Ground truthing non-traditional metal stable isotope palaeoredox proxies: Cases of Uranium and Thallium

Throughout the 4.6 billion years of Earth’s history and the nearly equally long history of life, the contents of molecular oxygen (O2) in Earth’s atmosphere and oceans have varied greatly – from effectively absent to a most important constituent. Molecular oxygen is the most energetic of terminal oxidants available for biological respiration and sustains 85% of the standing biomass on the modern Earth. Therefore, any attempt to write a history of life on this planet cannot be done without a concurrent history of oxygen. Central to writing this history of oxygen lies a novel line of geochemical evidence: signals preserved as the isotopic composition of redox-sensitive metals in sedimentary rocks. So-called isotopic palaeoredox proxies. 

The two metals Uranium and Thallium have proven to be valuable isotopic palaeoredox proxies. Yet, limits to our understanding of the dynamics of U and Tl in modern oceans and sediments constitute important challenges for the application of these proxy systems. In the case of U, biotic and abiotic reduction of soluble U(VI) to insoluble U(IV) imparts a strong enrichment of 238U relative to 235U in the reduced solid phase. Thus, the isotopic composition of residual seawater U, denoted as the 𝛿𝛿238𝑈𝑈 signature, tracks the extent of euxinic seafloor, as the heavy 238U isotope is preferentially removed in such environments. The challenge lies in the preservation of the seawater isotopic composition as a geochemical signal in the rock record. Marine precipitation of carbonate minerals effectively incorporates trace U without isotopic fractionation, such that carbonate rocks should constitute an ideal archive for geological δ238Useawater reconstructions. However, during shallow diagenesis of carbonate sediments, the primary δ238U signal is variably overprinted by authigenic U reduction in sediment porewaters. 

In the case of Tl, adsorption of dissolved Tl(I) onto deep-sea Mn oxide nodules is associated with an oxidation to Tl(III), which imparts a strong enrichment of 205Tl relative to 203Tl in the adsorbed phase. This isotopic fractionation manifests as a seawater isotopic Tl signal, denoted as the 𝜀𝜀205𝑇𝑇𝑇𝑇 signature, that is primarily dependant on the burial flux of Mn oxides, and by extension the seawater O2 content. This seawater Tl signal is effectively captured in sediments deposited under euxinic water columns where quantitative Tl drawdown inhibits isotopic fractionation. However, modern datasets of Tl are scarce, there is considerable uncertainty to the marine Tl budget, and many processes in the Tl cycle are barely elucidated. 

This thesis is an effort to drive forward the understanding of earth’s oxygenation history. Both by addressing the challenges to the Tl and U palaeoredox proxies as well as applying the proxies to illuminate central developments in earth’s oxidation history. It consists of three independent manuscripts, plus the outline for a fourth in the appendix, contributing with significant new insights into the marine geochemistry of U and Tl, and the oxygenation of earth’s oceans.

In Manuscript I, a chromatographic method of separating U(IV) and U(VI) from carbonate matrices is developed with the aim of overcoming syndepositional alteration of δ238U signals. This method is validated by comparison against HERFD-XANES oxidation state determination of bulk U. Application of the method on samples from a modern carbonate depositional environment shows a gradual increase of U(IV) with depth in the sediment. However, δ238U signatures are fractionated to heavier values already at the shallowest depths and we suggest that the sediments experience continued reoxidation of labile microbial U(IV) products across a shallow fluctuating redox cline to explain this observation.  

In Manuscript II, Tl phase associations and isotope dynamics are investigated across a continental margin sea. A suite of biogeochemically distinct sites within the sea were sampled for sediments and porewaters, while a set of rivers draining into the sea were sampled for waters and suspended particles. Sequential extractions of sediments and river particles, along with labelled isotope incubations of sediments, and sediment isotope signatures, indicate that Tl does not associate extensively with Mn in this environment. Rather, we find that leached TlA is likely associated with an aluminous phase of terrigenous origin. River particles have higher contents of this Al-associated Tl than marine sediments, and we observe a net efflux of dissolved Tl from several sites that cannot be ascribed to dissolution of marine authigenic phases. Therefore, we suggest that boundary exchange of continentally derived particles may constitute a source of Tl to global oceans. 

In Manuscript III, the ε205Tl paleoredox proxy is applied in concert with other trace metal palaeoredox proxies on marine shales deposited during the great oxidation event, at a time when sulfur isotope evidence suggests that the atmosphere oscillated between an oxic and an anoxic state. ε205Tl signatures and redox sensitive metal enrichments indicate that oxygenated seafloors with Mn oxide burial expanded globally at times of oxic atmospheric conditions and disappear again during a brief return to an anoxic atmosphere. This marks a decisive development in Earth’s redox history towards coupled marine-atmospheric oxygenation. 

In the outline for a fourth manuscript found in Appendix A, the dynamics of Tl enrichments and isotope signals are examined in the sediments of a continental shelf and margin setting exposed to a highly productive upwelling regime. We suggest that the open ocean setting, with continued renewal of the water column Tl inventory, drives strong Tl enrichments in sediments with a high rate of carbon turnover. In these high C-turnover sediments, isotopically unfractionated Tl is abiotically immobilized as a byproduct of sulphate reduction. We correlate an exhaustive dataset of marine sediment Tl concentrations with modelled total oxygen uptake (TOU) rates and find that high-TOU-sediments are greatly undersampled for Tl and possibly constitute a major and unappreciated Tl sink.