Relationships between the trace element composition of sedimentary rocks and upper continental crust

TLDR

Estimates of upper continental crust composition vary widely for many trace elements, and terrigenous sediment La/element ratios are relatively uniform for elements carried in clastic sediments. The authors aim to use the REE pattern of terrigenous sediment, which reflects upper crust composition, to derive robust estimates of upper crustal abundances directly from sedimentary data. They infer upper crustal abundances from sediment REE patterns and adjust lower crustal values by subtracting the revised upper crust from a bulk crust model. Revised upper crustal abundances are provided for Sc, Ti, V, Cr, Co, Ni, Nb, Cs, Ta, and Pb, while Rb, Zr, Ba, Hf, Th, K, U, and Rb remain unchanged, and these updates necessitate corresponding revisions to lower crustal abundances in the Taylor and McLennan model.

Abstract

Estimates of the average composition of various Precambrian shields and a variety of estimates of the average composition of upper continental crust show considerable disagreement for a number of trace elements, including Ti, Nb, Ta, Cs, Cr, Ni, V, and Co. For these elements and others that are carried predominantly in terrigenous sediment, rather than in solution (and ultimately into chemical sediment), during the erosion of continents the La/element ratio is relatively uniform in clastic sediments. Since the average rare earth element (REE) pattern of terrigenous sediment is widely accepted to reflect the upper continental crust, such correlations provide robust estimates of upper crustal abundances for these trace elements directly from the sedimentary data. Suggested revisions to the upper crustal abundances of Taylor and McLennan [1985] are as follows (all in parts per million): Sc = 13.6, Ti = 4100, V = 107, Cr = 83, Co = 17, Ni = 44, Nb = 12, Cs = 4.6, Ta = 1.0, and Pb = 17. The upper crustal abundances of Rb, Zr, Ba, Hf, and Th were also directly reevaluated and K, U, and Rb indirectly evaluated (by assuming Th/U, K/U, and K/Rb ratios), and no revisions are warranted for these elements. In the models of crustal composition proposed by Taylor and McLennan [1985] the lower continental crust (75% of the entire crust) is determined by subtraction of the upper crust (25%) from a model composition for the bulk crust, and accordingly, these changes also necessitate revisions to lower crustal abundances for these elements.