^(13)C^(18)O clumping in speleothems: Observations from natural caves and precipitation experiments

The oxygen isotope composition of speleothems is an important proxy of continental paleoenvironments, because of its sensitivity to variations in cave temperature and drip water δ^(18)O. Interpreting speleothem δ^(18)O records in terms of absolute paleotemperatures and δ^(18)O values of paleo-precipitation requires quantitative separation of the effects of these two parameters, and correcting for possible kinetic isotope fractionation associated with precipitation of calcite out of thermodynamic equilibrium. Carbonate clumped-isotope thermometry, based on measurements of Δ_47 (a geochemical variable reflecting the statistical overabundance of ^(13)C^(18)O bonds in CO_2 evolved from phosphoric acid digestion of carbonate minerals), potentially provides a method for absolute speleothem paleotemperature reconstructions independent of drip water composition. Application of this new technique to karst records is currently limited by the scarcity of published clumped-isotope studies of modern speleothems. The only modern stalagmite reported so far in the literature yielded a lower Δ_47 value than expected for equilibrium precipitation, possibly due to kinetic isotope fractionation. Here we report Δ_47 values measured in natural speleothems from various cave settings, in carbonate produced by cave precipitation experiments, and in synthetic stalagmite analogs precipitated in controlled laboratory conditions designed to mimic natural cave processes. All samples yield lower Δ_47 and heavier δ^(18)O values than predicted by experimental calibrations of thermodynamic equilibrium in inorganic calcite. The amplitudes of these isotopic disequilibria vary between samples, but there is clear correlation between the amount of Δ_47 disequilibrium and that of δ^(18)O. Even pool carbonates believed to offer excellent conditions for equilibrium precipitation of calcite display out-of-equilibrium δ^(18)O and Δ_47 values, probably inherited from prior degassing within the cave system. In addition to these modern observations, clumped-isotope analyses of a flowstone from Villars cave (France) offer evidence that the amount of disequilibrium affecting Δ_47 in a single speleothem can experience large variations at time scales of 10 kyr. Application of clumped-isotope thermometry to speleothem records calls for an improved physical understanding of DIC fractionation processes in karst waters, and for the resolution of important issues regarding equilibrium calibration of Δ_47 in inorganic carbonates