Abstract– We report the 53Mn-53Cr systematics of three dolomite grains from two different CI1 clasts contained within the Kaidun meteorite breccia. Three internal isochrones result in initial 53Mn/55Mn ratios of (4.2 ± 0.4) × 10−6, (4.6 ± 1.3) × 10−6, and (5.2 ± 1.1) × 10−6. These initial values are consistent with those measured for dolomite in the Orgueil CI1 chondrite (Hoppe et al. 2007; Petitat et al. 2009) but significantly lower than the initial ratio determined by Hutcheon et al. (1999) from a combination of different carbonate types within various lithologies of the Kaidun meteorite. We construct an accretion scenario for the Kaidun breccia by comparing the mineralogy and formation time scales of carbonates in the Kaidun CI1 lithologies to the analogous ones of the CI1 chondrite Orgueil. In Orgueil, dolomite precipitation precedes the formation of the first bruennerite grains by a few million years (Hoppe et al. 2007; Petitat et al. 2009). As the CI1 clasts in Kaidun lack breunnerite grains, and considering that aqueous alteration occurred prior to reaccretion of the various clasts onto the Kaidun parent body (e.g., MacPherson et al. 2009), we hypothesize that after rapid accretion and early aqueous alteration occurring within the first approximately 4 Myr after solar system formation, impact disruption of several asteroids and their reassembly into the Kaidun parent asteroid was complete within an additional approximately 2 Myr. This confirms that aqueous alteration, impact, and reaccretion of material in the asteroid belt were early processes that began contemporaneously with chondrule formation.
The Kaidun meteorite is a complex polymict breccia containing lithic clasts spanning a wide range of chondrite groups, including enstatite, ordinary, and carbonaceous chondrites. Also present in Kaidun are new types of CI1- and CM2-like lithologies, enstatite bearing clasts, Ca-rich achondrite material, impact melt products, and phosphide-bearing clasts (Zolensky and Ivanov 2003). Apparently, the Kaidun parent body accumulated materials from a wide region of the asteroid belt. Many of these materials were subjected to varying levels of physical processing such as heating, shock, melting, and aqueous alteration.
Major carbonaceous chondrite lithologies present in the Kaidun meteorite breccia are the CV3, CM2, CR2, and CI1 lithologies. They all contain carbonates, derived from aqueous alteration. Given that many other clasts are anhydrous, it has been proposed that aqueous alteration probably occurred before compaction (Zolensky and Ivanov 2003; MacPherson et al. 2009).
The extinct 53Mn-53Cr isotope system is based on the beta-decay of 53Mn to 53Cr (t1/2 = 3.7 Myr). The relatively long half-life of 53Mn among the now-extinct radionuclides, along with the fact that Mn and Cr are reasonably abundant elements (Lodders 2003), makes the 53Mn-53Cr decay system an important chronological tool to explore the time period from nebular events to accretion, differentiation, and parent-body processes such as aqueous alteration (Krot et al. 2006; Nyquist et al. 2009). Time scales of aqueous alteration can be deduced from carbonates, which crystallized from aqueous fluids on meteorite parent bodies (e.g., Endress et al. 1996; Hoppe et al. 2007; Petitat et al. 2009). In the first and only 53Mn-53Cr chronological study of Kaidun carbonates, Hutcheon et al. (1999) used an ion probe (ims 3f) to analyze a combination of several types of carbonate grains from three different lithologies. They obtained a good correlation of δ53Cr values with 55Mn/52Cr ratios implying an initial 53Mn/55Mn ratio of (9.1 ± 1.4) × 10−6 at the time of carbonate formation. As discussed further below, this 53Mn/55Mn ratio is identical to estimates of the initial 53Mn/55Mn ratio of the solar system (Shukolyukov and Lugmair 2006; Moynier et al. 2007; Nyquist et al. 2009) raising important issues regarding the start and duration of carbonate formation on meteorite parent bodies and the formation of the Kaidun meteorite.
In this study, we performed NanoSIMS analyses to determine 53Mn-53Cr internal isochrons on three dolomite grains in two CI1 clasts of the Kaidun meteoritic breccia to (1) date the formation time of individual carbonate grains found in the CI1 lithology of Kaidun and compare with the previous data of Hutcheon et al. (1999), and (2) constrain the time of compaction of the Kaidun parent body.
