteorites. In addition, the present-day 187Os/188Os
ratio of mantle rocks is ⬃0.13, which is similar to meteoritic
values. Thus, contribution of a significant mantle component
(e.g., ultramafic rocks) in an impact breccia or melt rock has to
be excluded. The PGE abundances in typical rocks from the
upper mantle are at least two orders of magnitude lower than
those in (chondritic and iron) meteorites. Thus, to achieve the
same disturbance of the Os isotope value, at least 100 times
higher mantle contributions (i.e., 10 wt% or more) than meteoritic contributions (e.g., 0.1 wt%) would be needed. The presence of such a significant mantle component would be easily
discernable from petrographic studies of the clast population in
breccias, and/or from geochemical data (especially Sr and Nd
isotopes) of melt rocks. Compared to the use of PGE elemental
abundances and ratios, the Os isotope method is superior with
610
C. Koeberl et al.
respect to detection limit and selectivity, as discussed in detail
(with several case histories) by Koeberl and Shirey (1997).
There are, however, a few disadvantages; the laboratory procedures (sample preparation, digestion, and measurement) are
complex, and the Os isotopic method does not allow determination of the projectile type.
Chromium isotopic method
This new method provides evidence not only for the presence of an extraterrestrial component in impactites, but also
information regarding the type of projectiles (Shukolyukov and
Lugmair, 1998). It is based on the determination of the relative
abundances of 53Cr, which is the daughter product of the extinct
radionuclide 53Mn (half-life ⳱ 3.7 m.y.). The 53Cr relative
abundances are measured as the deviations of the 53Cr/52Cr ratio
in a sample relative to the standard terrestrial 53Cr/52Cr ratio by
high-precision TIMS. These deviations are usually expressed
in e units (1 e is 1 part in 104, or 0.01%). Terrestrial rocks are
not expected to show (and do not reveal) any variation in 53Cr/
52
Cr ratio, because the homogenization of the Earth was completed long after all 53Mn (which was injected into, or formed
within, the solar nebula) had decayed. In contrast, most meteorite groups analyzed so far (Shukolyukov and Lugmair, 1998,
2000), such as carbonaceous, ordinary, and enstatite chondrites,
primitive achondrites, and other differentiated meteorites (including the SNC meteorites, which originated from Mars) show
a variable excess of 53Cr relative to terrestrial samples. The
range for meteorites is about Ⳮ0.1 to Ⳮ1.3 e, depending on
the meteorite type, except for carbonaceous chondrites, which
show an apparent deficit in 53Cr of ⬃ⳮ0.4 e. These differences
reflect a heterogeneous distribution of 53Mn in the early solar
system and early Mn/Cr fractionation in the solar nebula and
in meteorite parent bodies (Lugmair and Shukolyukov, 1998).
The negative e value for carbonaceous chondrites is an artifact
of using the 54Cr/52Cr ratio for a second-order fractionation correction (Lugmair and Shukolyukov, 1998), because these meteorites carry a pre-solar 54Cr component (Shukolyukov et al.,
2000). The actual, unnormalized 53Cr/52Cr ratio is similar to
that of other undifferentiated meteorites, and the apparent 53Cr
deficit in the carbonaceous chondrites is actually due to an excess of 54Cr. However, the presence of 54Cr excesses in bulk
carbonaceous chondrites allows us to distinguish clearly these
meteorites from the other meteorite classes.
Shukolyukov and Lugmair (1998) have used the Cr isotopic method on samples from the K-T boundary in Denmark
and Spain and found that ⬃80% of the Cr in these samples
originated from an impactor with a carbonaceous chondritic
composition. Shukolyukov et al. (2000) found a similar (carbonaceous chondritic) signal for some samples from Archean
spherule layers in the Barberton Mountain Land, South Africa.
Shukolyukov and Lugmair (2000) reported on analyses of
impact-melt rocks from the East Clearwater (Canada), Lappajärvi (Finland), and Rochechouart (France) impact structures.
In all three cases, the impactors were found to have had ordinary chondritic composition.
The Cr isotopic method thus has the advantage over both
the use of Os isotopes or PGE abundance ratios of being selective not only regarding the Cr source (terrestrial versus extraterrestrial), but also regarding the meteorite type. A disadvantage of this method is the complicated and time-consuming
analytical procedure. In addition, a significant proportion of the
Cr in an impactite, compared to the abundance in the target,
has to be of extraterrestrial origin. This point is illustrated in
Figure 2, which shows the detection limit of the Cr isotopic
method, assuming a chondritic composition of an extraterrestrial component. This detection limit is a function of the Cr
content in the (terrestrial) target rocks that were involved in the
formation of the impact breccias or melt rocks. For example, if
the average Cr concentration in the target is ⬃185 ppm (the
average Cr concentration in the bulk continental crust; Taylor
and McLennan, 1985), only an extraterrestrial component of
⬎1.2% can be detected.
Analytical procedures
For the present study, we selected impact-melt rocks and
target rocks from two large impact structures in South Africa,
Vredefort and Morokweng (see following for details on the
structures and samples). The Os isotopic analyses of the Vredefort rocks are those of Koeberl et al. (1996), where analytical
details are given. We analyzed 12 samples from the Morokweng
impact structure for Os isotopic composition, and the abundances of Os, Ir, and Pt were determined using methods described in detail by Ravizza and Pyle (1997) and Hassler et al.
