Preparation, Characterization, and Value Assignment of Carbon Dioxide Isotopic Reference Materials:  RMs 8562, 8563, and 8564

Three carbon dioxide isotopic reference materials (RM 8562, RM 8563, and RM 8564) have been prepared and characterized for composition, heterogeneity, and stability using dual-inlet isotope ratio mass spectrometry. RMs 8562−8564 are not certified, but their use as proxies for carbonate standards allows users to realize precisely the Vienna PDB scale and to provide comparability of stable carbon and oxygen isotope ratio data obtained by investigators in different laboratories. Samples (∼400 μmol each) are sealed in borosilicate glass tubes (9-mm outer diameter) for distribution by the National Institute of Standards and Technology (NIST) and the International Atomic Energy Agency (IAEA). δ13C and δ18O value assignments and uncertainties for the CO2 RMs were determined by two complementary methods: (1) through an international comparison exercise, to allow single-point standardization onto the VPDB scale and (2) through isotopically verified measurements of relative (Δδ) values between RMs (Δδ45CO2, Δδ46CO2, and Δδ47CO2). The intercomparison exercise involved thirteen international laboratories; measurements were normalized by carbonate and other RMs. The Δδ values were measured at NIST, where instrument design and measurement protocols allowed isotopically consistent determinations of δ45CO2, δ46CO2, and δ47CO2, a necessary condition of measurement accuracy. The value assignments (and combined standard uncertainties) for the δ13CVPDB and δ18OVPDB values are, respectively, −3.76 ‰ (0.03 ‰) and −8.45 ‰ (0.11 ‰) for RM 8562, −41.56 ‰ (0.06 ‰) and −23.72 ‰ (0.11 ‰) for RM 8563, and −10.45‰ (0.04 ‰) and +0.19 ‰ (0.10 ‰) for RM 8564. Within any batch (each consisting of over 1000 samples), sample-to-sample isotopic variations (standard deviations) were measured to be less than 0.007 ‰ and 0.026 ‰ for δ13C and δ18O, respectively. No conclusive evidence of long-term isotopic drift in these RMs has been observed after two years of monitoring.