The NanoSIMS at the Muséum National d’Histoire Naturelle (MNHN) in Paris was used for 53Mn-53Cr analyses by rastering a 1–3 nA 16O− primary beam over a 5 × 5 μm2 area on the polished sample. Raster analysis mode was preferred to spot analysis mode because during raster analyses, the 55Mn+ signal remained stable over time, whereas during spot analyses, the 55Mn+ count rates decreased with time. The secondary ion intensities of 52Cr+, 53Cr+, and 55Mn+ were measured by coupling multicollection to magnetic field peak-switching (two-step analysis based on two distinct magnet field settings) at high mass resolution (∼11,000 for 52Cr, ∼9000 for 53Cr, and ∼7000 for 55Mn) sufficient to resolve all molecular ion interferences, including hydrides.
The conversion of ion signals measured by secondary ion mass spectrometry (SIMS) to the 55Mn/52Cr and 53Cr/52Cr ratios in the sampled volume requires determination of a relative sensitivity factor (RSF) for the elemental ratio and of the instrumental mass fractionation (IMF) for the isotope ratio. As with all analytical techniques, this is accomplished by the measurement of standard materials made under the same analytical conditions as those used to measure the samples. In SIMS, there exists a complication whereby the ionization probability for any particular isotope can depend on the mineralogical and chemical composition of the sample, thereby potentially giving rise to a so-called “matrix effect” in the RSF and/or IMF calibration. An accurate determination of these parameters therefore requires the use of standards, which “closely” match the physical and chemical composition of the unknown analyzed. This represents a significant challenge for Mn-Cr analyses of carbonates, as has been recognized by several groups (e.g., Hutcheon et al. 1990; Hoppe et al. 2007). The problem is that it is not possible to find any natural carbonate that can be used as a standard for SIMS Mn-Cr analyses as, in all samples investigated so far, their Cr concentrations vary at the micrometer scale (Hoppe et al. 2007). Consequently, one needs to rely on other materials with homogeneous Mn/Cr to calculate a RSF, and we, like most other studies, have chosen to use silicate minerals and glasses for this purpose.
In previous 53Mn-53Cr investigations of extraterrestrial carbonates, NBS 611 (synthetic glass) and San Carlos Olivine (SCO; natural sample) were used as reference materials (Endress et al. 1996; Hutcheon and Phinney 1996; Brearley et al. 2001; Brearley and Hutcheon 2002; Hoppe et al. 2007; de Leuw et al. 2009). Other reference materials such as the Orgueil matrix have also been analyzed (Hoppe et al. 2007). The true Mn/Cr ratios of the NBS 611 and the SCO are 1.20 (GeoRem; Max Planck Institute database) and 7.80 (Ito and Ganguly 2006), respectively. However, carbonates in meteorites show elevated Mn/Cr ratios (e.g., Hoppe et al. 2007). Hence, we chose two glass standards having relatively high Mn/Cr ratios in addition to the NBS 611 and the SCO standards for the NanoSIMS measurements. T1-G is a silicate glass with Mn/Cr = 50.5 (GeoRem; Max Planck Institute database) and ATHO-G has a rhyolite composition with Mn/Cr = 130.8 (GeoRem; Max Planck Institute database). By incorporating these two new standards, a wider range of Mn/Cr ratios was covered (see Table 1), which is more comparable to the range of Mn/Cr of the meteoritic carbonates.
Table 1. Elemental compositions of the standards used in this study. They were used to calculate the relative sensitivity factor and the IMF of the instrument.
Cr (μg g−1)
Mn (μg g−1)
Ti (μg g−1)
V (μg g−1)
CaO (μg g−1)
Note: The composition data for ATHO-G, NBS 611, and T1-G are taken from the GeoRem database. The composition data for San Carlos Olivine are taken from Ito and Ganguly (2006).
