At the National Ocean Sciences AMS Facility (NOSAMS), samples containing 100-1000 µg C are treated routinely. Samples as small as 20 µg C can also be analyzed. Successful development of a method to analyze even smaller samples will open new areas of research, such as dating of ice core materials, studying black carbon and rock varnish, and studying atmospheric particles, to radiocarbon and expand existing ones, such as compound specific radiocarbon analysis (CSRA). For such applications we have developed a method in which a commercially available CO2 analyzer (Sable Systems International CA-2A) is used to quantify the dilution of samples. This replaces pressure-volume measurements that can be affected by trace impurities. Gases generated at variable levels during sample processing that can be inadvertently included in manometric measurements include N2O, HCl, and NO. Currie et al. (2000) developed a technique using on-line manometry/mass spectrometry to measure the dilution factor accurately and were able to measure the fm of samples containing approximately 3 µg C. Our analyzer uses non-dispersive infrared radiation (NDIR) to accurately and precisely measure CO2 non-destructively. The precision is well within the range required for our analyses and results in the error in the dilution ratio contributing negligibly to the error in fm. To test our system, we measured the ratio of 13C to 12C, the stable natural isotopes of carbon, on gas mixtures we prepared. Using 13C allowed us to confirm that the CO2 analyzer was measuring the dilution ratio properly and with high precision. Our data indicate the method is robust to a dilution factor of 6. Initial tests on two samples converted to graphite and analyzed on the AMS gave good results as well. These results suggest that we can successfully prepare samples containing as little as 4 µg C for radiocarbon analysis and our method should likely be useful for samples containing as little as 1-2 µg C. Consideration of all factors contributing to the uncertainty in the measurement limits the method to samples younger than 10,000 yrs old. The technique holds promise for environmental studies, for studies tracing bomb carbon in the natural carbon cycle, and for evaluating sample processing blanks.
Currie L.A., J.D. Kessler, J.V. Marolf, A.P. McNichol, D.R. Stuart and J.C. Donoghue (2000) Low-level (submicromole) environmental 14C metrology. Nucl. Instr. And Meth. In Phys. B 172. 440-448.
See more of Compound Specific Analysis and Preparation
See more of The 10th International Conference on Accelerator Mass Spectrometry (September 5-10, 2005)