STABLE ISOTOPE SYSTEMATICS OF CALCITE by ELLEN K. OLSEN A DISSERTATION Presented to the Department of Earth Sciences and the Division of Graduate Studies of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy June 2023 DISSERTATION APPROVAL PAGE Student: Ellen K. Olsen Title: Stable Isotope Systematics of Calcite This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the Department of Earth Sciences by: James M. Watkins Chairperson/Advisor Mark H. Reed Core Member Paul J. Wallace Core Member Daniel G. Gavin Institutional Representative and Krista Chronister Vice Provost for Graduate Studies Original approval signatures are on file with the University of Oregon Division of Graduate Studies. Degree awarded June 2023 2 © 2023 Ellen K. Olsen 3 DISSERTATION ABSTRACT Ellen K. Olsen Doctor of Philosophy Department of Earth Sciences June 2023 Title: Stable Isotope Systematics of Calcite The oxygen isotopic composition of calcite is widely used in paleoclimate studies to infer the temperatures of carbonate formation across a wide range of geologic environments including hydrothermal veins, caves, lakes, surface oceans, and in marine sediments. Carbonate-based temperature reconstructions depend on empirical d18O-T relationships that are affected by factors such as carbonate growth rate, solution composition, pH, and source(s) of dissolved inorganic carbon (DIC). We carried out calcite growth experiments over a range of pH (7.5-12.8), temperature (T = 10-25°C), ionic strength (I = 0.1-1.6; [NaCl] = 0-1.4 M) and concentration of the enzyme carbonic anhydrase ([CA] = 0-3 µM). The enzyme CA promotes isotopic equilibration of the DIC pool, which in turn, has a strong influence on the isotopic composition of the mineral. We divide the experimental results into two categories: (1) calcite grown from an equilibrated DIC pool, and (2) calcite grown from a non-equilibrated DIC pool. Results from (1) are used to determine the kinetic isotope effects (KIEs) attending the crystal growth reaction as a function of pH and ionic strength. No evidence of an ionic strength effect on oxygen isotope partitioning between calcite and DIC was found for NaCl concentrations up to 0.35 M, but in higher ionic strength solutions, NaCl was found to inhibit the efficacy of CA and prevent complete isotopic equilibration of the DIC pool, resulting in lower and more variable oxygen isotope fractionations. The oxygen isotope partitioning between calcite and water was found to systematically decrease with increasing pH. Results from (2) are used to determine the KIEs attending the CO2 hydration and hydroxylation reactions as a function of T, pH, ionic strength and [CA]. This study is the first to separately quantify 4 the kinetic fractionation factors (KFFs) for CO2 and OH- separately during CO2 hydroxylation. The experimental results have been used to develop a generalizable model of oxygen isotope effects in the CaCO3-DIC-H2O system. The model can be used to predict the d18O of calcites grown from variably-equilibrated DIC pools and can explain why different experimental setups have yielded different d18O-T relationships for inorganic calcite. This dissertation includes previously published and unpublished co-authored material. 5 CURRICULUM VITAE NAME OF AUTHOR: Ellen K. Olsen GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene Western Washington University, Bellingham, WA DEGREES AWARDED: Doctor of Philosophy, Earth Sciences, 2023, University of Oregon Bachelor of Science, Geology, Thesis Option, 2015, Western Washington University AREAS OF SPECIAL INTEREST: Isotope Geochemistry Experimental Petrology Paleoclimate PROFESSIONAL EXPERIENCE: Graduate Employee, University of Oregon, 2016-2022 Research Assistant, Western Washington University, 2016 GRANTS, AWARDS, AND HONORS: Earth Sciences Research Recognition Award, University of Oregon, 2022 Geological Society of America Graduate Student Research Grant, 2021-2022 Central Oregon Geoscience Society Student Research Grant, 2021-2022 UO General University Scholarship, University of Oregon, 2021-2022 Marthe E. Smith Memorial Scholarship, University of Oregon, 2021-2022 6 Earth Sciences Good Citizen Award, 2021 Harvey E Lee Grad Scholarship, University of Oregon, 2020-2021 ArtSci Oregon Research as Art Exhibition winner, University of Oregon, 2020 Baldwin Scholarship, University of Oregon, 2018-2019 Marthe E. Smith Memorial Scholarship, University of Oregon, 2017-2018 Elma Hendricks Scholarship, University of Oregon, 2017-2018 Johnston Scholarship, University of Oregon, 2016-2017 Department of Geology Outstanding Graduate Award, Western Washington University, 2014-2015 Geology Department Advance for Research, Western Washington University, 2013-2014 Research and Sponsored Programs Undergraduate Student Grant, Western Washington University, 2013-2014 Skagit Rock and Gem Club Scholarship, 2013-2014 James L. Talbot Scholarship, Western Washington University, 2012-2013 President’s Scholarship, Western Washington University 2011-2013 PUBLICATIONS: Olsen, E.K., Watkins, J.M., and Devriendt, L.S. (2022). Oxygen isotopes of calcite precipitated at high ionic strength: CaCO3-DIC fractionation and carbonic anhydrase inhibition. Geochimica et Cosmochimica Acta 325, 170- 186. Devriendt, L.S., Mezger, E.M., Olsen, E.K., Watkins, J.M., Kaczmarek, K., Nehrke, G., de Nooijer, L.J., and Reichart, G.-J. (2021). Sodium incorporation into inorganic CaCO3 and implications for biogenic carbonates. Geochimica et Cosmochimica Acta 314, 294-312. 7 ACKNOWLEDGMENTS I am incredibly grateful to my adviser, James Watkins, for his support, patience, and mentorship that have helped me become a better scientist. Thank you, Jim, for always providing encouragement and quality pep talks when I needed them. Thank you to Mark Reed for insightful aqueous geochemistry conversations and for providing a bounty of neat rocks to look at. I would also like to sincerely thank Paul Wallace and Daniel Gavin for their insights and feedback. Thank you to Marla Trox, Dave Stemple, and Sandy Thoms, as well the UO Earth Sciences department, for their support in getting to the finish line. Thank you to Andrew Ross and Jennifer McKay at the OSU CEOAS Stable Isotope Laboratory, and James Palandri at the UO Stable Isotope Laboratory for isotope analyses and helpful technical conversations over the years. Special thanks to Edward Vinis for help with modifying the experimental design and conducting early experiments of Chapter V. I am so grateful for the intellectual discussions, collaboration, support, and camaraderie of my lab mates over the years, with particular thanks to Marisa Acosta, Molly Pickerel, Laurent Devriendt, and Edward Vinis. I cannot thank my friends enough for their insightful geologic conversations, as well as endless support, without which this would not have been possible. Special thanks to those from my original cohort, including Anne Fulton, Michelle Muth, and Michael Hudak, as well those met along the journey, with particular recognition for Ryan Seward, Kellum Tate-Jones, Annika Dechert, Christina Cauley, and Kate Scholz. Thank you to my mom, my dad, my sister, and my aunts, for all of their support and encouragement. And finally, a very special thanks to my cat, Crumpet. This research was supported by NSF grant no. EAR1749183 awarded to J.M. Watkins. 8 For Crumpet 9 TABLE OF CONTENTS Chapter Page I. INTRODUCTION .................................................................................................. 19 II. CACO3-DIC-H2O SYSTEM BACKGROUND ...................................................... 22 1. Carbonates in Seawater ................................................................................. 22 1.1 Seawater Composition and pH ................................................................. 23 2. Isotopic Fractionation .................................................................................... 24 2.1 Equilibrium vs. Kinetic Fractionation ...................................................... 24 2.2 Isotope and Fractionation Factor Notation .............................................. 25 2.3 d13C and d18O of Common Geologic Reservoirs ........................................ 26 3. Dissolved Inorganic Carbon Speciation .......................................................... 26 4. CaCO3-DIC-H2O Chemical Reactions and Isotopic Fractionations ............... 30 4.1 CO2 Dissolution ........................................................................................ 31 4.2 CO2 (De)hydration and (De)hydroxylation .............................................. 31 4.3 HCO3- (De)protonation ............................................................................ 34 4.4 CaCO3 Precipitation/Dissolution ............................................................. 36 4.5 H2O Dissociation ...................................................................................... 37 5. Isotopic Composition of Calcite ..................................................................... 38 5.1 Oxygen Isotopes: d18O-T .......................................................................... 38 5.2 Clumped isotopes: ∆47-T .......................................................................... 40 6. Paleoenvironment Applications ...................................................................... 40 10 Chapter Page 6.1 Trace Elements ........................................................................................ 42 7. Bridge ............................................................................................................ 44 III. OXYGEN ISOTOPES OF CALCITE PRECIPITATED AT HIGH IONIC STRENGTH: CACO3-DIC FRACTIONATION AND CARBONIC ANHYDRASE INHIBITION ..................................................................................... 45 1. Introduction ................................................................................................... 45 2. Methods ......................................................................................................... 46 2.1 Calcite Growth Experiments .................................................................... 46 2.2 d18O, d13C, and [DIC] Measurements ....................................................... 50 3. Results ........................................................................................................... 51 3.1 Oxygen Isotope Fractionation .................................................................. 51 4. Model for Calcite Growth from a DIC Pool with Variable Level of Isotopic Equilibration ..................................................................................... 53 4.1 Chemical Reactions .................................................................................. 55 4.2 CaCO3 Flux .............................................................................................. 56 4.3 CO2 Flux .................................................................................................. 58 5. Model Results ................................................................................................ 59 6. Discussion ...................................................................................................... 61 6.1 The Effect of Ionic Strength on the Oxygen Isotope Fractionation Between Calcite and the EIC ......................................................................... 61 6.2 The Origin of d18Ocalcite Variability at High Ionic Strength ...................... 61 6.2.1 Effect of Dissolved Ions on d18Ow .............................................. 61 11 Chapter Page 6.2.2 Effect of Ion Pairing on Oxygen Isotope Fractionation Between DIC Species and H2O ........................................................... 61 6.2.3 Effect of DIC Speciation on d18Ocalcite ........................................ 62 6.2.4 Effect of NaCl on the Kinetics of Catalyzed CO2 (De)hydration ................................................................................... 62 7. Implications .................................................................................................... 64 7.1 Towards a General Model for Kinetic Oxygen Isotope Effects ................. 64 7.2 Application to Marine Calcifiers .............................................................. 66 8. Summary ........................................................................................................ 66 9. Bridge ............................................................................................................ 67 IV. EFFECT OF pH ON THE ISOTOPIC COMPOSITION OF CALCITE GROWN FROM CO2-FED SOLUTIONS: EXPERIMENTS AND MODELING ..... 68 1. Introduction ................................................................................................... 68 2. Methods ......................................................................................................... 69 3. Experimental Results ..................................................................................... 70 3.1 Carbon Isotopic Fractionation ................................................................. 75 3.2 Oxygen Isotopic Fractionation ................................................................. 77 4. Updating the Ion-by-Ion Model ..................................................................... 78 5. Summary ........................................................................................................ 84 6. Bridge ............................................................................................................ 84 12 Chapter Page V. EXTREME ISOTOPIC FRACTIONATIONS IN CACO3 FROM A HYPERALKALINE SPRING ANALOG: QUANTIFYING KINETIC FRACTIONATIONS ATTENDING CO2 HYDROXYLATION ............................. 86 1. Introduction ................................................................................................... 86 1.1 CO2 Captured as CaCO3 .......................................................................... 87 1.2 OH--H2O Fractionation ............................................................................ 89 1.3 Motivation ................................................................................................ 89 2. Methods ......................................................................................................... 90 3. Results and Discussion ................................................................................... 95 3.1 Experiment Results .................................................................................. 95 3.1.1 CaCO3 Precipitates .................................................................... 95 3.1.2 Impacting of Varying Experimental Parameters ....................... 96 3.2 Isotopic Results ........................................................................................ 99 3.2.1 CO2, DIC, and CaCO3 ............................................................... 99 3.2.2 Outliers ...................................................................................... 102 3.2.3 CaCO3 Precipitated via CO2 Hydroxylation .............................. 104 4. Box Model Approach ..................................................................................... 107 4.1 Model Assumptions .................................................................................. 108 4.2 Concentrations ......................................................................................... 109 4.3 Isotopes .................................................................................................... 109 4.4 Comparison to Previous Work ................................................................. 111 5. CO2 Hydroxylation KFFs .............................................................................. 114 5.1 Complications and Possible Causes of Variability .................................... 114 13 Chapter Page 6. Summary ........................................................................................................ 116 VI. SUMMARY ......................................................................................................... 118 APPENDICES ........................................................................................................... 120 A. APPENDIX A ................................................................................................ 120 B. APPENDIX B ................................................................................................. 130 C. APPENDIX C ................................................................................................. 