Figure 2. Detection limits for extraterrestrial component (ETC) in impactites (e.g., suevitic breccias or impact-melt rocks), using chromium
isotope method. Detection limit, assuming chondritic composition of
extraterrestrial component, is plotted (in percent) on y-axis as function
of Cr content in terrestrial target rocks that were involved in impactite
formation.
Comparison of the osmium and chromium isotopic methods
(2000). The samples were digested by NiS fire assay, followed
by transfer of volatile OsO4 by an Ar gas stream directly (without nebulizer) into the torch of a magnetic sector ICP-MS (Finnigan Element), allowing the rapid determination of the Os content and isotopic composition. The liquid residue after sparging
of Os can then be used for conventional inductively coupled
plasma mass spectroscopy (ICP-MS) measurement of the abundances of the PGEs (here, Ir, Pt, and Os). Due to the considerable analytical effort for the determination of the Cr isotopic
data, only two melt-rock samples from each of the two impact
structures were measured. The methods were described in detail
by Shukolyukov and Lugmair (1998) and Shukolyukov et al.
(2000).
VREDEFORT IMPACT STRUCTURE: BACKGROUND
The Vredefort structure in South Africa, centered ⬃140 km
southwest of Johannesburg, currently has a diameter of ⬃100
km, which is believed to represent the central uplift of this
impact structure. The entire impact structure could initially
have been as large as 300 km, comprising the entire Witwatersrand basin (Henkel and Reimold, 1998). The origin of the structure has been debated during most of the twentieth century; e.g.,
microscopic deformation features in quartz, with an appearance
similar to deformation lamellae observed in some tectonically
deformed rocks, were controversial because of their unusual
appearance at the optical level (cf. Grieve et al., 1990; Reimold,
1993). However, recent data confirmed the existence of impactcharacteristic shock-metamorphic effects in Vredefort rocks,
such as basal Brazil twins in quartz (Leroux et al., 1994) and
shock-characteristic planar deformation features (PDFs) in zircon (Kamo et al., 1996), supporting an impact origin as opposed
to an internal origin of the structure (see Reimold and Gibson,
1996, for a review). The Vredefort event is well dated as 2023
Ⳳ 4 Ma (Kamo et al., 1996). Dikes of granophyric rock, the
so-called Vredefort granophyre, occur in the basement core of
the structure and along the boundary between the core and the
supracrustal rocks of the collar. Previous studies indicated that
the granophyre could have formed by impact melting of granite,
shale, and quartzite. A major mafic contribution is unlikely due
to the scarcity of mafic clasts and because the granophyre composition can be perfectly modeled from mixing of felsic crustal
and supracrustal rocks (French et al., 1989; French and Nielsen,
1990; Reimold et al., 1990; Koeberl et al., 1996; Reimold and
Gibson, 1996). All these authors agreed that the Vredefort
granophyre represents an impact-melt rock, probably material
that gravitationally settled into impact-generated fractures in the
crater basement after the impact event.
MOROKWENG IMPACT STRUCTURE:
BACKGROUND
The Morokweng impact structure is centered at 23⬚32⬘ E
and 26⬚20⬘ S, close to the border with Botswana, in the North-
611
west Province of South Africa. The structure was recognized
as a circular positive magnetic anomaly of as much as 350 nT
above regional background. This anomaly forms a central 30km-diameter, near-circular area, which is surrounded by a concentric, magnetically quiet zone that is 20 km wide. Refined
processing of the gravity and aeromagnetic data revealed the
possible presence of a larger circular structure (Corner et al.,
1997). The discovery of impact-characteristic shock-metamorphic effects in rocks from the Morokweng area (e.g., Corner et
al., 1997; Koeberl et al., 1997; Hart et al., 1997) confirmed the
presence of a large meteorite-impact structure. The size of the
Morokweng impact structure is still debated (e.g., Reimold et
al., 1999; Andreoli et al., 1999), but new evidence (Reimold et
al., 2000) seems to favor a diameter of ⱖ70–80 km. Three
boreholes in the south-central area of the aeromagnetic anomaly
(MWF03, core depth 130.3 m; MWF04, 189.3 m; MWF05,
271.3 m) were sampled (see Koeberl et al., 1997; Reimold et
al., 1999, for detailed information on the three cores).
All three boreholes penetrated a top layer of the Tertiary
to Holocene Kalahari Group calcrete, which is directly underlain by a dark-brown melt rock having a thickness of about 125
m in borehole MWF05. Only in borehole MWF05 was the
lower contact of the melt rock intersected. In this hole, granitic
rocks were reached at 225 m depth, whereas the other holes
were terminated while still within the melt-rock unit. Most of
the melt rock appears fresh and homogeneous, except for a large
number of lithic clasts. The clast population includes many gabbro fragments, but microscopic studies reveal that felsic, clearly
granitoid-derived, clasts are dominant. The granites drilled below the melt body in core MWF05 are locally brecciated and
pervasively recrystallized. Some primary minerals are preserved and display shock deformation in the form of PDFs in
quartz, plagioclase, alkali feldspar, and K-feldspar (Koeberl et
al., 1997; Reimold et al., 1999). Some thin (⬍10 cm wide)
breccia veins occur in the granitoids of drill core MWF05,
which have since been identified as injections of melt, or as
partly recrystallized, locally produced, cataclastic material. It is
not clear yet whether granitoid basement has been reached at
the bottom of drill core MWF05, but it is possible that a (mega?)
breccia zone below the melt rock and above the basement was
intersected. SHRIMP ion probe dating of zircons from the melt
rock yielded an age of 146.2 Ⳳ 1.5 Ma, which is indistinguishable from that of the Jurassic-Cretaceous boundary (Koeberl et
al., 1997; Reimold et al., 1999).
