Assessing the effects of embedding resins on carbonate stable and clumped isotope analyses

Abstract

Rationale: Embedding resins are widely used to fix carbonates for high-precision sample preparation and high-resolution sampling. However, these embedding materials are difficult to remove after sample preparation and are known to affect the accuracy of carbonate stable isotope analyses. Nevertheless, their impact on clumped isotope analysis, which is particularly sensitive to contamination artifacts, has so far not been tested. The observation that running resin-containing samples decreased the reproducibility of clumped isotope values for internal laboratory carbonate standards and increased the external standard deviation (SD 0.061–0.088‰) compared to the long-term observations (0.034‰), prompted us to set up an experiment to test the influence of resin addition on instrument performance. Methods: Here we analyzed the stable and clumped isotope composition of a pure calcium carbonate standard (ETH-4) mixed with three types of embedding resins in 2:1 and 1:1 proportions. Our aim was to assess how resin addition affects isotope analyses. Results: We found that none of the stable isotopic values were significantly different. The δ13C values were −10.22 ± 0.07‰ (mean ± SD) for pure ETH-4, while the δ13C values of ETH-4 mixed with embedding resins in 2:1 and 1:1 proportions were −10.21 ± 0.06‰ and −10.18 ± 0.06‰, respectively (p > 0.05). The δ18O values were −18.82 ± 0.11‰ for pure ETH-4 versus −18.81 ± 0.09‰ and −18.82 ± 0.08‰ for 2:1 and 1:1 ETH-4:resin mixtures, respectively (p > 0.05). Given the large uncertainty in our results, we did not find significant differences between different mixtures in the carbonate clumped isotope values (Δ47), with 0.458 ± 0.107‰, 0.464 ± 0.086‰, and 0.417 ± 0.089‰ in pure ETH-4 and ETH-4 with 2:1 and 1:1 resin mixtures, respectively (p > 0.05). However, a resin-related bias in the results might be masked by the large uncertainty. The measured ETH-4 values in our study are similar to the InterCarb values (δ13C = −10.20‰, δ18O = −18.81‰, Δ47 = 0.450‰, InterCarb-Carbon Dioxide Equilibrium Scale). However, the external SD of Δ47 in sessions measuring ETH-4 with resins is higher than in sessions without deliberate resin addition for the same measuring period. Conclusions: We find that the potential contamination from the resin addition leads to a larger variability for Δ47 values in sessions measuring ETH-4 including resins. We therefore recommend purification of embedded samples using a contamination trap with Porapak prior to analysis, if possible, or avoiding resins during sample preparation and workup, as well as monitoring the measurement quality during and after sessions with samples containing embedding resins.