doi:10.1016/j.gca.2004.10.006 PGE, Re-Os, and Mo isotope systematics in Archean and early Proterozoic sedimentary systems as proxies for redox conditions of the early Earth C. SIEBERT, 1 J. D. KRAMERS, 1, *TH.MEISEL, 2 PH.MOREL, 1 and TH. F. NÄGLER 1 1 Isotope Geology, Institute of Geological Sciences, University of Berne, Erlachstr. 9a. CH-3012 Bern, Switzerland 2 General and Analytical Chemistry, University of Leoben, Franz-Josef-Str. 18, A-8700 Leoben, Austria (Received June 21, 2004; accepted in revised form October 11, 2004) Abstract—Re-Os data and PGE concentrations as well as Mo concentrations and isotope data are reported for suites of fine clastic sediments and black shales from the Barberton Greenstone Belt, South Africa (Fig Tree and Moodies Groups, 3.25–3.15 Ga), the Belingwe Greenstone Belt, Zimbabwe (Manjeri Formation, ca. 2.7 Ga) and shales from the Witwatersrand, Ventersdorp and Transvaal Supergroups, South Africa ranging from 2.95 to 2.2 Ga. Moderately oxidizing conditions are required to mobilize Re and Mo in the environment, Mo fractionation only occurs in solution, and these parameters thus have potential use as paleoredox proxies for the early Earth. PGE + Re abundance patterns of Barberton Greenstone Belt sediments are uniform and very similar in shape to those of komatiites. This indicates (1) that the PGE came from a source of predominantly ultramafic composition and, (2) that PGE were transported and deposited essentially in particulate form. Sediments from the younger Belingwe Greenstone Belt show more fractionated PGE + Re patterns and have Re/Os ratios 10 to 100 higher than those of Barberton sediments. Their PGE abundance patterns and Re/Os ratios are intermediate between those of the mid-Archean shales and Neoproterozoic to Recent black shales. They reflect scavenging of Re from solution in the sedimentary environment.  98/95 Mo values of black shales of all ages correlate with their concentrations. The Barberton Greenstone Belt samples have 1–3 ppm Mo, similar to a granitoid-basaltic source. This Mo has  98/95 Mo between -1.9 and -2.4‰ relative to present day mean ocean water molybdenum, MOMO and is thus not isotopically fractionated relative to such a source. Similar to the PGE this indicates transport in solid form. Sediments from the Belingwe Greenstone Belt show in part enhanced Mo concentrations (up to 6 ppm) and Mo isotope fractionation ( 98/95 Mo up to -1.4‰ relative to MOMO). The combined PGE + Re and Mo data show mainly reducing conditions in the mid-Archean and suggest that by 2.7 Ga, the atmosphere and oceans had become more oxidizing. Substantially younger samples from the Transvaal Supergroup (to ca. 2.2 Ga) surprisingly have mainly low Mo concentrations (around 1 ppm) and show no significant Mo isotope fractionation relative to the continental source. Among possible explanations for this are a return to reducing atmospheric conditions after 2.7 Ga, reservoir effects, or Mo removal by sulfide precipitation following sulfate reduction in early Proterozoic oceans. Copyright © 2005 Elsevier Ltd 1. INTRODUCTION There is a broad consensus in the literature that the Earth’s atmosphere in the Archean was poor in oxygen and that a major rise of oxygen levels occurred between 2.4 and 1.8 Ga (Kasting, 1993; Karhu and Holland, 1996; Canfield, 1998; Collerson and Kamber, 1999; Canfield et al., 2000; Kasting, 2001; Sreenivas et al., 2001). Arguments are mainly based on detrital uraninite and pyrite in clastic sediments (e.g., Schidlowski, 1981; Fleet, 1998), on studies of paleosols showing mobility of Fe as Fe 2+ (Holland et al., 1989; Holland and Beukes, 1990), and on considerations of Banded Iron Formations (e.g., Beukes, 1984). Further arguments include REE patterns with and without Ce anomalies in carbonate sediments (Bau and Dulski, 1996; Bau et al., 1999; Kamber and Webb, 2001) and mass-independent Sulfur isotope fractionation (Farquhar et al., 2000, 2002). However, the dissenting view of Ohmoto (1996) must be noted. He argues for a relatively early oxygenation of the atmosphere (1.5% of present-day value) be- tween 3 and 2.2 Ga based mainly on Fe/Ti ratios in paleosols, whereby the reductive dissolution of Fe occurred postdeposition- ally by hydrothermal fluids or organic acids formed by the decay of terrestrial organic matter. Within the framework of this broad consensus, there is still considerable debate about the nature and relative importance of various sources and sinks of oxygen, the timing of the rise of oxygen concentration, and the possible presence of anoxic/oxic “niches” in the Proterozoic ocean. Among sources of O 2 , photol- ysis of biogenic methane followed by thermal escape of H (Catling et al., 2001) has been considered in addition to photosynthesis. Consideration of the whole mantle as a potential oxygen sink via subduction (Kump et al., 2001) is contradicted by data on V and Cr contents of Archean basalts and komatiites which indicate that the oxygen fugacity of the upper mantle has not changed very much since the Archean (Canil, 1997). Clearly, however, the possibility of multiple oxygen sources and sinks allows in princi- ple for a highly complicated history of atmospheric oxygen in the Precambrian rather than a simple rise. Further, the different prox- ies show in part different histories. Bau et al. (1999) noted the first occurrence of a Ce anomaly in the Mooidraai Dolomite, Transvaal Supergroup, which they dated at 2394  26 Ma (although it may be as young as 2.2 Ga, Dorland et al.). The Campbellrand Sub- group, stratigraphically below it and dated at 2521  3 Ma, does not yet show a Ce anomaly (Kamber and Webb, 2001). Farquhar et al. (2000) have found that non–mass-dependent sulfur isotope * Author to whom correspondence should be addressed (kramers@ geo.unibe.ch). Geochimica et Cosmochimica Acta, Vol. 69, No. 7, pp. 1787–1801, 2005 Copyright © 2005 Elsevier Ltd Printed in the USA. All rights reserved 0016-7037/05 $30.00 + .00 1787