In 1999, we were able to sample and study a 3.5-km-long
KHK-1 drill core obtained by Anglogold Limited on the farm
Kelso 351 (⬃23⬚12⬘E, 26⬚40⬘S), ⬃38–40 km southwest of the
presumed center of the impact structure. The rocks recovered
from this core are considered to be representative of the Morokweng target rocks, especially the more mafic rocks that were
not well represented among the (predominantly granitoid) clast
population of the three central Morokweng drill cores. Preliminary stratigraphic results for the KHK-1 core were presented
by Reimold and Koeberl (1999), and further details on the drill
612
C. Koeberl et al.
core stratigraphy, as well as first results of chemical and chronological analysis of drill core samples, were reported by Reimold et al. (2000). The top 599.15 m (not recovered) represent
dolomite and chert of the 2.25–2.5 Ga Transvaal Supergroup.
This package continues until 889.1 m, to the contact with a
gabbroic intrusion. Some dolomite follows, above a package of
arenitic metasediments (to 1200 m). A series of mafic volcanic
flows was intersected to a depth of 1417.8 m, where a 1.2 m
cataclastic breccia that exhibited no shock-deformation features
was observed. The breccia is underlain by more metasediment,
including some diamictite also lacking shock-deformation features. Felsic volcanics and a felsic granophyre follow until
1784.2 m, below which more metasediment, including two thin
diamictite bands, was encountered. Below this depth, a thick
package of felsic granophyre was intersected to the depth of
2669.4 m, where it is terminated by a gabbroic intrusion. We
did not find any quartz clasts with PDFs in this felsic granophyre. A thick gabbro intrusion follows until 3012 m depth,
where locally pegmatoidal, but mostly micropegmatoidal, granitoids continue until the final depth of the borehole (3420 m).
Whereas no evidence of shock metamorphism could be detected
in any of the granophyric rocks of the drill core, there is ample
evidence that at least some of the granophyric material is of
secondary origin, formed from melting of a granitoid precursor.
RESULTS: METEORITIC COMPONENT AT
VREDEFORT
Vredefort granophyre has been interpreted either as an igneous intrusion or as impact-melt rock that was injected into
fractures in the floor of the impact structure (French et al., 1989;
French and Nielsen, 1990). In addition, French et al. (1989)
analyzed the Ir content of seven Granophyre samples and reported a range of 57–130 ppt Ir (Fig. 1). These authors also
analyzed a suite of country rocks and found that the Vredefort
granophyre is enriched in Ir by a factor of ⬃20–50 compared
to granitic country rocks (target rocks?). However, they also
found that some Witwatersrand shale samples have Ir contents
of 160–330 ppt (Fig. 1). French et al. (1989) suggested that all
Ir in the Vredefort granophyre could be explained by admixture
of Ir from shales and similar rocks during impact melting. However, this would require that at least one-third of the Vredefort
granophyre composition be derived from shale, which contradicts mixing calculations of French and Nielsen (1990), who
only found a 10% shale contribution. All other relevant country
rocks have Ir contents of 3–62 ppt. Thus, it seems that there is
excess Ir in the Vredefort granophyre; nevertheless, this ambiguity made the Vredefort granophyre, and Vredefort in general,
an ideal target for an Os isotopic study.
Such a study was undertaken by Koeberl et al. (1996) in
order to (1) search for a meteoritic component in Vredefort
granophyre, and (2) to confirm that the granophyre represents
an impact-melt rock. The Os abundances in the granophyre
range from 0.11 to 1.11 ppb, which is significantly higher than
the average of the source-rock values, indicating a distinct enrichment of Os in most of the granophyre samples compared to
the country rocks. In addition, the 187Os/188Os ratios of the
granophyre samples are significantly lower than those of the
supracrustal country rocks (Table 1). The 187Re/188Os and
187
Os/188Os ratios of the Vredefort granophyre scatter about a
2 Ga isochron, the majority of the initial 187Os/188Os ratios (at
2 Ga) ranging from 0.13 to 0.22 (Fig. 3). These values overlap
the meteoritic data range and indicate that all measured granophyre samples contain some meteoritic Os. In addition, the ReOs isotopic composition of the granophyre is significantly different from that of any of the target rocks. Shale samples were
also analyzed and yielded values of 37–162 ppt Os, but the Os
isotopic compositions are significantly different from those of
the granophyre samples.
Thus, a meteoritic component in the Vredefort granophyre
is the only explanation that is in agreement with these observations. This conclusion is supported by some enrichments in
Cr, Co, Ni, and Ir in the granophyre compared to the country
rocks, although these enrichments are not as unambiguous as
the Re-Os data. Assuming chondritic meteorite Os abundances
(⬃500 ppb), Koeberl et al. (1996) concluded that the Vredefort
granophyre contains ⱕ0.2% of a chondritic component. However, the meteoritic component is not homogeneously distributed, probably due to a nugget effect, as is evident from a spread
in 187Re/188Os ratios, in agreement with observations from other
impact-melt rocks (cf. Koeberl, 1998). Meteorites contain ⬃10
times less Re than Os, indicating that the Re contribution from
the meteoritic material to the Vredefort granophyre was ⱕ30%,
subordinate to the Os, which is almost exclusively of meteoritic
origin. The almost constant Re abundances found in the granophyre samples indicate homogenization during the impact.