ATHO-G (rhyolite, glass)
6.1 ± 1.4
798 ± 8
1470 ± 30
3.9 ± 0.3
1.7 ± 0.1
NBS 611 (silicate, glass)
394 ± 9
485 ± 10
426 ± 1
San Carlos Olivine
T1-G (silicate, glass)
20.9 ± 2
1010 ± 30
4400 ± 100
190 ± 11
7.1 ± 0.1
On a plot of the measured ion ratios against the known atomic ratio, the four standards are linearly correlated resulting in an RSF of 0.95 ± 0.09 (Fig. 1). This RSF is identical within errors to the one obtained from analyses of NBS611 and SCO (Hoppe et al. 2007) and also to the RSF recently obtained from a synthetic calcite standard (Fig. 2) (Sugiura et al. 2010). The latter result is approximately 15% lower than the RSF calculated from the calibration line in this study; however, the error bars overlap and the Sugiura et al. (2010) results are still considered preliminary and were made under somewhat different analytical conditions than those used here. Indeed, the 55Mn+, 52Cr+, and 53Cr+ count rates measured decrease with time demonstrating either a charging effect of the calcite standard or a changing ion extraction efficiency in their (spot) analysis mode. Therefore, although we recognize that a matrix effect could be present that would result in a systematic inaccuracy in our Mn/Cr determinations in carbonate, we note that the preliminary agreement in the RSF for carbonates and silicates at the approximately 15% level is reasonably close and, to compare easily with previous SIMS results obtained on carbonates, we will use here the RSF determined on our silicate standards. If future work demonstrates a clear matrix effect between silicates and carbonates in the Mn-Cr RSF, then the slopes of all SIMS-measured isochrones will need to be adjusted by this (approximately constant) factor.
As the IMF remained constant over the entire measurement session (Fig. 3), it was corrected externally, i.e., 53Cr+/52Cr+ ratios in the meteoritic carbonates were normalized by assuming that the IMF measured for the four standards during the same analytical session applied also to the Kaidun carbonates. The 53Cr excesses are reported as δ53Cr, expressed as the deviation, in parts per mil, from the reference 53Cr/52Cr value of 0.113457 ± 0.000001 (Birck and Allègre 1988). Given the magnitude of 53Cr excesses observed, possible matrix effects on the mass fractionation correction are negligible.
At the start of each measurement session, at least two standard measurements (NBS 611 and T1-G) with various 55Mn/52Cr ratios were run to evaluate the day-to-day reproducibility of the mass spectrometer. The 55Mn/52Cr ratios and δ53Cr values remained within approximately 3% and approximately 5‰, respectively. Isotope count rates were corrected for dynamic background and dead time. The dynamic background was quantified for each different sample by applying a −10 V offset relative to the peak center on the deflection plates in front of each electron multiplier (this is sufficient to remove the peak entirely from the detector). The averaged magnitude of the background (0.015 ± 0.008 cps) relative to the Cr counting rates is negligible as it only represents approximately 0.1% of the lowest 53Cr+ count rate. Prior to data collection, the analytical spot was sputtered with the primary beam for approximately 300 s, to stabilize the secondary ion signal and to remove possible surface contamination. Uncertainties are expressed as 2σ, taking into account the external reproducibility on the standards. The errors of the mean elemental 55Mn/52Cr ratios were calculated using the error propagation equation (Bevington and Robinson 2002).
The mineral chemistry of individual Kaidun carbonates was determined at MNHN and University Paris 7 (Jussieu) by using conventional secondary electron microscope (SEM) and electron microprobe analysis (EMPA) techniques.
Three Kaidun dolomites, embedded in two different CI1 lithic clasts within the Kaidun_3.10.i and Kaidun_cavity sections, were analyzed. Both lithologies are similar in mineral composition and size to the CI1 clasts described by Zolensky and Ivanov (2003). The Kaidun_3–10.i section (Fig. 4) consists of an enstatite chondrite (EC), a CM2, and a CI1 clast. The typical CI1 matrix is dominated by Mg-rich serpentine and saponite (this work; Zolensky and Ivanov 2003). Pyrrhotite, pentlandite, and magnetite are very abundant. The absence of anhydrous silicates, chondrules, and calcium-aluminum-rich inclusions (CAIs) and the presence of secondary minerals (i.e., carbonates) indicate a high degree of aqueous alteration.
The three carbonate grains analyzed range from 30 to 40 μm in diameter, and occur as isolated matrix grains, irregular in shape and sometimes showing an association with magnetite (Fig. 5). The MgO contents of these dolomites range from 17.0 to 21.2 wt% and their CaO contents reach 28.1 wt%. They contain significant amounts of iron (around 3.3 wt% FeO) and varying amounts of MnO (between 1.2 and 4.1 wt%). The elemental compositions of the three Kaidun dolomites analyzed plot within the compositional range of CI1 chondrite dolomites (Endress and Bischoff 1996; Hoppe et al. 2007; Petitat et al. 2009) (Fig. 6).