194 D. APPENDIX D ................................................................................................ 199 E. APPENDIX E ................................................................................................. 230 REFERENCES CITED ............................................................................................. 271 14 LIST OF FIGURES Figure Page Chapter II 1. DIC speciation in seawater solutions and in NaCl solutions ................................ 29 2. Schematic depiction of reactions of the CaCO3-DIC-H2O system ........................ 30 3. Carbon and oxygen isotopic fractionation for DIC species over solution pH ....... 32 4. Compilation of d18O vs. temperature calibrations for CaCO3 .............................. 39 5. Compilation of ∆47 vs. temperature calibrations for CaCO3 ................................ 41 Chapter III 1. Behavior of the calcite growth experiments ......................................................... 48 2. Variations in total alkalinity, dissolved inorganic carbon, and calcite saturation state during each calcite growth experiment ....................................................... 49 3. Calcite-water oxygen isotope fractionation expressed as 1000lnac/w .................... 54 4. Effect of solution saturation on calcite growth rate ............................................. 58 5. Model for the time-dependent behavior in experiments at T = 25°C, pH = 8.3 and [NaCl] = 0 M ................................................................................................ 60 6. The pH-stat assays of Nielsen and Frieden (1972) suggest an exponential dependence of CO2 (de)hydration reaction velocity on [NaCl] ............................. 63 7. Data-model comparison for the calcite-water oxygen isotope fractionation (1000lnac/w) in solutions with different NaCl concentrations (0-1.4 M) .............. 65 Chapter IV 1. Experiment headspace CO2 concentration, total alkalinity (TA), NaOH added, DIC d13C, DIC concentration, and calculated degree of supersaturation (W) ...... 71 15 Figure Page 2. Over the pH range 7.5-9.3, W increases, growth rate decreases, and ion activity ratio decreases ...................................................................................................... 74 3. Calcite-DIC carbon isotope fractionation expressed as 1000lnac-DIC .................... 76 4. Oxygen isotope fractionation between calcite and experimental solution expressed as 1000lnac-w ........................................................................................ 79 5. Updated ion-by-ion model of calcite growth at 25°C in CaCl2 solutions .............. 82 Chapter V 1. Experimental apparatus ....................................................................................... 92 2. Experimental headspace CO2 concentration and solution pH over the course of each experiment ................................................................................................... 94 3. SEM images of CaCO3 precipitates ..................................................................... 97 4. Isotopic compositions of precipitated carbonates and gas sources ....................... 100 5. The d13C of CaCO3 and DIC ................................................................................ 103 6. Carbon and bulk oxygen kinetic isotope fractionations (KIFs) for all experiments of this study ..................................................................................... 106 7. Schematic denoting the carbon reservoirs in our box model, and the fluxes and fractionations between them ................................................................................ 108 8. Model outputs showing how the KIEs change with increasing Ff relative to a fixed hydroxylation flux FH ................................................................................. 111 9. Box model results of e18EIC/CO2+H2O/e13EIC/CO2 slopes using different KFF values for 18O on CO2, and correspondingly on OH-, during CO2 hydroxylation with increasing degrees of CO2(aq) distillation .................................................... 112 10. Results from Clark et al. (1992) showing isotopic distillation of headspace CO2 during hydroxylation ........................................................................................... 113 11. Carbon and bulk oxygen KFFs over temperature from our open-air experiments and from past studies ....................................................................... 115 16 LIST OF TABLES Table Page Chapter II 1. d18O and d13C of Common Geologic Reservoirs ................................................... 27 2. Equilibrium Oxygen and Carbon Fractionation Factors ...................................... 33 3. Kinetic Oxygen and Carbon Fractionation Factors ............................................. 35 Chapter III 1. Experimental parameters and isotopic data for all experiments of this study ..... 52 2. Compilation of equilibrium fractionation factors (EFFs; T in Kelvin) ................ 53 3. Constants and parameters used in the model ...................................................... 57 Chapter IV 1. Experimental parameters ..................................................................................... 72 2. Isotopic data ........................................................................................................ 77 3. Model input parameters ....................................................................................... 81 4. Model parameters calculated from input parameters ........................................... 81 5. Oxygen and carbon fractionation factors used to generate the ion-by-ion model curves ................................................................................................................... 83 Chapter V 1. Experimental parameters, solution composition and pH for all experiments ....... 91 2. Direct experimental solution measurements ......................................................... 93 3. Isotopic data for all experiments .......................................................................... 101 17 Table Page 4. Carbon and bulk oxygen kinetic isotope fractionations ....................................... 107 5. Model parameters ................................................................................................ 108 6. Model curves varying KFFs for CO2 hydroxylation ............................................ 112 7. Literature compilation of KFFs during CO2 hydroxylation ................................. 114 18 CHAPTER I INTRODUCTION Calcite grows in a wide variety of geologic settings and conditions including caves, lakes, surface oceans, marine sediments, and hydrothermal veins. It has been a feature of ocean sediments since the Archean, as the oceans have been inferred to be supersaturated with respect to calcite for most of the earth’s history. It takes up trace element impurities into its structure, which can serve as a sink for metal contaminants in the environment. Trace impurities, as well as isotopic ratios of major and minor constituents, can also serve as paleoclimate archives, which are valuable in part because calcite is such a common mineral. Shifts in the oxygen and carbon isotopic compositions of seafloor carbonate sediments have been correlated to global climate signals (e.g. sea surface temperature, glaciations) and global biotic events (e.g. level of productivity, species diversifications). A more recent avenue of interest in carbonate minerals has been their CO2 sequestration potential for carbon capture and storage, which could be engineered at a scale so that significant amounts of CO2 may be removed from the atmosphere. While many paleoproxies based on trace element and/or isotopic composition of carbonate minerals have been studied, the δ18O-T relationship of carbonate formation and thereby the tem- perature of the aqueous solution from which it grew (e.g. seawater) has been studied the most extensively. When calcite grows slowly enough for isotopic equilibrium to be maintained between the mineral and the solution, the oxygen isotope partitioning between the phases is a function of solution temperature. However, isotopic equilibrium is not maintained in the majority of cases for either natural or synthetic calcite, and therefore factors such as mineral growth rate, solution composition, pH, and dissolved inorganic carbon (DIC) source(s) may have an effect on the oxy- gen isotope fractionation between calcite and solution. As an additional complexity, the majority of calcic seafloor sediments are biogenic, meaning calcifying organisms mediated the precipita- tion of CaCO3 which imparts “vital effects” on the isotopic and trace element composition of the mineral that are poorly understood. These vital effects may involve a combination of enzymatic and metabolic processes that act to regulate the pH of the calcifying fluid, increase the degree of supersaturation, and exclude growth inhibitors. This work aims to better understand how the stable isotope composition of calcite is affected by inorganic and biogenic processes by carrying out well-controlled calcite growth experiments over a range of pH, temperature, ionic strength, and concentration of the enzyme carbonic anhydrase (CA). The enzyme CA catalyzes CO2 (de)hydration reactions, which acts to promote isotopic equilibration of the dissolved inorganic carbon (DIC) pool since those reactions are some of the 19 slowest in the DIC-H2O system. When calcite grows from an equilibrated DIC pool, any kinetic isotope effects (KIEs) recorded in the calcite may be attributed to processes occurring at the mineral-solution interface as the mineral grows. In Chapter II, I provide a background of the CaCO3-DIC-H2O system. I introduce the concepts of isotopic fractionation, both equilibrium and kinetic, and provide equations and notation relevant to understanding the isotopic data presented in the later chapters. I discuss the DIC system reactions and corresponding isotopic fractionations, as well as how various factors affect the isotopic composition of inorganic and biogenic calcite. I close the chapter with brief descriptions of a range of trace element and/or isotopic-based paleoproxies and their application to paleoenvironment interpretations. This chapter sets the foundation of concepts that are explored more in-depth in the following chapters. The first two suites of experiments (Chapters III and IV) were carried out under near-isotopic- equilibrium growth conditions, systematically varying the solution ionic strength and pH across a range that mimics that of natural calcite precipitation environments (e.g. seawater, estuaries, freshwater lakes, alkaline lakes). The results from these chapters give insight into the fairly small magnitude nonequilibrium isotope effects typical of natural and experimental calcite that may be viewed as deviations from the empirical δ18O-T equilibrium. In the third suite of experiments (Chapter V), carbonate minerals grew under uncommon, far-from-isotopic-equilibrium conditions and recorded large kinetic isotope effects (KIEs). While equilibrium-based paleoproxies aren’t applicable, these carbonates may still record useful paleoenvironmental information. In Chapter III, calcite growth experiments were performed over a range of ionic strength (I = 0.1-1.6; [NaCl] = 0-1.4 M) at 25°C and pH 8.3 in order to examine the effect of solution composition and background electrolytes on oxygen isotope partitioning between calcite and water. No evidence of an ionic strength effect on oxygen isotope partitioning between calcite and DIC was found for NaCl concentrations up to 0.35 M. In higher ionic strength solutions, however, NaCl was found to inhibit the efficacy of carbonic anhydrase (bCA) and prevent complete isotopic equilibration of the DIC pool, resulting in lower and more variable oxygen isotope fractionations between calcite and water. We then modeled the experimental results with an isotopic box model, and quantified the salt inhibition of bCA, which has implications for kinetic isotope effects in biogenic calcite since biocalcifiers use carbonic anhydrase in their calcifying fluid. In Chapter IV, calcite growth experiments were carried out over pH 7.5-9.3 at 25°C and low ionic strength to investigate the effect of pH on isotope partitioning between calcite and water. The oxygen isotopic composition of calcite decreased with increasing pH. We updated the ion-by- ion model to reflect these new experimental constraints, which is a generalizable model for calcite growth that can be applied to a wide range of natural and experimental calcites. In Chapter V, we carry out experiments that mimic the travertine precipitation that occurs upon contact with atmospheric CO2 at hyperalkaline springs upwelling through peridotites. In these experiments, high pH, DIC-free, CaCl2 solutions rapidly precipitate carbonate minerals upon contact with gaseous CO2, with the large kinetic isotopic fractionations for both carbon and oxygen 20 likely reflecting unidirectional CO2 hydroxylation. We successfully reproduced the isotopic frac- tionations observed in both natural and laboratory settings, and are the first experimental study to separately quantify the kinetic fractionation factors (KFFs) of oxygen on CO and OH−2 during CO2 hydroxylation. Chapter III is co-authored with James M. Watkins and Laurent S. Devriendt, and is published in Geochimica et Cosmochimica Acta. Chapters IV and V are also co-authored with James M. Watkins and Laurent S. Devriendt, and are in preparation for submission to Geochemistry, Geophysics, Geosystems. 21 CHAPTER II CACO3-DIC-H2O SYSTEM BACKGROUND 1 Carbonates in seawater Calcium carbonate minerals have been a mainstay of oceanic sediments since the Archean, long before the rise of biocalcifiers that dominate calcic sediments in the modern ocean (Grotzinger and Kasting, 1993; Zeebe and Wolf-Gladrow, 2001). The main CaCO3 polymorphs are calcite and aragonite, characterized by rhombohedral and orthorhombic crystal structures, respectively, which leads to differences between the minerals including solubility, crystal habit, trace element uptake, and isotopic composition (Zeebe and Wolf-Gladrow, 2001). The Mg/Ca ratio of the host solution generally determines which inorganic CaCO3 polymorph will precipitate, with low-Mg calcite at Mg/Ca < 2, both high-Mg calcite and aragonite for the range 2 < Mg/Ca < 5.3, and aragonite at Mg/Ca > 5.3 (Hardie, 1996). Mg increases the surface energy of calcite, thereby promoting precipitation of aragonite instead (Sun et al., 2015). CaCO3 polymorph formation is also influenced by temperature, with lower temperatures favoring calcite (Folk, 1994; Morse et al., 1997). Experiments by Morse et al. (1997) found that calcite precipitated at < 6°C in modern seawater (Mg/Ca = 5). Globally, aragonite and high-Mg calcite are found in warm ocean waters, while low-Mg calcite is typical at higher latitudes or in deep bottom waters (Morse et al., 1997). Mg-calcites grown in cool seawater and seawater with low carbonate ion concentration ([CO2−3 ]) contain less Mg than calcite grown in warmer seawater (Major and Wilber, 1991). While the Mg/Ca, and to a lesser extent water temperature, dictate inorganic CaCO3 poly- morph precipitation, biocalcifiers are not restricted as such. In the modern ocean, corals generally build their skeletons out of aragonite, while foraminifera and coccolithophores typically precipitate calcite tests and shells. Calcifying organisms impart “vital effects” on the isotopic and trace ele- ment composition of the mineral that are not yet fully understood, and are likely species-dependent. These vital effects are inferred to be due to a combination of different processes, including enzy- matic activity and metabolic processes involved in the regulation of the calcifying fluid composition (McConnaughey, 1989; Chen et al., 2018). Biocalcifiers do not typically create their skeletons di- rectly from seawater, but from a calcifying fluid modified from seawater composition via preferential pumping of Ca2+ in and H+ out of the calcifying space, creating an environment of higher pH and increased degree of supersaturation with respect to CaCO3 (Chen et al., 2018). 