Carbon dioxide levels in Earth?s atmosphere, and consequently ocean temperatures, are rising.  How high and how fast ocean temperatures can rise can be learned from temperature measurements of ancient oceans. One of the most promising techniques for measuring ancient ocean temperatures is based on the co-enrichment of rare heavy carbon and heavy oxygen isotopes (*termed clumped isotopes) in calcium carbonate (CaCO₃) like that of fossil shells.  Unfortunately, such paleotemperature determinations made using fossil shells can be reset when sediments, buried by younger sediments, are exposed to higher temperatures at depth in the Earth. Our study examines the factors controlling the rate of this clumped isotope resetting (termed reordering) using experimental and theoretical methods. These efforts to better quantify rates of clumped isotope reordering will lead to more accurate knowledge of ocean temperature extremes.

Our experiments started with the synthesis of mineral powders in the laboratory (Figure 1). These powders were then heated at temperatures as high as 385 ?C (725 ?F) for different durations up to 96 hours and measured for alteration of the original clumped isotope composition (Figure 2).  The samples were carbonate minerals with either trigonal structure (calcite, CaCO₃) or rhombohedral structure (aragonite, CaCO₃; strontianite, SrCO₃; witherite, BaCO₃). Experiments yielded the following progression of reordering rates, from fastest to slowest: strontianite, witherite, aragonite, and calcite (Figure 3). These experiments demonstrate the importance of crystal structure (specifically, the connectivity between atoms in minerals) and chemical composition in controlling clumped isotope reordering.

In concert with the experiments, we attempted to simulate the reordering process on supercomputers considering interactions of electrons, atoms, and molecules at the atomic level. The approaches used, density functional theory (DFT) and ab-initio molecular dynamics (AIMD), permit testing of the effect of chemical impurities in the mineral structure, especially water, on clumped isotope reordering. Such tests can help validate trends in experimental data and direct future experiments. Furthermore, theoretical models provide an atomistic view of possible movements of heavy oxygen atoms in the crystal structure, revealing the structural changes preceding clumped isotope reordering. Rearrangements in the connectivity of atoms across multiple possible transition states is examined to decipher the most likely pathway for reordering.  

First-principles simulations reveal that substitution of magnesium and water in the calcite structure facilitated clumped isotope reordering by lowering the energy needed for exchange of heavy oxygen between CaCO₃ molecules. The steps involved in water-mediated oxygen diffusion are shown in Figure 4.  We also discovered a bi-tetrahedral intermediate structure involved in oxygen exchange between molecules (Figure 5). Lastly, AIMD models show that the free energy changes required to reach the transition state for oxygen exchange follow the experimentally-determined sequence of clumped isotope reordering (from fastest to slowest): strontianite, witherite, and calcite.  These results will help scientists identify the best fossils for paleotemperature studies and thus produce more accurate reconstructions of ancient ocean temperatures, setting constraints on modern global warming.

This study has broader impacts beyond providing valuable insight into a process important for measurement of ancient ocean temperatures. Accurate clumped isotope reordering rates are also essential for the application of clumped isotopes to reconstruct the thermal histories of sedimentary basins, information critical for petroleum exploration. Several graduate students have worked on different aspects of cross-disciplinary study, learning and exchanging knowledge in isotope geochemistry, experimental geochemistry, mass spectrometry, density functional theory, and molecular dynamics.  The results of this study have resulted in four presentations at national and international conferences, and two manuscripts (one in press and one in revision).

*Isotopes are atoms with the same number of protons but different numbers of neutrons. In our study, the rare heavy isotopes are oxygen-18 and carbon-13, whereas the common light isotopes are oxygen-16 and carbon-12. Thus, for clumped isotopes we measure the relative abundance of Ca¹³C¹⁸O¹⁶O₂.

Last Modified: 05/31/2023
Modified by: Ethan L Grossman