Thus, the Vredefort granophyre is a mixture of a large amount
of low-Os, high-Re material (crustal rocks) with a small contribution of high-Os, low-Re meteoritic material. The Os isotope results showed that all analyzed Vredefort granophyre
samples contain some meteoritic Os, confirming that the Vredefort granophyre is an impact-melt rock.
To confirm the findings from the Os isotope study, and for
comparison of the two isotopic techniques, we analyzed two
Vredefort granophyre samples for their Cr isotopic composition. In contrast to the Os isotope study, the Cr isotope study
did not reveal the presence of extraterrestrial Cr: the two Vredefort granophyre samples yielded 53Cr/52Cr values of ⳮ0.01
Ⳳ 0.06 e and Ⳮ0.03 Ⳳ 0.03 e (2rmean), which is indistinguishable from the terrestrial mean value (0 e) (Table 2). The average
Cr content of the Vredefort granophyre is ⬃420 ppm (Koeberl
et al., 1996; Reimold et al., 1999), whereas Cr contents in the
various target rocks range from 6 to 750 ppm; a possible average Cr content is ⬃300 ppm. Based on the Cr data, the upper
limit for a chondritic component in these samples is ⬃2% (Fig.
2), which is consistent with the better defined limit (⬃0.2%)
provided by the Os isotopes. An abundance of terrestrial Cr
masks the cosmic Cr and, in this case, the sensitivity of the Cr
Comparison of the osmium and chromium isotopic methods
613
TABLE 1. COMPOSITION OF VREDEFORT GRANOPHYRE AND COUNTRY ROCK AND
MOROKWENG DRILL CORE SAMPLES
Sample
Type
Vredefort samples
AV81-22
Granophyre
AV81-59
Granophyre
AV81-63
Granophyre
BG-7/2
Granophyre
BG-9
Granophyre
BG-4/2
Granophyre
BG-4/1
Granophyre
AV81-35
Alkali granite
AV81-5
Ventersdorp Lava
AV81-12
Witwatersrand shale
UP-63
Ventersdorp Lava
SNE
Witwatersrand shale
S1/2
Witwatersrand shale
Morokweng drill core samples
MO-15
Melt rock
MO-20
Melt rock
MO-35cl
Clast
MO-37a
Melt rock
MO-48
Melt rock
MO-63a
Breccia
MO-65
Granite
MO-69b
Breccia
MO-70
Breccia
1006
Mafic rock
1536.7
Mafic rock
2617.8
Mafic rock
Ir
(ppb)
0.13
0.094
0.105
n.d.
n.d.
n.d.
n.d.
0.0011
0.016
0.290
n.d.
n.d.
n.d.
9.6
7.7
11.2
4.0
10.4
0.029
0.016
0.004
0.016
0.014
⬍0.002
⬍0.002
Pt
(ppb)
Os
(ppb)
Os/188Os
187
n.d.
3.0
1.6
n.d.
n.d.
n.d.
n.d.
⬍0.5
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
0.111
0.228
0.356
1.11
n.d.
n.d.
n.d.
0.0158
0.162
0.0368
n.d.
n.d.
n.d.
0.558
0.410
0.286
0.196
n.d.
n.d.
n.d.
6.272
1.013
1.572
18.1
16.4
25.1
11.0
20.6
⬍0.040
0.907
⬍0.040
n.d.
0.0059
0.184
⬍0.040
7.87
6.31
8.11
3.45
8.84
0.34
0.29
0.035
0.036
0.042
0.032
0.009
0.1316
0.1341
0.1449
0.1319
0.1320
0.2027
0.2417
1.383
0.1283
9.533
12.41
0.6423
Note: Vredefort Os content and Os isotope data from Koeberl et al. (1996); Ir and Pt data from French
et al. (1989). All Morokweng data (this work) by isotope dilution magnetic sector (inductively coupled
plasma mass spectroscopy) (Finnigan Element). Os analyses done by sparging of OsO4 in Ar directly
into the ICP-MS (Hassler et al., 2000). n.d. ⳱ not determined.
isotopic method is not sufficient to resolve the meteoritic
component.
RESULTS: METEORITIC COMPONENT AT
MOROKWENG
With the exception of a few obviously altered melt rock
samples, the Morokweng melt body is extremely homogeneous
in composition. Variations for major elements do not exceed 2–
5 relative%. The Morokweng melt rock contains relatively high
proportions of CaO (on average, 3.41 wt%), MgO (3.70 wt%),
and Fe2O3 (5.87 wt%), and an average SiO2 content of 65.75
wt%. Siderophile elements are consistently enriched (the variation between samples is less than a factor of 2) in the meltrock samples in comparison with rocks of such major element
composition (granodioritic to dioritic), with average values of
440 ppm for Cr, 50 ppm for Co, 780 ppm for Ni, and 32 ppb
for Ir. No variation with depth and no differences between drill
cores were found (Koeberl et al., 1997; Reimold et al., 1999).
Koeberl et al. (1997) noted that it was highly unlikely that mafic
to ultramafic country rocks, or other mantle-derived sources,
were responsible for these high siderophile element and Ir concentrations. The abundances of the PGEs in the various Morokweng impact-melt rock samples analyzed were found to
have almost chondritic ratios, and, depending on the values
used for normalization, ⬃2–5 wt% of a chondritic component
is present in the melt rocks (Koeberl et al., 1997). In contrast
to melt rocks from most other impact structures (Koeberl,
1998), the meteoritic component has a high abundance and is
uniformly distributed in the Morokweng impact-melt rocks. To
confirm the presence of this component, we performed Cr and
Os isotopic studies, the preliminary results of which were reported in abstract form by Shukolyukov et al. (1999) and Koeberl et al. (2000), respectively.