22 Zachos et al. (2001) presents carbon and oxygen isotopic data of foraminiferal carbonate sed- iments from the seafloor that spanned the last 65 million years, correlating shifts in the isotopic compositions to global climate signals for oxygen and global biotic events for carbon. Since the vast majority of calcic seafloor sediments are biogenic (Zeebe and Wolf-Gladrow, 2001), understanding biocalcification processes is of high priority since those biogenic sediments serve as an important paleoclimate archive. 1.1 Seawater composition and pH The geochemical composition and pH of seawater have varied over geologic time. Shifts in seawater composition can be inferred from variable evaporite mineralogy, with aragonite and MgSO4 salts appearing together (Precambrian, Permian, modern), and K, Mg, Ca-bearing salts appearing with calcite (Cambrian, Silurian, Cretaceous) (Hardie, 1996; Lowenstein et al., 2001). The modern ocean is an “aragonite sea” characterized by high [Na+], high Mg/Ca, and low Ca2+/SO2−4 (Hardie, 1996; Lowenstein et al., 2001). The major ions in seawater (Na+, Cl−, Mg2+, SO2−4 , Ca 2+, K+) have been consistent over geologic time but their relative proportions, and therefore overall seawater composition, have varied (Lowenstein et al., 2001; Turchyn and DePaolo, 2019). The relative ion proportions in seawater plays an important role in CaCO3 polymorph control, trace element partitioning into CaCO3, evaporite mineralogy and precipitation sequence, and more (Ichikuni, 1973; Folk, 1994; Lowenstein et al., 2001; Coggon et al., 2010). It is challenging to reconstruct paleo-ocean salinity and overall geochemical composition. One technique is based on fluid inclusions trapped in halite, which only precipitates after ∼ 90% of seawater has evaporated, so ancient ocean composition is back-calculated with many limitations since some of the major ions will have been depleted prior to onset of halite precipitation (McCaffrey et al., 1987; Lowenstein et al., 2001; Brennan et al., 2004). Other techniques to reconstruct ancient ocean salinity and composition involve box models or calculations of fluxes of ions into the oceans (e.g. river water input, terrestrial silicate weathering processes, hydrothermal brines from mid- ocean ridges) and fluxes of ions out (e.g. precipitation of stable minerals, evaporation), which generally correlate significant shifts in seawater composition to global tectonics and climate (Hardie, 1998; Turchyn and DePaolo, 2019). While the pH of modern ocean surface waters typically falls within a narrow range of 8.1-8.3, the pH of the oceans varies with depth (pH 7.8-8.1), and has also varied over geologic time (Grotzinger and Kasting, 1993; Beck et al., 2005; Halevy and Bachan, 2017). The pH of the ocean continues to change as a result of human activities, with projections under the “business-as-usual” scenario predicting the average surface ocean pH to decrease to 7.73 by the end of the 21st century (Jiang et al., 2019). The δ18O of foraminiferal calcite has been observed to vary with seawater carbonate ion concentration, recording isotopically lower values with increasing [CO2−3 ] or pH (Zeebe, 1999). Increasing ocean pH by 0.2-0.3 would correspondingly result in a decrease of δ18O of calcite by ∼ 0.22-0.33‰, which would typically be interpreted as an increase in ocean temperature (see section 5.1; Zeebe, 1999). An independent measure of seawater pH may help constrain the δ18O kinetic 23 isotope effects that are due to pH and not temperature. The δ11B of marine biocalcifiers is a promising paleo-pH proxy that does not seem to be significantly affected by diagenetic processes (Zeebe and Wolf-Gladrow, 2001; Edgar et al., 2015; Mavromatis et al., 2015; Uchikawa et al., 2015). 2 Isotopic fractionation Before going into further depth on the CaCO3-DIC-H2O system and its reactions, I will take a step back to introduce some key concepts of isotope geochemistry, including the principles of isotopic fractionation and relevant notation. 2.1 Equilibrium vs. kinetic fractionation Isotopes are atoms of the same element that contain different amounts of neutrons, thus resulting in different atomic masses. Isotopic fractionation is the enrichment of one isotope relative to the other isotopes of that element among coexisting phases, which may occur through various physical, chemical, and biological processes. The majority of isotopic fractionations are due to the difference in mass of the isotopes. The percent of mass difference determines the extent of fractionation, with elements with large relative mass differences exhibiting greater fractionation. This partitioning of isotopes may be described as either equilibrium or kinetic isotopic fractionation. Equilibrium isotopic fractionation occurs when reversible reactions have equal forward and backward reaction rates. The equilibrium fractionation is the distribution of isotopes that minimizes the overall Gibbs free energy of the system. Heavier isotopes are preferentially fractionated into phases in which they are most strongly or stably bound, due to vibrational energies that correlate negatively with atomic mass. Equilibrium fractionation is usually temperature-dependent. Due to commonly strong relationships between isotopic signature and temperature for equilibrium isotopic fractionation, many “isotope thermometers” have been developed for a wide range of geologic environments and conditions. Kinetic isotopic fractionation occurs during unidirectional reactions, during reversible reactions that have not yet reached equilibrium, during diffusion, or during differential bond-breaking. It may be a result of reservoir distillation including Rayleigh processes, or during physical processes including freezing or evaporation. Biological processes resulting in isotopic fractionation are lumped together as “vital effects”, which are poorly understood but may relate to enzymatic activity, metabolism, or other processes. Heavier isotopes move more slowly, and therefore react more slowly, than lighter isotopes. More energy is required to break bonds involving heavy isotopes due to their greater mass, which is another reason that reaction rates involving heavy isotopes proceed more slowly than for reactions with light isotopes. Kinetic isotopic fractionation is common with chemical reactions that have not yet reached equilibrium and for which the reaction rates for the heavy and light isotopes are different. Different environmental factors (e.g. temperature) or the types of phases involved (e.g. solids, liquids, gases, solid-liquid, etc.) promotes or inhibits isotopic equilibration of the system. Generally, 24 kinetic effects are less common at higher temperatures because isotopic exchange is more rapid at elevated temperatures, so equilibrium isotope partitioning is more readily established. Additionally, isotopes exchange more quickly in gases than in liquids, and in liquids than in solids. Exchange in solid phases at low temperatures is very slow and depends primarily on the diffusion coefficients and the molecular structure of the solid. The diffusion rate of oxygen in calcite at low to moderate temperatures is so slow that oxygen isotope ratios in calcite tend to be preserved even in shells millions of years old. A kinetic fractionation factor (KFF) describes the partitioning of isotopes that would occur during an instantaneous, unidirectional reaction of one reactant to one product, and therefore represents the kinetic limit or fully expressed kinetic isotope effect for that reaction. At isotopic equilibrium, the forward and backward reaction rates are equal, so the resulting equilibrium frac- tionation factor (EFF) is the intermediate value between opposing KFFs for the forward and backward reactions. The EFF may be calculated as the ratio between the forward KFF and back- ward KFF αk+1/αk−1 = αeq. The isotopic signature of many natural and experimental calcites falls somewhere between the EFF and KFF because it is uncommon to have purely unidirectional reactions (KFF), and for low temperature carbonate precipitation, isotopic equilibrium is often not achieved (EFF) either. 2.2 Isotope and fractionation factor notation We report isotope ratios using standard(δ-notation usi)ng the equation: δ− Rsample − 1 · 1000 (1) Rstandard where R is the isotope ratio of the heavy, rarer isotope over the lighter, more common isotope (e.g. 18O/16O, 13C/12C, 44Ca/40Ca). Carbon isotope compositions are reported relative to the Vienna Pee Dee Belemnite (VPDB) standard, while oxygen isotope compositions are typically reported relative to Vienna Standard Mean Ocean Water (VSMOW) standard. Isotopic data may be converted between the two scales by the following relationship from Coplen et al. (1983): δ18OVPDB = (δ 18OVSMOW − 30.91)/1.03091. (2) The atmospheric geoscience community monitors the isotopic composition of atmospheric CO2, reporting values using the VPDB-CO2 scale. For carbon, there is no offset between the VPDB- CO2 and VPDB scales, but for oxygen there is an approximately 10‰ offset that requires conversion between the two scales using the equation (Swart et al., 1991; Srivastava and Verkouteren, 2018): δ18O 18VPDB−CO2 = (δ OVPDB − 10.25)/1.01025 · (3) 25 The extent of isotope partitioning between two co-existing phases A and B is described by the isotope fractionation factor α, and is typically expressed as: (18O/16O) 18A δ OA + 1000 αA/B = (18O/16 = O) δ18 · (4) B OB + 1000 Positive αA/B values indicate that phase A is isotopically heavier than phase B (i.e. relatively enriched in the heavy isotope), with negative αA/B values indicating the opposite. We often choose to report fractionations as 1000lnα values so fractionation values are given in permil (‰) rather than in smaller and less intuitive decimal values (e.g. αcalcite/water = 1.030197 is equivalent to 1000lnαcalcite/water = 29.75‰). Equilibrium isotope fractionations are typically expressed in terms of 1000lnα, while kinetic fractionations are often expressed as epsilon values (‰): A/B = [αA/B − 1 ] · 1000. (5) 2.3 δ13C and δ18O of common geologic reservoirs The commonly used standard for carbon isotopic data, Vienna Pee Dee Belemnite (VPDB), normal- izes carbon isotope data against the carbon isotopic composition of a Cretaceous marine belemnite fossil that was remarkably enriched in 13C. By convention, the δ13C of this belemnite is now consid- ered 0‰ VPDB, resulting in samples commonly having negative δ13C compositions when reported relative to this standard. Vienna Standard Mean Ocean Water (VSMOW) is the typical standard for oxygen isotopes, and is based on a mixture of distilled (i.e. pure, salt-free) seawater samples from around the globe. By convention, typical seawater is considered 0‰ VSMOW. The typical range in δ13C and δ18O of many common geologic reservoirs are listed in Table 1. The δ18O of marine carbonates is typically around 30‰ isotopically heavier than the ∼ 0 ‰ ocean water, while the offset between the δ13C of shallow foraminifera biocalcifiers (-2 to 2‰) and shallow oceanic DIC (0 to 2.4‰) is considerably smaller (Zeebe and Wolf-Gladrow, 2001; Mackensen and Schmiedl, 2019). DIC in the surface ocean is typically enriched in 13C relative to the deep ocean due to the preferential uptake of 12C by organisms living in the photic zone (Zeebe and Wolf- Gladrow, 2001). Though there is both regional and seasonal variability in the isotopic composition of atmospheric CO2, it broadly has shifted to isotopically lighter δ 13C ≈ -9.5 to -7.0‰ from the pre-industrial value of -6.3‰, while the δ18O of ∼ 41‰ is approximately in isotopic equilibrium with ocean water (Trolier et al., 1996; Zeebe and Wolf-Gladrow, 2001; NOAA Global Monitoring Laboratory). 3 Dissolved inorganic carbon speciation Inorganic carbon in solution occurs as CO2(aq), H2CO3, HCO − 3 , and CO 2− 3 , which collectively constitute the dissolved inorganic carbon (DIC) pool. The relative proportions of the DIC species are primarily a function of solution pH (Fig. 1). To a lesser extent, DIC speciation is also depen- 26 Table 1: δ18O and δ13C of common geologic reservoirs Reservoir Value Reference Oxygen δ18O ‰ (VSMOW) atmospheric CO2 39.4 to 42.6 Trolier et al., 1996; Zeebe and Wolf-Gladrow, 2001; NOAA Global Monitoring Laboratory glacial ice -20 to -55 Zeebe and Wolf-Gladrow, 2001 marine carbonates 27 to 32 Zeebe and Wolf-Gladrow, 2001 meteoric water - Eugene, -12 to -11 My experimental solutions OR meteoric water - global -30 to 0 Bowen and Wilkinson, 2002 ocean water 0 Zeebe and Wolf-Gladrow, 2001 organic matter 15 to 35 Zeebe and Wolf-Gladrow, 2001 Carbon δ13C ‰ (VPDB) anthropogenic CO2 -28 to -26 Mackensen and Schmiedl, 2019 atmospheric CO2 -9.5 to -7.0 Trolier et al., 1996; Zeebe and Wolf-Gladrow, 2001; NOAA Global Monitoring Laboratory C3 plants -21 to -32 O’Leary, 1988; Zeebe and Wolf-Gladrow, 2001 C4 plants -10 to -16 O’Leary, 1988; Zeebe and Wolf-Gladrow, 2001 foraminifera - planktic -2 to 2 Zeebe and Wolf-Gladrow, 2001; Mackensen (shallow) and Schmiedl, 2019 foraminifera - benthic -4 to 1 Mackensen and Schmiedl, 2019 (deep) marine organic matter -19 to -31 Zeebe and Wolf-Gladrow, 2001; Mackensen and Schmiedl, 2019 oceanic DIC (shallow) 0 to 2.4 Zeebe and Wolf-Gladrow, 2001; Mackensen and Schmiedl, 2019 oceanic DIC (deep) -0.8 to 1.3 Zeebe and Wolf-Gladrow, 2001; Mackensen and Schmiedl, 2019 volcanic CO2 -5 Mackensen and Schmiedl, 2019 27 dent on temperature, pressure, solution salinity and composition (Zeebe and Wolf-Gladrow, 2001; Millero et al., 2006; Millero et al., 2007). The DIC species, or ion pairs including DIC species, are the only reservoirs of inorganic carbon in solution. Isotopic equilibration of aqueous carbon only requires equilibrium to be reached among the DIC species, and as a result carbon isotope equili- bration timescales are relatively rapid (< 30 seconds regardless of solution pH) (Zeebe et al., 1999; Zeebe and Wolf-Gladrow, 2001). In contrast, oxygen isotope equilibration requires equilibrium to be achieved between the DIC species and water (H2O, OH −) (Zeebe and Wolf-Gladrow, 2001; Uchikawa and Zeebe, 2012). While OH−-H2O equilibration is rapid, the only reactions through which DIC exchanges with water (CO2 hydration and hydroxylation, see section 4.2) proceed rather slowly. Therefore, the timescale for establishing oxygen isotope equilibrium among the DIC species and water is much longer than that required to establish carbon isotope equilibrium between the DIC species, ranging from ∼10 minutes to equilibrate oxygen isotopes among DIC-H2O in solutions of pH < 5.5, to ∼1-16 hours over the pH range 7.5 to 9.3 (Zeebe and Wolf-Gladrow, 2001; Uchikawa and Zeebe, 2012). The equilibrium constant (K) for the dissociation of a weak acid (HA) may be calculated as follows: a + · a − K = H Aa , (6) aHA where the activities (a) of each species are a product of the activity coefficient (γ) and the ion concentration: γ + −H+ [H ] · γA− [A ] Ka = · (7) γHA[HA] The activity coefficients (γ) are folded into the expression when equilibrium constants are deter- mined experimentally (i.e. as functions of T, P, solution composition and concentration, etc.), and are thus presented as K∗ values. When modeling the CaCO3-DIC-H2O system, it is important to select pK values and other constants that are appropriate to the specifics of your system. Just as pH = - log ([H+10 ]), pK = - log10(K), where K values are equilibrium constants. K1 and K2 are equilibrium constants for the first and second dissociations of carbonic acid, calculated by: [HCO −][H+3 ] K ∗1 = (8) [CO2] and [CO 2−∗ 3 ][H +] K2 = · (9) [HCO −3 ] Another important aqueous geochemistry quantity is the total alkalinity (TA), which may be de- scribed as the slight charge imbalance between positive and negative “conservative ions” in solution (i.e. Na+, Ca2+, Cl−), whose concentrations are unaffected by changes in pH, pressure, and tem- perature (unlike HCO−3 , CO 2− 3 , etc.) (Zeebe and Wolf-Gladrow, 2001). In natural waters, a variety of nonconservative ions may be present (e.g. B(OH)−4 , PO 3− 4 ), but in the simple CaCl2 ± NaCl ± NH4Cl solutions of this dissertation, the total alkalinity may be approximated by the carbonate 28 0 100 100 A B C CO32- 90 90 HCO3- HCO3- CO2 HCO3- CO32- 80 80 70 70 60 60 2 50 50 40 40 30 30 20 CO32- 20 10 10 CO2 CO2 4 0 0 4 5 6 7 8 9 10 11 12 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 7 8 9 10 11 12 pH [NaCl] (M) pH 0 100 100 D III F V 90 E 90 HCO3- CO32-HCO3- IV CO2 HCO3- CO32- 80 80 70 III 70 IV V 60 60 2 50 50 [NaCl] (M) 40 40 1.5 30 30 1.0 20 20 0.5 0.0 10 CO32- 10 CO2 CO2 4 0 0 4 5 6 7 8 9 10 11 12 