U-Pb age of the oldest achondrite points to 26Al heterogeneity in the early Solar System

doi:10.21203/rs.3.rs-1957266/v1

Download PDF

Article

U-Pb age of the oldest achondrite points to ²⁶Al heterogeneity in the early Solar System

https://doi.org/10.21203/rs.3.rs-1957266/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 29 Aug, 2023

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

The homogeneity of distribution of radioisotope ²⁶Al in the early solar nebula, a major heat source for early planetary differentiation and foundational assumption to high resolution cosmochronology, remains debatable. Here we report a precise Pb-isotopic age of 4565.56±0.12 million years (Ma) for the recently discovered andesitic achondrite Erg Chech 002. Combining this age with published high-precision ²⁶Al-²⁶Mg data, we demonstrate that the initial ²⁶Al/²⁷Al in the source material of this achondrite was distinctly higher than in several well preserved and precisely dated achondrites. We argue that the current data clearly indicate spatial heterogeneity of ²⁶Al in the precursor molecular cloud or the protoplanetary disk of the Solar System, likely associated with the late infall of stellar materials with freshly synthesized radionuclides.

The presence of aluminium-26 (²⁶Al, T_1/2=0.705 Ma ¹) in the early Solar System, and its possible role as a heat source for planetary and asteroidal melting and early metamorphism, was predicted ² more than 60 years ago. The discovery ³ of excesses of ²⁶Mg in Allende CAIs correlated with the Al/Mg ratio gave evidence that ²⁶Al was indeed present in the early Solar System at the time of CAI formation and decayed in situ. Subsequent studies ^4–6 showed ubiquity of ²⁶Al, but the uniform abundance of ²⁶Al in the early Solar System, with the notable exception of FUN CAIs and isotopically anomalous refractory mineral grains in chondrites ^4,7,8, was commonly assumed rather than demonstrated. The age interval between formation of CAIs and chondrules measured with the Pb-Pb chronometer ⁹ was found to be broadly consistent with the time intervals determined with the ²⁶Al-²⁶Mg chronometer ^4,5, reinforcing the notion of uniform initial distribution of ²⁶Al throughout early Solar System materials at the level of ~ 5×10^− 5 measured in CAIs, referred to as the “canonical ratio” in cosmochemical studies.

The difference in initial abundance of ²⁶Al between the CAI formation region and the rest of the protoplanetary disk where chondrules, achondrites, and planets formed was proposed ¹⁰ on the basis of precisely measured mass-independent Mg isotopic variations in meteoritic materials with low Al/Mg ratios. However, the reliability of this approach was questioned ^11,12, because construction of a CAI isochron by including genetically unrelated amoeboid olivine aggregates (AOAs) can produce apparent initial ²⁶Mg isotopic variations in low-Al/Mg meteorite minerals. Recent work ¹³ highlighted the limitation of the approach ¹⁰ based on the then sparse Al-Mg data set for AOAs.

A more reliable assessment of homogeneity in ²⁶Al distribution can be achieved by comparing ²⁶Al-²⁶Mg and Pb-Pb ages for a pair of meteorites or their components where both chronometers date the same event (i.e., rapid crystallisation of the melt), and remained closed to migration of parent and daughter elements. Progress in meteorite chronology achieved in the last 15 years due to increased precision of ²⁶Al-²⁶Mg ^14,15 and Pb-Pb ^16,17 dating, discovery of ²³⁸U/²³⁵U variability in meteorites ¹⁸, and improving accuracy of Pb-Pb ages by using measured rather than assumed ²³⁸U/²³⁵U ^19,20 enabled revisiting the distribution patterns of ²⁶Al and other short-lived radionuclides in the early Solar System ²¹. Application of refined dating methods to recently found very old and more pristine achondrites – volcanic angrites ^{15–17, 22} and ungrouped achondrites ^21,23−27 – yielded a detailed record of magmatism on the asteroids during the first ~ 5 Ma of the Solar System. The age differences between CAIs and achondrites were found to be typically longer when measured with the ²⁶Al-²⁶Mg chronometer than with the Pb-Pb chronometer. This disagreement has been interpreted ²² as indicating ~ 4 times lower ²⁶Al/²⁷Al ratio in the zone of the solar protoplanetary disk where the angrite parent body formed, compared to the “canonical” ratio in the formation zone of CAIs. Further enhancements in sensitivity of Pb-isotopic analyses made it possible to precisely date individual chondrules ²⁸. Comparison of Pb-isotopic and ²⁶Al-²⁶Mg dates for the same chondrules from LL3.1 chondrite North West Africa (NWA) 5697 and CV3 chondrite Allende ²⁹ also showed a mismatch between the CAI – chondrule apparent age differences, and suggested that two groups of chondrules were formed with initial ²⁶Al/²⁷Al ~ 3 times and ~ 11 times lower than the “canonical” ratio.

However, interpretation of the disagreement between ²⁶Al-²⁶Mg and Pb-isotopic dates as evidence for initial ²⁶Al/²⁷Al heterogeneity has been disputed. The age data for CAIs and chondrules were called into question ^30–32. Among the suggested possible causes of age inaccuracy are biases in Pb-isotopic ages derived from multi-step partial dissolution that was used in chondrule dating ²⁸, age variability in CAIs ^19,33,34, and variability of ²⁶Al/²⁷Al ratios in CAIs ³⁵. These effects, or their combination, could potentially mimic the patterns of heterogeneous ²⁶Al/²⁷Al distribution.

Better candidates for reference meteorite materials that can be reliably dated with both ²⁶Al-²⁶Mg and Pb-Pb methods (see supplementary text 1 for discussion of reliability of isotopic dates) can be found among achondrites. For example, volcanic angrites, a group of rapidly cooled, volatile depleted basaltic rocks that are unbrecciated and unshocked ³⁶, are made of mineral assemblages that facilitate precise ²⁶Al-²⁶Mg and Pb-Pb dating. Some of these meteorites are sufficiently large and homogeneous to be studied by multiple methods and by several research groups. Therefore, volcanic angrites are better suited for reliable and verifiable age dating than chondrules and CAIs that are small and heterogeneous. Volcanic angrites studied so far, including D’Orbigny, Sahara 99555, and NWA 1670, have uniform Pb-isotopic ages ^16,17,37, ²³⁸U/²³⁵U ratios ^20,38 and ²⁶Al/²⁷Al ratios at the time of crystallisation ^15,22,39. The largest and best studied volcanic angrite, D’Orbigny, is widely used as a time anchor in early Solar System chronological studies with short-lived radionuclides ^{20,25,27,40−43}. Combined precise ²⁶Al-²⁶Mg and Pb-isotopic age data with supporting ²³⁸U/²³⁵U ratios have been also reported for ungrouped achondrites NWA 2976 ²¹, NWA 6704 ^23,25, and Asuka 881394 ²⁷.

Recent discovery of andesitic achondrite Erg Chech 002 (EC 002) and its exceptionally old age according to ²⁶Al-²⁶Mg systematics ^44,45 provides an opportunity to further explore the initial distribution of ²⁶Al. Combined regression of 19 plagioclase and 11 pyroxene Al-Mg SIMS (secondary-ion mass spectrometry) analyses ⁴⁴ yielded an isochron corresponding to ²⁶Al/²⁷Al = (5.72 ± 0.07)×10^− 6. The MC-ICPMS (multi collector inductively coupled plasma mass spectrometry) analysis of six pyroxene and plagioclase fractions and one whole rock ⁴⁵ yielded a steeper isochron corresponding to ²⁶Al/²⁷Al = (8.89 ± 0.09)×10^− 6. The absence of excess dispersion in both regressions – which include multiple analyses of two minerals with different rates of Mg diffusion ^46,47 and cover a wide range of Al/Mg ratios – indicates that the event dated by ²⁶Al-²⁶Mg isochrons was very rapid, i.e., probably crystallisation rather than cessation of diffusion upon cooling. It also suggests that the minerals remained close to migration of Al and Mg at micron scale ⁴⁴ ever since. The difference in the slope between the isochrons based on SIMS ⁴⁴ and MC-ICPMS ⁴⁵ data is probably caused by uncorrected systematic analytical uncertainty in Al/Mg ratios measured in SIMS due to using Ca-rich plagioclase as a standard in analysis of Na-rich plagioclase. Such analytical biases due to imperfect matrix matching are common in SIMS ⁴⁸. If the shallower slope of the SIMS isochron was caused by slow cooling ⁴⁵, then the SIMS data for plagioclase would depend on the crystal size and SIMS spot location, and would have been widely dispersed below the MC-ICPMS isochron, rather than forming a tight array without excess scatter ⁴⁴. The ²⁶Al/²⁷Al ratio yielded by isochrons for EC 002 is ca. 4–5 times higher than the initial ²⁶Al/²⁷Al in the previously known oldest achondrites NWA 11119 ⁴² and Asuka 881394 ²⁷.

