Precise carbon isotopic ratio analyses of micro amounts of carbonate and non-carbonate in basalt using continuous-flow isotope ratio mass spectrometry.

RATIONALE: Continuous-flow isotope ratio mass spectrometry (CF-IRMS) is a specialized technique used to quickly analyze very small amounts of sample. We have used CF-IRMS to assess the influences of sample weight and relative carbon content on the accuracy and precision of the δ13 C values of micro amounts of carbonate and non-carbonate in silicate rocks. METHODS: The analytical work was performed on a Gasbench II (GB) sample preparation device and on an Elemental Analyzer (EA), which were both interfaced to CF-IRMS instruments. Potential silicate matrix effects on the carbon isotopic analyses were investigated by measuring mixtures of calibrated carbon reference materials and quartz powder. The calibration lines, established by the measured raw values and the known values of three reference materials mixed with quartz powder, were used to calibrate the δ13 C values of basalt samples from eastern China. RESULTS: The δ13 C values measured by GB-CF-IRMS of one national carbonate reference material, GBW04416, deviate slightly from the known value for approximately 20-70 µg of carbonate contained in 4.5-mL vials; the smaller the sample size, the lower the measured δ13 C values. External precision better than 0.1‰ (1σ, n = 26) is achieved at a signal intensity for mass 44 of between 868 and 1614 mV, corresponding to a sample weight of 30.8-50.2 µg, whereas it is reduced to 0.27‰ (1σ, n = 34) at a signal intensity between 519 and 1614 mV, corresponding to a sample weight of 21.1-50.2 µg. In the EA-CF-IRMS experiments for non-carbonate carbon, at high carbon concentration (greater than 800 ppm) and at optimum sample weights, the accuracy and precision are both better than 0.2‰. For carbon concentrations less than 500 ppm, the measured δ13 C values deviate from the average by up to -1.2‰ and the precision is 0.74‰. CONCLUSIONS: The measured δ13 C values decrease substantially at lower carbon concentration and higher sample weights, and poorer precision is obtained. Suggestions are made to measure repeatedly the same carbon concentration of sample and reference materials in order to obtain not only reproducible, but also accurate carbon isotope ratios.