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 7 8 9 10 11 12 pH [NaCl] (M) pH Figure 1: (A-C) DIC speciation in seawater solutions using pK values from Millero et al. (2006), and (D-F) in NaCl solutions using pK values from Millero et al. (2007). The bjerrum plots (A, D) depict speciation for fresh (dark blue), saline (light blue), and supersaline (yellow) waters. On panels E-F, DIC speciation calculated using the R-package of PHREEQC with the Pitzer database (Charlton and Parkhurst, 2011; De Lucia and Kühn, 2013) is demonstrated to provide similar DIC speciation as Millero et al. (2007). Vertical dashed lines depict the ranges in pH covered by calcite growth experiments in this dissertation. Chapter III performs experiments at pH 8.3 and a range of 0-1.4 M [NaCl]. Chapter IV experiments range from pH 7.5-9.3 at low ionic strength. Experiments in Chapter V have initial pH values from 11.0-12.8. alkalinity ( = [HCO−] + 2[CO2−3 3 ]) (Zeebe and Wolf-Gladrow, 2001). There are six key quantities that describe the DIC system: [CO − 2− +2], [HCO3 ], [CO3 ], [H ], [DIC], and TA. Because of the relationships among them, if any two are specified, the other four quantities are accordingly fixed and may be calculated (Zeebe and Wolf-Gladrow, 2001). In the experiments of this dissertation, we measure solution pH continuously with a probe, periodically titrate the solution to obtain total alkalinity, and also periodically take solution samples that are later analyzed for [DIC], thereby obtaining quantities for [H+], TA, and DIC, and enabling us to calculate [CO2], [HCO − 3 ], and [CO 2− 3 ]. Due to the effects of solution salinity and composition, the equilibrium DIC speciation at a given pH is very different for typical seawater than in simple NaCl solutions. Millero et al (2006) determined robust pK (where [CO ]=[HCO−]) and pK (where [HCO−]=[CO2−1 2 3 2 3 3 ]) values in natural 29 log concentration log concentration % of total DIC % of total DIC % of total DIC % of total DIC seawater by systematically varying temperature and salinity, while Millero et al (2007) quantified pK values in NaCl solutions of variable ionic strength. DIC speciation shifts as a function of salinity in both seawater and NaCl solutions (Fig. 1A, D). Increasing salinity shifts both pK1 and pK2 to lower pH, though shifts for pK1 are much more modest compared to shifts for pK2. Therefore, the proportion of CO2 is much less affected than the proportion of CO 2− 3 for solutions of intermediate pH. Conversely, decreases in temperature shift pK values to higher pH (Zeebe and Wolf-Gladrow, 2001; Bajnai and Herwartz, 2021). The experiments addressed in this dissertation were performed in simple solutions containing CaCl2 ± NaCl ± NH4Cl, which may be adequately described by the pK values determined by Millero et al. (2007) in simple NaCl solutions. We juxtapose this DIC speciation with the distribu- tion of DIC species in seawater (Millero et al., 2006) across the same range in pH and [NaCl] (Fig. 1). Experiments in Chapters III and IV were conducted in 25°C solutions of 30 mM CaCl2 and 5 mM NH4Cl, at pH 8.3 across a range in solution [NaCCl]O=2(g0)-1.4 M (Ch. III) or over a range of pH (7.5-9.3) at low ionic strength and minimal [NaCl] (Ch. IV). Experiments in Chapter V were conducted in 10-25°C solutions of 10-30 mM CaCl2 at high pH 11.0-12.8. HCO - OH- + CO2(aq) + H2O HCO3- + H+3 4 CaCO3-DIC-H2O chemical reactions and isotopic fractionations CO32- + H+ H2CO3 This section goes into more detail regarding reactions between species of the CaCO3-DIC-H2O system, as shown in Fig. 2, as well as the equilibrium isotopic fractionations between species (EFFs: Table 2; Figure 3) and isotopic fractionations attending unidirectional reactions (KFFs: Table 3). CO2(g) CO2(aq) HCO3- OH- + + H O HCO3- + H+2 H+ + CaCO3 Ca2+ + CO32- + H+ CaCO3 Ca2+ + H2CO3 Figure 2: Schematic depiction of reactions of the CaCO3-DIC-H2O system. When CO2 gas dissolves into solution, the resulting aqueous CO2 may react with H2O or OH −. DIC speciation depends primarily on solution pH, but is affected to a lesser degree by other factors including temperature, pressure, and solution composition. 30 4.1 CO2 dissolution As CO2 gas diffuses from the atmosphere into the oceans, lakes, or other bodies of water, it dissolves into solution and exchanges isotopes with dissolved inorganic carbon (DIC) by the reaction: K0 CO2(g) −←−−→− CO2(aq), (10) during which the isotopes are fractionated according to mass-dependent processes. During equilib- rium isotopic fractionation, the heavier isotopes tend to be enriched in the phase in which they are more strongly bound, and in this case carbon and oxygen have opposite fractionation effects where 13C prefers CO (g) and 182 O prefers CO2(aq). At 25°C, the resulting CO2(aq) is depleted in 13C by ∼1.1-1.4‰ (Vogel et al., 1970; Zhang et al., 1995; Yumol et al., 2020), and enriched in 18O by ∼0.3‰ (Beck et al., 2005; Barkan and Luz, 2012). Owing to rapid equilibration between aqueous and gaseous CO2, the kinetic fractionation factors (KFFs) have not been fully explored, with no studies to my knowledge reporting KFFs for CO2 degassing. One study quantified the carbon isotope KFF during CO2 dissolution and found the aqueous CO2 to be further depleted by ∼1‰ from the equilibrium fractionation for a total carbon depletion of ∼2.1‰ (Zhang et al., 1995). Vogel et al. (1970) quantified the oxygen isotope KFF during CO2 dissolution and found CO2(aq) was enriched in 18O by 0.8‰, with limitations that this value was applicable ∼ 0°C. 4.2 CO2 (de)hydration and (de)hydroxylation Upon dissolving into solution, the aqueous CO2 undergoes hydration: −k−+→1CO (aq) + H O←−− HCO − + H+2 2 3 (11) k−1 and/or hydroxylation: − −k−+→4CO2(aq) + OH ←−− HCO −3 · (12) k−4 The relative production of HCO−3 from hydration vs. hydroxylation depends primarily on solution pH. At low pH, hydration dominates, while the opposite is true at high pH. At 25°C, the hydration and hydroxylation fluxes are equal at approximately pH 8.5 (Sade and Halevy, 2017, Figure 1). For the pH range of typical ocean water (surface pH 8.1-8.3, deep ocean pH 7.7-7.9) and many lakes or rivers (average pH 6-8), both hydration and hydroxylation appreciably contribute to HCO−3 production. The contribution from CO2 hydration may be considered negligible in low ionic strength solutions at 25°C at or above pH 10.5 (Devriendt et al., 2017; Bajnai and Herwartz, 2021). CO2 hydration and hydroxylation are the only pathways for DIC to exchange isotopes with H2O, and are also the slowest DIC exchange reactions (Zeebe and Wolf-Gladrow, 2001; Sade and Halevy, 2017, Figure 1). Over the pH range 7-10, both hydration and hydroxylation reactions are several orders of magnitude slower than H O-OH− and HCO−-CO2−2 3 3 exchange (Sade and Halevy, Figure 31 10 HCO3- 40 CO2 8 CO32- 38 6 36 4 DIC 2 34 DIC 0 32 HCO3- 2 CO2 30 4 28 6 26 8 CO32- 24 10 3 4 5 6 7 8 9 10 11 12 13 3 4 5 6 7 8 9 10 11 12 13 pH pH Figure 3: Carbon (left) and oxygen (right) isotopic fractionation for the DIC species over solu- tion pH 3-13. Isotopic fractionation is expressed as 1000lnαx−DIC for carbon isotopes, and as 1000lnα 13 18x−water for oxygen isotopes. The δ C and δ O of DIC over the pH range was calculated using the Zeebe (2007) expression, Millero et al. (2007) pK values, and equilibrium fractionation factors from Mook (1986) and Zhang et al. (1995) for carbon, and Beck et al. (2005) for oxygen. 1). Kinetic isotope effects (KIEs) attending these rate-limiting reactions are therefore more likely to be recorded in the EIC (Equilibrated Inorganic Carbon = HCO−3 + CO 2− 3 ) and precipitated carbonate minerals than KIEs attending the other, more rapid DIC-H2O system reactions (Sade and Halevy, 2017). Isotopic equilibration of the DIC can be sped up by the addition of carbonic anhydrase (CA), an enzyme that catalyzes CO2 hydration and dehydration reactions, but does not affect equilibrium DIC speciation (Kupriyanova and Pronina, 2011). Uchikawa and Zeebe (2012) found that oxygen isotopes of DIC species equilibrated twice as quickly with the addition of 3.7·10−9 M CA. Utilizing CA to speed up CO2-HCO − 3 interconversion should promote rapid DIC equilibration and minimize or eliminate KIEs in calcite stemming from DIC-H2O disequilibrium, thereby allowing kinetic effects occurring at the mineral surface (calcite-DIC disequilibrium) to be isolated. However, the addition of carbonic anhydrase does not promote DIC equilibration at high pH (>10) where CO2 hydroxylation dominates over hydration, because CA does not catalyze (de)hydroxylation reactions. While most researchers treat the DIC system as consisting of three species, CO2(aq) + HCO − 3 + CO2−3 , there is technically a fourth species: H2CO3. Often, the carbonic acid species is ignored, or rather, lumped in with CO2(aq) due to its much lower concentration than CO2(aq) and the fact that H2CO3 is not chemically separable from CO2(aq) (Zeebe and Wolf-Gladrow, 2001). Depending on 32 1000ln13 x-DIC 1000ln18 x-water Table 2: Equilibrium oxygen and carbon fractionation factors Fractionation Equation α (25°C) Notes factor Oxygen αeq 17.611CO (g)−w +0.9821 1.0412 Zeebe (2007)2 TK αeqCO (aq)−w exp( 2520 2 +0.01212) 1.0413 Beck et al. (2005) 2 TK αeqHCO −w exp( 2590 2 +0.00189) 1.0315 Beck et al. (2005)3 TK αeqCO −w exp( 2390 2 -0.00270) 1.0245 Beck et al. (2005)3 TK αeq exp(( 17747c−w - 29.777)/1000) 1.0302 Coplen (2007); Watkins et al.TK (2013) αeq eq eqc−HCO αc−w/αHCO −w 0.9987 Beck et al. (2005); Copen3 3 (2007); Watkins et al. (2013) αeq αeq eqc−CO c−w/αCO −w 1.0056 Beck et al. (2005); Copen3 3 (2007); Watkins et al. (2013) Carbon αeqCO (g)−HCO 1/((-0.1141·TC + 10.78)/1000 + 1) 0.9921 Zhang et al. (1995)2 3 αeqCO (aq)−CO (g) (0.0049·TC - 1.31)/1000 + 1 0.9988 Zhang et al. (1995)2 2 αeq −867CO −HCO ( + 2.52)/1000 + 1 0.9996 Mook (1986)3 3 TK αeq −9866CO (aq)−HCO ( + 24.12)/1000 + 1 0.9910 Mook (1986)2 3 TK αeq exp((-2.4612 + 7666.3CO (g)−c - 2988000 2 )/1000) 0.9897 Bottinga (1968) 2 TK TK αeq 1/(αeq eqc−HCO CO (g)−c/αCO (g)−HCO ) 1.0025 Bottinga (1968); Zhang et al.3 2 2 3 (1995) αeq eq eqc−CO αc−HCO /αCO −HCO 1.0029 Bottinga (1968); Mook3 3 3 3 (1986); Zhang et al. (1995) the reaction mechanism and pathway, CO2(aq) may follow Eq. 11, or may first convert to H2CO3 via the reaction: k+2 CO2(aq) + H2O −←−−→− H2CO3 (13) k−2 before converting to HCO−3 by: k+3 H CO ←−−−→− HCO − +2 3 3 + H , (14) k−3 but the dissociation of H2CO3 is many orders of magnitude faster than the hydration and hydrox- ylation reactions, so its effect on DIC speciation is limited (Zeebe and Wolf-Gladrow, 2001; Sade and Halevy, 2017). Isotopic fractionations attending reactions that involve multiple species with the same element, such as oxygen during CO2 (de)hydration and (de)hydroxylation reactions (Eq. 11, 12), are com- plicated to quantify because there are different fractionations attending the different species. The KFF for carbon during CO2 hydration would be represented by  13 − , while for oxygenHCO3 /CO2(aq) 33 two separate KFFs are needed: 18 18− and  − . However, it is not always possibleHCO3 /CO2(aq) HCO3 /H2O to separate the KFFs attending the different species, so often a “bulk oxygen” KFF is reported instead, which involves a weighted sum based on how many oxygen atoms each species contributes (see Eq. 11, 12; Dietzel et al., 1992; Christensen et al., 2021). Equilibrium HCO−3 is ∼9‰ heavier in carbon and ∼9.5‰ lighter in oxygen than aqueous CO2(aq) (Fig. 3, Table 2). The KFFs attending CO2 (de)hydration have been studied both experimentally (Clark and Lauriol, 1992; Yumol et al., 2020) and theoretically (Guo, 2009; Zeebe, 2014; Guo and Zhou, 2019; Guo, 2020), with no clear consensus. The CO2 hydration carbon and bulk oxygen KFFs have been estimated between -17.6 to -33‰ and -13 to -18.8, respectively (Clark et al., 1992; Zeebe, 2014; Yumol et al., 2020). CO2 dehydration KFF estimates are larger for carbon (-30 to -32‰, Clark and Lauriol, 1992; Guo, 2009; Guo, 2020) and smaller for oxygen (Clark and Lauriol, 1992; Guo and Zhou, 2019) than that of CO2 hydration, which is in agreement with the opposite behavior observed in equilibrium isotopic partitioning between HCO−3 and CO2(aq) (Fig. 3). Notably, Guo and Zhou (2019) calculated separate KFFs attending CO2 dehydration for oxygen between H O-HCO−2 3 and CO2-HCO − 3 , with a weighted sum that is in approximate agreement with an earlier experimental study (Clark and Lauriol, 1992). Unidirectional CO2 hydroxylation is favored under conditions of high pH and when precipitation of carbonate minerals follows, thus hindering back-reaction by removing the DIC from solution. As with (de)hydration reactions, bulk oxygen KFFs are typically reported as previous studies have been unable to tease apart the oxygen KFFs for HCO−3 -CO2 and HCO −-OH−3 separately (Zeebe, 2020). Previous studies have examined travertines from natural hyperalkaline springs as well as undertaken laboratory experiments in an effort to quantify the KFFs for CO2 hydroxylation (Clark et al., 1992; Böttcher et al., 2018; Christensen et al., 2021), though these efforts are complicated by the uncertainty in the equilibrium oxygen isotope fractionation between OH− and H2O (see section 4.5; Green and Taube, 1963; Böttcher et al., 2018; Zeebe, 2020; Bajnai and Herwartz, 2021). Many studies agree the carbon KFF is quite large, approximately -17‰ with little to no temperature dependence, while the bulk oxygen KFF appears to have a stronger temperature-dependence, and an approximate value at 25°C of -7.3‰ when expressed relative to the weighted sum of CO2+OH− (Zeebe; 2020; Christensen et al., 2021, 2022). Theoretical models of CO2 dehydroxylation suggest a carbon KFF of -22.5‰ and a bulk oxygen KFF of -16.7‰, which is in agreement with a later calculation of separate oxygen KFFs (Guo, 2009; Guo and Zhou, 2019; Guo, 2020). 4.3 HCO−3 (de)protonation Bicarbonate deprotonates to carbonate by the reaction: k+ HCO −3 ←−−−→ 5 − CO 2− + H+3 · (15) k−5 HCO−3 -CO 2− 3 exchange is many orders of magnitude faster than hydration hydroxylation over the pH range carbonates precipitate (Eigen, 1964; Pinsent et al., 1956; Zeebe and Wolf-Gladrow, 2001; 34 Table 3: Kinetic oxygen and carbon fractionation factors Carbon Oxygen Reaction  (‰) Reference/Note  (‰) Reference/Note CO2(g) → CO2(aq) -2.1 Zhang et al. (1995) +0.8 Vogel et al. (1970), 0°C CO2(aq) → CO2(g) ? ? CO (aq)+H O → HCO− + H+2 2 3 -23 to -33 Zeebe (2014) -13 to -15 Zeebe (2014) -17.6 Yumol et al. (2020) -18.8 Yumol et al. (2020), from CO2(g) -19.7 calculated after Clark and Lauriol (1992), 0°C; Zhang et al. (1995); Yu- mol et al. (2020) HCO− + H+3 → CO2(aq) + H2O -32 Clark and Lauriol -5.5 Clark and Lauriol (1992); 0°C (1992), 0°C -30.0 Guo (2009); Guo (2020) -4.5 calculated from Guo and Zhou (2019) -22.9; H −2O-HCO3 ; CO2- +4.7 HCO−3 , Guo and Zhou (2019) CO2(aq) + OH − → HCO−3 -17.1 Christensen et al. (2021) -7.3 Christensen et al. (2021, 2022) HCO−3 → CO −2(aq) + OH -22.5 Guo (2009); Guo (2020) -16.7 Guo (2009) -75.2; OH−-HCO−3 ; CO2- +12.6 HCO−3 , Guo and Zhou (2019) (in agreement with Guo, 2009) HCO−3 → CO 2− + 3 + H ? -5 Devriendt et al. (2017); calculated after Kim et al. (2006) CO2−3 + H + → HCO−3 ? ? Ca2+ + HCO− +3 → CaCO3 + H 0 Watkins and Hunt -4 Watkins et al. (2014) (2015) CaCO + H+ → Ca2+3 + HCO−3 +2.5 Bottinga (1968); Mook -1.8 Coplen (2007); Wang et (1986); Zhang et al. al. (2013); Watkins et (1995); Watkins and al. (2014) Hunt (2015) Ca2+ + CO2−3 → CaCO3 0 Watkins and Hunt -2 Watkins et al. (2014) (2015) -0.5 Devriendt et al. (2017); calculated after Kim et al. (2006) CaCO → Ca2+3 CO2−3 +2.1 Bottinga (1968); Zhang -5.3 Coplen (2007); Wang et et al. (1995); Watkins al. (2013); Watkins et and Hunt (2015) al. (2014) 35 Sade and Halevy, 2017) so that instantaneous isotopic equilibration of HCO−3 -CO 2− 3 is typically assumed (Zeebe and Wolf-Gladrow, 2001; Devriendt et al., 2017; Christensen et al., 2021). Equilib- rium CO2−3 is -0.4‰ and -6.8‰ isotopically lighter than HCO − 3 for carbon and oxygen, respectively (Mook, 1986; Beck et al., 2005; Christensen et al., 2021). Due to the rapid equilibration of HCO−3 -CO 2− 3 , the KFFs attending the forward and backward reactions of this exchange have not been fully explored. Kim et al. (2006) found their carbonates grown from the quantitative DIC pool recorded isotopically heavier oxygen isotope compositions than carbonates grown from a fraction of the DIC pool (3-91%), and attributed this to preferential deprotonation of isotopically light HCO−3 . Devriendt et al. (2017) calculated an oxygen KFF of 18 2− − = -5‰ based on the Kim et al. (2006) study.CO3 /HCO3 4.4 CaCO3 precipitation/dissolution Isotopic equilibrium is only achieved in rare cases of natural low-temperature calcite growth be- cause commonly, mineral precipitation reactions outpace both the reverse dissolution reactions and isotopic equilibration of DIC species with the much larger H2O oxygen reservoir. CaCO3 mineral precipitation proceeds via both HCO − 3 and CO 2− 3 pathways. Still, some models assume only CO2−3 contributes to calcite growth (e.g. Devriendt et al., 2017), though there is increasing evidence that both HCO− and CO2−3 3 participate. Despite HCO − 3 being the dominant DIC species over the range of pH, temperature, and solution composition conditions in which the majority of natural carbonates precipitate (Millero et al., 2006; Millero et al., 2007), CO2−3 is still the main DIC contributor to the mineral lattice even when CO2−3 constitutes <0.5% of total DIC pool (Devriendt et al., 2017; Sade et al., 2020). The direct participation of bicarbonate in carbonate mineral growth proceeds by: HCO −3 + Ca 2+ −←−−→− CaCO3 + H+, (16) while the carbonate ion pathway follows: CO 2− 2+3 + Ca − k ←− + −→ c − CaCO3 · (17) k−c Previous studies of calcite growth kinetics suggest that the rate-limiting process during growth is the dehydration of its constituent ions, which must shed their hydration spheres in order to attach at the mineral surface (Andersson et al., 2016; Zuddas and Mucci, 1998). Calcium ions have higher charge density than carbonate ions, and are therefore more strongly hydrated and take longer to shed their hydration sphere (Helgeson and Kirkham, 1976). The dehydration rate of Ca2+ has been estimated at 1-2 orders of magnitude slower than that of anions such as CO2−3 (Larsen et al., 2010). However, in our experimental solutions, as well as seawater and many natural waters, the concentration of Ca2+ is three or more orders of magnitude higher than CO2−3 , which suggests the dehydration of (bi)carbonate may be rate-limiting instead (Larsen et al., 2010). 