OXYGEN-ISOTOPE ANALYSES OF CHEMICALLY AND MICROBIALLY PRODUCED MANGANESE OXIDES AND MANGANATES

Understanding the formation of metal deposition in the geological reco rd depends in part on understanding some of the basic reactions that c ould have occurred when those ore deposits were formed. For manganese oxides and manganates, the delta(18)O value might reflect the conditio ns that influenced the oxidation pathway during deposition, which incl ude temperature, the delta(18)O values of H2O or dissolved O-2, and mi crobial catalysis. Mn(IV) 10 Angstrom manganates were, therefore, prep ared by three different procedures with different sources of water hav ing distinct delta(18)O-H2O values, while keeping the delta(18)O-O-2 c onstant. Manganates were prepared (1) chemically (i.e., abiotically) u nder strong alkaline conditions at 3 degrees C, (2) biologically with Mn(II)-oxidizing, dormant spores of a marine bacterium, Bacillus sp. s train SG-1, and (3) biologically with metabolically active cells of a Mn(II)-oxidizing marine bacterium, strain SI85-9A1. The SG-1- and SI85 -9A1-produced manganates were formed under environmentally relevant co nditions of pH (7.6) and temperature (20-25 degrees C) in natural filt ered seawater containing 100 mu M Mn(II). X-ray diffraction analysis a nd oxidation state measurements demonstrated that the chemically preci pitated mineral is characterized by a collapsible (and expandable) 10 Angstrom peak (buserite), whereas those produced by SG-1 and SI85-9A1 had noncollapsible 10 Angstrom peaks. Oxygen isotope analysis of the M n(IV) minerals revealed significant incorporation of molecular O-2 fro m each synthesis (32-50%), which is consistent with direct oxidation o f Mn(II) to Mn(IV) in a single step, rather than by disproportionation of lower valence state intermediates. Estimates of the kinetic isotop e fractionation values for molecular O-2 for the chemical and SG-1 man ganates were -13 and -22 parts per thousand, respectively, whereas val ues of -5 and +1 parts per thousand were estimated for H2O. In contras t, the SI85-9A1-produced minerals showed only a small negative fractio nation for either oxygen source. Application of these results to natur al deposits of manganese oxides indicates a microbial origin for a fre shwater manganese nodule from Oneida Lake, NY, USA, as well as a 50% d issolved oxygen signal. Oxygen isotopic values of Mn(IV) manganates fr om the Kaikata seamount, however, appear to reflect the delta(18)O val ue of seawater exclusively. Therefore, Mn(IV) manganates in nature may not always contain a dissolved oxygen signal, presumably a result of different oxidation pathways or postdepositional alteration.