WHAT CAN δ18O ANALYSES OF UNRECRYSTALLIZED EARLY QUARTZ CEMENTS IN 1.9 GA IRON FORMATIONS TELL US ABOUT PALEOPROTEROZOIC SEAWATER?

Chemical sediments can preserve geochemical signals reflecting the waters from which they precipitate, but these values are often modified by diffusion or recrystallization during diagenesis or later metamorphism. δ¹⁸O analyses of silica in the form of quartz, one of the main constituents of iron formations, have been used to estimate Precambrian ocean conditions. However, almost all quartz in banded iron formations (BIFs) is recrystallized from amorphous precursors that precipitated directly from ocean water. The same is true of most of the silica in granular iron formations (GIFs) with one notable exception: many pores between sand-size clasts (peloids and/or ooids) are 0.2 to 0.8 mm across and filled with drusy quartz cements. Synsedimentary rip-up clasts and high minus-cement porosities indicate these cements precipitated near the depositional interface. To get the purest primary signal, we focused on quartz cements in low-grade GIFs from the 1.9 Ga Gunflint and Sokoman Formations. Growth bands identified in quartz cement crystals using cathodoluminescence imaging (CL) help confirm their void-filling and unrecrystallized nature, making them more likely to retain their original δ¹⁸O. Using an ion microprobe, we analyzed primary quartz cements (46 spots in 5 samples) and recrystallized quartz in adjacent peloids (9 spots in 4 samples). In the Gunflint, δ¹⁸O ranges from 23.7 to 26.4‰ VSMOW in cements and 21.3 to 26.8‰ in peloids. Sokoman values were slightly lower with 18.9 to 23.8‰ in cements and 19.5‰ in the one peloid analyzed. While the overall ranges overlap, the δ¹⁸O of peloids vs. cements in a given sample tend to differ. In individual pores, cement values varied slightly from edge to center, but no consistent increasing or decreasing trends were observed. Overall, the δ¹⁸O of the cements are similar to the most elevated δ¹⁸O reported from GIFs, but span a substantially smaller range. This suggests most of the lower values reported previously may represent post-depositional modifications instead of primary water conditions. The variability in δ¹⁸O within single pores and cement crystals in our samples strongly suggests our values reflect the original isotopic composition of and fluctuations in pore waters very close to Precambrian seafloor.