Improved accuracy of oxygen and hydrogen isotope analysis in speleothem fluid inclusions: the importance of crusher temperature

Speleothem fluid inclusion isotope analysis provides the oxygen and hydrogen isotope composition of the parent water from which the carbonate precipitated (d¹⁸O_FL; d²H_FL). In contrast to carbonate isotopes, it is not affected by kinetic isotope effects. Fluid inclusion isotopes can be analyzed by crushing heated speleothem fragments and measuring the isotopic composition of the released water vapor. However, during this process, analytical artifacts related to pre-crushing evaporation and/or post-crushing adsorption can occur, potentially skewing the isotope values away from their origin and biasing temperatures calculated from the combination of d¹⁸O_FL and d¹⁸O_Cc. In d²H-d¹⁸O cross-plots, analytical (pre-crushing) evaporation has been suggested to induce very shallow slopes down to 1.4, lower than trends induced by evaporation under natural conditions in the soil or inside the cave.

            In this study, we used a Picarro L2140i isotope analyzer with an artificially generated moist background setup to examine the effect of analytical evaporation by quantifying the water loss prior to analysis when applying different crushing temperatures. We targeted two layers with different calcite fabrics from a flowstone of Touhami Cave (GTOF2), Morocco, as well as speleothems from Scladina Cave, Belgium and Bloukrantz Cave, South Africa.

The samples have different water contents and show different isotope effects of analytical evaporation that highly depend on the crushing temperature. Our results indicate that high water content samples (>1-2 µL/g) are generally more reliable compared to low yield samples (<0.5 µL/g), although high yield samples can be altered significantly by in crusher evaporation. In contrast to crushing at 110°C or 125°C, crushing at 90°C prevents most analytical evaporation in the samples we analyzed, increasing the sample water yield by up to 50%. Furthermore, for our low water content samples different crushing temperatures of 110°C and 125°C result in different evaporation slopes. At 110°C, the evaporation slope can even be parallel to the global meteoric water line. A potential explanation for these different evaporation slopes involves various amounts of adsorption of water to freshly crushed calcite powder, although this requires further exploration.

These findings have crucial implications, especially for low yield samples, because data that plot in the vicinity of the global meteoric water line are generally regarded as trustworthy. In our experiments, crushing at 90°C produces accurate d¹⁸O_FL, d²H_FL, and d-excess values for all high yield samples. Realistic cave air temperatures from combined d¹⁸O_Cc and d¹⁸O_FL analysis is retrieved from all samples analyzed, supported by consistent TEX₈₆ temperatures and modern-day drip water isotope compositions.