Radiocarbon dating small carbonate samples with the gas ion source of AixMICADAS

International audienceSince the beginning of AMS, ⁠ 14 C analyses in foraminifera have been widely used to date deep-sea cores and to estimate reservoir ages of surface and deep ocean layers. Classical AMS requires acid hydrolysis of carbonate samples followed by graphitization of solid targets. For dating very small carbonate samples, an alternative route is to bypass the graphitization step by means of a CO ⁠ 2 gas ion source coupled to a carbonate hydrolysis system. We report tests and developments performed over 3 years with a carbonate handling system (CHS) coupled to AixMICADAS. Gas source AMS measurements were performed after different pretreatments applied to various carbonate samples including blanks, reference materials, corals and foraminiferal samples ranging from about 5 μg to 100 μg of carbon. In parallel, the CHS coupled to the automated graphitization system (AGE3, IonPlus AG) was used to develop a similar leaching procedure for carbonates dated with solid targets. The CHS-AGE3 system can be used efficiently and precisely for samples larger than 300 μgC and less precisely down to 100 μgC by applying a constant contamination correction. The gas method was thoroughly investigated in order to characterize its performances and limitations. Based on measuring numerous IAEA-C1 blank samples pretreated with hydrochloric acid, an equivalent age of about 52,000 yr BP is routinely achieved, though an even lower background close to 55,000 yr BP is reached when the system is flushed to suppress the memory effect from sample to sample. The long-term reproducibility is about 7‰ based on replicated analyses of a modern coral and the accuracy of the gas method was also confirmed by measuring coral and foraminifera samples previously dated by conventional AMS. The background correction is negligible for samples down to 30 μgC and can be effectively corrected for samples down to 10 μgC by applying a constant contamination correction though the recommended samples size is between 30 μgC and 100 μgC to achieve a reasonable precision. Our methods were applied to date planktonic foraminifera from two deep-sea cores. Several species with different visual aspects were tested for sample sizes ranging from single shells (ca. 10 μgC) to standard size sample (ca. 100 μgC) for the gas source as well as larger samples (ca. 500 μgC) measured as graphite targets. The dataset does not reveal any large age bias, but illustrates how gas CO ⁠ 2 measurements on small foraminiferal samples could improve the reliability of deep-sea core dating by providing key information on sample heterogeneities due to sediment mixing and diagenesis