Kinetic clumped isotope fractionation in the DIC-H2O-CO2 system: Patterns, controls, and implications

The carbonate clumped isotope thermometer is a powerful tool in paleoclimate research, because it constrains carbonate formation temperature based on the extent of 13C and 18O clumping within the carbonate lattice (Δ47) and thus does not require knowledge of the isotopic composition of the water from which carbonates precipitate. However, there is growing evidence that kinetic processes in the precipitating solution, particularly the slow inter-conversion between CO2 and HCO3−, can cause deviations of the clumped isotope composition of dissolved inorganic carbon (DIC) from their expected equilibrium values, leading to disequilibrium clumped isotope composition in the carbonate precipitates. Such disequilibrium effects impede the application of the carbonate clumped isotope thermometer to important climate archives (e.g., speleothems and corals), and hinder the realization of the full potential of this novel thermometer. Here, I systematically examine the patterns and controls of kinetic clumped isotope fractionations (i.e., Δ47, Δ48 and Δ49) in the DIC-H2O-CO2 system through first-principles theoretical calculations and numerical modeling (IsoDIC), and explore their implications for the carbonate clumped isotope thermometry. I show that, in contrast to the large carbon and oxygen isotope effects, intrinsic clumped isotope fractionations associated with CO2 hydration and hydroxylation and their reverse reactions (i.e., HCO3− dehydration and dehydroxylation) will lead to only relatively small disequilibrium Δ47 effects in the HCO3− that are directly produced or consumed by these reactions. Instead, the disequilibrium Δ47 effects observed in most natural and laboratory-synthesized carbonates arise in large part from the mixing or removal of DIC pools with similar Δ47 compositions but distinct bulk carbon and oxygen isotope compositions. Further, my model simulations show characteristic evolutions of the clumped isotope composition of DIC during three common isotope fractionation processes, i.e., DIC-H2O isotope exchange, CO2 degassing, and CO2 absorption, and predict correlated enrichments in δ13C, δ18O and Δ48 but depletions in Δ47 and Δ49 of DIC during the early stage of CO2 degassing and vice versa during CO2 absorption, yielding typical disequilibrium Δ47/δ18O, Δ48/δ18O and Δ47/Δ48 slopes of about −0.03, 0.03, and −1.0 for CO2 degassing and about −0.02, 0.04, and −0.6 for CO2 absorption. While both the physicochemical condition (e.g., T, pH, [DIC], pCO2) and the isotopic composition (e.g., δ13C, δ18O, Δ47, Δ48, Δ49) of the aqueous solution and air CO2 can affect the magnitudes of these disequilibrium isotope effects, the correlations among these disequilibrium effects are relatively insensitive to changes in most environmental parameters except the isotopic composition of the aqueous solution and air CO2. The quantitative agreements between my model results and the existing observations from laboratory experiments and natural carbonates suggest that the model captures the key processes governing clumped isotope fractionations in the DIC-H2O-CO2-CaCO3 system, and support the development of novel approaches, e.g., coupled Δ47-Δ48 measurements, to discern the kinetic processes involved in carbonate formation and to correct for disequilibrium clumped isotope effects in carbonate archives and derive accurate estimates of their formation temperatures.