36 Equilibrium calcite records an oxygen isotope composition 5.5‰ heavier than CO2−3 , but 1.3‰ lighter than HCO−3 (Beck et al., 2005; Coplen, 2007; Watkins et al., 2013). Kinetic fractionations for CaCO3 precipitation were fit to the ion-by-ion model results of Watkins et al. (2014) and found to be 18 = -2‰ and 182− − = -4‰. Using these KFFs paired with the Copen (2007)CaCO3/CO3 CaCO3/HCO3 calcite-water equilibrium fractionation and the Wang et al. (2013) equilibrium fractionations for HCO−3 -water and CO 2− 3 -water, the kinetic fractionations for CaCO3 dissolution were calculated as 18 18 CO2− = -5.3‰ and  = -1.8‰. 3 /CaCO3 HCO − 3 /CaCO3 4.5 H2O dissociation Another reaction within the CaCO3-DIC-H2O system is the dissociation of water: −KH2O←− w −→− H+ + OH− · (18) The timescale of the dissociation of water is short, comparable to the rate of HCO− 2−3 -CO3 exchange (Zeebe and Wolf-Gladrow, 2001), and is therefore regarded as having attained isotopic equilibrium when dealing with the DIC-H2O system and the much slower equilibration of CO2- HCO−3 . Although dissociation of water is commonly written as Eq. (18), most hydrogen is hydrated in complexes such as H3O +, H5O + 2 , etc., though for simplicity we write it in its most reduced form as seen here. Most hydrogen ion does not actually exist as “free hydrogen” in aqueous solutions, though it can be convenient to treat it as such when writing expressions. The equilibrium oxygen isotopic fractionation between hydroxide and water is still debated. Experimental studies suggest that 1000lnαOH−−H O is between -45 to -49‰ at 15°C, and ∼-42.5‰2 at 25°C (Green and Taube, 1963; Bajnai and Herwartz, 2021), while theoretical calculations suggest much smaller values of -19.1 to -23.5‰ at 25°C, depending on the number of water molecules involved in the reactions (Zeebe, 2020). This ∼20‰ offset between experimental and theoretical studies is cause for concern when applying these values to quantitative problems and is in need of further study. Some researchers postulate a 1000lnαOH−−H O ≈ 0‰, based on the observation that there is2 little difference in the oxygen isotopic fractionations between several calcium-bearing minerals and water, compared to the corresponding dissolved species and water (e.g. CO2−3 -water and CaCO3(s)- water, or SO2−4 -water and CaSO4(s)-water) and that if this trend holds true for portlandite then 1000lnαOH−−H O ≈ 0 because 1000lnα2 Ca(OH)2(s)−H O is small (Böttcher et al., 2018). Zeebe (2020)2 disagrees, citing that for carbonate and sulfate minerals, the covalent bonds of both the aqueous species (CO2−3 and SO 2− 4 ) and solid minerals are similar in nature and dominate the isotopic fractionation, with differences between the aqueous and solid environments being more minor. Zeebe (2020) explains that the O-H bond of H2O and OH − is considerably weaker than the O-C bond of dissolved and solid carbonate, and is expected to exhibit larger isotopic differences between the aqueous and solid phases. 37 5 Isotopic composition of calcite 5.1 Oxygen isotopes: δ18O-T The oxygen isotopic composition of calcite exhibits a strong temperature-dependence (Urey, 1947; McCrea, 1950; Kim and O’Neil, 1997; Coplen, 2007; Watkins et al., 2013; Daëron et al., 2019) and is commonly used in paleoclimate studies to reconstruct past sea surface temperatures. Equilibrium oxygen isotope partitioning between calcite and water is found only in extremely slowly grown natural calcites (Coplen, 2007; Daëron et al., 2019). Previous calibrations based on both natural and experimental calcite that have precipitated at or near isotopic equilibrium are systematically one to several permil isotopically lighter than the extremely slowly grown natural subaqueous calcites grown in Devils Hole cave (Coplen, 2007) and Laghetto Basso lake within Corchia Cave (Daëron et al., 2019). Laghetto Basso calcite is a lower temperature (7.9°C) parallel to the previously well-established Devils Hole calcite (33.7°C), creating a more robust equilibrium oxygen isotopic temperature calibration (Daëron et al., 2019), of which has nearly the same slope (∼0.2‰ decrease per unit Kelvin) as the calibration from experimental and common natural calcites (Kim and O’Neil, 1997) but which is offset ∼1.5‰ to lower values. This experimental calibration using Devils Hole and Laghetto Basso cave calcites is indistinguishable from both the equation of Watkins et al. (2013) and the original prediction made in Coplen (2007) based solely on the Devils Hole calcite (Fig. 4). Most natural and experimental calcite grows too quickly to maintain equilibrium between the CaCO3-DIC-H2O reservoirs, in which case factors including growth rate, pH, degree of supersatu- ration (Ω), and solution composition affect the oxygen isotope signature recorded in the mineral. Most natural calcites record kinetic oxygen isotope effects of at least 1-2‰ (Daëron et al., 2019). Despite these strong pervasive kinetic effects, calcite-water oxygen isotope thermometry works rea- sonably well because many natural carbonates precipitate under narrow ranges in pH and growth rates (Watkins et al., 2014). Higher degrees of supersaturation correspond to faster growth rates (Nielsen et al., 2012), which in turn leads to lower oxygen isotope composition (Dietzel et al., 2009) as isotopic fractionation moves away from equilibrium partitioning and towards the kinetic limit (Lemarchand et al., 2004; Tang et al., 2008; Nielsen et al., 2012). Coplen (2007) suggested that rapid growth rates result in the preferential incorporation of isotopically light CO2−3 , while very low degrees of supersaturation (Ω=0.16-0.21) correlated with very slow growth rates. Gabitov et al (2012) systematically changed growth rate and found that slow growth allowed the crystal surface to reach equilibrium with the bulk calcite lattice, but that fast growth “captured” disequilibrium crystal surface isotopic signatures that were isotopically lighter than the bulk lattice. These experiments suggest that the near-surface region of calcite may be depleted in 18O relative to the bulk lattice under equilibrium conditions (Gabitov et al, 2012). Increasing growth rate resulting in isotopically lighter calcite supports the Watson (2004) growth entrapment model (GEM). While Watson (2004) models calcite growth by the diffusivity of the near-surface layers competing with 38 36 30.2 Daëron et al. (2019) 34 30.0 29.8 29.6 32 Watkins et al. (2013) 29.4 23 24 25 26 27 Temperature (°C) 30 Coplen (2007) 28 26 Kim and O’Neil (1997) 24 0 5 10 15 20 25 30 35 40 Temperature (°C) Figure 4: Compilation of δ18O vs. temperature calibrations for a range of natural and synthetic, inorganic and biogenic CaCO3. The calibrations of Coplen (2007), Watkins et al. (2013), and Daëron et al. (2019) are in agreement (see inset), and are generally accepted in the low temperature carbonate community to represent equilibrium oxygen isotope fractionation between calcite and water. The calibration of Kim and O’Neil (1997) has the same slope as the above calibrations with an approximately 2 ‰ offset to isotopically lower values. Many natural and synthetic calcites do not achieve isotopic equilibrium due to a variety of factors, and often record isotopic compositions along the Kim and O’Neil (1997) calibration. crystal growth rate, another growth model treats the isotopic signature of a mineral as a reflection of the competition between attachment and detachment rates at the crystal surface, yielding isotopic compositions ranging from equilibrium to fully kinetic (DePaolo, 2011). Impurity ions in the calcite lattice may also affect the δ18O of the calcite. Experiments carried out over a range in temperature (25 – 80°C) and solution Mg/Ca found that increasing Mg content of the calcite resulted in heavier δ18O such that at 25°C, 5 mol% MgCO3 increased 1000lnαc/w by 0.88‰ (Mavromatis et al., 2012). Chapter III explores the effect of ionic strength on isotopic fractionation in inorganic calcite, while Chapter IV investigates the role of solution pH. 39 1000ln c/w 1000ln c/w 5.2 Clumped isotopes: ∆47-T Clumped isotopologues are molecules in which heavy isotopes are bonded, or “clumped”, together. In carbonates, 18O and 13C result in a mass 47 CO2 molecule of 13C18O16O, or a mass 63 CO2−3 molecule of 13C18O16O16O. The clumping of heavy isotopes is temperature-dependent, with a greater degree of clumping occurring at lower temperatures and fading to a stochastic distribu- tion with increasing temperature (Ghosh et al., 2006; Eiler, 2007). This is a self-contained isotopic thermometer in calcite, not requiring isotopic analysis (or estimation) of the water the calcite grew from, unlike the oxygen isotope thermometer (Eiler, 2007). The first ∆47-temperature calibration was presented in Ghosh et al. (2006) based on analysis of mass 47 CO2 produced from phosphoric acid digestion of synthetic calcite and biogenic aragonite. In less than two decades, the clumped isotope carbonate community has exploded with dozens of revised ∆47-T calibrations based on theoretical models and calculations (Guo, 2009; Passey and Henkes, 2012), experimentally-grown carbonates (Zaarur et al., 2013; Tang et al., 2014; Kluge et al., 2015; Levitt et al., 2018), slowly-grown natural inorganic calcite (Daëron et al., 2019), a wide variety of biogenic carbonates including from otoliths (Ghosh et al., 2007), foraminifera (Wacker et al., 2014 – also brachiopod, bivalve, ostrich egg) and even separate mineral-specific calibrations such as for siderite (van Dijk et al., 2019) (Figure 5). Despite the abundance of slightly different proposed equilibrium ∆47-T calibrations (Fig. 5), the clumped isotope community seems to be converging on a single equilibrium calibration re- sulting from efforts in cross-correlating results between laboratories and accounting for how slight differences in methods may affect the calibration equations (e.g. acid digestion temperature, CO2 purification process, selection of reference frame, selection of standards, use or lack of carbonic anhydrase when growing synthetic calcites, etc.) (Tripati et al., 2015; Kelson et al., 2017). As the community nears an equilibrium ∆47-T calibration, studies have begun to tease apart KIEs in ∆47 of carbonate minerals. Solution pH is expected to influence isotope clumping in fast- growing carbonates (Hill et al., 2014; Watkins and Hunt, 2015; Tripati et al., 2015; Kluge et al., 2018; Daëron et al., 2019). While I do not report ∆47 values in this dissertation, experiments of both Chapters IV and V grew CaCO3 across a range of pH. Growth rates for calcite in Chapter IV are fast, but unlikely to record particularly large KIEs in ∆47. CaCO3 of Chapter V, however, mimicked processes occurring at hyperalkaline springs from which travertines have recorded large ∆47 enrichments up to 0.2‰ above expected equilibrium values (Falk et al., 2016). We therefore expect our experimental CaCO3 from Chapter V to preserve significant KIEs in ∆47 as well, which is an avenue of continued study. 6 Paleoenvironment applications The isotopic composition of major elements in calcite, as well as the concentration and isotopic composition of trace elements, has been an important topic of study over the last several decades. Some studies quantify trace element incorporation and isotopic fractionation using natural inorganic 40 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0 5 10 15 20 25 30 35 40 Temperature (°C) Figure 5: Compilation of ∆47 vs. temperature calibrations for a range of natural and synthetic, inorganic and biogenic CaCO3. Arranged from light yellow to dark blue: Ghosh et al., 2006; Ghosh et al., 2007; Guo, 2009; Dennis and Schrag, 2010; Passey and Henkes, 2012; Zaarur et al., 2013; Tang et al., 2014; Wacker et al., 2014; Kluge et al., 2015; Tripati et al., 2015; Kelson et al., 2017; Levitt et al., 2018; Peral et al., 2018; Daëron et al., 2019; van Dijk et al., 2019. Notably, the outlying calibration at lower ∆47 values is the van Dijk et al., 2019 calibration for siderite, while all other calibrations are for CaCO3. Earlier calibrations (yellow to light green colors) tend to report lower ∆47 values, with more recent calibrations (dark green to blue colors) trending towards higher ∆47. The slopes of the ∆47-T calibrations are generally consistent. or biogenic calcite, while other studies experimentally precipitate calcite under controlled laboratory conditions. Many paleoproxies involving the isotopic and/or trace element concentration of calcite have been recognized and quantified, with additional potential proxies being a continuing forefront of research in the low temperature geochemistry community. Shifts in the isotopic composition of speleothem calcite in caves along the South African coast were interpreted as climate signals, with lower δ18O values signaling weaker summer rains and stronger winter rains, and lower δ13C values corresponding with more C3 grasses, which are iso- topically lighter than C4 grasses (Table 1) (Bar-Matthews et al., 2010). Biogenic lacustrine calcite 41 ∆47 was found to have the highest δ18O and δ13C when it was deposited along with gypsum, interpreted as low lake levels and an arid climate, and the lowest δ18O and δ13C when it was deposited along with clay sediments during higher lake levels with a moister climate (Escobar et al., 2012). Mg/Ca and Sr/Ca concentrations of calcite from veins in mid-ocean ridge flank basalts were used to re- construct past seawater composition (Coggon et al., 2010). Veins that formed at low temperatures (<6°C) were found to be reliable records of seawater chemistry due to minimal seawater-basalt interaction at that temperature, while veins that formed at warmer temperatures required calcu- lation of the evolved fluid back to pre-basalt interaction in order to tease out the seawater signal (Coggon et al., 2010). 