Bulk Earth element composition

來自專欄高維度穩定同位素

作者:Huiming Bao

--Another high-dimensional thinking

This piece is inspired by a talk presented by Prof. 汪在聰on June 27, 2018 at Institute of Geochemistry, CAS.

One of the top Earth and planetary science problems is the elemental composition of the bulk Earth, as this information is critical to a range of first-order problems including the accretion and differentiation processes and history of the early Earth. Today, the Earth has already differentiated into different layers (e.g. core, mantle, and crust) and these layers are not only different among but also heterogeneous within in terms of elemental composition. Direct sampling is not only impossible but also useless to the problem. Thus, scientists construct Earth formation models to fit observational and experimental data. For example, abundances and ratios of many non-volatile siderophile elements (e.g., W, Mo, Fe, Ni, Co, Cr, V) and the light element abundances in the core can be explained by metal–silicate partitioning data and models of heterogeneous accretion and core formation at high pressure–temperature plus a changing oxygen fugacity during accretion. In this business, one of the most important pieces of observation is the depletion trend of volatile elements in the bulk Earth, particularly for the siderophile volatile elements when normalized to the assumed original source material of Earth, the CI chondrite.

However, many lithophile and siderophile elements of similar volatility show very different magnitudes of depletion. These discrepancies could be due to poor measurement or could be resolved with the existing CI-chondrite based compositional model by adjusting parameters on core formation, impact-induced losses, and late addition of materials.

After analytically confirming the observed elemental abundance pattern, especially the siderophile volatile elements In and Cd in the bulk silicate Earth and their partitioning in the mantle, Prof. Zaicong WANG (汪在聰) and his colleagues approached this problem by considering not just the partition coefficients but the ratios of partition coefficients, i.e., the relative siderophile behavior of Zn, Cd and In at a range of core formation conditions (Wang et al., 2016). This approach helps to eliminate or cancel out many factors that may affect individual partition coefficients. The ratios of In, Cd, and Zn in bulk silicate Earth and bulk Earth indicate that some siderophile volatile elements indeed do not follow the assumed volatile depletion trend of the Earth. They therefore proposed that Earth』s main building materials are now extinct and differ from known chondritic meteorites.

Here the spirit of high-dimensional approach is loud and clear. Different accretion and core formation conditions, e.g. P, T, oxygen fugacity, and light elements in metal (particularly S), affect metal–silicate partition coefficients, but not the relative siderophile behaviors of In, Cd and Zn. Wang and colleagues』 argument hinges heavily on this 「high-dimensional」 signature. Although I applaud the 「high-dimensional」 approach, I would like to point out that this relative siderophile behavior assumption, however, is not as robust as the relative behavior among different isotopes of the same element, as our 「traditional」 four types of high-dimensional parameters have been focusing on. After all, we are talking about three different elements here. Siderophiles are just a group of elements residing at the center of the periodic table that love to bond with iron. The relative consistency in their behavior at different conditions is not firmly established. That said, Wang et al. (2016) can certain bet on the expected consistency and draw their very interesting conclusion.

Wang, Z. C., et al. (2016). "Earths moderately volatile element composition may not be chondritic: Evidence from In, Cd and Zn." Earth and Planetary Science Letters 435: 136-146.