Thesis: The first goal of this project was to develop a methodology based on the study of the sulfur isotopic composition enabling the determination of iron sulfides origin, biotic or abiotic, within the corrosion products layers (CPL). Then, the aim was to apply this methodology to real corrosion systems in order to determine the mechanisms of iron sulfides formation. Sulfur isotopic analyses methodologies, adapted to micrometric iron sulfides layers observed in real corrosion systems, were developed in nanoSIMS and ToF-SIMS. The study of iron sulfides formed in anoxic carbonated medium with or without sulphate-reducing bacteria validated the use of these methods for the determination of iron sulfides origin. The application of these methods coupled with the precise characterization of irons sulfides formed in the real corrosion systems show two kind of corrosion pattern. In pattern 1, the iron sulfides are localized in the external part of the CPL. They result from the Fe2+ migration from the metal surface to areas rich in biotic S2-. In this pattern, corrosion rates are lower than 20 μm/year for laboratory systems, and lower than 5 μm/year for archaeological objects. In pattern 2, the large presence of conductive phases in the CPL results in the delocalization of electrons, and so a disequilibrium of the charges at the metal’s surface. That leads to the migration of biotic S2- in the CPL till the metal where they precipitate in iron sulphides. This pattern shows high corrosion rates (~100 μm/an) that might be resulting from the accumulation of biocorrosion and chloride corrosion mechanisms.