Sulfur isotope analysis of microcrystalline iron sulfides using SIMS imaging: Extracting local paleo-environmental information from modern and ancient sediments.

RATIONALE: Sulfur isotope ratio measurements of bulk sulfide from marine sediments have often been used to reconstruct environmental conditions associated with their formation. In-situ microscale spot analyses by secondary ion mass spectrometry (SIMS) and laser ablation multiple-collector inductively coupled plasma mass spectrometry (LA-MC-ICP-MS) have been utilized for the same purpose. However, these techniques are not often suitable for studying small (≤10 μm) grains or for detecting intra-grain variability.

METHODS: Here, we present a method for the physical extraction (using lithium polytungstate heavy liquid), and subsequent sulfur isotope analysis (using SIMS; CAMECA IMS 7f-GEO) of microcrystalline iron sulfides. SIMS sulfur isotope ratio measurements were made via Cs+ bombardment of raster squares with sides of 20-130 μm, using an electron multiplier (EM) detector to collect counts of 32 S- and 34 S- for each pixel (128x128 pixel grids) for between 20 and 960 cycles.

RESULTS: The extraction procedure did not discernibly alter pyrite grain-size distributions. The appparent inter-grain variability in 34 S/32 S in 1-4 μm-sized pyrite and marcasite fragments from isotopically homogeneous hydrothermal crystals was ~ ±2‰ (1σ), comparable with the standard error of the mean for individual measurements (≤ ±2‰, 1σ). In contrast, grain-specific 34 S/32 S ratios in modern and ancient sedimentary pyrites and marcasites can have inter- and intra-grain variability >60‰. The distributions of intra-sample isotopic variability are consistent with bulk 34 S/32 S values.

CONCLUSIONS: SIMS analyses of isolated iron sulfide grains yielded distributions that are isotopically representative of bulk 34 S/32 S values. Populations of iron sulfide grains from sedimentary samples record the evolution of the S-isotopic composition of pore water sulfide in their S-isotopic compositions. These data allow past local environmental conditions to be inferred.