We measured 12 samples from the MWF03, MWF04,
MWF05, and KHK-1 cores by magnetic sector ICP-MS for the
Os isotopic composition and the abundances of Ir, Os, and Pt.
The samples included four impact-melt rocks, one meltsaturated clast within a melt rock, three vein breccia samples
(Reimold et al., 1999), a granite clast, and three mafic target
rocks. Stratigraphic and petrographic information on these samples was given by Koeberl et al. (1997) and Reimold et al.
(1999, 2000). The analytical results are given in Table 1. The
elemental abundances of the three PGEs measured in this study
agree well with data from Koeberl et al. (1997), Andreoli et al.
(1999), and McDonald et al. (2001). Iridium abundances in the
melt-rock samples range up to 11.2 ppb (Fig. 1), and there are
corresponding high abundances of Os and Pt. PGE abundances
614
C. Koeberl et al.
in granitic and mafic target rocks are very low; Ir values in the
mafic rocks are ⬍2–14 ppt, and 4–29 ppt in the felsic rocks or
vein breccias, and abundances of Os and Pt are somewhat
higher (Table 1). The target rock values measured here are below the detection limit of all other studies cited above. The Ir
and Os data are almost identical to data obtained by NiS fire
assay and/or neutron activation analysis (Koeberl et al., 1997;
Koeberl and Reimold, 2002), but Pt data in the present data set
are somewhat lower. It is interesting that the Ir abundances
measured by instrumental neutron activation analysis in two
different laboratories (Koeberl et al., 1997; Koeberl and Rei-
Figure 3. 187Os/188Os vs. 187Re/188Os diagram for Vredefort granophyre samples and some country rocks that are assumed to be representative of target rocks, compared to field for chondrites and iron
meteorites (after Koeberl et al., 1996). All but one of Vredefort granophyre samples plot close to 2 Ga reference isochron, which intersects
meteorite data array. This shows that Vredefort granophyre samples
have narrow range of initial 187Os/188Os (at 2 Ga) ratios that overlap
range of meteoritic initial ratios, which, in absence of any significant
mantle rock component in Vredefort granophyre (Koeberl et al., 1996),
confirms presence of meteoritic component in these samples, and that
Vredefort granophyre is impact-melt rock.
mold, 2002; Andreoli et al., 1999) are higher by a factor of
two or more compared to the ICP-MS data. The chondritenormalized abundance patterns of the PGEs (and other siderophile elements), however, have similar shapes, at elevated
absolute abundances. The reasons for this discrepancy are
unknown, but other comparative studies (neutron activation
versus ICP-MS) also found a similar result (Dai et al., 2001).
Nevertheless, the chondrite-normalized abundance patterns of
the impact-melt rock PGE data given in Table 1 are flat and
indicate the presence of ⬃2% of a meteoritic component,
whereas the normalized patterns of the various felsic and mafic
target rocks are highly fractionated (for a more detailed discussion, see Koeberl and Reimold, 2002).
The results for Os show high contents of as much as ⬃9
ppb in the impact-melt rock samples, and correspondingly low,
but remarkably uniform, isotopic ratios (187Os/188Os 0.1316–
0.1341). The breccias show a much wider variation in both
isotope ratio and Os abundance. One of the breccias has a low
Os abundance, but a fairly low isotope ratio. Considering that
this breccia (MO70) is of granitic composition, and other granites or granitic breccias may have lower PGE contents, it is
possible that this sample contains a small, but dominating, extraterrestrial Os component. Figure 4 shows the Os content versus the Os isotopic composition. The MO70 breccia sample
plots close to the hyperbolic mixing line that connects the meteorite and target rock fields. The impact-melt rock samples plot
close to this line at higher meteorite contribution values. The
mafic rock samples from the KHK-1 deep drill core, representing the mafic contribution to the Morokweng impact-melt
rocks, have very low Os abundances (9–42 ppt) and high isotope ratios (187Os/188Os to 12). This result clearly indicates that
the mafic target rocks of the area did not contribute measurably
TABLE 2. CHROMIUM ISOTOPE DATA FOR VREDEFORT
GRANOPHYRE AND MOROKWENG DRILL CORE SAMPLES
Sample
Type
Cr
(ppm)
Cr/52Cr
(e)
53
Vredefort samples
AV81-22
Granophyre
BG-9
Granophyre
BG-4/2
Granophyre
AV81-35
Alkali granite
UP-63
Ventersdorp Lava
SNE
Witwatersrand shale
S1
Witwatersrand shale
394
419
429
2.7
223
142
138
n.d.
ⳮ0.01 Ⳳ 0.06
Ⳮ0.03 Ⳳ 0.03
n.d.
n.d.
n.d.
n.d.
Morokweng drill core samples
MO-15
Melt rock
MO-48
Melt rock
359
408
Ⳮ0.24 Ⳳ 0.04
Ⳮ0.27 Ⳳ 0.03
Note: Vredefort Cr content from Koeberl et al. (1996); Morokweng Cr
content from Koeberl and Reimold (2002). n.d. ⳱ not determined.
Figure 4. Plot of Os isotope ratio vs. Os abundance for 12 samples
from Morokweng impact structure, South Africa. Target rocks (including those from deep drill core outside of structure) define low Os
abundance, radiogenic field, whereas impact-melt rocks have clear meteoritic component (high Os concentration, nonradiogenic). Mixing
between two components follows mixing hyperbola (or set of hyperbolas) that passes close to vein breccia point just below target rock
field.