We determined the Pb-isotopic age of EC 002 by analysing 15 pyroxene, 7 whole rock and 1 plagioclase fractions for Pb isotopic composition (analytical procedures, data table, and detailed presentation of results are in supplementary materials). Five of these fractions (two pyroxenes, two whole rocks, and plagioclase) were also analysed for U and Pb concentrations to assess concordance of the U-Pb system. The isochron, which includes 25 analyses of final leachates and residues, yields the Pb-Pb age of 4565.56 ± 0.12 Ma, MSWD = 1.3 (Fig. 1A). The accuracy of the Pb-Pb isochron age is confirmed by the agreement with the weighted average ²⁰⁷Pb*/²⁰⁶Pb* model age of 4565.53 ± 0.11 Ma (Fig. 1B), calculated assuming that ²⁰⁴Pb after blank subtraction is the initial Pb of primordial isotopic composition ^49,50. The age is calculated using ²³⁸U/²³⁵U = 137.8288 ± 0.0054 (MSWD = 2.1, n = 5) measured in acid leached whole rock fraction (supplementary materials). The uncertainties of the isochron age and weighted average ²⁰⁷Pb*/²⁰⁶Pb* model age shown in Fig. 1 include uncertainties of the U isotope composition.

The measured ²³⁸U/²³⁵U of EC002 is higher than in other studied achondrites ^{20,23,27,34,38,51} and terrestrial rocks, both mafic and felsic ^51,52, but is similar to U isotope composition in terrestrial zircon and other uranium-rich accessory minerals ^53,54. It is currently unclear whether the elevated ²³⁸U/²³⁵U in EC 002 is caused by nucleosynthetic heterogeneity, by isotope fractionation during melting and/or magma fractionation, or by their combination. These possibilities could be distinguished when meteorites cogenetic to EC 002 but more mafic are analysed. The difference in ²³⁸U/²³⁵U between EC 002 and other achondrites emphasizes the importance of supporting each critical Pb-isotopic date in cosmochemistry by direct measurement of uranium isotopic composition. Using average meteoritic ²³⁸U/²³⁵U = 137.786 ± 0.013 ³⁴ instead of a directly measured value would make the age of EC 002 appear 0.42 Ma younger. This change is 3.5 times larger than the uncertainty of the Pb-isotopic age.

Multiple lines of evidence indicate that the age of 4565.56 ± 0.12 Ma derived from the Pb-Pb isochron for EC 002 can be reliably interpreted as the time of its crystallisation. The isochron passes through the point of primordial Pb, thereby eliminating uncertainty in isotopic composition of initial Pb (Fig. S9b). In U-Pb concordia diagram (Fig. S11), the residues and final leachates that were analysed for Pb and U concentrations plot on a discordia line that passes through zero, and has upper intercept identical to the Pb-isotopic isochron age, confirming that the U-Pb system remained closed to migration of parent and daughter nuclides from crystallization till acid leaching in the lab. Furthermore, the U-Pb systems in the residues are nearly concordant (discordance within ± 2%, Fig. S11 and Data S1), demonstrating closed system behaviour even after acid leaching. The reliability of the ²³⁸U/²³⁵U ratio is confirmed by consistency between the ratios measured on separate dissolutions of the rock with and without acid leaching (Fig. S14). The interpretation of the Pb-isotopic age as the time of crystallisation is supported by slower diffusion of Pb in pyroxene ⁵⁵ compared to Mg diffusion in plagioclase ⁴⁷, which indicates that the U-Pb chronometer closed either before or simultaneously with the ²⁶Al-²⁶Mg chronometer ²⁷, and the latter has been shown above to date crystallisation or very rapid cooling. Reliability and high precision of both U-Pb and ²⁶Al-²⁶Mg ages makes EC 002 a promising candidate for assessment of ²⁶Al distribution in the early Solar System.

The initial ²⁶Al/²⁷Al ratios for ungrouped achondrites, volcanic angrites, and CAIs in comparison with their ²⁰⁷Pb-²⁰⁶Pb ages are plotted in Fig. 2. The initial ²⁶Al/²⁷Al in EC 002 ⁴⁵ is clearly resolved from other studied achondrites (except NWA 2976, which has large uncertainty of initial ²⁶Al/²⁷Al) but overlaps with the “canonical” value in the CAIs. Well resolved difference in initial ²⁶Al/²⁷Al between EC 002 and volcanic angrites persists if the value ²⁶Al/²⁷Al in EC 002 measured by SIMS ⁴⁴ is accepted instead of the value measured by MC-ICPMS ⁴⁵. This is reliable evidence that ²⁶Al was heterogeneously distributed among the feeding zones of parent asteroids of achondrites. This evidence is independent from any data for CAIs and chondrules, and is free from ambiguities concerning their U-Pb and Al-Mg systematics. The ²⁶Al heterogeneity can also be shown (supplementary materials text 2 and Fig. S15) as projections of initial ²⁶Al/²⁷Al to either the age of CAIs ³⁴ (4567.30 ± 0.16 Ma) or the age of D’Orbigny (4563.48 ± 0.14 Ma). Fig. S15 clearly shows that there is a significant gap in projected initial ²⁶Al/²⁷Al between EC002, on the one hand, and volcanic angrites, on the other.

Precise initial δ²⁶Mg*₀ = -0.009 ± 0.005 derived from the y-intercept of the Al-Mg isochron ⁴⁵ indicates that the source reservoir of the parent asteroid of EC 002 had separated from the nebula that evolved with solar (= CI chondrite) ²⁷Al/²⁴Mg=0.101⁵⁶ from the initial δ ²⁶Mg*_o = -0.0379 (Fig. S16). Substantial increase of the Al/Mg ratio occurred during rapid partial melting of the protolith and/or fractional crystallisation of the magma on the parent body shortly after its accretion. Our model (details in supplementary materials text 3 and Fig. S16) suggests the time of accretion remains unresolved from the EC 002 crystallisation time.

To explore the relationship between the time of accretion and magmatism, initial ²⁶Al/²⁷Al, and nucleosynthetic affinity, we determined the isotopic abundances of ⁵⁰Ti, ⁵⁴Cr, ⁸⁴Sr, and ¹⁷O in the same aliquot of EC 002, thereby assuring internal consistency of the data (the procedures and results are in supplementary materials). The average ε⁵⁰Ti=-1.08 ± 0.30 (ε unit in geochemistry is a part per 10,000 deviation from the standard value), ε⁵⁴Cr=-0.65 ± 0.10, determined from two separate dissolutions, ε⁸⁴Sr= 0.06 ± 0.24, and Δ΄¹⁷O = -0.145 ± 0.009 (UCLA) or -0.114 ± 0.023 (UG) (defined as a part per 1,000 vertical deviation from the terrestrial fractionation line in three oxygen isotope plot; the procedures and results are in supplementary materials), unequivocally point to the non-carbonaceous affinity (Fig. 3), in agreement with classification of EC 002 as a non-carbonaceous achondrite ⁴⁴ on the basis of thulium distribution. The ε⁵⁴Cr-Δ¹⁷O isotope systematics place EC 002 at the border of HED, angrites, and brachinites regions (Fig. 3d). The variations in initial ²⁶Al/²⁷Al are thus uncorrelated with formation in the inner vs. outer Solar System, and suggest a more nuanced pattern of ²⁶Al distribution. The integrated patterns of ²⁶Al/²⁷Al and stable nucleosynthetic isotope variations will provide a crucial test for the models of collapse of the protosolar molecular cloud, addition of freshly synthesized nuclides, and asteroid and planet formation in the solar protoplanetary disk.

Our data suggest, analogous with the observed ²³⁸U/²³⁵U variations ¹⁸, that we should exercise caution with using any single value of the initial ²⁶Al/²⁷Al in the studies of meteorite chronology, and in modelling thermal evolution of asteroids. Developing a generalised approach for isotopic dating with Al-Mg and other extinct isotope chronometers that accounts for heterogeneous distribution of the parent radionuclide would allow to produce more accurate and reliable age data for meteorites and asteroidal and planetary materials and make further steps towards understanding formation of our Solar System.

U-Pb analytical procedures

Washing procedure and separation for U-Pb analysis

The U-Pb systematics of Erg Chech 002 were studied at the Australian National University, and at the University of California, Davis.

Analyses at SPIDE²R lab, RSES, ANU, include two analytical sessions, A167 and A170. A167 mineral fractions were analysed for Pb isotopes and U and Pb concentrations; solutions after U + Pb separation were also used for Sr isotope analysis. Fractions from batch A170 were analysed only for Pb isotopes.

The rock chips were crushed in a boron carbide mortar and sieved into < 100 µm and 100–250 µm fractions. The 100–250 µm fraction was washed with ethanol and part of it was separated into mineral fractions by handpicking under a binocular microscope. All mineral or whole-rock fractions were ultrasonically cleaned with ethanol and distilled acetone 2–3 times, ~ 15 min each time. After drying the fractions were weighed and transferred in Teflon vials for step dissolution.

The step dissolution procedure is similar to that described in ⁴⁰. The first step is leaching in 0.5 M HNO₃ of ambient temperature with ultrasonic agitation (wash 1, W1). The second step includes leaching in hot (110°C) 7M HNO₃ (W2a) and hot 6M HCl (W2b) for 2 hours in each acid with ultrasonic agitation for 15 min after each hour. The leached minerals are finally dissolved in a mixture of concentrated HF and HNO₃ (residue fraction, R) overnight. All fractions were spiked with a mixed ²⁰²Pb-²⁰⁵Pb-²³³U-²³⁶U spike.

