Manganese nodules are a type of ferromanganese chemical precipitate containing a plethora of economically important metals increasingly being used in the high- and green-technology sectors. By measuring oxygen isotope ratios in these materials, we can increase our understanding of how they form, and constrain which metals are likely to be present in them. There is a lack of suitable reference materials for use in oxygen isotopic studies of samples dominated by Mn oxide minerals (i.e., Mn nodules and crusts) and, therefore, a need for their development. The samples NOD-A-1 and NOD-P-1 (United States Geological Survey), Mn nodules from the Atlantic and Pacific oceans, respectively, provide the best opportunity for such a reference material but their application has been hampered by their incomplete mineralogical characterization and lack of consensus on how to pre-treat (if at all) these type of samples for δ18O analysis. We have performed a detailed mineralogical study using scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS), powder X-ray diffraction (PXRD), and Raman spectroscopy on these reference materials. In sample NOD-A-1 todorokite, goethite and/or hematite, Mg-calcite, phillipsite or chabazite, quartz and barite were identified. The minerals todorokite, birnessite, asbolane, vernadite, goethite, pyrochroite, quartz, phillipsite or chabazite, anorthite, kaolinite and barite were identified in NOD-P-1. We conducted experiments to assess the effect on δ18O and mineralogy resulting from acidification with hydrochloric acid (HCl), pre-fluorination with bromine pentafluoride (BrF5) at ambient temperature, or both acidification and pre-fluorination treatments. Results from samples given both treatments were deemed most meaningful to the study of Mn crusts and nodules, and therefore nodule research in general, by providing δ18O values representative of the relatively pure oxide-silicate fraction only. We report a δ18O value of 11.0 ± 1.0‰ for the oxide-silicate fraction of NOD-A-1, and a value of 11.1 ± 0.7‰ for the same fraction in NOD-P-1, with results showing a δ18O shift of up to 10.3‰, as a function of nodule and the pre-treatment used. The δ18O values of samples given both treatments converge and are substantially 18O-depleted relative to that of corresponding untreated material, which was found to contain various 18O-enriched phases such as carbonate minerals and non-stoichiometric water. Based on our results, we recommend that for all future analyses of Mn nodules and crusts, the samples receive both treatments to ensure accurate and consistent results.
Read full abstract