Comparison of the osmium and chromium isotopic methods
to the high siderophile element abundances observed in the melt
rocks. It is evident that the Os isotope data confirm the presence
of a meteoritic component in these melt rocks.
Because the meteoritic component in the Morokweng
impact-melt rocks seems to be fairly abundant, we performed
a Cr isotopic study. This study had the following goals: first,
we wanted to confirm the extraterrestrial nature of the siderophile element anomaly indicated by the PGE data of Koeberl
et al. (1997) and Hart et al. (1997), and second, to use the
potential of the Cr isotope signal to provide information about
the type of projectile involved. The results (Table 2) indicated
that about half of the Cr in the melt rock is of extraterrestrial
origin. It was also possible to show that only an ordinary chondritic source, rather than a carbonaceous chondritic source, can
explain the Cr isotope data. A similar conclusion, based only
on PGE abundance data, was obtained by Koeberl et al. (1997),
who concluded that the PGE data fit best an ordinary chondrite
source (probably H-chondrite). However, using the Cr isotope
data together with trace element data (Fig. 5; Shukolyukov et
al., 1999), we conclude that the Morokweng bolide was most
likely of L-chondritic composition. This interpretation agrees
well with the conclusions based on PGE abundances (McDonald et al., 2001).
CONCLUSIONS
It is clear that there are advantages and disadvantages of
both the Os and Cr isotopic methods for the determination of
a meteoritic component in impact breccias, relative to each
other and in comparison with PGE abundance measurements.
PGE abundance studies may indicate the presence of an extra-
615
terrestrial component, and the normalized patterns can give an
indication of the impactor type. However, it is not possible to
distinguish between terrestrial and extraterrestrial PGEs. This
determination is possible in principle by both the Os and Cr
isotopic methods. In the Os isotopes, it is necessary to distinguish between a possible mantle signal and an extraterrestrial
signal, because the Os isotopic ratios for mantle and meteoritic
material are very similar. There is a significant difference in the
total abundance of Os in mantle rocks (⬃1–4 ppb Os) and meteorites (typical chondrites have ⬃400–800 ppb Os). Thus, at
least 100 times more mantle than meteoritic material needs to
be added to normal crustal rocks (e.g., in an impact breccia) to
result in the same Os isotopic ratio of the bulk rock. Detailed
field investigations and petrographic studies, if necessary combined with trace element and Rb-Sr and/or Sm-Nd isotopic
analyses, will show if significant amounts of ultramafic materials are present.
However, the Cr isotopic composition allows the distinction between terrestrial and extraterrestrial Cr, and there is a
difference between some meteorite groups (Shukolyukov and
Lugmair, 1998). For example, ordinary chondrites (types H and
L) have isotopic characteristics that are slightly different from
those of enstatite chondrites and are, even more significantly,
different from those of carbonaceous chondrites. Thus, assuming that a certain (fairly significant) percentage of the chromium
in an impact melt of an ejecta layer is of extraterrestrial origin,
it should be possible by measuring the Cr isotopic composition
to determine not only that an extraterrestrial component is present, but also which meteorite type might have been involved.
This determination is not possible for iron meteorites, which do
not carry significant amounts of Cr, but basaltic achondrites
(such as the eucrites, which have very low PGE abundances
and, thus, low Os) have high Cr abundances. Therefore, the Cr
isotopic method may be the only reasonable way to identify
achondritic impactors. The analytical effort for Cr isotope measurements is substantial. The differences in isotopic composition are very small, requiring extremely precise and timeconsuming measurements, severely limiting the number of
samples that can be measured. In addition, a substantial amount
of chromium has to be of extraterrestrial origin to show an
effect in the Cr isotopic composition that is larger than the precision of the measurement. Thus, while being more selective
than the Os isotopic method, the Cr isotopic method is less
sensitive.
ACKNOWLEDGMENTS
Figure 5. Correlation between Cr and Ir contents in impact-melt rock
samples from Morokweng impact structure. Position of projectile point
was determined using concentration of Ir (from Koeberl et al., 1997;
Koeberl and Reimold, 2002) and proportion of cosmic Cr (obtained
from Cr isotopic composition) in samples MO15 and MO48. Cr and
Ir concentrations for ordinary chondrites are from Wasson and Kallemeyn (1988).
Supported by the Austrian Fonds zur Förderung der wissenschaftlichen Forschung, Y58-GEO (Koeberl), the U.S. National
Science Foundation (Peucker-Ehrenbrink), the National Research Foundation of South Africa (Reimold), and NASA grant
5-8172 (Shukolyukov and Lugmair). We thank Anglogold Limited and their Consulting Geologist G. Cantello for permission
to study the KHK-1 borehole and to publish these results. We
616
C. Koeberl et al.
also thank M. Humayun, F.T. Kyte, and K. MacLeod for helpful
suggestions on the manuscript, and Kyte for the hint that a
eucrite signal could be detected by using the Cr isotopic
method. This is Wits Impact Cratering Research Group contribution 16, and Woods Hole Oceanographic Institution contribution 10410.
REFERENCES CITED
Alvarez, L.W., Alvarez, W., Asaro, F., and Michel, H.V., 1980, Extraterrestrial
cause for the Cretaceous-Tertiary extinction: Science, v. 208, p. 1095–
1108.