Conversion to a soluble form and chemical separation of Pb and U followed the procedures described in detail in ⁵⁷ and ²³. Chemical separation of Pb has been done using Eichrom AG1x8 anion exchange resin. All solutions passed through columns containing 0.05 ml of the resin twice. U was separated after Pb using Eichrom UTEVA resin. All data were corrected for the average procedure blanks measured for each analytical session. The average procedure blanks for the residues were in the range of 1.2–1.6 pg Pb, and for W2b fractions − 0.7–0.9 pg Pb.

Lead and U isotope ID-TIMS measurements at ANU

The Pb samples were loaded on filaments of zone refined Re with Aldrich silica gel ⁵⁸ and measured on a modified MAT 261 thermal ionization mass spectrometer in static multi-collector mode using Faraday cups or peak jumping mode using secondary electron multiplier (SEM). The choice of the detector depends on the isotopic composition of a sample. Thus, all R aliquots have a high ²⁰⁶Pb/²⁰⁴Pb ratio up to 10,000–20,000 and were measured with an array of Faraday cups for getting accurate ratios of all isotopes except ²⁰⁴Pb; less abundant ²⁰⁴Pb was measured using SEM after lowering filament current or using the three-isotope method measuring only ^{202–204−205}Pb using SEM.

The reproducibility and accuracy of measurements was monitored using SRM 981 Pb isotopic standard and the EarlyTime (ET1x) standard, a synthetic mixture of enriched Pb and U isotopes that imitates radiogenic U-Pb system in a sample with the age of 4560 Ma and containing small quantity of primordial Pb ⁵⁹. Because of the concentration of Pb that we detected in the batch A167 analyses was higher than expected, and therefore higher than in the standards that were run with the batch A167, for the analytical session A170 we prepared and measured standard loads that contained six times more Pb than our regular standard loads of 0.3–0.5 ng. Results of the SRM 981 and Early Time runs are shown in Fig. S1. The ²⁰⁷Pb/²⁰⁶Pb ratio of regular SRM 981 loads is 0.91450 ± 0.00041 (2SD, n = 11) for Faraday cups, 0.91481 ± 0.00047 (n = 12) for SEM. For large SRM 981 standards the ²⁰⁷Pb/²⁰⁶Pb ratio is 0.91459 ± 0.00007 (n = 9) for Faraday cups, and 0.91523 ± 0.00106 (n = 9) for SEM. The most precise ²⁰⁷Pb/²⁰⁶Pb ratios can be obtained from a combination of Faraday and SEM measurements. Such combined measurements yielded the ²⁰⁷Pb/²⁰⁶Pb value of 0.91460 ± 0.00007 (n = 9) for the large standard loads. This value is both precise and in perfect agreement with the certified ²⁰⁷Pb/²⁰⁶Pb ratio of 0.91464 ± 0.00033 for the SRM 981. Most of the radiogenic residue fractions of EC 002 were measured by a combination of SEM and Faraday cups detectors.

For regular loads of ET1x, the most radiogenic standard in the EarlyTime series, ²⁰⁷Pb/²⁰⁶Pb ratio is 0.62356 ± 0.00099 (n = 13) for Faraday cups and 0.62389 ± 0.00038 (n = 13) for SEM. For large ET1x ²⁰⁷Pb/²⁰⁶Pb ratio is 0.62360 ± 0.00012 (n = 10) for Faraday cups and 0.62427 ± 0.00053 (n = 14) for SEM. Combination of Faraday and SEM measurements yield ²⁰⁷Pb/²⁰⁶Pb ratio of 0.62366 ± 0.00013 (n = 10).

Summary plot (Fig. S1) for ²⁰⁷Pb/²⁰⁶Pb ratios (which are crucial for accurate age determination) revealed systematic bias between Faraday cups (or FAR + SEM combined measurements) and SEM measurements, this effect is especially pronounced for large standards. The shift corresponds to the ²⁰⁷Pb/²⁰⁶Pb model age correction of approximately − 1.5 Ma for ET1x standards, therefore all SEM-only measurements (mainly W2b fractions) were corrected for consistency with the FAR + SEM combined measurements, which are in agreement with certified values. Uncertainties of the standards measurements were propagated during data reduction using an AnySpike excel spreadsheet ⁶⁰.

Lead isotope ID-TIMS measurements at UCD

Three pyroxene fractions of 6.43 to 11 mg and one whole-rock fraction of 24.13 mg were separated from EC 002, then rinsed and sonicated three times in distilled acetone. After rinsing, all four fractions underwent three acid-leaching stages—referred to as “washes”—to separate terrestrial Pb contamination from radiogenic and non-radiogenic Pb original to EC 002. The leftover liquid following these washes was collected for Pb isotope analysis. For Wash 1, all four fractions were sonicated in 0.5N HNO₃ five times for ten minutes at a time. For Wash 2, the fractions were heated to 110°C in 6N HNO₃ on a hotplate for one hour, twice. Wash 3 is identical to Wash 2, but with 6N HCl in place of 6N HNO₃. The washes were spiked with a ²⁰²Pb-²⁰⁵Pb-²³³U-²³⁶U tracer for Pb isotopic measurements and U concentration calculations, then dried down and re-dissolved in 2.5N HCl to convert any fluorides remaining in the washes to chlorides. The washes were dried down again and re-dissolved in 0.5N HBr.

Following the acid-leaching stages, the remaining solid residues for each fraction were spiked with the same ²⁰²Pb-²⁰⁵Pb-²³³U-²³⁶U tracer and completely dissolved in a 2:1 solution of concentrated HF and HNO₃, respectively. The dissolved residues were then dried down, re-dissolved in concentrated HNO3, and dried down again. They were then dissolved in 6N HCl to convert any remaining fluorides in the residues to chlorides, then dried down and re-dissolved in 0.5N HBr. Following dissolution in 0.5N HBr, Pb was separated from matrix elements in the washes and residues via anion exchange column chromatography. For Pb isotope analysis, each sample was loaded with silica gel onto a zone-refined Re filament, and measurements were performed on a Triton Plus thermal ionization mass spectrometer (TIMS) at UC Davis.

The Triton Plus TIMS is equipped with nine Faraday cups and one secondary electron multiplier (SEM) and has a total of ten amplifiers out of which six have a 10¹³ Ω and four a 10¹¹ Ω resistor, which can be connected to any of the Faraday cups through a relay matrix. Prior to measurements, amplifier gain, and baseline calibrations were performed and the yield between the SEM and Faraday cup determined. For all samples, at least a few cycles were analysed in peak jumping mode on the SEM. For the larger samples, Pb was also measured with ²⁰⁴Pb in the axial SEM, while all other isotopes (²⁰²Pb, ²⁰⁵Pb, ²⁰⁶Pb, ²⁰⁷Pb, and ²⁰⁸Pb) were collected simultaneously in Faraday cups coupled with 10¹³ Ω resistors. The real time SEM yield compared to the Faraday cups was determined by additional measurement of ²⁰⁵Pb in the axial SEM. Procedural blanks were between 1–2 pg for residues fractions.

During the study, the SRM 981 standard was repeatedly analysed to ensure analytical consistency between analytical sessions at ANU and UCD. ²⁰⁷Pb/²⁰⁶Pb ratio of the SRM 981 standards measured at UCD using SEM is 0.91489 ± 0.00032 (2SD, n = 9) and has shown small but systematic bias relative to the certified value and the combined FAR + SEM measurements at ANU (Fig. S1). UCD data also were normalised to ANU FAR + SEM measurements of the SRM 981 standard: (²⁰⁷Pb/²⁰⁶Pb)_norm = (²⁰⁷Pb/²⁰⁶Pb)_sample/(²⁰⁷Pb/²⁰⁶Pb)_SRM-981 × 0.91460, where 0.91460 is ²⁰⁷Pb/²⁰⁶Pb ratio measured at ANU. Uncertainties of the standards measurements were also propagated during data reduction, similar to ANU data.

U and Pb distribution, concentration, and ratios

Batch A167 includes six fractions: two whole-rock, three pyroxene, and one plagioclase. Batch A170 includes ten smaller fractions of pyroxene and four fractions of the whole rock. Four fractions were measured at UCD: three pyroxenes and one whole-rock. For the batch A167, U and Pb concentrations were measured along with Pb isotopes, while for the A170 and UCD batches only Pb isotopes were measured. U-Pb and Pb-Pb isotopic data are presented in Data Table S1. Concentrations are reported relative to the weights before leaching.

Uranium and Pb distribution between minerals and whole-rock leachates is presented in Fig. S2A and S2B, ²⁰⁶Pb/²⁰⁴Pb ratios are shown in Fig. S2C. Uranium mainly concentrates in W2a and W2b leachates. Augitic pyroxene of brown colour has the highest U concentrations up to 320 ppb in W2b leachate, while orthopyroxene of green colour has the concentrations of U in the range of 25–45 ppb in W2a, W2b and R fractions. Plagioclase has the lowest U concentrations of 8–13 ppb in W2a, W2b and R fractions. Silica oxide minerals, like tridymite and cristobalite, were not analysed in this study but usually contain a relatively low amount of U ⁶¹ because U⁴⁺ is too large (103 pm ⁶²) to substitute for the relatively small (40 pm ⁶²) Si⁴⁺. The whole-rock fractions show U concentration up to 85–95 ppb in W2a and W2b leachates.

