Beyond temperature: Clumped isotope signatures in dissolved inorganic carbon species and the influence of solution chemistry on carbonate mineral composition
"Clumped-isotope" thermometry is an emerging tool to probe the temperature history of surface and subsurface environments based on measurements of the proportion of ^(13)C and ^(18)O isotopes bound to each other within carbonate minerals in ^(13)C^(18)O^(16)O_2^(2−) groups (heavy isotope "clumps"). Although most clumped isotope geothermometry implicitly presumes carbonate crystals have attained lattice equilibrium (i.e., thermodynamic equilibrium for a mineral, which is independent of solution chemistry), several factors other than temperature, including dissolved inorganic carbon (DIC) speciation may influence mineral isotopic signatures. Therefore we used a combination of approaches to understand the potential influence of different variables on the clumped isotope (and oxygen isotope) composition of minerals. We conducted witherite precipitation experiments at a single temperature and at varied pH to empirically determine ^(13)C–^(18)O bond ordering (Δ_(47)) and δ^(18)O of CO_3^(2−) and HCO_3^− molecules at a 25 °C equilibrium. Ab initio cluster models based on density functional theory were used to predict equilibrium ^(13)C–^(18)O bond abundances and δ^(18)O of different DIC species and minerals as a function of temperature. Experiments and theory indicate Δ_(47) and δ^(18)O compositions of CO_3^(2−) and HCO_3^− ions are significantly different from each other. Experiments constrain the Δ_(47)–δ^(18)O slope for a pH effect (0.011 ± 0.001; 12 ⩾ pH ⩾ 7). Rapidly-growing temperate corals exhibit disequilibrium mineral isotopic signatures with a Δ_(47)–δ^(18)O slope of 0.011 ± 0.003, consistent with a pH effect. Our theoretical calculations for carbonate minerals indicate equilibrium lattice calcite values for Δ_(47) and δ^(18)O are intermediate between HCO_3^− and CO_3^(2−). We analyzed synthetic calcites grown at temperatures ranging from 0.5 to 50 °C with and without the enzyme carbonic anhydrase present. This enzyme catalyzes oxygen isotopic exchange between DIC species and is present in many natural systems. The two types of experiments yielded statistically indistinguishable results, and these measurements yield a calibration that overlaps with our theoretical predictions for calcite at equilibrium. The slow-growing Devils Hole calcite exhibits Δ_(47) and δ^(18)O values consistent with lattice equilibrium. Factors influencing DIC speciation (pH, salinity) and the timescale for DIC equilibration, as well as reactions at the mineral–solution interface, have the potential to influence clumped-isotope signatures and the δ^(18)O of carbonate minerals. In fast-growing carbonate minerals, solution chemistry may be an important factor, particularly over extremes of pH and salinity. If a crystal grows too rapidly to reach an internal equilibrium (i.e., achieve the value for the temperature-dependent mineral lattice equilibrium), it may record the clumped-isotope signature of a DIC species (e.g., the temperature-dependent equilibrium of HCO_3^−) or a mixture of DIC species, and hence record a disequilibrium mineral composition. For extremely slow-growing crystals, and for rapidly-grown samples grown at a pH where HCO_3^− dominates the DIC pool at equilibrium, effects of solution chemistry are likely to be relatively small or negligible. In summary, growth environment, solution chemistry, surface equilibria, and precipitation rate may all play a role in dictating whether a crystal achieves equilibrium or disequilibrium clumped-isotope signatures.
© 2015 Elsevier. Received 28 July 2014; accepted in revised form 18 June 2015; available online 7 July 2015. AKT thanks the reviewers and editor for their comments, as well as Kate Ledger, Hagit Affek, Tobias Kluge, James Watkins, Jim Rustad, William Casey, Oleg Pokrovsky, Ian Fairchild, Bruce Watson, Henry Teng, Philippe Van Cappellen, Adrian Villegas-Jimenez, Andreas Luttge, Bernhardt Trout, Michael Reddy, Dan Breecker, Andrew Dickson, Gideon Henderson, Frank Millero, Justine Kimball, and Chris Roberts for discussions relevant to this work. AKT and PSH thank Ben Elliott, Anastassia Alexandrova, and Ben Schwartz for input, and Fernando R. Clemente of Gaussian, Inc., for helpful suggestions with Gaussian09. AKT acknowledges support from the Department of Energy through BES grant DE-FG02-13ER16402, a UCLA Career Development Award, a Hellman Fellowship, NSF grants EAR-0949191, EAR-1325054, ARC-1215551, and ACS grant #51182-DNI2. RAE and JBR acknowledge support from NSF grant OCE-1437166. REZ was supported by NSF grant OCE09-27089. JBR acknowledges support from NSF grant OCE-1357665 and NOAA grant NA13OAR4310186. The support of the U.S. Geological Survey National Research Program made this article possible. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Supplemental Material - mmc1.xlsx