Andreoli, M.A.G., Ashwal, L.D., Hart, R.J., and Huizenga, J.M., 1999, A Niand PGE-enriched quartz norite impact melt complex in the Late Jurassic
Morokweng impact structure, South Africa, in Dressler, B.O., and Sharpton, V.L., eds., Large meteorite impacts and planetary evolution 2, Geological Society of America Special Paper 339, p. 91–108.
Corner, B., Reimold, W.U., Brandt, D., and Koeberl, C., 1997, Morokweng
impact structure, Northwest Province, South Africa: Geophysical imaging
and some preliminary shock petrographic studies: Earth and Planetary
Science Letters, v. 146, p. 351–364.
Creaser, R.A., Papanastassiou, D.A., and Wasserburg, G.J., 1991, Negative
thermal ion mass spectrometry of osmium, rhenium and iridium: Geochimica et Cosmochimica Acta, v. 55, p. 397–401.
Dai, X., Koeberl, C., and Fröschl, H., 2001, Determination of PGEs in impact
melt breccias using NAA and USN-ICP-MS after anion exchange preconcentration: Analytica Chimica Acta, v. 436, p. 79–85.
Fehn, U., Teng, R., Elmore, D., and Kubik, P.W., 1986, Isotopic composition
of osmium in terrestrial samples determined by accelerator mass spectrometry: Nature, v. 323, p. 707–710.
French, B.M., and Nielsen, R.L., 1990, Vredefort bronzite granophyre: Chemical evidence for an origin as meteorite impact melt: Tectonophysics, v.
171, p. 119–138.
French, B.M., Orth, C.J., and Quintana, C.R., 1989, Iridium in the Vredefort
bronzite granophyre: Impact melting and limits on a possible extraterrestrial component, in Proceedings, 19th Lunar and Planetary Science Conference: New York, Cambridge University Press, p. 733–744.
Grieve, R.A.F., Coderre, J.M., Robertson, P.B., and Alexopoulos, J., 1990, Microscopic planar deformation features in quartz of the Vredefort structure:
Anomalous but still suggestive of an impact origin: Tectonophysics, v.
171, p. 185–200.
Hart, R.J., Andreoli, M.A.G., Tredoux, M., Moser, D., Ashwal, L.D., Eide,
E.A., Webb, S.J., and Brandt, D., 1997, Late Jurassic age for the Morokweng impact structure, southern Africa: Earth and Planetary Science Letters, v. 147, p. 25–35.
Hassler, D.R., Peucker-Ehrenbrink, B., and Ravizza, G.E., 2000, Rapid determination of Os isotopic composition by sparging OsO4 into a magneticsector ICP-MS: Chemical Geology, v. 166, p. 1–14.
Henkel, H., and Reimold, W.U., 1998, Integrated geophysical modelling of a
giant, complex impact structure: Anatomy of the Vredefort Structure,
South Africa: Tectonophysics, v. 287, p. 1–20.
Kamo, S.L., Reimold, W.U., Krogh, T.E., and Colliston, W.P., 1996, A 2.023
Ga age for the Vredefort impact event and a first report of shock metamorphosed zircons in pseudotachylitic breccias and Granophyre: Earth
and Planetary Science Letters, v. 144, p. 369–388.
Koeberl, C., 1998, Identification of meteoritical components in impactites, in
Grady, M.M., Hutchison, R., McCall, G.J.H., and Rothery, D.A., eds.,
Meteorites: Flux with time and impact effects: Geological Society [London] Special Publication 140, p. 133–152.
Koeberl, C., and Reimold, W.U., 2002, Geochemistry and petrography of impact breccias and target rocks from the 145 Ma Morokweng Impact Structure, South Africa: Geochimica et Cosmochimica Acta, v. 66, in press.
Koeberl, C., and Shirey, S.B., 1993, Detection of a meteoritic component in
Ivory Coast tektites with rhenium-osmium isotopes: Science, v. 261, p.
595–598.
Koeberl, C., and Shirey, S.B., 1997, Re-Os systematics as a diagnostic tool for
the study of impact craters and distal ejecta: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 132, p. 25–46.
Koeberl, C., Reimold, W.U., and Shirey, S.B., 1996, A Re-Os isotope and
geochemical study of the Vredefort Granophyre: Clues to the origin of
the Vredefort structure, South Africa: Geology, v. 24, p. 913–916.
Koeberl, C., Armstrong, R.A., and Reimold, W.U., 1997, Morokweng, South
Africa: A large impact structure of Jurassic-Cretaceous boundary age:
Geology, v. 25, p. 731–734.
Koeberl, C., Peucker-Ehrenbrink, B., and Reimold, W.U., 2000, Meteoritic
component in impact melt rocks from the Morokweng, South Africa,
impact structure: An Os isotopic study: Lunar and Planetary Science, v.
31, Abstract #1595, CD-ROM.
Leroux, H., Reimold, W.U., and Doukhan, J.C., 1994, A T.E.M. investigation
of shock metamorphism in quartz from the Vredefort dome, South Africa:
Tectonophysics, v. 230, p. 223–239.
Luck, J.M., and Turekian, K.K., 1983, Osmium-187/Osmium-186 in manganese nodules and the Cretaceous-Tertiary boundary: Science, v. 222, p.
613–615.
Lugmair, G.W., and Shukolyukov, A., 1998, Early solar system timescales according to 53Mn-53Cr systematics: Geochimica et Cosmochimica Acta, v.
62, p. 2863–2886.
McDonald, I., Andreoli M.A.G., Hart, R.J., and Tredoux, M., 2001, Platinumgroup elements in the Morokweng impact structure, South Africa: Evidence for the impact of a large ordinary chondrite projectile at the
Jurassic-Cretaceous boundary: Geochimica et Cosmochimica Acta, v. 65,
p. 299–309.