6.1 Trace elements Trace element partitioning between calcite and solution is commonly affected by factors such as temperature, precipitation rate, solution ionic strength, impurity ion concentration, or the concen- tration of other ions in solution. While solid solution incorporation with the impurity ion replacing Ca2+ in the calcite lattice is commonly assumed, there are several other mechanisms of trace element incorporation including surface adsorption, occlusion, and separate phase formation (Pingitore et al., 1988). Some impurity ions are smaller than Ca2+, such as Mg2+ and Mn2+, while others such as Sr2+ and Ba2+, are larger. Both Mg/Ca and Sr/Ca in calcite are widely used to reconstruct past sea surface temperatures. Cléroux et al. (2008) report Mg/Ca and Sr/Ca temperature calibrations derived from foraminiferal calcite, with both Mg and Sr uptake into calcite increasing with temperature. Lopez et al. (2009) also found an increase in [Mg] with temperature that was mostly insensitive to to precipitation rate and degree of supersaturation, and suggested that temperature-dependence of Mg incorporation may be related to the changes in calcite growth mechanisms they observed with temperature. Other studies of Mg partitioning found increased Mg uptake as both growth rate and degree of supersatu- ration increased (Mavromatis, et al., 2013). Despite the strong Sr/Ca temperature relationship, Sr incorporation into calcite is also affected by growth rate, solution composition ([Sr], [Ca], [CO2−3 ]), ionic strength, and the Mn and Mg content of the calcite (Ichikuni, 1973; Carpenter and Lohmann, 1992; Cléroux et al., 2008). The uptake of smaller trace elements (i.e. Mn2+, Mg2+, Fe2+) may distort the calcite lattice and subsequently allow for greater incorporation of larger ions such as Sr2+ (Ichikuni, 1973), as the co-variation of trace element concentrations is commonly observed (Carpenter and Lohmann, 1992; Cléroux et al., 2008). Carbonate mineral growth rate commonly affects trace element partitioning, as well as the isotopic composition of the trace element incorporation. Sr incorporation is primarily affected by growth rate, with a lesser dependence on ionic strength (Tang et al., 2012). While biogenic and inorganic marine calcites incorporate similar concentrations of Mg, the [Sr] of biogenic calcite is consistently offset to higher values compared to inorganic calcite due to high growth rates (Carpen- ter and Lohmann, 1992). Experimental calcite and biogenic calcite were found to have similar Sr partition coefficients, suggesting they precipitate at similar rapid rates, while slowly-grown, natural 42 inorganic calcite is a better candidate for calculating Sr partition coefficients that approach equi- librium (Carpenter and Lohmann, 1992). Sr isotope fractionation in calcite has also been demon- strated to be strongly dependent on growth rate, and that equilibrium ∆88/86SrCaCO −Sr2+ ≈ 03 (Böhm et al., 2012). Trace element uptake also affects calcite crystal growth morphology. Mg2+ incorporation into calcite has also been shown to affect the morphology of the crystals by differential step interaction, which leads to roughened step edges and the development of pseudofacets that alter overall crystal shape (Davis et al., 2004). Folk (1974) found that high Mg/Ca solutions resulted in more elongate to aciciular calcite crystals, while low Mg/Ca solutions yielded more equant to rhombic crystal morphologies. Intrasectoral zoning of trace element concentrations has been observed for Mg, Sr, and Mn in calcite, indicating that certain step geometries preferentially incorporate these trace elements (Paquette and Reeder, 1995; Davis et al., 2004). Li+ incorporation into calcite has also been experimentally demonstrated to stabilize the basal (0001) face, thereby modifying crystal shape (Wang et al., 2011). Trace elements may also affect the cathodoluminescence (CL) of a calcite crystal, which provides insight into textures and zoning that are not visible on an optical microscope. Mn2+ and trivalent REE (Sm3+, Eu3+) are the main CL activators in carbonates, while Fe2+ is the main CL quencher (Budd et al., 2000; Habermann, 2002; Major and Wilber, 1991; Mason, 1987; Richter et al., 2003). Mg and Sr do not demonstrate a relationship with CL in carbonates (Mason, 1987). Cathodoluminescence (CL) has primarily been applied to diagenetic calcite cements to infer pH and redox potential of diagenetic environments, though stalagmites and biogenic calcifiers have also been imaged (Machel, 1985). A few studies on synthetic doped calcite have been conducted. Since Mn incorporation depends inversely on growth rate, natural calcites precipitated more slowly in cooler, deeper marine environments or recrystallized at slower rates may contain more Mn and be cathodoluminescent, in contrast to quickly precipitated calcite in shallow, warm seawater (Major and Wilber, 1991). Growth banding in diagenetic carbonates has been interpreted to be due to shifts in pH and redox conditions since Mn and Fe, the primary activator and quencher, have multiple valence states and their incorporation into calcite is most favored in less oxidizing waters where a greater portion of the ions are present in the 2+ valence state (Major and Wilber, 1991; Mason, 1987). The δ11B of calcite is used as a paleo-pH proxy (Zeebe and Wolf-Gladrow, 2001; Mavromatis et al., 2015; Uchikawa et al., 2015). While δ18O, δ13C, and Mg/Ca in calcite are strongly affected by diagenesis, δ11B does not appear to be, which suggests that the boron isotopic composition of calcite may be a reliable paleo-pH indicator for even diagenetically altered calcic sediments (Edgar et al., 2015; Fantle, 2015). Boron in the oceans is present as both boric acid, B(OH)3, and borate ion, B(OH)−4 , characterized by pH-dependent speciation in which [B(OH)3] = [B(OH) − 4 ] at ∼ pH 8.6. Assuming that the δ11B of average seawater does not substantially change over time, then the consistent isotopic offset between the two boron species and the hypothesis that only B(OH)−4 is incorporated into calcite allow for solution pH to be reconstructed from δ11B of calcite (Zeebe 43 and Wolf-Gladrow, 2001; Mavromatis et al., 2015; Uchikawa et al., 2015). However, experiments indicate kinetic effects primarily due to calcite growth rate and solution [B]/[DIC] which may affect validity of this paleo-proxy (Uchikawa et al., 2015). 7 Bridge In the preceding Chapter (II), I introduced important concepts including equilibrium and kinetic isotopic fractionation, the CaCO3-DIC-H2O system and its reactions, and factors affecting the trace element and isotopic composition of carbonate minerals. While the oxygen isotope composition of calcite exhibits a strong temperature dependence, most natural and experimental carbonates pre- cipitate too rapidly to maintain isotopic equilibrium between the mineral and the growth solution, and therefore factors other than temperature may affect the isotopic composition of the resulting carbonate. In Chapter III, we conduct well-controlled calcite growth experiments to isolate non- equilibrium kinetic isotope effects caused by the solution ionic strength. Many past studies conduct experiments in low ionic strength solutions, with conclusions that may not be as directly applicable to carbonates from saline settings. This work has significant implications for natural inorganic and biogenic calcite growing in seawater or even more saline environments. 44 CHAPTER III OXYGEN ISOTOPES OF CALCITE PRECIPITATED AT HIGH IONIC STRENGTH: CACO3-DIC FRACTIONATION AND CARBONIC ANHYDRASE INHIBITION From Olsen, E.K., Watkins, J.M., and Devriendt, L.S. (2022). Oxygen isotopes of calcite precipitated at high ionic strength: CaCO3-DIC fractionation and carbonic anhydrase inhibition. Geochimica et Cosmochimica Acta 325, 170-186. 1 Introduction Laboratory-controlled calcite and aragonite precipitation experiments have been used to determine isotopic fractionation factors between carbonate minerals and dissolved carbonate species or water. When crystals grow slowly, near chemical equilibrium conditions, oxygen isotope partitioning is expected to depend solely on temperature, providing a theoretical foundation for oxygen isotope thermometry (Bigeleisen and Mayer, 1947; Urey, 1947). It is often the case, however, that crystals grow fast enough that isotopic exchanges between phases do not reach equilibrium, leading to kinetic isotopic effects (KIEs) that depend on variables other than temperature and arise from either or both of the following: (1) a dissolved inorganic carbon (DIC = CO2 + HCO − + CO2−3 3 ) pool that is not fully equilibrated with water (Usdowski et al., 1991; Zeebe and Wolf-Gladrow, 2001; Devriendt et al., 2017b) and/or (2) transport of ions to and from the mineral surface and reaction of ions at the mineral surface (DePaolo, 2011; Watkins et al., 2014, 2017). Until recently, it was difficult to separate these two different sources of KIEs. The enzyme carbonic anhydrase from bovine erythrocytes (bCA) is commercially available and can be used in carbonate precipitation experiments to reduce or eliminate KIEs arising from homogeneous chemical reactions between DIC species and water (Uchikawa and Zeebe, 2012; Watkins et al., 2013, 2014). Addition of bCA in calcite growth experiments is useful for investigating isotopic fractionations under conditions that mimic the secretion of biogenic carbonates and for isolating surface reaction-controlled KIEs. In the few calcite growth experiments where bCA has been employed for the purpose of equili- brating the DIC pool, there is a resolvable pH- and growth rate-dependence to the KIEs (Watkins et al., 2013, 2014). This suggests that surface reaction-controlled kinetic effects may be sensitive to the proportion of HCO−3 versus CO 2− 3 participating in growth (pH effect) and there is a mass 45 dependence to the reaction rate constants (growth rate effect) for the reactions: k Ca2+ f + HCO −3 ←−−−→− CaCO3 + H+ (1) kb k 2+ 2− −−→fCa + CO3 ←−− CaCO3 (2) kb where the kf ’s and k b’s are forward (precipitation) and backward (dissolution) rate constants, respectively. Previous experiments with bCA were done in low ionic strength solutions, but back- ground electrolytes are known to affect the solubility product of calcite (Mucci, 1983), calcite growth and dissolution kinetics (Ruiz-Agudo et al., 2010, 2011; Hong and Teng, 2014), calcite growth mor- phology (Wang et al., 2011), and solution speciation (e.g. Millero et al., 2006, 2007). Hence, solution composition may influence KIEs and be partly responsible for oxygen isotope variability in natural carbonates. Here, we performed calcite-growth experiments under constant temperature (25ºC) and pH (8.3) but with distinct salinities (0.0 < [NaCl] < 1.4 M; I = 0.1-1.6) to assess the effect of ionic strength on calcite-DIC oxygen isotope fractionation and on the activity of bCA. We found no evidence for a salinity effect on the oxygen isotope fractionation between calcite and the DIC species up to a NaCl concentration of ∼ 0.35 M. However, NaCl significantly lowers the activity of the enzyme bCA, giving rise to highly variable isotopic results at higher NaCl concentrations. We use these well- controlled experiments to adapt and refine a previously published isotopic box model for kinetic isotope effects in the CaCO3-DIC-H2O system (Chen et al., 2018; Christensen et al., 2021). The model is used to constrain the functional form for the salt effect on the enzyme-catalyzed rate constant for CO2 hydration. 2 Methods 2.1 Calcite growth experiments We use the same experimental setup as Watkins et al. (2013, 2014). Although the methods were described previously, we review them here because the details are used to develop a quantitative model reflecting our experimental conditions (Section 4). We measured the δ18O of input CO2, which matters if the DIC pool is not fully isotopically equilibrated, and monitored [DIC] and total alkalinity 2-4 times per day. Solutions were prepared by dissolving CaCl2·2H2O (30 mM), NH4Cl (5 mM) and variable amounts of NaCl (0-1.4 M) in 1.7 L of deionized water. The beaker with the prepared solutions was submerged in a water bath containing both heating and cooling elements for precise temperature control (25 ± 0.2°C). The beaker was sealed by a lid that has ports for a pH probe, NaOH dripper, sampling syringe, and gas bubbler (Watkins et al., 2013). DIC was added to the solution by contin- uous bubbling of a CO2-in-N2 gas mixture through a diffusion stone at 0.5 standard cubic feet per hour using a SmartTrak100 gas flow controller (Sierra Instruments). The pCO2 of the headspace 46 was recorded by a K-30 USB CO2 Probe Data Logger (CM0039 from CO2meter.com). Prior to the start of an experiment, the gas was fluxed through the solution for 1-6 hours, during which time the headspace pCO2 decreased from ∼500 ppm to 240-280 ppm as the lab air was replaced by the gas tank mixture. To begin an experiment, the autotitrator was activated and between 0.3-1.2 mL of 1 M NaOH was dispensed to bring the pH up to a setpoint of 8.3 ± 0.02. The increase in pH caused CO2 from the bubbles to partition into solution, which led to an abrupt decrease in the pCO2 of the headspace. Variable amounts of bCA (0-3µM) were added after the solution had reached a pH of 8.3. The total alkalinity (TA) as measured by Gran titration increased from 0 to ∼0.2-0.7 mEq/L following the addition of NaOH. The TA was measured each time samples were collected for [DIC] during the course of an experiment. From 22 titrations of the same solution carried out over the course of one week, we determined a 95% confidence interval of ± 0.03 and a standard error of 0.003 for our reported TA values, which accounts for any error due to pH probe calibration drift, issues with the autotitrator dispensing HCl accurately, and subjectivity in the Gran method calculation. All of our experiments behaved similarly to the example shown in Figure 1a. Each experiment had a pre-precipitation period (Stage I) during which DIC, TA, and pCO2 of the headspace increased as CO2 dissolved in solution. Eventually, enough DIC was added for calcite crystals to nucleate and grow on the beaker walls, marking the onset of the calcite precipitation period (Stage II). The duration of Stages I and II, as well as the rate of NaOH addition, varies somewhat between experiments (Fig. 1b), which we attribute to fluctuations in the average size of bubbles emanating from the diffusion stone and random nucleation processes leading to differences in the initial surface area of the first crystals. The growth rate of calcite (mmol/h) was calculated from the rate of NaOH addition during Stage II (cf. Watkins et al., 2013). The TA and [DIC] were measured at regular intervals and results from all experiments are plotted together in Figs. 2a and 2b. Solution ionic strength and free activities of Ca2+ and CO2−3 were calculated through the R-package of PHREEQC using the Pitzer database (Charlton and Parkhurst, 2011; De Lucia and Kühn, 2013). The degree of supersaturation (Fig. 2c, Ω = αCa2+ * αCO2−/K sp) was determined by using these ionic activities and the solubility product of calcite3 at 25°C (K −8.42sp = 10 , Jacobson and Langmuir, 1974). Additional considerations regarding DIC speciation as a function of salinity and solution composition are provided in Appendix A. In all experiments, a short period of rapid crystal growth at high Ω (5-11) was followed by a long period of crystal growth under relatively steady conditions at Ω = 3.7 ± 1.5. Following the final measurements, experimental solutions were discarded and crystals adhered to the beaker walls were rinsed 3 times with de-ionized water and allowed to air-dry. Precipitates from each experiment were imaged on the FEI Quanta 200 Environmental Scanning Electron Microscope (ESEM) at the Center for Advanced Materials Characterization (CAMCOR) at the University of Oregon (Appendix B). ESEM imaging found only rhombohedral to rhombo-scalenohedral CaCO3 crystals present. Additionally, X-Ray Diffraction (XRD) analysis of select experiments spanning the range in [NaCl] conducted at the Oregon State University X-Ray Diffraction Facility confirmed 47 5 8.6 A Stage I Stage II CO2 addition Calcite growth and CO2 addition 8.5 4 8.4 3 8.3 2 8.2 [Ca2+] removed to form CaCO3 1 8.1 TA 0 8.0 0 10 20 30 40 50 60 70 time (hours) 6 B 5 4 3 [NaCl] (M) 1.5 2 1.0 1 0.5 0.0 0 0 10 20 30 40 50 60 70 80 90 100 time (hours) Figure 1: Behavior of the calcite growth experiments. (A) Example of an experimental run (S11). During Stage I, the increase in TA matches the amount of NaOH added to offset the addition of CO2 to solution. At the onset of Stage II, the change in TA is due to addition of NaOH as well as removal of Ca2+ to CaCO3. During Stage II, the difference between NaOH added and TA can be used to calculate the CaCO3 flux (mmol/h). (B) Rate of NaOH addition for all experiments. The range of slopes during Stage II translates to a range in the CaCO3 precipitation rate from 0.04 to 0.15 mmol/h. 48 NaOH (mEq/L) NaOH (mEq/L) pH 1.6 A 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1.0 B 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 12 11 C [NaCl] (M) 1.5 10 9 1.0 8 0.5 7 6 0.0 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 time (hours) Figure 2: Variations in total alkalinity (TA, A), dissolved inorganic carbon (DIC, B) and calcite saturation state (Ω, C) during the course of each calcite growth experiment. Black dots are mea- sured points (or calculated from measured points) and curves are splines. This figure shows the variabilities in TA, DIC and Ω are not related to the different NaCl concentrations. Typically, Ω increased to between 5 and 9 around the 15-20 hour mark before decreasing to a quasi-steady state value of Ω = 2 to 4 after 30 hours. 