Pb mainly concentrates in R fractions (Fig. S2B). The concentrations here are from 19 ppb in plagioclase up to 94 ppb in orthopyroxene. W1a leachates of whole-rock fractions have relatively large proportions of Pb, indicating efficient removing of terrestrial Pb and/or dissolution of phosphates. The Pb concentrations here are 23–24 ppb.

²⁰⁶Pb/²⁰⁴Pb ratios (corrected for spike, fractionation and blank) have large variations (Fig. S2С) and indicate that early leaching steps effectively removed non-radiogenic Pb. All W1a, W1b, and W2a leachates have relatively low ²⁰⁶Pb/²⁰⁴Pb ratios of 29–156, while W2b aliquots have ²⁰⁶Pb/²⁰⁴Pb ratios in the range of 671–3580 and R aliquots up to ~ 12280 and ~ 50870 in clinopyroxene and orthopyroxene, respectively. These ²⁰⁶Pb/²⁰⁴Pb ratios are similar to ones reported for angrites ^16,37.

Isochron calculation

The adequate choice of points for isochron regression is critical to obtain precise and accurate Pb-Pb age. Usually, first washes (W1a, W1b and W2a) have low ratios of radiogenic Pb to other Pb components, and contain a large amount of terrestrial Pb that pulls the isochron towards the modern terrestrial Pb isotopic composition (Fig. S3). Therefore, all first washes should be removed from isochron calculation (Fig. S4). Further rejection is based on consideration of the following geochemical and/or analytical criteria. First, plagioclase in most cases has low ²³⁸U/²⁰⁴Pb and contains a large proportion of non-radiogenic Pb, and therefore usually excluded from the regression. In EC 002 plagioclase is relatively radiogenic but shows open U-Pb system behaviour (Fig. S5), so we excluded these two points. In addition, albitic plagioclase in EC 002 has lamellae of K-feldspar ⁶³, which can cause incongruent dissolution of plagioclase and biased ²⁰⁷Pb/²⁰⁶Pb ratios ⁶⁴. The next step is elimination of points with a small amount of radiogenic Pb because they are more affected by variations of Pb blank and its isotopic composition. In order to choose the right cut-off value, we analyse the distribution of ²⁰⁶Pb* (* denotes radiogenic) in W2b and R fractions (Fig. S6).

Most fractions that contain less than 0.5×10^− 13 mol of radiogenic ²⁰⁶Pb* are W2b fractions from A170 analytical session. Their low content of radiogenic ²⁰⁶Pb* is explained by small mass of these fractions and resistance of orthopyroxene to hot 6M HCl acid. Exclusion of these 11 points yields the isochron with the age of 4565.49 ± 0.18 Ma, MSWD = 2.8 (Fig. S7). A regression through the points with ²⁰⁶Pb* > 1×10^− 13 mol yielded the age of 4565.51 ± 0.17 Ma, MSWD = 2.4 (Fig. S8). Further increase of the cut-off amount of ²⁰⁶Pb* up to 2×10^− 13 mol yields an isochron with the age of 4565.56 ± 0.12 Ma, MSWD = 1.3 (Fig. S9). Using points with a larger amount of ²⁰⁶Pb* > 5×10^− 13 mol gave a similar result with the age of 4565.55 ± 0.12 Ma, MSWD = 1.5 (Fig. S10). In any case, even using data with a low amount of radiogenic Pb does not affect the accuracy significantly. All ages shown in the Figures S7-S10 are identical within their uncertainties, showing that the choice of cut-off points is not critical. The regression based on the data points with ²⁰⁶Pb* > 2×10^− 13 mol (Fig. S9) appears to be optimal. It yields the best precision and, unlike regressions that include smaller fractions, this isochron shows no dispersion of the points in excess of analytical uncertainties. It does not require further rejections of the points based on other criteria. We thus consider the age of 4565.56 ± 0.12 Ma our best estimate of crystallization age of EC 002.

Concordance of the U-Pb system

The U-Pb data for fractions W2b and R from the session A167 were plotted in a Wetherill-type Concordia diagram (Fig. S11). The discordance of the U-Pb system calculated as $\left(\frac{\frac{207\text{P}\text{b}}{206\text{P}\text{b}} \text{d}\text{a}\text{t}\text{e}- \frac{238\text{U}}{206\text{P}\text{b}} \text{d}\text{a}\text{t}\text{e}}{\frac{207\text{P}\text{b}}{206\text{P}\text{b}} \text{d}\text{a}\text{t}\text{e}}\right)$× 100. All residue fractions show discordance values from − 1.9–2.0%, indicating that residues have almost closed U-Pb system even after acid leaching. The W2b leachates have discordance values of 92–95% for the whole rock and pyroxene fractions, and 63% for plagioclase, indicating removal of U relative to Pb in this leaching step. All other leachates also have positive discordance values of 8–97% (Data Table S1). Discordia regression for R and W2b fractions (Fig. S11) yields an upper concordia intercept at 4565.56 ± 0.34 Ma (corrected for U isotopic composition) which is in agreement with the Pb-Pb isochron age of 4565.56 ± 0.12 Ma within their uncertainties. The zero (within uncertainties) intercept of the discordia line suggests that fractionation between U and Pb is recent and caused by acid leaching.

Uranium isotope analysis, methods, and results

Washing procedure and separation for U isotopic composition

Our procedure of sample preparation for U isotope analysis (batch A169) is similar to those described in refs ²³ and ⁶⁵. The sample was crushed in a boron carbide mortar and pestle. Sieved whole-rock fractions were pre-cleaned by ultrasonic agitation in ethanol, MQ water, and distilled acetone 2–3 times. After this, each fraction was divided into two parts: 1) 30–40% for direct dissolution after pre-cleaning (thereafter “bulk” fractions), and 2) 60–70% for two-step acid leaching. The larger fractions were leached twice in 1 mL of 0.5 HNO₃ at 120°C on a hot plate for 30 min with following ultrasonic agitation for 10 min to dissolve possible phosphates (thereafter “wash” fractions). These two washes were combined (2 mL in total) for subsequent chemical separation. The residue and untreated bulk fractions were then dissolved in a mixture of concentrated HF and HNO₃ in the proportion of 3:1 for two days at 120°C on a hot plate. To convert fluorides into soluble salts, all aliquots were evaporated and re-dissolved in 6–7 mL of concentrated HNO₃ at 120°C for 12 days with 30 min ultrasonic agitation each day. During this period samples were evaporated and re-dissolved again in concentric HNO₃ four times (every 3 days) to facilitate dissolution of fluorides. After evaporation, all aliquots were re-dissolved in 5–6 mL of 6M HCl at 190°C in Parr bombs. After complete dissolution, the solutions were spiked with IRMM-3636 ^233–236U double spike ⁶⁶, evaporated and then again re-dissolved in 6M HCl for homogenization. To be sure that spike and samples are fully homogenized, all aliquots were kept on a hot plate at 120°C for three days.

All aliquots were processed through a two-step chemical separation. For the first step, we used Bio-Rad AG-1x8 anion exchange resin (200–400 mesh) to separate U and Fe from the sample matrix. The matrix elements were eluted with 6M HCl, and U and Fe were eluted with 0.5M HNO₃. To reduce the amount of organics, pre-filter resin was put at the bottom of each column. In the second step, we processed U + Fe aliquots through UTEVA resin. Iron was eluted with 3M HNO₃ and U was collected in a mixture of 0.02M HNO₃ and 0.2M HF. For MC-ICPMS analysis, aliquots were evaporated and re-dissolved in 0.5M HNO₃.

Uranium isotope MC-ICPMS measurements

The U isotopic composition was determined on a Thermo Neptune Plus MC-ICPMS at the Australian National University (ANU), Research School of Earth Sciences (RSES), in static multi-collector mode. The samples were introduced to the plasma with Aridus desolvating nebulizer. Faraday cups used for measuring ²³⁸U and monitoring ²²⁹Th and ²³²Th were connected to the amplifiers with 10¹¹ Ω resistors, while the cups used for measuring ²³³U, ²³⁵U, and ²³⁶U were connected to the amplifiers with 10¹² Ω resistors. We achieved the signal intensity of ~ 1V per 1 ppb of ²³⁸U with sample solution uptake of 0.1 mL/min, which corresponds to a total ion yield of ~ 1.5%. Measurements quality was monitored by bracketing with the IRMM-184 uranium isotopic standard ⁶⁷. The mass spectrometer background was measured before each 3–4 sample measurements and was subtracted from a sample signal. All data were corrected for instrumental fractionation assuming the exponential law of fractionation. The data for secondary standards: SRM 960 (equivalent to CRM-112a), and terrestrial basalts BCR-2 and BHVO-2, and for EC 002 were adjusted for consistency with the accepted value of ²³⁸U/²³⁵U = 137.683 ± 0.020 in IRMM-184 standard ⁶⁷. Standards were analysed at various intensities (²³⁸U signal between ~ 1 and 20 V) and with the various sample to spike ratios to match the concentrations and sample to spike ratios of the unknowns and confirm consistency of the data obtained under varying analytical conditions. All EC 002 fractions were analysed at the intensity of 15–17 V.

