Abstract:
There is growing evidence from redox sensitive elements and stable isotope tracers (i.e. Mn contents, Mo and Fe isotope variations and multiple S isotope systematics) that shallow-water sediments of the 2.95 Ga Mozaan Group (Pongola Supergroup) in South Africa were deposited in a so- called oxygen oasis, i.e. in oxidizing shallow seawater. Moreover, previously reported stable Cr isotopic fractionation in Archean paleosols and iron formations (IF) have been interpreted as a signature of oxidative weathering of Cr(III) to Cr(VI) in soils by free atmospheric oxygen, and delivery of isotopically heavy Cr(VI) into the oceans. The oldest reported isotopically heavy Cr fingerprints of this process were found in the Ijzermijn IF, which is also located within the 2.95 Ga Mozaan Group (Pongola Supergroup) in South Africa. These heavy Cr isotopic signatures thus point to a globally oxygenated atmosphere at the time of the Pongola Supergroup sedimentation rather than the existence of localized oxygen oasis as suggested from Mo and Fe isotope systematics. However, mineralogical, elemental and sulfur isotopic evidence in Mozaan Group sediments is largely in favor of the concurrent atmosphere having remained reducing. Furthermore, fractionated stable Cr isotopic signatures have only been found to date in surface outcrop samples of the Ijzermijn IF from the White Mfolozi Inlier exposed along the White Mfolozi River Gorge. During the course of this thesis, this outcrop was resampled along with two drill cores of the Ijzermijn IF and a drill core of the Scotts Hill IF to represent multiple exposures of Mozaan Group IF with different states of preservation. A detailed geochemical comparison on bulk samples of different units was undertaken using stable Cr, Mo and Fe isotopes as well as (234U/238U) activity ratios coupled with trace and major elements. Plausibility of an oxygen oasis during the deposition of the Ijzermijn IF was also investigated and the findings (i.e. the stable Fe isotopic composition) were compared with a ferruginous, oxic modern environment, Arvadi Spring in Engadin, Switzerland, that mimics the depositional conditions during high oxygen intervals before the Great Oxidation Event (GOE).
Drill core samples from the Ijzermijn IF are characterized by high contents in the iron carbonate mineral siderite (FeCO3) and thus reveal large loss on ignition (LOI) and Fe(II)/Fetot ratios. These samples also have (234U/238U) activity ratios that are in secular equilibrium (i.e. value of 1), and have unfractionated δ53/52Cr values that are within the range of the igneous inventory. Outcrop samples from the White Mfolozi River gorge, on the other hand, have no siderite and thus show low LOI and very low Fe(II)/Fetot ratios, have disequilibrium (234U/238U) activity ratios, unusually high U/Th ratios (up to 12.6), and partly show fractionated δ53/52Cr values outside of the igneous inventory (up to 0.418 ‰). The outcrop IF and shale samples of the Sinqeni Formation from the White Mfolozi River bed show very strong deviations from the secular equilibrium (234U/238U) activity ratio. (234U/238U) values for
Ijzermijn IF samples range from 1.05 to 1.77, similar to U dissolved in modern river water. (234U/238U) activity in more U-rich, overlying Vlakhoek Member shales range from 0.89 to 0.96, similar to values of actively weathering saprolitic basaltic soil profiles. Outcrop samples also show enrichments of other elements (W, Tl, As, MREE) that far exceeding that observed in correlative drill core units. These data indicate that, modern weathering effects superimposed geochemical fingerprints on the original paleoredox signatures at the White Mfolozi River bed outcrop. Overall rare earth element and yttrium (REE+Y) mixing models agree well with previous studies, confirming that they were minimally disturbed by weathering and are consistent with a high magnitude of continental solutes delivered in a near– shore depositional environment, with a minor contribution of hydrothermally derived fluids that upwelled into shallower depositional setting. Ijzermijn IF outcrop samples tend towards heavier δ56/54FeIRMM-014 (- 1.729 to - 0.438 ‰) and δ98/95MoNIST3134+0.25 (- 0.476 to + 1.160 ‰) values than drill core ones (- 2.522 to - 0.753 ‰ and - 0.546 to + 0.382 ‰, respectively). Neither metal stable isotope system of outcrop samples closely correlates with (234U/238U), thus a near-pristine sedimentary origin for these signatures cannot be excluded. Nevertheless, for any interpretation of redox-conditions during the sedimentation of the Mozaan Group in the Pongola basin, Fe and Mo isotopic results of outcrop samples from this and previous studies must be regarded with great caution.