The isotopic and elemental ratios measured for different areas in the different carbonates are presented in Table 2. Kaidun_3.10.i-2, Kaidun_3.10.i-3, and Kaidun_cavity-5 carbonates were analyzed, respectively, at two, three, and three different regions, each of them showing large enrichments in 53Cr relative to 52Cr with δ53Cr up to approximately 600‰. Their respective 55Mn/52Cr ratios range up to 14,000 and are linearly correlated with δ53Cr constituting strong evidence for in situ 53Mn decay (Fig. 7). Within a single analysis, the chromium content often varied at a micrometer scale resulting in a drop or an increase in the elemental and/or isotopic ratios. When during one measurement, the isotopic or elemental ratios differed by at least 20% of the initial ratio measured, they were then divided into different submeasurements, e.g., the internal isochron of Kaidun_3.10.i-2 is made up of three points even though measurements were performed in only two distinct areas. A best-fit line forced through the origin yields a slope corresponding to initial 53Mn/55Mn ratios at the time of carbonate formation of (4.20 ± 0.43) × 10−6, (4.55 ± 1.32) × 10−6, and (5.16 ± 1.07) × 10−6 for the three samples analyzed.
Table 2. Internal Mn-Cr results for three dolomite grains within two CI1 clasts in the Kaidun meteorite breccia.
(53Mn/55Mn) × 10−6
2σ × 10−6
Note: δ53Cr values are corrected for IMF and elemental 55Mn/52Cr ratios with the relative sensitivity factor (RSF) from the calibration line. Error bars are 2σ, but do not include the possible error on the RSF of approximately 10–15% (see Hoppe et al. 2007). Measurements that have been divided into different submeasurements are represented in bold.
Assuming an homogeneous distribution of 53Mn at the start of the solar system (for another view, see Gounelle and Russell 2005), relative 53Mn-53Cr ages (Δt) can be calculated from the initial 53Mn/55Mn ratios obtained from the slopes of the isochrones according to
To determine the absolute time scale of carbonate crystallization, the initial 53Mn/55Mn ratio of the solar system must be known. CAIs are considered to represent the first rocks formed in the solar system and as such have been used to anchor some extinct radionuclide chronometers. However, it has been known for a long time that the 53Mn-53Cr systematics has been disturbed in CAIs and hence it is not possible to use them to establish the initial 53Mn/55Mn ratio of the solar system (Birck and Allègre 1988; Papanastassiou et al. 2005). Hence, one needs to rely on other samples and correlations with other long- and short-lived chronometers to estimate the initial 53Mn/55Mn ratio of the solar system (Birck and Allègre 1988; Shukolyukov and Lugmair 2006; Yin et al. 2007). This problem has been recently considered in detail by Nyquist et al. (2009) who derived a best estimate for (53Mn/55Mn)0 of the solar system as (9.1 ± 1.7) × 10−6 based primarily on relative concordancy of Al-Mg and 207Pb/206Pb inferred ages for the D’Orbigny-clan angrites compared with these systems in CAIs. This value is compatible with initial ratios of 53Mn/55Mn = (8.5 ± 1.5) × 10−6 based on the analysis of a suite of carbonaceous chondrites (Shukolyukov and Lugmair 2006; Moynier et al. 2007), which might represent nebular Mn-Cr fractionation. Thus, we take the value of Nyquist et al. (2009) of 9.1 × 10−6 as an estimate of the (53Mn/55Mn)0 at 4568 (±1) Ma from which we calculate relative formation ages assuming homogeneity of 53Mn/55Mn across the nebular zones where these meteoritic components formed and accreted into early planetesimals.
Under these assumptions, dolomite in the CI1 lithology of Kaidun formed within the first 4 Myr after the start of the solar system, at 3.0 ± 0.9, 3.7 ± 1.3, and 4.1 ± 0.9 Myr, respectively. As previously suggested (e.g., Endress et al. 1996), this implies that aqueous alteration initiated very early after the start of the solar system. However, our data solve a problem of formation of carbonates contemporaneous with CAIs (Fig. 9), which was implied by the bulk isochrone obtained by Hutcheon et al. (1999). As there is petrographic evidence that aqueous alteration occurred before final assembly of Kaidun (MacPherson et al. 2009), this suggests that break-up of a CI-like protolith and accretion onto the Kaidun parent body took place at least 3 Myr after the start of the solar system.
Constraints on the time scales of accretion are important for dynamical models of the early solar system, but can be difficult to obtain from primitive meteorites since most chemical fractionation events that can be dated from these objects occurred in the nebula. The study of meteoritic breccia offers the possibility to gain new insights into accretion and differentiation of early-formed planetesimals (Bischoff et al. 2006). The Kaidun breccia contains carbonaceous chondrite (CI1, CM2, CR2, CO3, CV3), ordinary chondrite, and EC lithologies as well as differentiated meteorite components. It remains unknown when these different meteorite groups formed, but clearly, they (or very similar materials) underwent different histories on their respective parent bodies before accretion and compaction into the Kaidun meteorite.