Measurements of the IRMM-184 isotopic standard with standard to spike ratios of 3, 10, and 30 revealed a linear correlation between ²³⁸U/²³⁵U and ²³⁶U/²³⁸U ratios (Fig. S12a). To resolve the potential reason for the correlation, we measured the isotopic composition (²³⁵U/²³⁶U and ²³⁸U/²³⁶U ratios) of the IRMM-3636 spike using solutions with concentrations of 1, 3, 10, and 20 ppb (Fig. S13). Y-axis intercepts yielded a ²³⁵U/²³⁶U ratio of 4.8 × 10^− 5 and a ²³⁸U/²³⁶U ratio of 2.4 × 10^− 4. Usage of these measured ratios ²³⁶U/²³⁵U = 20674 and ²³⁸U/²³⁵U = 4.8873 instead of certified values of 21988 and 5.1629, respectively ⁶⁶, eliminated correlation between ²³⁸U/²³⁵U and ²³⁶U/²³⁸U ratios in the different sample/spike mixtures (Fig. S12b). All data were processed with the measured isotopic composition of the IRMM-3636 spike. Results of the standards and EC 002 measurements are presented in Table S1.

Normalized ²³⁸U/²³⁵U ratios of secondary standards, SRM 960 and terrestrial basalts, agree with published values ^52,67, confirming the reliability of the procedure. EC 002 is found to have heterogeneous isotopic composition of U. The bulk and residue fractions, which are dominated by intrinsic U, have consistent ²³⁸U/²³⁵U ratios (Fig. S14), which are higher than the average Solar system and terrestrial basalts values and which we interpret as the isotopic composition of the meteorite before it was exposed to a terrestrial environment. The isotopic composition of the leachate is displaced towards the average terrestrial uranium (Fig. S14), and we interpret it as a mixture between intrinsic uranium, and uranium introduced by weathering.

⁵⁴Cr isotope systematics

Dissolution, purification, and ε⁵⁴Cr isotopic analysis of Erg Chech 002 were conducted at the University of California, Davis (UC Davis). An interior, fusion-crust-free chip of EC 002 was crushed to a powder using an agate mortar and pestle. A 3 mg aliquot of this whole-rock powder was set aside for oxygen isotope analysis. The remaining mass was prepared for dissolution. The whole-rock powder was placed in a PTFE Parr bomb digestion capsule, along with twice-distilled (ultrapure), concentrated HF and HNO₃ in a 3:1 ratio. The PTFE capsule was sealed in a stainless-steel jacket, and heated in an oven at 190°C for 96 h. This high-pressure, high-temperature environment is necessary to ensure complete dissolution of highly refractory phases, such as chromite.

Once completely dissolved, Cr was separated from the sample matrix via a three-column separation procedure according to ⁶⁸ using one column with an anion-exchange resin (Bio-Rad AG1-X8, 200–400 mesh) and two columns with cation-exchange resin (Bio-Rad AG50W-X8, 200–400 mesh) in a PicoTrace class 10–100 clean lab.

Following elemental separation, the isotopic composition of the purified Cr fractions was measured using a Thermo Triton Plus thermal ionization mass spectrometer (TIMS). The Cr fractions were loaded onto four outgassed tungsten filaments, with 3 µg sample per filament. The four sample filaments were bracketed with filaments (two before and two after) loaded with an equal amount (3 µg) of NIST SRM 979 Cr isotopic standard.

Each filament analysis consisted of 1200 ratios, with an 8 s integration time for each ratio. The signal intensity for ⁵²Cr was set to 10 V (± 15%). A detector gain calibration was completed at the start of each filament analysis, and the amplifiers were rotated and baseline-measured after every block of 25 ratios. The instrumental mass fractionation was corrected using an exponential fractionation law and a ⁵⁰Cr/⁵²Cr ratio of 0.051859 ⁶⁹. The ⁵⁴Cr/⁵²Cr ratio is expressed in ε-notation, or parts per 10,000 deviation from the measured NIST SRM 979 standard:

$$\epsilon {}^{\text{x}}\text{C}\text{r}= \left[\frac{{\left(\frac{{}^{\text{x}}\text{C}\text{r}}{{}^{52}\text{C}\text{r}}\right)}_{\text{E}\text{C} 002}}{{\left(\frac{{}^{\text{x}}\text{C}\text{r}}{{}^{52}\text{C}\text{r}}\right)}_{\text{S}\text{R}\text{M} 979}}-1\right]\times {10}^{4}$$

where ^xCr is either ⁵³Cr or ⁵⁴Cr.

Cr isotopic composition was measured in two aliquots (A and B), results are presented in Table S2.

⁸⁴Sr isotope systematics

Strontium isotope compositions were measured in the fractions from the batches A167 (after for U-Pb separation) and A169 (after U separation). All separation and measurement procedures are described in ⁷⁰, Sr isotopic composition was measured on TritonPlus thermal ionization mass-spectrometer at ANU, RSES. The ⁸⁴Sr/⁸⁶Sr ratios are reported as relative deviations from the SRM 987 Sr standard as ε notation according to the following formula:

$${\epsilon }{}^{84}\text{S}\text{r}= \left[\frac{{\left(\frac{{}^{84}\text{S}\text{r}}{{}^{86}\text{S}\text{r}}\right)}_{\text{S}\text{a}\text{m}\text{p}\text{l}\text{e}}}{{\left(\frac{{}^{84}\text{S}\text{r}}{{}^{86}\text{S}\text{r}}\right)}_{\text{S}\text{R}\text{M} 987}}-1\right]\times {10}^{4}$$

Results of ⁸⁷Sr/⁸⁶Sr and ⁸⁴Sr/⁸⁶Sr are presented in Table S3.

⁵⁰Ti isotope systematics

A 1.1 g interior chip of EC 002, free from fusion crust, was crushed in an agate mortar. The powdered sample was sieved through 100 µm and 250 µm mesh screens, and the 100–250 µm fraction was analysed for Ti isotopes without mineral separation. A 24 mg aliquot of this fraction was digested with a concentrated HF–HNO₃ mixture at 200 ˚C using a Parr bomb^®. The digested sample was converted to a soluble form by repeated evaporation with concentrated HNO₃, and dissolved in 6 M HCl. The chemical separation of Ti was performed following the procedure described by ⁷¹. First, the sample in 6 M HCl was loaded onto the column packed with Bio-Rad AG1-X8 anion exchange resin (200–400 mesh), in which Ti is eluted while Fe and U are retained by the resin. Second, Ti was separated from matrix elements including Cr and Ca using Eichrom TODGA resin (50–100 µm mesh). Finally, Ti was purified using the AG1-X8 resin, in which the remaining matrix elements were eluted in 4 M HF as well as 0.4 M HCl + 1 M HF, followed by Ti elution in 1 M HCl + 2% H₂O₂.

The isotopic composition of the purified Ti fraction was measured using a Thermo Fisher Scientific Neptune Plus multiple collector inductively coupled plasma mass spectrometer (MC-ICPMS) at the University of Tokyo. The sample diluted to a concentration of 100 ppb was introduced to the MC-ICPMS using a CETAC Aridus II desolvating nebulizer with a sample uptake rate of ~ 0.15 mL/min. Measurements were performed using a Jet sample cone and an X skimmer cone with high mass resolution, which resulted in ⁴⁸Ti signal intensities of ~ 2.5×10^–10 A. In addition to five Ti isotopes, ⁴³Ca, ⁵¹V and ⁵³Cr were monitored to correct for isobaric interferences on Ti isotopes from ⁴⁶Ca, ⁴⁸Ca, ⁵⁰V, and ⁵⁰Cr. Data were obtained in dynamic mode from 40 cycles, 2 lines/cycle, 8.4 s integration/line, and 4 s idle time between lines. Instrumental mass fractionation was corrected with the exponential law by assuming ⁴⁹Ti/⁴⁷Ti = 0.749766⁷². The sample measurement was bracketed by analyses of an Alfa Aesar Ti standard solution. The ⁵⁰Ti/⁴⁷Ti ratio of the sample is expressed in ε-notation defined as follows:

$${{\epsilon }}^{50}\text{T}\text{i}=\left[\frac{{\left(\frac{{}_{ }{}^{50}\text{T}\text{i}}{{}_{ }{}^{47}\text{T}\text{i}}\right)}_{\text{E}\text{r}\text{g} \text{C}\text{h}\text{e}\text{c}\text{h} 002}}{{\left(\frac{{}_{ }{}^{50}\text{T}\text{i}}{{}_{ }{}^{47}\text{T}\text{i}}\right)}_{\text{A}\text{l}\text{f}\text{a} \text{A}\text{e}\text{s}\text{a}\text{r}}}-1\right]\times \text{10,000}$$

Analytical uncertainty on the sample ε⁵⁰Ti combined the internal precision (2 SE) and the reproducibility of the standard analyses (2SD), added in quadrature. For EC 002 we obtained ε⁵⁰Ti value of -1.08 ± 0.30.

