Solar-system abundances of the elements

A new abundance table has been compiled, based on a critical review of all C1 chondrite analyses up to mid-1982. Where C1 data were inaccurate or lacking, data for other meteorite classes were used, but with allowance for fractionation among classes. In a number of cases, interelement ratios from meteorites or lunar and terrestrial rocks as well as solar wind were used to check and constrain abundances. A few elements were interpolated (Ar, Kr, Xe, Hg) or estimated from astronomical data (H, C, N, O, He, Ne). For most elements, the new abundances differ by less than 20% from those of Cameron (1982a). In 14 cases, the change is between 20 and 50% (He, Ne, Be, Br, Nb, Te, I, Xe, La, Gd, Tb, Yb, Ta and Pb) and in 5 others, it exceeds 50% (B, P, Mo, W, Hg). Some important interelement ratios ( Na K, Se Te, Rb Sr, Kr Xe, La W, Th U, Pb U, etc.) are significantly affected by these changes. Three tests were carried out, to see how closely C1 chondrites approximate primordial solar system abundances. (1) A plot of solar vs Cl abundances shows only 7 discrepancies by more than twice the nominal error of the solar abundance: Ga, Ge, Nb, Ag, Lu, W and Os. Most or all apparently reflect errors in the solar data or f-values. (2) The major cosmochemical groups (refractories, siderophiles, volatiles, etc.) show no significant fractionation between the Sun and C1's, except possibly for a slight enrichment of volatiles in Cl's. (3) Abundances of odd-A nuclides between A = 65 and 209 show an almost perfectly smooth trend, with elemental abundances conforming to the slope defined by isotopic abundances. There is no evidence for systematic fractionation of the major cosmochemical groups from each other. Small irregularities (10-15%) show up in the Ag-Cd-In and Sm-Eu regions; the former may be due to a ~ 15% error in the Ag abundance and the latter, to a 10-20% fractionation of Eu during condensation, to contamination of C1 chondrites with interplanetary dust during regolith exposure, or to a change from s-process to r-process dominance. It appears that the new set of abundances is accurate to at least 10%, as irregularities of 5-10% are readily detectable. Accordingly, Cl chondrites seem to match primordial solar-system matter to ≤ 10%, with only four exceptions. Br and I are definitely and B is possibly fractionated by hydrothermal alteration, whereas Eu seems to be enriched by nebular condensation or regolith contamination.