49 [DIC] (mM) TA (mEq/L) that calcite is the only phase present. 2.2 δ18O, δ13C, and [DIC] measurements Exetainers for all experimental water samples were prepared on a GasBench II in the Stable Isotope Laboratory at the University of Oregon, by flushing them with helium gas for five minutes. The isotopic composition of CO2 in the CO2-in-N2 mixtures (Table 1) was measured in the Stable Isotope Lab at the University of Oregon after cryogenic separation from the N2 gas on a vacuum line that is otherwise and usually used for mineral fluorination. The gas tank with a regulator was connected to the inlet port of the vacuum line and set to 1 bar gauge pressure. Gas was allowed to flow into the line and atmospheric air was flushed out by cyclical opening and closing of valves at least three times to completely fill the line with the CO2-in-N2 gas mixture and exclude air. Dewar flasks containing liquid nitrogen (LN2) were placed at three traps along the line for at least 20 minutes to allow the small proportion of CO2 in the gas mixture to freeze. The N2 in the line was pumped away, and then LN2 dewars at the two upstream traps were removed and heated to allow the CO2 to collect at the yield measurement trap that was still submerged in LN2. Yield was obtained by heating the last and third trap that was isolated between valves, and observing the pressure on a digital manometer. An initial low digital gauge pressure reading was common due to the small fraction of CO2 present in the gas (typically 200 ppm), often necessitating multiple cycles of gas mixture inlet and freezing of CO2 at the traps. After all the upstream valves were closed, the last LN2 dewar was removed, the trap heated, and downstream valve opened, allowing the purified CO2 to flow from the vacuum line to a Thermo-Finnigan MAT 253 mass spectrometer. The isotopic composition of the sample CO2 gas was measured using the mass spectrometer dual inlet system, and a CO2 reference gas of known composition. Though the standard analytical error (1σ) is ±0.010‰ for δ18O and ±0.005‰ for δ13C, we suggest that our analyses may have greater uncertainty for several other reasons. Isotopic fractionation could be occurring at the gas tank regulator as the gas flows from the tank into the fluorination line, as well as during the preferential freezing of isotopically heavy CO2. Additionally, cryogenic separation procedures and mass spectrometer analyses suggest that some of our CO2-in-N2 gas tanks contained a small amount of H2O, which decreases our confidence in the measurements. Taking these uncertainties into account, we conclude the isotopic analysis of our CO2 gas to be within ± 1‰, which is sufficiently accurate for our purposes. The isotopic composition of calcite, aqueous solution, and DIC were analyzed at the Stable Isotope Laboratory in the College of Earth, Ocean, and Atmospheric Sciences (CEOAS) at Oregon State University. Data are presented in standard delta notation. Water samples for DIC and δ13C were taken 2-4 times per day during an experiment, while water for δ18O was taken only at the end of most experiments since previous work showed constant δ18O values over the course of an experiment (Baker, 2015). For DIC analysis, 3.5 mL of water was injected through rubber septa into He-flushed exetainers. At CEOAS, 0.1 mL of 85% orthophosphoric acid was added to the exetainers. After equilibrating 50 for 4 hours, the gas headspace was analyzed by continuous-flow mass spectrometry on a GasBench- DeltaV system. A known concentration in-house NaHCO3 standard was analyzed with the samples, from which a calibration curve was determined and DIC concentration in our samples was approx- imated. Data are reported in standard delta notation relative to VPDB (Appendix A, Table 2). The standard analytical error (1σ) is ±0.15‰ for δ13C of DIC. Water samples for δ18O were collected in He-flushed exetainers and filled almost completely (∼11 mL) to minimize headspace and allow for replicate analyses. The δ18O was analyzed using the CO2-equilibration method whereby the CO2 headspace is equilibrated with 5 mL of water while agitated in an 18°C water bath. The CO2 was then analyzed by dual inlet mass spectrometry on a DeltaPlus XL and data are reported in standard delta notation relative to VSMOW (Table 1). The standard analytical error (1σ) is ±0.05‰ for δ18O of H2O. Calcite samples were reacted with 105% orthophosphoric acid in a Kiel III preparation device for 8 minutes at 70°C. Evolved CO2 and H2O gases were condensed and CO2 was separated and transferred into a MAT 252 mass spectrometer for analysis via dual inlet mass spectrometry. Data are reported in standard delta notation relative to VPDB (Table 1). The δ18O data were converted to the VSMOW scale by the following relationship: δ18O = 1.03091·δ18VSMOW OVPDB + 30.91 (Coplen et al., 1983). The standard analytical error (1σ) is ±0.05‰ for δ18O and ±0.03‰ for δ13C. 3 Results 3.1 Oxygen isotope fractionation The fractionation factor between two phases or compounds, for example calcite (c) and water (w), are related to delta values as follows: (18O/16O) 18c δ Oc + 1000 αc/w = (18 = O/16O) δ18 (3) w Ow + 1000 With [NaCl] = 0 and [bCA] ≥ 0.2 µM, the data form a tight cluster (1000lnαc/w = 28.0 ± 0.1‰, n = 4, Table 1, Fig. 3) that agrees with previous results from calcite growth experiments carried out at 25°C, pH 8.3 and with bCA (Watkins et al., 2013, 2014; Baker, 2015). These experiments indicate 0.2 µM bCA is sufficient to maintain an isotopically equilibrated DIC pool in low salinity experiments. Results from experiments with added NaCl (0.2-1.4 M) and [bCA] ∼ 0.2 µM vary from 25.2 to 27.6‰ in 1000lnαc/w (Table 1, Fig. 3) and are comparatively lower than results from the low salinity experiments. Repeat NaCl experiments but with higher bCA concentrations (up to 2.9 µM) display 1000lnαc/w values consistently higher than those from experiments with [bCA] ∼ 0.2 µM, indicating higher [bCA] are required for isotopic equilibration of the DIC pool in solutions of high ionic strength. Specifically, experiments with [NaCl] from 0.18 to 0.35 M and [bCA] ∼ 1 µM resulted in 1000lnαc/w 51 52 Table 1: Experimental parameters and isotopic data for all experiments of this study. Experiment [NaCl] Salinitya bCA [bCA] R log R δ18O CO δ1810 gas 2 O 18 w δ O 18 c δ Oc 1000lnαc/w (M) (g/kg) (mg) (µM) (mmol/h) (mol/m2/s) (VSMOW) (VSMOW) (VPDB) (VSMOW) (VSMOW) S2b 0.52 35 11.35 0.22 - - 12.89 -11.65 -16.39 14.02 25.64 S3b 0.52 35 10.48 0.21 - - 12.89 -11.52 -16.43 13.98 25.47 S4b 0.18 15 10.06 0.20 - - 12.89 -11.42 -14.24 16.23 27.59 S5 0.35 25 10.89 0.21 0.053 -6.10 12.89 -11.32 -14.34 16.13 27.38 S6 0 3.5 10.6 0.21 0.044 -6.45 12.74 -11.31 -13.66 16.83 28.06 S7 0.18 15 10.04 0.20 0.057 -6.20 12.74 -11.34 -14.34 16.13 27.40 S8 0 3.5 10.48 0.21 0.053 -6.27 12.74 -11.39 -13.84 16.65 27.96 S9 0.35 25 10.21 0.20 0.070 -6.24 12.74 -11.40 -15.95 14.47 25.83 S10 0.69 45 10 0.20 0.071 -6.34 12.74 -11.39 -16.52 13.88 25.24 S11 0.35 25 10.01 0.20 0.093 -6.27 12.74 -11.62 -16.51 13.89 25.48 S12 1.37 85 10.03 0.20 0.094 -6.17 22.20 -11.57 -16.06 14.36 25.89 S13 1.03 65 10.27 0.20 0.040 -6.54 22.20 -11.72 -15.61 14.82 26.50 S14 0.86 55 10.24 0.20 0.103 -6.30 22.20 -11.82 -16.62 13.78 25.57 S15 1.20 75 10.41 0.20 0.101 -6.31 22.20 -11.82 -16.21 14.20 25.99 CA1 0.52 35 0 0.00 0.101 -6.31 24.30 -11.83 -16.01 14.41 26.20 CA2 0.52 35 20.35 0.40 0.102 -6.26 24.30 -11.86 -15.84 14.58 26.40 CA3 0.52 35 0.97 0.02 0.102 -6.27 24.30 -11.91 -16.43 13.98 25.86 CA4 0 3.5 0 0.00 0.101 -6.07 24.30 -11.68 -16.72 13.68 25.33 CA5 0.52 35 50.13 0.98 0.090 -6.31 24.30 -11.17 -14.18 16.29 27.39 CA6 0.52 35 100 1.96 0.105 -6.34 24.30 -11.18 -14.16 16.31 27.42 CA7 0.26 20 50 0.98 0.095 -6.30 24.30 -11.33 -13.67 16.82 28.07 CA9 0.95 60 100 1.96 0.089 -6.37 13.45 -11.57 -15.93 14.49 26.02 CA12 0.69 45 75.82 1.49 0.085 -6.44 13.45 -11.76 -14.91 15.54 27.24 CA13 1.03 65 150 2.94 0.082 -6.35 13.45 -11.65 -15.21 15.24 26.84 CA14 0.18 15 49.01 0.96 0.101 -6.29 13.45 -11.85 -14.16 16.31 28.09 CA15 0.35 25 50 0.98 0.100 -6.31 13.45 -11.86 -14.16 16.31 28.11 CA18c 0 3.5 20 0.39 0.152 -6.28 23.50 -11.88 -14.26 16.21 28.03 CA20 0 3.5 10.12 0.20 0.089 -6.34 13.63 -11.95 -14.34 16.12 28.02 a Salinities given are approximate. b Calcite growth rate could not be calculated for early experiments due to use of an unreliable NaOH autotitrator. c CA18 used an 800 ppm CO2-in-N2 gas tank, while all other experiments used 200 ppm CO2-in-N2 tanks. Flow rate was scaled back so that all experiments have a constant flux of 0.12-0.13 mmol CO2/h. values (28.1 ± 0.05‰) indistinguishable from the low salinity experiments (28.0 ± 0.1‰). These results suggest that, for these experiments, (1) the DIC pool remained isotopically equilibrated with ∼ 1 µM of bCA and (2) [NaCl] of up to 0.35 M has no significant effect on 1000lnαc/w. However, experiments with [NaCl] > 0.35 M resulted in 1000lnαc/w values significantly lower than 28.0‰ despite very high [bCA] (up to 2.9 µM). The cause of the lower 1000lnαc/w values at high ionic strength is investigated in Section 6.2. 4 Model for calcite growth from a DIC pool with variable level of isotopic equilibration The variability in 1000lnαc/w as a function of [bCA] is a manifestation of kinetic effects arising from a variably equilibrated DIC pool. Recent progress has been made on the development of numerical models that quantify kinetic isotope effects in the CaCO3-DIC-H2O system (Chen et al., 2018; Christensen et al., 2021; Uchikawa et al., 2021). In this section, we present a model adapted from that of Chen et al. (2018), and use it to evaluate the oxygen isotopic variations observed. This is aided by the constraints we have on the CO flux, δ182 O of CO2 gas, and carbonate growth rates. We begin with a 1.7 L solution at 25°C, pH = 8.3, [Ca2+] = 30 mM and [DIC] ∼ 0.01 mM. The DIC is initially isotopically equilibrated. Equilibrium fractionation factors used in the model are provided in Table 2. As CO2 bubbles through, some of it partitions into solution, constituting a net flux of CO2(aq) (i.e., DIC). The incoming CO2(g) has the isotopic composition of the gas tank, which is out of equilibrium with the dissolved CO2(aq) and water. We assume there is no isotopic fractionation of CO2 as it gets converted from the gaseous to dissolved state (equilibrium fractionation between CO2(g) and CO2(aq) <0.2‰, Brenninkmeijer et al., 1983; Beck et al., 2005; diffusive isotope effects < 0.7‰; O’Leary, 1984). The CO2 that enters solution undergoes hydration and hydroxylation reactions to produce HCO−3 , a fraction of which deprotonates to form CO 2− 3 . As the concentration of CO2−3 increases, the degree of supersaturation (Ω) also increases, and calcite grows at a rate that depends on the degree of supersaturation. Fast calcite growth draws down [DIC], thus slowing calcite growth in a negative feedback. For the model to be informative, it should: (1) capture the behavior that a quasi-steady state is reached in the system whereby the influx of DIC from CO2 (FCO2) is balanced by the outflux Table 2: Compilation of equilibrium fractionation factors (EFFs; T in Kelvin). Compounds Equation α (25 °C) References CO2(g)-H2O 17.611 T −1 + 0.9821 1.0412 Zeebe (2007) CO −22(aq)-H2O exp(2520 T + 0.01212) 1.0413 Beck et al. (2005) HCO−3 -H2O exp(2590 T −2 + 0.00189) 1.0315 Beck et al. (2005) CO2−-H O exp(2390 T−23 2 - 0.00270) 1.0245 Beck et al. (2005) Calcite-H O exp(( 177472 T - 29.777)/1000) 1.0302 Coplen (2007), Watkins et al. (2013) 3 OH−-H O (1 + [-4.4573 + 10.3255·10 - 0.5976·10 6 ]/1000)−12 T 2 0.9771 Zeebe (2020)T 53 HCO3- 31 DIC 28.4 30 [NaCl]=0 Coplen (2007) 28.2 28.0 29 0.2 0.3 0.4 bCA ( M) Kim and O’Neil (1997) 28 bCA (µM) 2µM 2.0 0.2µM 1.5 1µM 27 1.5µM 3µM 1.0 0.5 0.4µM 0.0 26 0µM 25 CO32- 24 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 [NaCl] (M) Figure 3: Calcite-water oxygen isotope fractionation expressed as 1000lnαc/w. Results from this study’s experiments (circles) at low salinity ([NaCl]=0, n=4) agree with previous calcite growth experiments conducted at pH 8.3 and 25°C with bovine carbonic anhydrase (Watkins et al., 2014 – triangles; Baker, 2015 – diamonds), as well as with the expected value obtained using the Kim and O’Neil (1997) calibration. Experiments with [NaCl] > 0.35 M show lower and variable 1000lnαc/w despite the use of high bovine carbonic anhydrase concentrations. Equilibrium fractionation factors are listed in Table 2 (Beck et al., 2005; Coplen, 2007). We use the Zeebe (2007) expression to calculate oxygen isotope fractionation between the sum of DIC and water, substituting Millero et al. (2007) pK values for simple NaCl solutions and equilibrium fractionation factors from Beck et al. (2005). 54 1000ln c/w 1000ln c/w of DIC to calcite (FCaCO3), (2) reach steady state fluxes that agree with the inferred growth rate (mmol/h), and (3) provide insights into the isotopic results. For this latter point, we specifically seek to explain the 3‰ variability among the experiments despite the use of high [bCA]. 4.1 Chemical reactions The model tracks the following chemical and isotope exchange reactions (Chen et al., 2018; Chris- tensen et al., 2021): k+1 CO2 + H2O←−−−→− HCO − +3 + H (4) k−1 − −k−+→4CO2 + OH ←−− HCO −3 (5) k−4 a CO + H18 −− + 2 O←−−−− 1− 2 −−→− HC18OO − + H+2 (6) 1/3a−1 18 − −−−a+CO2 + OH ←−−− 4 −−−→− HC18OO −2 (7) 1/3a−4 b C18 +1 OO + H O←−−−−−−−−−→− HC18OO − + H+2 2 (8) 2/3b−1 − −−−b+4C18OO + OH ←−−−−−−→− HC18OO −2 (9) 2/3b−4 The k ’s are rate constants for the (de)hydration and (de)hydroxylation reactions. Rate constants for the 18O-substituted species are denoted a or b, with a representing substitution on H O or OH−2 , and b representing substitution on CO2. The ratio of these rate constants is equal to the equilibrium constant for each reaction, as given in Table 3. We solve numerically these five coupled ordinary differential equations: d[CO2] = -k [CO ] + k [EIC]χ[H+ − FCO+1 2 −1 ] k+4[CO2][OH−] + k−4[EIC]χ+ 2 (10) dt V d[EIC] FCaCO = k+1[CO2]− k−1[EIC]χ[H+] + k+4[CO ][OH−2 ]− k−4[EIC]χ− 3 (11) dt V d[C18OO] 2 = -b+1[C 18OO] + b [18EIC]18χ[H+]− b [C18OO][OH−−1 +4 ] dt 3 2 18 + b [18EIC]18 + FCO R2 COχ 2−4 (12) 3 V 55 d[18EIC] 1 = a [CO ]r − a [18EIC]18χ[H+] + a [CO ][18 1+1 2 w −1 +4 2 OH−]− a 18 18−4[ EIC] χ dt 3 3 18 − 2 18 18 + 2+ b+1[C OO] b−1[ EIC] χ[H ] + b [C18OO][OH−]− b [18+4 −4 EIC]18χ 3 3 − FCaCO3 [ 18EIC] αc/EIC (13) V [EIC] and d[Ca2+] −FCaCO= 3 (14) dt V where FCO2 and F −1 − 2− CaCO3 are fluxes (moles s ), V is volume (L), and HCO3 and CO3 are written together as EIC (short for “equilibrated inorganic carbon”), assuming instantaneous equilibrium between these two species (Chen et al., 2018). The factors of 1/3 and 2/3 are needed for oxygen isotope mass balance; for every mole of HC18OO−2 that undergoes dehydrations, ∼2/3 goes to C18OO and ∼1/3 goes to H182 O. Kinetic isotope fractionation between calcite and EIC (i.e., αc/EIC ) is dependent on pH and growth rate, as described by the ion-by-ion model (Watkins et al., 2014) and a parameterized analytical expression (Devriendt et al., 2017b). Here, we adopt the ion-by-ion model of Watkins et al. (2014) but note that the Devriendt et al. (2017b) formulation produces nearly identical outputs at pH 8.3 and calcite growth rate = 10−6.3±0.3 mol m−2 s−1. The free parameters of the model are FCO2 and FCaCO3 , and constraints on the functional form of each are discussed below. 