Oxygen isotope analysis

The triple oxygen isotope composition is expressed in form of the classical δ-notation with

$${{\delta }}^{\text{17,18}}\text{O}=\left(\frac{{\left(\frac{{}^{\text{17,18}}\text{O}}{{}^{16}\text{O}}\right)}_{\text{S}\text{a}\text{m}\text{p}\text{l}\text{e}}}{{\left(\frac{{}^{\text{17,18}}\text{O}}{{}^{16}\text{O}}\right)}_{\text{S}\text{M}\text{O}\text{W}}}-1\right)\text{*}1000$$

The oxygen isotope anomaly is expressed in the ${\varDelta }^{{\prime }}$¹⁷O notation with:

$${\varDelta }^{{\prime }}{}^{17}\text{O}=1000\text{*}\text{ln}\left(\frac{{\delta }{}^{17}\text{O}+1}{1000}\right)-0.528\text{*} 1000\text{*}\text{l}\text{n}\left(\frac{{\delta }{}^{18}\text{O}+1}{1000}\right)$$

Or using the traditional “non-prime” notation with:

$${\varDelta }^{17}\text{O}={\delta }{}^{17}\text{O}-0.52\text{*}{\delta }{}^{18}\text{O}$$

Oxygen isotopic analysis at the University of Göttingen

The oxygen isotope composition of Erg Chech 002 was determined at the University of Göttingen by infrared (IR) laser fluorination ⁷³. An aliquot of approximately 2 mg was loaded into a small, 2-pit sample holders together with San Carlos olivine and then placed into an air lock system. The all-metal air lock was pumped down to ~ 3 x 10^–6 mbar and heated using heating tape to ~ 75°C for 24 h. The air lock system is attached via a CF40 gate valve (with Kel-F gaskets) to the sample chamber. The empty fluorination chamber was evacuated and heated to about 60°C for 24 h before being exposed to ~ 100 mbar BrF₅ for ~ 15 min in order to remove any moisture that adheres to the inner chamber walls after the chamber. The moisture is adsorbed to the walls and window when the sample chamber is opened, as necessary for cleaning or changing the BaF₂ window. To minimize contamination, the sample chamber is filled with Ar while being open. After the air lock cooled to room temperature, the sample holder containing Erg Chech 002 and San Carlos olivine was introduced through a gate valve into the fluorination chamber. Samples were then exposed to BrF₅ (100 mbar) and fluorinated by scanning the laser beam across the sample pit at increasing laser energy up to 45 W. Following fluorination, sample O₂ gas was transferred through cold traps and NaCl (for F₂ removal) to a 5 Å molecular sieve trap. From this trap, sample O₂ was transferred via He gas stream (10 mL min^–1) through a 5 Å molecular sieve packed gas chromatography column (3 m, 1/8”, 50°C) into a second 5 Å molecular sieve trap located in front of a Thermo 253 Plus mass spectrometer. After evacuation of He from this trap, sample O₂ was expanded at 50°C into the bellows of the mass spectrometer. Samples were measured relative to reference gas that was calibrated using O₂ released from San Carlos olivine (δ¹⁸O = 5.23‰, Δ^’17O_0.528 = − 0.052‰; ⁷⁴). The entire fluorination, gas transfer and measurement procedure were timed using LabVIEW to avoid any user-specific effects. On the basis of replicate analyses of San Carlos olivine, the analytical uncertainty was assumed to be ~ 0.3‰ for δ¹⁸O and ~ 0.016‰ for Δ^’17O (2σ SD).

Oxygen isotopic analyses at UCLA

A whole-rock oxygen isotopic measurement of Erg Chech 002 was performed at UCLA. The sample was loaded into a 316L stainless steel chamber for analysis. The sample was heated with an infrared lamp through the ZnSe window to ~ 120 °C while pumping for ~ 3 hours to eliminate surface absorbed water. The background pressure in the sample chamber before fluorination was ~ 10^− 7 bar.

Oxygen in the form of O₂ gas was extracted from the sample using laser-heating assisted fluorination. Approximately 90 mbar of doubly-distilled F₂ was loaded into the sample chamber as the fluorinating agent ⁷⁵. The sample was melted in the presence of F₂ with a 20 W CO₂ laser pulsed at 10 Hz. After complete fluorination, the product O₂ was purified by passing through a heated KBr trap with cold traps at liquid nitrogen temperature located on either side. The KBr serves as a getter for excess F₂, releasing bromine that is trapped at liquid nitrogen temperature. The traps also sequester SiF₄. The extracted O₂ was then collected into 13X molecular sieve cooled by liquid N₂. The gases are released from the 13X mol sieve at 200°C for 30 minutes, and then transferred into a sample vial by freezing onto a silica gel substrate.

The oxygen isotope ratios were determined on O₂ using a high-mass-resolution, double-focusing gas-source mass spectrometer at UCLA (Nu Instruments Panorama 001). The mass resolving power (M/DM) of 40,000 used for these measurements is sufficient to resolve mass interferences (e.g., NF⁺, ⁷⁶). The ³²O₂⁺, ³³O₂⁺, and ³⁴O₂⁺ ion beams were measured using Faraday cups with amplifier resistors of 10¹⁰ Ω, 10¹³ Ω, and 10¹¹ Ω, respectively. Analyses were obtained from 20 blocks. Each block comprised 30 cycles of sample/reference gas comparisons, and each cycle consists of 30 s integration. The reference gas was calibrated using O₂ purified by gas-chromatography from air. Accuracy is checked additionally by analysing two geostandards for oxygen isotope ratios, San Carlos olivine (SC olivine) and Gore Mountain Garnet. For this study, the Δ’¹⁷O of Erg Chech 002 was anchored to the composition of San Carlos olivine that was used in Göttingen and by Ziegler et al. at UNM ⁷⁷. Results are presented in Table S4.

Acknowledgments:

Funding: Australian Research Council grant DP190100002 “The history of accretion in our Solar System” (YA).

Author contributions:

U-Pb chronology: EK, YA, PC, MHH, QZY

U isotopic composition and Sr isotope systematics: EK and YA

Cr isotope systematics: SD, AM, QZY

Ti isotope systematics: YH

O isotope analyses: HT, EDY, AP, TDR

Writing – original draft: YA, EK

Writing – review & editing: YA, QZY, EK, PC, MHH, AM, SD, YH, HT, EDY, AP, TDR

Competing interests: Authors declare that they have no competing interests

Data and materials availability: All data are available in the main text or the supplementary materials

