Correction for the 17O interference in δ(13C) measurements when analyzing CO2 with stable isotope mass spectrometry

Measurements of δ(¹³C) determined on CO₂ with an isotope-ratio mass spectrometer (IRMS) must be corrected for the amount of ¹⁷O in the CO₂. For data consistency, this must be done using identical methods by different laboratories. This report aims at unifying data treatment for CO₂ IRMS by proposing (i) a unified set of numerical values, and (ii) a unified correction algorithm, based on a simple, linear approximation formula. Because the oxygen of natural CO₂ is derived mostly from the global water pool, it is recommended that a value of 0.528 be employed for the factor λ, which relates differences in ¹⁷O and ¹⁸O abundances. With the currently accepted N(¹³C)/N(¹²C) of 0.011 180(28) in VPDB (Vienna Peedee belemnite) reevaluation of data yields a value of 0.000 393(1) for the oxygen isotope ratio N(¹⁷O)/N(¹⁶O) of the evolved CO₂. The ratio of these quantities, a ratio of isotope ratios, is essential for the ¹⁷O abundance correction: [N(¹⁷O)/N(¹⁶O)]/[N(¹³C)/N(¹²C)] = 0.035 16(8). The equation [δ(¹³C) ≈ ⁴⁵δ_VPDB-CO2 + 2 ¹⁷R/¹³R (⁴⁵δ_VPDB-CO2 – λ⁴⁶δ_VPDB-CO2)] closely approximates δ(¹³C) values with less than 0.010 ‰ deviation for normal oxygen-bearing materials and no more than 0.026 ‰ in extreme cases. Other materials containing oxygen of non-mass-dependent isotope composition require a more specific data treatment. A similar linear approximation is also suggested for δ(¹⁸O). The linear approximations are easy to implement in a data spreadsheet, and also help in generating a simplified uncertainty budget.