Protocols for in situ measurement of oxygen isotopes in goethite by ion microprobe

Abstract

We present protocols for high-spatial resolution measurement of oxygen isotope ratios of goethite (α-FeOOH) with the Sensitive High Mass Resolution Ion Microprobe – Stable Isotopes (SHRIMP-SI) and propose a natural sample as a potential goethite reference material (RM) for ion microprobe analysis. We assess the effects of goethite chemical composition, crystallographic orientation, and texture on the accuracy and repeatability of SHRIMP-SI δ18O (δ18OSIMS) results. Synthetic goethites evaluated as potential δ18OSIMS RM are powdery, porous, and finely crystalline; they do not yield repeatable results. A dense colloform goethite from the Capão topaz mine, Minas Gerais, Brazil, fulfills major prerequisites: it is stoichiometrically relatively pure, yields repeatable oxygen isotope results, and occurs in abundance to produce a RM for long-term use. We use an average laser fluorination δ18OLF-VSMOW value of −17.3 ± 0.3‰ (1SD) obtained for five aliquots of this RM to normalize all δ18OSIMS measurements. Multiple δ18OSIMS analyses of a large fragment of the Capão L4 (CL4) RM analyzed in three different runs yield an overall repeatability of −17.3 ± 0.5‰ (2SD, n = 294) for all three runs combined. Natural variability and crystal orientation effects are the main reasons for the excess spread of the δ18OSIMS results compared to the spot internal precision (ca. 0.2‰). All δ18OSIMS analyses (n = 1027) in various aliquots of CL4, randomly oriented and analyzed in 26 sessions during eight distinct runs, yield an overall repeatability of −17.3 ± 0.7‰ (2SD), confirming that CL4 is a suitable SIMS RM. After ascertaining its suitability as a RM, we used CL4 to standardize analyses of other natural goethite samples with the SHRIMP-SI and compared δ18OSIMS and laser fluorination results to test the relationship between natural properties (e.g., porosity, minor elements substituting for Fe), preparation procedures (e.g., polish and relief), instrument conditions, and the overall precision and accuracy of the SIMS analyses. Samples containing minor elements substituting for Fe (e.g., Al, Mn, Cu, Zn, etc.) or as contaminants (e.g., Si, P) require significant matrix corrections. Because we could not find homogenous natural goethite samples showing a large range in metal concentrations, we extrapolate our calibration curves beyond the composition of our calibration goethite samples. δ18OSIMS results corrected for instrument mass fractionation (using CL4) and compositionally dependent matrix effects (using several calibration goethites of known elemental composition) are less precise but statistically indistinguishable from their laser fluorination results. However, porous samples are unsuitable for SHRIMP-SI δ18O analysis. Dense colloform samples yield repeatable results for individual growth bands, showing that the high spatial resolution, moderate precision, and speed of analysis of the SHRIMP-SI can resolve variations in oxygen isotope composition acquired during sample growth. (U[sbnd]Th)/He geochronology of equivalent aliquots from the same goethite samples reveal that the combination of the two methods permits the extraction of temporal variation in the isotopic compositions of meteoric solutions in the geological past.