Norris, T. L., Gancarz, A. J., Rokop, D. J. & Thomas, K. W. Half-life of 26Al. Journal of Geophysical Research: Solid Earth 88, B331-B333, doi:https://doi.org/10.1029/JB088iS01p0B331 (1983).
Urey, H. C. The cosmic abundances of potassium, uranium, and thorium and the heat balances of the Earth, the Moon, and Mars. Proceedings of the National Academy of Sciences 41, 127, doi:10.1073/pnas.41.3.127 (1955).
Lee, T., Papanastassiou, D. A. & Wasserburg, G. J. Demonstration of 26 Mg excess in Allende and evidence for 26 Al. Geophysical Research Letters 3, 41–44, doi:https://doi.org/10.1029/GL003i001p00041 (1976).
MacPherson, G. J., Davis, A. M. & Zinner, E. K. The distribution of aluminum-26 in the early Solar System—A reappraisal. Meteoritics 30, 365–386, doi:https://doi.org/10.1111/j.1945-5100.1995.tb01141.x (1995).
Russell, S. S., Srinivasan, G., Huss, G. R., Wasserburg, G. J. & MacPherson, G. J. Evidence for Widespread 26 Al in the Solar Nebula and Constraints for Nebula Time Scales. Science 273, 757–762, doi:doi:10.1126/science.273.5276.757 (1996).
Srinivasan, G., Goswami, J. N. & Bhandari, N. 26Al in Eucrite Piplia Kalan: Plausible Heat Source and Formation Chronology. Science 284, 1348–1350, doi:doi:10.1126/science.284.5418.1348 (1999).
Ireland, T. R. Presolar isotopic and chemical signatures in hibonite-bearing refractory inclusions from the Murchison carbonaceous chondrite. Geochimica et Cosmochimica Acta 54, 3219–3237, doi:https://doi.org/10.1016/0016-7037(90)90136-9 (1990).
Park, C. et al. Calcium-aluminum-rich inclusions with fractionation and unidentified nuclear effects (FUN CAIs): II. Heterogeneities of magnesium isotopes and 26Al in the early Solar System inferred from in situ high-precision magnesium-isotope measurements. Geochimica et Cosmochimica Acta 201, 6–24, doi:https://doi.org/10.1016/j.gca.2016.10.002 (2017).
Amelin, Y., Krot, A. N., Hutcheon, I. D. & Ulyanov, A. A. Lead Isotopic Ages of Chondrules and Calcium-Aluminum-Rich Inclusions. Science 297, 1678–1683, doi:doi:10.1126/science.1073950 (2002).
Larsen, K. K. et al. Evidence for magnesium isotope heterogeneity in the solar protoplanetary disk. The Astrophysical Journal 735, L37, doi:10.1088/2041-8205/735/2/l37 (2011).
Kita, N. T. et al. 26Al-26Mg isotope systematics of the first solids in the early solar system. Meteoritics & Planetary Science 48, 1383–1400, doi:https://doi.org/10.1111/maps.12141 (2013).
Wasserburg, G. J., Wimpenny, J. & Yin, Q.-Z. Mg isotopic heterogeneity, Al-Mg isochrons, and canonical 26Al/27Al in the early solar system. Meteoritics & Planetary Science 47, 1980–1997, doi:https://doi.org/10.1111/maps.12014 (2012).
Gregory, T., Luu, T.-H., Coath, C. D., Russell, S. S. & Elliott, T. Primordial formation of major silicates in a protoplanetary disc with homogeneous 26 Al/ 27 Al. Science Advances 6, eaay9626, doi:doi:10.1126/sciadv.aay9626 (2020).
Jacobsen, B. et al. 26Al–26Mg and 207Pb–206Pb systematics of Allende CAIs: Canonical solar initial 26Al/27Al ratio reinstated. Earth and Planetary Science Letters 272, 353–364, doi:https://doi.org/10.1016/j.epsl.2008.05.003 (2008).
Schiller, M., Baker, J. A. & Bizzarro, M. 26Al–26Mg dating of asteroidal magmatism in the young Solar System. Geochimica et Cosmochimica Acta 74, 4844–4864, doi:https://doi.org/10.1016/j.gca.2010.05.011 (2010).
Amelin, Y. U–Pb ages of angrites. Geochimica et Cosmochimica Acta 72, 221–232, doi:https://doi.org/10.1016/j.gca.2007.09.034 (2008).
Connelly, J. N., Bizzarro, M., Thrane, K. & Baker, J. A. The Pb–Pb age of Angrite SAH99555 revisited. Geochimica et Cosmochimica Acta 72, 4813–4824, doi:https://doi.org/10.1016/j.gca.2008.06.007 (2008).
Brennecka, G. et al. 238U/235U variations in meteorites: Extant 247Cm and implications for Pb-Pb dating. Science 327, 449–451 (2010).
Amelin, Y. et al. U–Pb chronology of the Solar System's oldest solids with variable 238U/235U. Earth and Planetary Science Letters 300, 343–350, doi:https://doi.org/10.1016/j.epsl.2010.10.015 (2010).
Brennecka, G. A. & Wadhwa, M. Uranium isotope compositions of the basaltic angrite meteorites and the chronological implications for the early Solar System. Proceedings of the National Academy of Sciences 109, 9299–9303 (2012).
Bouvier, A., Spivak-Birndorf, L. J., Brennecka, G. A. & Wadhwa, M. New constraints on early Solar System chronology from Al–Mg and U–Pb isotope systematics in the unique basaltic achondrite Northwest Africa 2976. Geochimica et Cosmochimica Acta 75, 5310–5323, doi:https://doi.org/10.1016/j.gca.2011.06.033 (2011).
Schiller, M., Connelly, J. N., Glad, A. C., Mikouchi, T. & Bizzarro, M. Early accretion of protoplanets inferred from a reduced inner solar system 26Al inventory. Earth and Planetary Science Letters 420, 45–54, doi:https://doi.org/10.1016/j.epsl.2015.03.028 (2015).
Amelin, Y. et al. U-Pb, Rb-Sr and Ar-Ar systematics of the ungrouped achondrites Northwest Africa 6704 and Northwest Africa 6693. Geochimica et Cosmochimica Acta 245, 628–642, doi:https://doi.org/10.1016/j.gca.2018.09.021 (2019).
Nyquist, L. E., Reese, Y., Wiesmann, H., Shih, C. Y. & Takeda, H. Fossil 26Al and 53Mn in the Asuka 881394 eucrite: evidence of the earliest crust on asteroid 4 Vesta. Earth and Planetary Science Letters 214, 11–25, doi:https://doi.org/10.1016/S0012-821X(03)00371-6 (2003).
Sanborn, M. E. et al. Carbonaceous achondrites Northwest Africa 6704/6693: Milestones for early Solar System chronology and genealogy. Geochimica et Cosmochimica Acta 245, 577–596, doi:https://doi.org/10.1016/j.gca.2018.10.004 (2019).
Wadhwa, M. et al. Ancient relative and absolute ages for a basaltic meteorite: Implications for timescales of planetesimal accretion and differentiation. Geochimica et Cosmochimica Acta 73, 5189–5201, doi:https://doi.org/10.1016/j.gca.2009.04.043 (2009).
Wimpenny, J. et al. Reassessing the origin and chronology of the unique achondrite Asuka 881394: Implications for distribution of 26Al in the early Solar System. Geochimica et Cosmochimica Acta 244, 478–501, doi:https://doi.org/10.1016/j.gca.2018.10.006 (2019).
Bollard, J. et al. Early formation of planetary building blocks inferred from Pb isotopic ages of chondrules. Science Advances 3, e1700407, doi:10.1126/sciadv.1700407 (2017).
Bollard, J. et al. Combined U-corrected Pb-Pb dating and 26Al-26Mg systematics of individual chondrules – Evidence for a reduced initial abundance of 26Al amongst inner Solar System chondrules. Geochimica et Cosmochimica Acta 260, 62–83, doi:https://doi.org/10.1016/j.gca.2019.06.025 (2019).
Fukuda, K. et al. A temporal shift of chondrule generation from the inner to outer Solar System inferred from oxygen isotopes and Al-Mg chronology of chondrules from primitive CM and CO chondrites. Geochimica et Cosmochimica Acta 322, 194–226, doi:https://doi.org/10.1016/j.gca.2021.12.027 (2022).
Pape, J., Mezger, K., Bouvier, A. S. & Baumgartner, L. P. Time and duration of chondrule formation: Constraints from 26Al-26Mg ages of individual chondrules. Geochimica et Cosmochimica Acta 244, 416–436, doi:https://doi.org/10.1016/j.gca.2018.10.017 (2019).
Siron, G., Fukuda, K., Kimura, M. & Kita, N. T. New constraints from 26Al-26Mg chronology of anorthite bearing chondrules in unequilibrated ordinary chondrites. Geochimica et Cosmochimica Acta 293, 103–126, doi:https://doi.org/10.1016/j.gca.2020.10.025 (2021).
Bouvier, A. & Wadhwa, M. The age of the Solar System redefined by the oldest Pb–Pb age of a meteoritic inclusion. Nature Geoscience 3, 637–641, doi:10.1038/ngeo941 (2010).
Connelly, J. N. et al. The absolute chronology and thermal processing of solids in the solar protoplanetary disk. Science 338, 651–655 (2012).
Kawasaki, N., Wada, S., Park, C., Sakamoto, N. & Yurimoto, H. Variations in initial 26Al/27Al ratios among fine-grained Ca-Al-rich inclusions from reduced CV chondrites. Geochimica et Cosmochimica Acta 279, 1–15, doi:https://doi.org/10.1016/j.gca.2020.03.045 (2020).
Keil, K. Angrites, a small but diverse suite of ancient, silica-undersaturated volcanic-plutonic mafic meteorites, and the history of their parent asteroid. Chemie der Erde - Geochemistry 72, 191–218, doi:https://doi.org/10.1016/j.chemer.2012.06.002 (2012).
Amelin, Y. The U–Pb systematics of angrite Sahara 99555. Geochimica et Cosmochimica Acta 72, 4874–4885, doi:https://doi.org/10.1016/j.gca.2008.07.008 (2008).
Tissot, F. L. H., Dauphas, N. & Grove, T. L. Distinct 238U/235U ratios and REE patterns in plutonic and volcanic angrites: Geochronologic implications and evidence for U isotope fractionation during magmatic processes. Geochimica et Cosmochimica Acta 213, 593–617, doi:https://doi.org/10.1016/j.gca.2017.06.045 (2017).
Spivak-Birndorf, L., Wadhwa, M. & Janney, P. 26Al–26Mg systematics in D’Orbigny and Sahara 99555 angrites: Implications for high-resolution chronology using extinct chronometers. Geochimica et Cosmochimica Acta 73, 5202–5211, doi:https://doi.org/10.1016/j.gca.2009.02.038 (2009).
