Mn and Fe contents of cathodoluminescent (CL) calcite cements that formed in shallow fresh-water aquifers can be used to infer the redox potential of the pore fluids from which they formed. A semiquantitative geochemical model relating these factors is based upon the assumption that the Mn2+ and Fe2+ concentrations of the ground waters were controlled by the Eh-dependent solubility of manganese and iron minerals and that the Mn and Fe concentrations of calcite cements precipitating from these solutions can be related to their concentrations in solution by distribution coefficients. Although the qualitative model of Oglesby (1976), Carpenter and Oglesby (1976), and Frank and others (1982) correctly relates calcite CL to the redox conditions of precipitation, numerous inconsistencies arise when it is compared to hydrogeochemical data for modern ground waters and to petrographic observations of CL calcite-cement sequences. First, the model defines redox conditions for calcite precipitation incompatible with those reported from modern ground waters. Second, the wide range of diagenetic conditions proposed by the model for precipitation of brightly luminescent calcite is incompatible with the observation that this is volumetrically the least important cement type. This model also fails to account for the Fe contents reported for many nonluminescent calcite cements. Finally, the Eh predicted from the Mn content of the calcite cements is considerably more oxidizing than that predicted from the Fe content. In part, these inconsistencies reflect the difficulty of estimating ancient redox conditions from calcite Mn and Fe contents because of the variable composition, crystallinity, and thermodynamic stability of Mn- and Fe-oxyhydroxide minerals.
A geologically reasonable, internally consistent model of the redox conditions of calcite cementation can be derived by assuming that the ground-water aMn2+ and aFe2+ were controlled by equilibrium with a poorly crystalline ferric hydroxide [ΔGf0Fe(OH)3 ≈ -696.5 kJ mol-1] and a fictive MnO2* with an apparent ΔGf0 ≈ -549.3 kJ mol-1. These modifications to the existing model make it internally consistent so that similar Eh conditions of calcite precipitation are defined on the basis of Mn and Fe. Furthermore, the revised geochemical model defines Mn2+ and Fe2+ concentrations and redox conditions that are compatible with those reported from ground waters. Modern carbonate aquifers can thus provide a useful analog to interpret the distribution of CL-zoned calcite cements in ancient carbonate rocks, allowing a more refined understanding of paleoaquifer systems and the prediction of diagenetic facies and cement distribution in ancient carbonate sequences.