Résumé
Oxidation of Fe2+ by anoxic water in the subsurface is a key geochemical process, contributing to the formation of natural dihydrogen (H2). The development and application of effective tools to accurately characterize the content and speciation of iron in samples is thus a major concern for H2 prospection. Traditionally, the study of iron has been conducted through either time-consuming analyses at the micrometer scale or faster analyses at the bulk rock scale, raising concerns about the accuracy and representativeness of the characterization depending on the chosen approach. Moreover, most techniques are typically limited to determining either Fe distribution or Fe speciation, thus necessitating a full series of analyses to reach a comprehensive understanding of the sample. This approach does not align with the need for rapid and numerous characterizations required in H2 prospection programs. In this study, we investigated the relevance of using Mössbauer Spectroscopy (MS) on complex mineral assemblage, by characterizing five Fe-rich natural samples. Among others, we conclude, based on the quality of the resulting spectra fitting that room-temperature (295K) data collection is more effective than low-temperature (6K) data collection, due to the challenges in deconvoluting the complex spectra of mixed mineral assemblages at low temperature. Fe2+/ΣFe ratios obtained from MS are compared with those derived from conventional Fe2+ titration on the same samples. The comparison shows a great correlation between MS and titration results with an average deviation of 0.04 on the Fe2+/ΣFe ratio. This confirms the reliability of MS, providing at the same time insights into both Fe distribution (i.e., Fe mineralogy) and Fe speciation, contrary to titration that only gives access to bulk Fe2+/ΣFe ratio. Finally, results show that the accuracy of MS spectra fitting is significantly influenced by prior knowledge of the sample mineralogy, which can be easily leveraged by rapid and routinely performed characterization techniques (e.g., multispectral mineral imaging).