Koefoed, P. et al. U–Pb and Al–Mg systematics of the ungrouped achondrite Northwest Africa 7325. Geochimica et Cosmochimica Acta 183, 31–45, doi:https://doi.org/10.1016/j.gca.2016.03.028 (2016).
MacPherson, G. J., Nagashima, K., Krot, A. N., Doyle, P. M. & Ivanova, M. A. 53Mn–53Cr chronology of Ca–Fe silicates in CV3 chondrites. Geochimica et Cosmochimica Acta 201, 260–274, doi:https://doi.org/10.1016/j.gca.2016.09.032 (2017).
Srinivasan, P. et al. Silica-rich volcanism in the early solar system dated at 4.565 Ga. Nature Communications 9, 3036, doi:10.1038/s41467-018-05501-0 (2018).
Zhu, K. et al. Chromium Isotopic Evidence for an Early Formation of Chondrules from the Ornans CO Chondrite. The Astrophysical Journal 873, 82, doi:10.3847/1538-4357/aafe79 (2019).
Barrat, J.-A. et al. A 4,565-My-old andesite from an extinct chondritic protoplanet. Proceedings of the National Academy of Sciences 118, e2026129118, doi:10.1073/pnas.2026129118 (2021).
Fang, L. et al. Half-life and initial Solar System abundance of 146Sm determined from the oldest andesitic meteorite. Proceedings of the National Academy of Sciences 119, e2120933119, doi:doi:10.1073/pnas.2120933119 (2022).
Cherniak, D. J. & Dimanov, A. Diffusion in Pyroxene, Mica and Amphibole. Reviews in Mineralogy and Geochemistry 72, 641–690, doi:10.2138/rmg.2010.72.14 (2010).
Van Orman, J. A., Cherniak, D. J. & Kita, N. T. Magnesium diffusion in plagioclase: Dependence on composition, and implications for thermal resetting of the 26Al–26Mg early solar system chronometer. Earth and Planetary Science Letters 385, 79–88, doi:https://doi.org/10.1016/j.epsl.2013.10.026 (2014).
Ireland, T. R. in Treatise on Geochemistry (Second Edition) (ed Karl K. Turekian) 385–409 (Elsevier, 2014).
Blichert-Toft, J., Zanda, B., Ebel, D. S. & Albarède, F. The Solar System primordial lead. Earth and Planetary Science Letters 300, 152–163, doi:https://doi.org/10.1016/j.epsl.2010.10.001 (2010).
Tatsumoto, M., Knight, R. J. & Allegre, C. J. Time Differences in the Formation of Meteorites as Determined from the Ratio of Lead-207 to Lead-206. Science 180, 1279–1283, doi:10.1126/science.180.4092.1279 (1973).
Goldmann, A., Brennecka, G., Noordmann, J., Weyer, S. & Wadhwa, M. The uranium isotopic composition of the Earth and the Solar System. Geochimica et Cosmochimica Acta 148, 145–158, doi:https://doi.org/10.1016/j.gca.2014.09.008 (2015).
Tissot, F. L. H. & Dauphas, N. Uranium isotopic compositions of the crust and ocean: Age corrections, U budget and global extent of modern anoxia. Geochimica et Cosmochimica Acta 167, 113–143, doi:https://doi.org/10.1016/j.gca.2015.06.034 (2015).
Hiess, J., Condon, D. J., McLean, N. & Noble, S. R. 238U/235U systematics in terrestrial uranium-bearing minerals. Science 335, 1610–1614, doi:10.1126/science.1215507 (2012).
Livermore, B. D., Connelly, J. N., Moynier, F. & Bizzarro, M. Evaluating the robustness of a consensus 238U/235U value for U-Pb geochronology. Geochimica et Cosmochimica Acta 237, 171–183, doi:https://doi.org/10.1016/j.gca.2018.06.014 (2018).
Cherniak, D. J. Pb diffusion in Cr diopside, augite, and enstatite, and consideration of the dependence of cation diffusion in pyroxene on oxygen fugacity. Chemical Geology 177, 381–397, doi:https://doi.org/10.1016/S0009-2541(00)00421-6 (2001).
4.4 Abundances of the elements in the Solar System: Datasheet from Landolt-Börnstein - Group VI Astronomy and Astrophysics · Volume 4B: "Solar System" in SpringerMaterials (https://doi.org/10.1007/978-3-540-88055-4_34) (Springer-Verlag Berlin Heidelberg).
Iizuka, T., Amelin, Y., Kaltenbach, A., Koefoed, P. & Stirling, C. H. U–Pb systematics of the unique achondrite Ibitira: Precise age determination and petrogenetic implications. Geochimica et Cosmochimica Acta 132, 259–273, doi:https://doi.org/10.1016/j.gca.2014.02.017 (2014).
Huyskens, M. H., Iizuka, T. & Amelin, Y. Evaluation of colloidal silicagels for lead isotopic measurements using thermal ionisation mass spectrometry. Journal of Analytical Atomic Spectrometry 27, 1439–1446, doi:10.1039/C2JA30083D (2012).
Connelly, J. N. & Condon, D. J. in Goldschmidt Conference Abstracts, A448 (2014).
Schmitz, M. D. & Schoene, B. Derivation of isotope ratios, errors, and error correlations for U-Pb geochronology using 205Pb‐235U‐(233U)‐spiked isotope dilution thermal ionization mass spectrometric data. Geochemistry, Geophysics, Geosystems 8 (2007).
Blundy, J. & Wood, B. Mineral-Melt Partitioning of Uranium, Thorium and Their Daughters. Reviews in Mineralogy and Geochemistry 52, 59–123, doi:10.2113/0520059 (2003).
Shannon, R. D. Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallographica Section A 32, 751–767, doi:https://doi.org/10.1107/S0567739476001551 (1976).
Mikouchi, T. & Zolensky, M. in Lunar and Planetary Science Conference. 2457.
Ito, K. T. M., Hibiya, Y., Homma, Y., Mikouchi, T. & Iizuka, T. The promise and potential pitfalls of acid leaching for PbPb chronology. Chemical Geology 525, 343–355, doi:https://doi.org/10.1016/j.chemgeo.2019.07.035 (2019).
Merle, R. et al. Exploring the efficiency of stepwise dissolution in removal of stubborn non-radiogenic Pb in chondrule U-Pb dating. Geochimica et Cosmochimica Acta 277, 1–20, doi:https://doi.org/10.1016/j.gca.2020.03.010 (2020).
Verbruggen, A. et al. Preparation and certification of IRMM-3636, IRMM-3636a and IRMM-3636b. JRC Scientific and Technical Reports (2008).
Richter, S. et al. New average values for the n(238U)/n(235U) isotope ratios of natural uranium standards. International Journal of Mass Spectrometry 295, 94–97, doi:https://doi.org/10.1016/j.ijms.2010.06.004 (2010).
Yamakawa, A., Yamashita, K., Makishima, A. & Nakamura, E. Chemical Separation and Mass Spectrometry of Cr, Fe, Ni, Zn, and Cu in Terrestrial and Extraterrestrial Materials Using Thermal Ionization Mass Spectrometry. Analytical Chemistry 81, 9787–9794, doi:10.1021/ac901762a (2009).
Shields, W. R., Murphy, T. J., Catanzaro, E. J. & Garner, E. L. Absolute isotopic abundance ratios and the atomic weight of a reference sample of chromium. Journal of Research of the National Bureau of Standards. Section A, Physics and Chemistry 70, 193 (1966).
Di, Y., Krestianinov, E., Zink, S. & Amelin, Y. High-precision multidynamic Sr isotope analysis using thermal ionization mass spectrometer (TIMS) with correction of fractionation drift. Chemical Geology 582, 120411, doi:https://doi.org/10.1016/j.chemgeo.2021.120411 (2021).
Hibiya, Y., Iizuka, T., Yamashita, K., Yoneda, S. & Yamakawa, A. Sequential Chemical Separation of Cr and Ti from a Single Digest for High-Precision Isotope Measurements of Planetary Materials. Geostandards and Geoanalytical Research 43, 133–145, doi:https://doi.org/10.1111/ggr.12249 (2019).
Niederer, F. R., Papanastassiou, D. A. & Wasserburg, G. J. The isotopic composition of titanium in the Allende and Leoville meteorites. Geochimica et Cosmochimica Acta 45, 1017–1031, doi:https://doi.org/10.1016/0016-7037(81)90128-9 (1981).
Sharp, Z. D. A laser-based microanalytical method for the in situ determination of oxygen isotope ratios of silicates and oxides. Geochimica et Cosmochimica Acta 54, 1353–1357, doi:https://doi.org/10.1016/0016-7037(90)90160-M (1990).
Pack, A. Isotopic Traces of Atmospheric O2 in Rocks, Minerals, and Melts. Reviews in Mineralogy and Geochemistry 86, 217–240, doi:10.2138/rmg.2021.86.07 (2021).
Young, E. D., Coutts, D. W. & Kapitan, D. UV laser ablation and irm-GCMS microanalysis of 18O/16O and 17O/16O with application to a calcium-aluminium-rich inclusion from the Allende meteorite. Geochimica et Cosmochimica Acta 62, 3161–3168, doi:https://doi.org/10.1016/S0016-7037(98)00216-6 (1998).
Young, E. D. et al. Oxygen isotopic evidence for vigorous mixing during the Moon-forming giant impact. Science 351, 493–496, doi:doi:10.1126/science.aad0525 (2016).
Gattacceca, J. et al. The Meteoritical Bulletin, No. 109. Meteoritics & Planetary Science 56, 1626–1630 (2021).
Begemann, F. et al. Call for an improved set of decay constants for geochronological use. Geochimica et Cosmochimica Acta 65, 111–121, doi:https://doi.org/10.1016/S0016-7037(00)00512-3 (2001).
Schoene, B. in Treatise on Geochemistry (Second Edition) (ed Karl K. Turekian) 341–378 (Elsevier, 2014).
Amelin, Y. et al. in Goldschmidt Conference Abstracts 57.
Lodders, K. Solar System Abundances and Condensation Temperatures of the Elements. The Astrophysical Journal 591, 1220–1247, doi:10.1086/375492 (2003).
Stacey, J. S. & Kramers, J. D. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letters 26, 207–221, doi:https://doi.org/10.1016/0012-821X(75)90088-6 (1975).

There is NO Competing Interest.

DataS1.xlsx
Erg Chech 002 U-Pb (A167) and Pb (A170, UCD) isotopic data
SupplementaryTextandFigs.docx

Download PDF

Journal Publication

published 29 Aug, 2023

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

U-Pb age of the oldest achondrite points to ²⁶Al heterogeneity in the early Solar System

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results Of U-pb Analyses And Age Calculation

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1