Different meteorite groups show very different degrees of secondary (i.e., asteroidal) processing, such as aqueous alteration, with the CI1 chondrite Orgueil representing the most aqueous altered sample that we currently have in meteorite collections. Among the minerals formed by aqueous alteration in Orgueil, dolomite ((Ca,Mg)CO3) is the most abundant carbonate type followed by breunnerite ((Mg,Fe)CO3) and finally calcite (CaCO3) (e.g., Endress and Bischoff 1996). The dolomites and calcites in Orgueil occur as randomly distributed grains, whereas breunnerites are both randomly distributed grains and aggregates within the matrix (e.g., Frederiksson and Kerridge 1988; Endress and Bischoff 1996; Petitat and Gounelle 2010). Based on oxygen and carbon isotopic compositions, Zito et al. (1998) postulated that calcite precipitated first, followed by dolomite and finally breunnerite. This hypothesis is supported by 53Mn-53Cr chronology of Orgueil carbonates (Petitat et al. 2009) with dolomite forming on average at approximately 3 Myr and breunnerite at approximately 11 Myr after the start of the solar system. Moreover, the earliest formed breunnerite grain analyzed had 53Mn/55Mn = 2.2 × 10−6, corresponding to approximately 7.5 Myr after the start of the solar system (Petitat et al. 2009).
In contrast to Orgueil, the analogous CI1 lithologies in Kaidun do not contain any grains of breunnerite (this work; Zolensky and Ivanov 2003). The absence of breunnerite in the CI1 clasts of the Kaidun meteorite breccia may indicate that the compaction of Kaidun occurred before the formation of breunnerite on a CI1 parent body. This would imply that accretion, impact disruption, and reassembly of several meteorite parent bodies into the Kaidun parent-asteroid should have occurred within the first approximately 7.5 Myr of the solar system. A plausible timeline would then be that aqueous alteration on the carbonaceous chondrite-like parent bodies initiated very soon after their accretion, probably by approximately 2–3 Myr after the start of the solar system. This is consistent with other estimates of accretion time scales of carbonaceous chondrites (e.g., Kleine et al. 2005; Markowski et al. 2006). Impacts disrupted a portion of at least one of these CI-like asteroids before approximately 7.5 Myr when aqueous alteration might have continued to form breunnerite. Although the age of anhydrous clasts in Kaidun is not directly constrained in this scenario, it seems likely that they could not be much younger (or else younger CI-like material might have been incorporated as well). What is also interesting is that 53Mn-53Cr ages of chondrules from unequilibrated ordinary chondrites overlap this period of accretion, impact disruption, and reaccretion witnessed by the unique collection of asteroid pieces that came to Earth as the Kaidun breccia.
This 53Mn-53Cr study presents the first three internal isochrones for three dolomite grains from two CI1 lithic clasts of the Kaidun meteorite. Our new data constrain the time formation of these carbonate grains. They also constrain the formation period of the Kaidun meteorite. Our conclusions are
1 The three dolomite grains analyzed in this study have identical initial 53Mn/55Mn ratios and diverge from the previous data of Hutcheon et al. (1999). The new data resolve the problem of the contemporaneous formation of carbonates and CAIs (Hutcheon et al. 1999).
2 Dolomites from CI1 clasts of Kaidun formed within the first approximately 4 Myr after the start of the solar system. Aqueous alteration is an early parent-body process that seems to have been initiated contemporaneously with chondrule formation in the protoplanetary disk.
3 Based on the comparison between the CI1 lithologies found in the Kaidun meteorite and the Orgueil CI1 chondrite, a scenario for the compaction of the Kaidun parent body was developed. The absence of breunnerite in the CI1 clasts of the Kaidun meteorite breccia indicates that its compaction might have occurred before the formation of breunnerite in analogous CI1 meteorite samples such as Orgueil. This constrains the accretion of the Kaidun parent body to within the first 7.5 Myr after the start of the solar system (Petitat et al. 2009).
Acknowledgments— We wish to thank the associate editor G. Srinivasan and two anonymous reviewers for their comments and discussions. This work is based on the PhD thesis of Manuel Petitat at the Laboratoire de Minéralogie et de Cosmochimie du Muséum, Muséum National d'Histoire Naturelle, Paris and was supported by the Origin network (Marie Curie European Fellowship).