For drill core DDN-1 and PMH24/L1 samples, a moderate positive correlation in δ56/54FeIRMM-014 with log [Fe/Mn] and a lack of a negative correlation between δ56/54FeIRMM-014 and δ98/95MoNIST3134+0.25 values are observed. These observations point to preferential adsorption of isotopically light Mo onto MnO2 particles in surface water, hence oxic surface water conditions in the Pongola Epicontinental Sea, as was already proposed in previous studies. Reduction of the Mn-oxides together with microbial Fe reduction in the reduced sediment pile of the Pongola basin led to the formation of Mn- and Fe- carbonates during early diagenesis storing the observed stable Fe and Mo isotopic signatures. The so- called MnO2-shuttle first proposed for the Neoarchean 2.46 Ga old Koegas IF is thus extended to late Mesoarchean shallow-marine depositional settings. From the stable Fe isotopic perspective, depositional model for the Ijzermijn IF requires the upwelling and partial oxidation of deep water Fe(II)aq directly by seawater O2 and MnO2 particles, leading to the progressive depletion of heavy Fe isotopes in the remaining Fe(II)aq pool at the shallower water depths, following an open system Rayleigh trend.
Iron concentrations and δ56/54FeIRMM-014 values of water samples (Fe(II)aq) and Fe-precipitates (red flocs; Fe(III)ppt) were determined from various sampling locations at the modern analog Arvadi Spring flow system. All results are from samples recovered during two sampling campaigns in October 2014 and June 2018. The source of the Arvadi water has a δ56/54FeIRMM-014 value of - 0.359 ‰ and becomes isotopically lighter in the center of the pond and towards the outflowing creek with minimum
δ56/54FeIRMM-014 values of - 1.482 ‰ and - 1.598 ‰, respectively. Iron concentrations of Arvadi waters along the flow path correlate with their respective δ56/54FeIRMM-014 values (R2= 0.66) and decrease from
0.436 ppm at the source to 0.082 ppm at the creek and become as low as 0.011 ppm after mixing with the natural forest brook. The iron concentration and δ56/54FeIRMM-014 values of red flocs also decrease along the flow path, from the source to the creek, from 32.69 wt % to 1.80 wt % and 0.323 ‰ to
- 0.467 ‰, respectively. However, unlike the water samples, the Fe isotopic compositions and Fe concentrations of red flocs lack a strong correlation. The stable Fe isotopic composition of the water and precipitates in Arvadi is a result of abiotic and biotic oxidation of Fe(II)aq to Fe(III)ppt under oxic conditions. The oxidation of Fe(II)aq to Fe(III)ppt follows an open system Rayleigh trend from the source to the creek, leading to progressively lighter Fe isotopic compositions for both, the remaining Fe(II)aq pool and the resulting Fe(III)ppt, along the flow path. The ongoing microbial dissimilatory iron reduction in anoxic niches along the flow path of the Arvadi Spring enhances the trend towards lighter Fe isotopic compositions in the Fe(II)aq pool, while causing the remaining Fe(III)ppt pool to become slightly heavier. This combination of abiotic and biotic Fe oxidation in the water and microbial Fe reduction in the sediment causes the observed decoupling of Fe concentrations and δ56/54FeIRMM-014 values in the red flocs. As such, the complex mechanism of Fe redox cycling in the Arvadi Spring gives important insight about Fe removal from water to sediment and allowing more accurate models for IF deposition in Mesoarchean to Paleoproterozoic oxygen oasis.