The nature of Mesoarchaean seawater and continental weathering in 2.85 Ga banded iron formation, Slave craton, NW Canada. (1st December 2016)

• Record Type: Journal Article
• Authors: Haugaard, Rasmus; Ootes, Luke; Creaser, Robert A.; Konhauser, Kurt O.
• Is Part Of: Geochimica et cosmochimica acta. Volume 194(2016:Dec. 01)
• Page Start: 34
• Page End: 56
• Publication Date: 2016-12-01
• Subjects: Slave craton -- Archaean -- BIF -- Seawater chemistry -- Sm−Nd isotopes; Geochemistry -- Periodicals; Meteorites -- Periodicals; Géochimie -- Périodiques; Météorites -- Périodiques; Geochemie; Astrochemie; Electronic journals; 551.905
• DOI: 10.1016/j.gca.2016.08.020
• Languages: English
• ISSNs: 0016-7037

Abstract

Abstract: Banded iron formations (BIF) have been extensively used as proxies to infer the chemical composition of ancient bulk seawater. However, their proximity to ancient crust suggests that they might also be used to reveal the composition of emergent continental landmass at the time of their deposition. Here we use the combination of geochemistry and Sm–Nd isotopes on a layer-by-layer basis to interpret the relative contributions of hydrothermal, hydrogenous and terrestrial input to one of the oldest documented Superior-type BIF in the world. The ∼2.85 Ga Central Slave Cover Group BIF is deposited within a rift basin related to a continental margin and is found associated with basement gneisses, as well as shoreline and shallow-shelf type facies, such as fuchsitic quartzite and pebble-to-cobble conglomerate, that confirm a near-shore depositional setting for the BIF. The BIF ranges from a pure chemical oxide (magnetite)−silicate (grunerite + actinolite) sediment with low Al2 O3 (<1 wt.%) into a mixture of chemical and clastic sediment characterized by higher Al2 O3 (⩽10 wt.%) and the occurrence of ferro-hornblende, biotite and garnet. The silica bands have low trace metal content (e.g., Ni), low ∑REE (average of 6 ppm) and a shale-normalized rare earth and yttrium (REY) pattern that is HREE-to-LREE enriched with (Pr/Yb)SN values reaching <0.2. The iron bands are more enriched, with average ∑REE of 26 and with a more uniform and less fractionated REY pattern (average (Pr/Yb)SN of 0.5). During active rifting of the basement, excess of Eu 2+ impacted the basin yielding seawater with Eu anomalies [(Eu/Eu ∗ )SN ] as high as 3.85 (average 2.75), larger than similarly-aged BIF. High-resolution geochemistry shows that there is more silica (19.4 wt.% SiO2 ) in the iron bands than iron (8.7 wt.% Fe2 O3 ) in the silica bands, implying that dissolved Fe 2+ came to the BIF site in pulses and that silica likely represents background deposition. Consistently radiogenic ε Nd ( t ) values for the iron bands (average +1.7) show that the dissolved REY in the source water during ferric iron precipitation was provided by submarine hydrothermal fluids with relatively uniform 143 Nd/ 144 Nd. The silica bands, by contrast, reveal high variation in seawater 143 Nd/ 144 Nd as evident from the bimodal ε Nd ( t ) distribution with one segment exhibiting negative ε Nd ( t ) values averaging −1.1 and another with positive ε Nd ( t ) values averaging +2.5. This suggests input of dissolved REY into the upper seawater from weathering of isotopically different crustal components in the source region. Collectively, we speculate that the low REY in the upper seawater and the overall low Ni content implies a highly weathered crustal surface that was unable to contribute a significant dissolved load to the shelf environment.