Modal abundances of CAIs: Implications for bulk chondrite element abundances and fractionations

Modal abundances of Ca,Al-rich inclusions (CAIs) are poorly known and reported data scatter across large ranges. We combine reported CAI modal abundances and our own set, and present a complete list of CAI modal abundances in carbonaceous chondrites. This includes (in area%): CV: 2.98, CM: 1.21, Acfer 094: 1.12, CO: 0.99, CK/CV (Ningqiang & DaG 055): 0.77, CK: 0.2, CR: 0.12 and CB: 0.1. CAIs are Poisson distributed and if only small areas (<1000 mm2) are studied, the data are probably not representative of the true CAI modal abundances, explaining their reported large scatter in a single chondrite group. Carbonaceous chondrites have excess bulk Al concentrations when compared to the CI-chondritic value. We find a correlation between this excess and CAI modal abundances and conclude that the excess Al was delivered by CAIs. The excess Al is only a minor fraction (usually ~10 rel%, but 25 rel% in case of CVs) of the bulk chondrite Al and cannot have contributed much 26Al to heat the chondrite parent body. Ordinary, enstatite, R- and K-chondrites have an Al deficit relative to CI chondrites and only very low CAI modal abundances, if any are present at all. Carbonaceous chondrites also had an initial Al deficit if the contribution of Al delivered by CAIs is subtracted. Therefore all chondrites probably lost a refractory rich high-T component. Only minor amounts of CAIs are present in the matrix or have been present in the chondrule precursor aggregates. Most CAI size distributions contain more than one size population, indicating that CAIs from within a single meteorite group had different origins.

💡 Research Summary

The paper by Hezel et al. addresses a long‑standing problem in meteoritics: the poorly constrained modal abundances of calcium‑aluminum‑rich inclusions (CAIs) in chondritic meteorites. Reported values in the literature span orders of magnitude even for a single meteorite group, hampering any quantitative use of CAIs in bulk‑composition or thermal‑history models. The authors combine previously published data with a new, systematic survey based on X‑ray elemental maps (Al, Ca, Mg, Si, Fe) of thin sections, identifying 2 049 CAI candidates and confirming that >95 % of the X‑ray‑identified objects are genuine CAIs.

A central methodological advance is the statistical treatment of CAI spatial distribution. The authors argue that CAIs are rare, randomly mixed particles within a matrix of far more abundant material, and therefore follow a Poisson distribution. Using a custom Mathematica 5.1 model they simulate a 100 mm × 100 mm area containing a realistic number of CAIs (≈3 500). By subdividing the area into square cells of varying edge length, they generate histograms of “cellular” CAI modal abundances and compare them with the theoretical Poisson probability density function. The key result is that the variance of measured CAI abundances is a strong function of the sampled area: cells smaller than ~2 500 mm² (≈50 mm × 50 mm) produce highly variable estimates, whereas larger cells converge to the true mean with 95 % confidence. This explains why many earlier studies, which typically examined thin sections of ~100 mm², reported a wide spread of values.

The authors then present their own measurements for a suite of carbonaceous chondrites (CV, CM, CO, CK, CR, CB) and the ungrouped Acfer 094. Reported area‑% CAI abundances are: CV 2.98 %, CM 1.21 %, Acfer 094 1.12 %, CO 0.99 %, CK/CV‑like 0.77 %, CK 0.20 %, CR 0.12 %, CB 0.10 %. Within each group, individual specimens still show variability, but the new data fall within the range of previously published values once the Poisson sampling bias is accounted for.

To link CAI abundance to bulk chemistry, the authors examine the excess aluminum (Al) in carbonaceous chondrites relative to the CI chondritic benchmark. They employ two independent approaches: (1) assuming all chondrites started with CI‑like Al, any excess Al must be supplied by CAIs; (2) estimating the Al contribution of chondrules and matrix, subtracting this from the measured bulk Al, and attributing the remainder to CAIs. Both methods converge on a modest contribution: CAIs supply roughly 10 % of the total Al in most carbonaceous groups, rising to ~25 % in CVs. Consequently, the radiogenic ²⁶Al carried by CAIs would have been insufficient to provide a dominant heat source for parent‑body metamorphism.

In contrast, ordinary, enstatite, R‑, and K‑chondrites display an Al deficit relative to CI and virtually no CAIs, implying that a high‑temperature refractory component was lost from these bodies. Even carbonaceous chondrites, after correcting for CAI‑derived Al, still show a small residual Al deficit, supporting a universal loss of a refractory, high‑temperature phase during early solar‑nebula processing.

Size‑distribution analysis reveals that most chondrite groups contain more than one CAI size population, often fitting a log‑normal distribution but with secondary peaks. This suggests that CAIs within a single meteorite group originated from multiple nebular reservoirs rather than a single homogeneous source.

Overall, the paper provides (i) a robust statistical framework for determining representative CAI modal abundances, (ii) a comprehensive, updated inventory of CAI area‑% across carbonaceous chondrite groups, (iii) quantitative evidence that CAIs contribute only a minor fraction of bulk Al and cannot dominate ²⁶Al‑driven heating, and (iv) a broader implication that all chondrite parent bodies experienced loss of a refractory, high‑temperature component. These findings refine our understanding of early solar‑system material mixing, the role of CAIs in bulk chemistry, and the thermal histories of chondritic parent bodies.