Morgan, J.W., 1978, Lunar crater glasses and high-magnesium australites:
Trace element volatilization and meteoritic contamination: Proceedings
of the 9th Lunar and Planetary Science Conference: New York, Pergamon
Press, p. 2713–2730.
Palme, H., 1982, Identification of projectiles of large terrestrial impact craters
and some implications for the interpretation of Ir-rich Cretaceous/Tertiary
boundary layers, in Silver, L.T., and Schultz, P.H., eds., Geological implications of impacts of large asteroids and comets on Earth: Geological
Society of America Special Paper 190, p. 223–233.
Palme, H., Janssens, M.-J., Takahasi, H., Anders, E., and Hertogen, J., 1978,
Meteorite material at five large impact craters: Geochimica et Cosmochimica Acta, v. 42, p. 313–323.
Palme, H., Göbel, E., and Grieve, R.A.F., 1979, The distribution of volatile and
siderophile elements in the impact melt of East Clearwater (Quebec):
Proceedings of the 10th Lunar and Planetary Science Conference: New
York, Pergamon Press, p. 2465–2492.
Palme, H., Grieve, R.A.F., and Wolf, R., 1981, Identification of the projectile
at the Brent crater, and further considerations of projectile types at terrestrial craters: Geochimica et Cosmochimica Acta, v. 45, p. 2417–2424.
Papike, J., editor, 1998, Planetary materials: Reviews in Mineralogy, v. 36,
915 p.
Ravizza, G., and Pyle, D., 1997, PGE and Os isotopic analyses of single sample
aliquots with NiS fire assay preconcentration: Chemical Geology, v. 141,
p. 251–268.
Reimold, W.U., 1993, A review of the geology of and deformation related to
the Vredefort Structure, South Africa: Journal of Geological Education,
v. 41, p. 106–117.
Reimold, W.U., and Gibson, R.L., 1996, Geology and evolution of the Vredefort impact structure, South Africa: Journal of African Earth Sciences,
v. 23, p. 125–162.
Reimold, W.U., and Koeberl, C., 1999, The deep borehole into the Morokweng
impact structure [abs.]: Meteoritics and Planetary Science, v. 34, p. A97.
Reimold, W.U., Horsch, H., and Durrheim, R.J., 1990, The “Bronzite”-Granophyre from the Vredefort structure: A detailed analytical study and re-
Comparison of the osmium and chromium isotopic methods
flections on the origin of one of Vredefort’s enigmas: Proceedings of the
20th Lunar and Planetary Science Conference: Houston, Texas, Lunar
and Planetary Institute, p. 433–450.
Reimold, W.U., Koeberl, C., Brandstätter, F., Kruger, F.J., Armstrong, R.A.,
and Bootsman, C., 1999, The Morokweng impact structure, South Africa:
Geologic, petrographic, and isotopic results, and implications for the size
of the structure, in Dressler, B.O., and Sharpton, V.L., eds., Large meteorite impacts and planetary evolution 2: Geological Society of America
Special Paper 339, p. 61–90.
Reimold, W.U., Armstrong, R.A., and Koeberl, C., 2000, New results from the
deep borehole at Morokweng, North West Province, South Africa: Constraints on the size of the J/K boundary age impact structure: Lunar and
Planetary Science, v. 31, Abstract #1074, CD-ROM.
Shukolyukov, A., and Lugmair, G.W., 1998, Isotopic evidence for the
Cretaceous-Tertiary impactor and its type: Science, v. 282, p. 927–929.
Shukolyukov, A., and Lugmair, G.W., 2000, Extraterrestrial matter on Earth:
Evidence from the Cr isotopes [abs.], in Catastrophic events and mass
extinctions: Impacts and beyond: Houston, Texas, Lunar and Planetary
Institute, LPI Contribution No. 1053, p. 197–198.
Shukolyukov, A., Lugmair, G.W., Koeberl, C., and Reimold, W.U., 1999, Chromium in the Morokweng impact melt rocks: Isotope evidence for extra-
617
terrestrial component and type of the impactor [abs.]: Meteoritics and
Planetary Science, v. 34, p. A107–A108.
Shukolyukov, A., Kyte, F.T., Lugmair, G.W., Lowe, D.R., and Byerly, G.R.,
2000, The oldest impact deposits on earth: First confirmation of an extraterrestrial component. in Gilmour, I., and Koeberl, C., eds., Impacts
and the early earth: Heidelberg, Germany, Springer-Verlag, Lecture Notes
in Earth Sciences, v. 91, p. 99–116.
Taylor S.R., and McLennan, S.M., 1985, The continental crust: Its composition
and evolution: Oxford, Blackwell, p. 57–72.
Völkening, J., Walczyk, T., and Heumann, K.G., 1991, Osmium isotope ratio
determinations by negative thermal ionization mass spectrometry: International Journal of Mass Spectrometry and Ion Processes, v. 105, p. 147–
159.
Wasson, J.T., 1985, Meteorites: New York, W.H. Freeman and Company, 267 p.
Wasson, J.T., and Kallemeyn, G., 1988, Compositions of chondrites: Philosophical Transactions of the Royal Society of London, ser. A, v. 325, p.
535–544.
MANUSCRIPT SUBMITTED OCTOBER 24, 2000; ACCEPTED
MARCH 22, 2001
BY THE
SOCIETY
Printed in the U.S.A.