www.acsprf.org

Summary  Our PRF-funded research focuses on the carbonate clumped-isotope thermometer, specifically on determining rates of solid-state reordering of C-O bonds in calcite.  Over the past year, graduate student Greg Henkes and I have conducted intensive laboratory work and modeling to develop the first thorough look at the kinetics of C-O bond reordering in calcite.  Our findings,  presented at the 2^nd Clumped Isotope Workshop held in London in August 2011, set the stage for thinking about this important but complex process.  The results are currently being prepared for publication.  During the course of the project, Greg has mastered laboratory skills related to high-temperature reaction experiments, and to high-precision clumped isotope mass spectrometry.  He has also made his first foray into chemical kinetics as necessary for understanding and modeling our rate data.  This project has contributed to the successful ‘break-in’ of our clumped-isotope mass spectrometry lab, which is now recognized as one of a handful of labs globally that is successfully making the clumped isotope measurement.  Greg’s training has allowed him to assist laboratory visitors, and thus his support from PRF indirectly helps to boost the national research infrastructure.  Based on preliminary findings, we crafted a NSF proposal intended to support a major direction of Greg’s PhD research.  The proposal was not funded, but received six ‘excellents’, one ‘very good’ and one ‘fair’, and we are confident that this proposal will find success upon future resubmissions.  The PRF funding allowed us to build a modest hydrothermal lab for conducting high T, high P experiments.  This equipment allows us to conduct the proposed research, and provides the opportunity to expand our research program into other aspects of high T, high P geochemistry.

Scientific Findings The carbonate clumped isotope thermometer represents a significant advance over the traditional carbonate oxygen isotope thermometer in that it can simultaneously record the temperature of carbonate mineralization, and the oxygen isotopic composition of the fluid in which the mineral formed.   The ability to constrain T and d¹⁸O_w during multiple generations of carbonate mineralization / diagenesis will allow burial, maturation, and exhumation histories of sedimentary basins to be teased apart in unprecedented detail.  However, in order to fully realize the potential of clumped isotope thermometry, we must understand the processes that can alter patterns of isotopic clumping subsequent to mineralization.  One broad class of such diagenesis is solid-state reordering of C-O bonds in the calcite lattice via self-diffusion of C and O.  Questions we seek to answer include: at what burial temperatures will a mineral begin to loose its primary temperature information?  What are the closure temperatures of carbonate minerals, and how do they relate to cooling rate?  Do rates of C-O bond reordering vary according to water content of the system?

            In order to address these questions, we are performing laboratory experiments to determine rates of solid state isotopic reordering reactions in calcite summarized as:

Ca¹²C¹⁸O¹⁶O₂ + Ca¹³C¹⁶O₃ <--> Ca¹³C¹⁸O¹⁶O₂ + Ca¹²C¹⁶O₃     (1)

Where the first species on the right-hand side is the ‘clumped’ isotopologue, the abundance of which decreases with increasing temperature.  Our approach follows traditional methods of determining reaction rates:  we react (= heat) samples at a number of different temperatures for different lengths of time [this reaction proceeds from right to left as depicted in (1)]; we determine the rate constant of the reaction at each reaction temperature; finally, we use these rate constants to determine Arrhenius parameters (activation energy E_a and frequency factor A_o) that allow prediction of reaction rates at different temperatures.

            To date we have studied two calcites in detail: a low temperature groundwater spar calcite (13 reaction temperatures and 86 temperature-time combinations), and an optical-grade calcite (5 reaction temperatures and 32 temperature-time combinations).  The reactions were carried out at low pressure (< 3 ATM) in a pure CO₂ atmosphere.  The clumped isotope composition of each sample was determined by gas source mass spectrometry using CO₂ as the analyte (CO₂ is generated from calcite by reaction with phosphoric acid).   As expected, the rate of reaction (1) increases with increasing temperature.  What was not expected, and which constitutes a key result of our studies, is that the reaction does not follow simple first-order kinetics:  the initial reaction is faster, and the later reaction is slower than would be predicted by a first-order reaction progress model.  Second- and third-order reaction progress models also fail to reproduce the data.

            We have developed two models to explain the non-first-order behavior of the system.  In the first model, the rate “constant” changes during the course of each reaction because of progressive annealing of lattice defects.  Specifically, the rate constant decreases with time as the lattice defects necessary to promote C-O bond reordering are annihilated or trapped.  In the second model, the calcites are actually mechanical mixtures of calcites with differing defect chemistry, and each kind of calcite has a unique susceptibility to C-O bond reordering.  Using this concept, we are able to accurately model our experimental data as a three-component system, where each component follows a first-order rate law but differs in activation energy and frequency factor.  In reality, both models are plausible and probably contribute to the overall kinetics of C-O reordering:  it is well-known that rates of self-diffusion of O and C may differ greatly in annealed and unannealed minerals (first model), and our own experimental results show that two different calcites have different susceptibilities to C-O reordering, consistent with subtle differences in chemistry and structure leading to differing defect chemistry and hence rates of C-O reordering (second model).

            We have also completed an initial set of experiments to determine rates of C-O reordering under high-pressure (100 MPa) water-saturated conditions.  These experiments were conducted using hydrothermal equipment purchased with PRF funds.  The salient result from these experiments is that reaction rates are identical for wet/high pressure reactions and dry/low pressure reactions.  This result has implications for the mechanism(s) responsible for increased O self-diffusion in minerals under wet conditions, and is a topic that we will pursue during the next year.