4.2 CaCO3 flux The surface area normalized growth rate of CaCO3 (moles m −2 s−1) is dependent on the degree of supersaturation through a commonly used rate law (Nancollas and Reddy, 1971; Berner and Morse, 1974; Morse, 1978): R = k(Ω− 1) n (15) or in logarithmic form log10R = log10k + n log10(Ω− 1) (16) Zuddas and Mucci (1998) performed seeded calcite growth experiments using CaCl2-NaCl solu- tions that closely match our solution compositions. Their data, which span a wide range of ionic strengths, are shown in Figure 4. A linear regression using all of the Zuddas and Mucci (1998) data yields n = 1.6 and k = 10−7.38. These values are used as a starting point, but the values can be adjusted to some degree, as permitted by the scatter of the data in Figure 4. 56 57 Table 3: Constants and parameters used in the model. Symbol Meaning Value Reference/Note Part I: Model Parameters V Volume of solution (L) 1.7 - FCO CO2 flux into solution FCO = mFCaCO + b m and b to fit [DIC] data2 2 3 RCaCO Carbonate precipitation rate (mol/m 2/s) R n 3 CaCO = k(Ω - 1) Zuddas and Mucci (1998)3 n = 1.6 k = 10−7.38 [Ca2+][CO2−] Ω = 3Ksp Sp Specific surface area (m2/mol) 30 Tang et al. (2008) SA Total reactive surface area (m2) Sp · nCaCO -3 FCaCO Carbonate precipitation flux (mol/s) SA · R3 CaCO -3 Part II: Reaction rate constants χ Fraction of HCO−3 in EIC χ = (1 + K2 )−1[H+] K 2 from Millero et al. (2006) k+1 Rate const. CO2 hydration (s −1) log10k+1 = 329.85 - 110.541log10(TK) - 17265.4 T Pinsent et al. (1956) andK Uchikawa and Zeebe (2012) k−1 Rate const. CO2 dehydration (M −1 s−1) k−1 = k+1/K 1 K 1 from Millero et al. (2007) k Rate const. CO hydrox (M−1 s−1+4 2 ) log10k+4 = 13.635 - 2895 T Pinsent et al. (1956) andK Uchikawa and Zeebe (2012) k− Rate const. CO dehydrox (s −1) k− = k ( Kw 4 2 4 +4 K ) Kw from DOE (1994)1 Part III: Isotopic parameters r 18O/16CO2 O ratio of CO2 - Isotope ratio r 18 16EIC O/ O ratio of EIC - Isotope ratio R [C18CO OO]/[CO2] 2rCO Isotopologue ratio2 2 R [18EIC EIC]/[ 16EIC] 3rEIC Isotopologue ratio K2·α 2− − 18χ Fraction of HC18OO− in EIC 18 CO3 /HCOχ = (1 + 3 −12 [H+] ) K 2 from Millero et al. (2007) a , b Rate consts for hydration (s−1+1 +1 ) a+1/k+1 = 1.0000 a Yumol et al. (2020) b+1/k+1 = 0.9812 a a−1, b−1 Rate consts for dehydration (M −1 s−1) a+1/k+1 = K 1 · αHCO−/H O Equilibrium constant3 2 b+1/k+1 = K 1 · αHCO−/CO Equilibrium constant3 2 a+4, b+4 Rate consts for hydrox. (M −1 s−1) a+4/k+4 = 0.9988 b Christensen et al. (2021) b /k = 1.0000b+4 +4 α − a −1 +4 K HCO /H Oa 2−4, b−4 Rate consts for dehydrox. (s ) k = 1 3 K · Equilibrium constant+4 w αOH−/H2O b+4/k = K1 +4 K · αHCO−/CO Equilibrium constantw 3 2 αc/EIC Growth rate-dependent isotopic fractionation f (T, pH, [Ca 2+], [HCO−], [CO2−3 3 ]) Ion-by-ion model of Watkins et al. (2014) a These values yield a bulk KFF that is consistent with Yumol et al. (2020). b These values yield a bulk KFF that is consistent with Christensen et al. (2021). −5.0 Zuddas & Mucci (1998) −5.5 y = 1.57x - 7.38 R2 = 0.77 −6.0 −6.5 −7.0 −7.5 −8.0 I = 0.10 I = 0.34 −8.5 I = 0.55 I = 0.93 −9.0 −1.0 −0.8 −0.6 −0.4 −0.2 0.0 0.2 0.4 0.6 0.8 1.0 log10(Ω−1) Figure 4: Effect of the solution saturation (Ω) on calcite growth rate (R) in simple CaCl2-NaCl solutions (data from Zuddas and Mucci, 1998). Seeded calcite precipitation experiments at varying ionic strength provide an empirical relationship between Ω and the surface area normalized growth rate R used in the model. The CaCO3 flux is related to the growth rate through the relationship: FCaCO3 = SA ·R (17) where SA is the reactive surface area (m2). At the onset of an unseeded experiment, SA = 0 and then it increases as calcite nucleates and grows. To calculate SA, we use a specific surface area, Sp, which is the surface area of calcite crystals at the end of an experiment (28 ± 5 m2/mol for an average particle size of ∼10 µm as determined from SEM images; Tang et al., 2008). Using the model, we estimate that 35% of the mass of crystals forms during the period of fast growth and 65% forms during the period of steady-state growth. The growth rate values in Table 1 are based on the steady state growth rate and thus represent minimum estimates. 4.3 CO2 flux The flux of CO (moles s−12 ) into solution is expressed as FCO2 = kp ([CO2]eq − [CO2]) (18) where [CO2]eq is calculated from Henry’s constant and the pCO2 of the gas tank (200 µatm in most experiments). The parameter kp (kg-soln s −1) describes the efficiency of gas transfer to solution, which varies between experiments depending on fluid dynamics (stirring) and the variable average bubble size produced by the diffusion stone. We therefore adjust kp as needed to satisfy 58 log10R (moles/m2/s) the constraint that the CO2 influx matches the DIC outflux (i.e., the calcite growth rate) at steady state. For experiment S8, a value for kp = 0.0026 kg-soln s −1 satisfies this constraint. 5 Model results Model outputs are compared to data from two low salinity experiments (exp. S8 and CA4 with no NaCl) in Figure 5. For the curves labeled “Model 1,” the reactive surface area increases monoton- ically (Fig. 5a), which leads to a steadily decreasing [DIC] and Ω during Stage II. For the curves labeled “Model 2,” the surface area of crystals is constant after the period of rapid nucleation and growth. This is akin to a seeded crystallization experiment where the reactive surface area is determined by the size distribution of seed crystals and assumed to be unchanging as overgrowth is added (e.g. Zuddas and Mucci, 1998). The different treatments of reactive surface area between Models 1 and 2 have little effect on the resulting isotopic composition of calcite. The FCO2 and FCaCO3 curves (Fig. 5b and 5c) illustrate the negative feedback that leads to steady state behavior. The flux of CO2 is at a maximum value initially because [CO2(aq)] ∼ 0 and it decreases steadily during Stage I as [CO2(aq)] accumulates in solution (Eq. 18). The accumulation of [CO2(aq)] leads to an increase in FCaCO3 into Stage II. This acts to draw down the [CO2(aq)], which is then compensated by an increase in FCO2 at the beginning of Stage II. A few hours into Stage II, FCaCO3 and FCO2 are nearly in balance. The model kp parameter that yields a match to the steady state growth rate (Fig. 5c) also matches the evolution of [DIC] (Fig. 5d), suggesting that the rate law derived from the seeded experiments of Zuddas and Mucci (1998) is valid for our unseeded experiments. In the absence of bCA enzyme (i.e. uncatalyzed experiment), the 1000lnα of CO2(aq) begins at the equilibrium value and then steeply drops (Fig. 5e) because of the input of bubbled CO2(g) with δ18O ∼ 24.3‰ vs VSMOW (∼5‰ lower than the equilibrium value). The light CO2 undergoes hydration and hydroxylation to form isotopically light EIC initially (Fig. 5f), but as the DIC residence time in solution increases due to increasing [DIC], the CO2 hydration reaction becomes increasingly bi-directional (Rb/Rf → 1, where Rb is the backward rate and Rf is the forward rate; Fig. 5g), leading to a progressively isotopically heavier EIC pool during Stage I that approaches the EIC equilibrium value. Around the 21-hour mark, the CaCO3 flux is sufficiently high, leading to higher Rf/Rb for the hydration reaction and isotopically lighter EIC pool until steady state is reached during Stage II. Without any enzyme added, the time-integrated 1000lnαc/w values are 25.3 for Model 1 and 25.6 for Model 2, which agree well with the measured 25.3 value. With a [bCA] of 0.2 µM, the rate constant of CO2 hydration (k+1) is increased by a factor of 200 (Uchikawa and Zeebe, 2012) and the 1000lnαEIC/w is equilibrated for all but the first couple hours of the simulated experiments (Fig. 5f). The 1000lnαc/w values are 28.0 for Model 1 and 27.9 for Model 2 (Fig. 5h), which match the measured value of 28.0 and is ∼1.5-2‰ lower than the equilibrium value due to the growth rate-dependent calcite-CO2−3 and calcite-HCO − 3 fractionation. 59 0.10 A 0.0600.08 Model 1 B Model 2 0.055 0.06 0.04 0.050 0.02 0.045 0.00 0.040 0.10 0.4 0.08 C D S80.3 0.06 0.2 0.04 experimental steady state value 0.1 Stage I Stage II0.02 0.00 0.0 42 32 E Catalyzed (0.2 μM bCA) 30 F Catalyzed (0.2 μM bCA) 40 28 26 38 Uncatalyzed (no bCA) Uncatalyzed (no bCA) 24 36 22 20 34 18 1.2 32 G Catalyzed (0.2 μM bCA) 30 H equilibrium calcite1.0 28 S8 0.8 Catalyzed (0.2 μM bCA) Uncatalyzed (no bCA) 26 CA4 0.6 24 Uncatalyzed (no bCA) 22 0.4 Rb/Rf = k-1[HCO-3][H+]/(k+1[CO2]) 20 0.2 18 0 10 20 30 40 50 0 10 20 30 40 50 time (hours) time (hours) Figure 5: Model for the time-dependent behavior in experiments S8 and CA4 at T = 25°C, pH = 8.3 and [NaCl] = 0 M. Model 1 assumes that the reactive surface area of carbonate crystals is proportional to the mass of carbonate precipitated whereas Model 2 treats the steady state portion of Stage II as a seeded experiment with fixed reactive surface area. The two different approaches yield similar isotopic outputs and are in good agreement with measured 1000lnαc/w values. 60 F (mmol/h) SAcarb (m2Rb/Rf δ18 ) O CO2(aq) CaCO3 FCO2 (mmol/h) 1000lnα 1000lnα DIC (mM)c/w EIC/w 6 Discussion 6.1 The effect of ionic strength on the oxygen isotope fractionation between calcite and the EIC Mineral-anion KFFs may be dependent on solution composition. The kinetics of calcite growth and dissolution have been postulated to depend NaCl concentration (Zuddas and Mucci, 1998). Hence it could be expected that ionic strength affects the calcite dissolution/precipitation ratio and by extension the isotopic fractionation between calcite and CO2−3 (and HCO − 3 ) (Devriendt et al., 2017b). The observation that the maximum measured 1000lnαc/w for a given [NaCl] does not vary up to [NaCl] of 0.35 M suggests that the calcite-CO2−3 (and possibly calcite-HCO − 3 ) KFF(s) is/are independent of change in ionic strength caused by Na+ and Cl− ions, contrary to the hypothesis proposed by Devriendt et al. (2017b). The decrease in the maximum 1000lnαc/w above [NaCl] = 0.35 M could be attributed to a changing calcite-EIC KFF but that would require an abrupt change in crystal growth mechanism or surface speciation for which there is no independent evidence. 6.2 The origin of δ18Ocalcite variability at high ionic strength Experiments with [NaCl] > 0.35 M display lower and more variable 1000lnαc/w values despite the very high [bCA] (up to 3 µM) used in these experiments. In the following subsections, we consider the possible explanations for this variability, including the δ18O of free water, the effect of ion pairs, the effect of NaCl on DIC speciation, and the effect of dissolved salts on the enzyme kinetics. 6.2.1 Effect of dissolved ions on δ18Ow Isotopically heavy oxygen is concentrated in the hydration spheres of some cations or anions in solution, resulting in an oxygen isotope composition of free water that is potentially lighter than the δ18O of the overall solution (Taube, 1954). However, the hydration spheres of Na+ and Cl− do not fractionate oxygen isotopes, regardless of solution molality (Taube, 1954). The other cations in our solutions exhibit opposing behavior, with Ca2+ ions having isotopically heavy hydration spheres (Sofer and Gat, 1972) and NH+4 ions having isotopically light hydration spheres (Taube, 1954). Although we use Ca2+ and NH+4 , our solutions are far too dilute with respect to these ions to expect a resolvable effect. A salt effect on hydration spheres is therefore not responsible for the 1000lnαc/w variations. 6.2.2 Effect of ion pairing on oxygen isotope fractionation between DIC species and H2O The equilibrium fractionation factors in the CO2-H2O system (αCO2−H2O, αHCO−−H O, and α3 2 CO 2− 3 −H O ) 2 are based on freshwater solutions (Beck et al., 2005). The addition of dissolved salts can lead to significant fractions of bicarbonate and carbonate ions existing as cation-CO2−3 complexes such as 61 NaHCO03 and NaCO − 3 , which may shift the isotopic composition of the HCO − and CO2−3 3 com- pounds reacting with Ca2+ to form CaCO3. Kim et al. (2014) showed that variable amounts of dissolved NaCl (I = 0 to 0.7) have a negligible effect on the equilibrium fractionation factors in the Na-Cl-CO2-H2O system. Similarly, Uchikawa and Zeebe (2013) found no discernible effects of MgCO03 on oxygen isotope equilibrium in the Mg-Cl-CO2-H2O system over a large range of MgCO03 abundances (0 to 40% of total CO 2− 3 ). Hence, isotope partitioning among ion pairs is not responsible for the 1000lnαc/w variations. 6.2.3 Effect of DIC speciation on δ18Ocalcite The relative proportions of CO2− − 183 and HCO3 within the EIC may affect δ Ocalcite where HCO − 3 attachment to the calcite surface is significant (Wolthers et al., 2012; Watkins et al., 2014). This is because CO2−3 and HCO − 3 have distinct oxygen isotope compositions (Fig. 3; Beck et al., 2005). Increasing salinity shifts the solution to higher CO2−3 /DIC (e.g. Millero et al., 2007), which de- creases the δ18O of the EIC. Moreover, the DIC equilibration time increases at higher CO2−3 /DIC due to lower [CO2] (Usdowski et al., 1991; Uchikawa and Zeebe, 2012). In simple NaCl solutions, the CO2−3 /DIC ratio increases from 1.8% to 5.3% between [NaCl] = 0 and 0.5 M, with very little change above [NaCl] = 0.5 M (Millero et al., 2007, Fig. S1.1e). However, this increase in CO2−3 /DIC only lowers the δ 18O of isotopically equilibrated DIC by about 0.3‰ (Fig. 3). In addition, the change in speciation translates to a minor increase in the DIC equilibration time that is accounted for in our model. Hence, a salt effect on solution speciation and equilibration time is not responsible for the 1000lnαc/w variations. 6.2.4 Effect of NaCl on the kinetics of catalyzed CO2 (de)hydration The enzyme bCA increases the CO2 (de)hydration rate constants k+1 and k−1. According to the Michaelis-Menten kinetic model, the expression for the enzyme-catalyzed rate constant for CO2 hydration is: k ∗+1 = k+1 · kcat · [CA] (19) KM where k cat is the turnover number and KM is the Michaelis-Menten constant. Uchikawa and Zeebe (2012) determined k /K = 2.7 x 107 M−1cat M s −1 for bCA in NaCl-free solutions. In the absence of an inhibitor, addition 0.2 µM of bCA is expected to increase k+1 by a factor of 200 in freshwater solutions. We performed assays after Uchikawa and Zeebe (2012) in an effort to directly measure any inhibitory effect of NaCl on bCA, but discovered many complications and inconsistencies with the procedure (Appendix C). Many inorganic and organic compounds are known to inhibit the activity of various forms of carbonic anhydrase (Bertucci et al., 2009, 2011a; De Simone and Supuran, 2012; Del Prete et al., 2014). Previous studies have found that several anions (including Cl−, Br−, and NO−3 ) affect the activity of CA. Nielsen and Frieden (1972) found these anions to have an inhibitory effect on the 62 activity of both oyster CA and bovine CA, while Dionisio-Sese and Miyachi (1992) studied a variety of marine algae and found that the inhibitory or catalytic effects of these anions on CA activity was species-dependent, with a proposed mechanism for inhibition being that these anions may displace the hydroxyl group bound to the zinc at the enzyme active site. Enzyme inhibition is typically expressed in terms of an inhibition constant, K I (mM), which is the concentration of inhibitor at which the rate of the uninhibited reaction is reduced by a factor of 2. The larger the value of K I, the weaker the inhibitor. Nielsen and Frieden (1972) carried out pH-stat assays and reported K I = 170 mM for bCA at T = 6°C and pH = 7.5. They showed the CO2 (de)hydration reaction velocity (V ) decreases exponentially with increasing [NaCl] (Figure 6). They define V as the difference between the catalyzed versus uncatalyzed reaction velocity, which in turn, we infer to be directly proportional to k cat/KM. Their results can be used to write an equation for k cat/KM that depends explicitly on [NaCl]: kcat kcat = ( )[NaCl]=0 exp(A · [NaCl]) (20) KM KM where the exponent A describes the strength of inhibition. Our fit to the data from Nielsen and Frieden (1972) gives A = -3.82 (Fig. 6). To assess whether NaCl as a mild inhibitor can account for our observations, Figure 7 shows model outputs using the functional form for k cat/KM from Eq. 20. Model outputs are obtained using the same parameters from experiment S8 and CA4 (Fig. 5) but with k∗+1 (Eq. 19 and 20) instead of k+1 in reactions (4), (6) and (8), and by varying [bCA] and [NaCl]. Model results using A = -3.82 (K I = 170 mM, based on data from Nielsen and Frieden, 1972) are shown on the left panels of Fig. 7 (panels a to d). Measured data and model outputs are compared in two separate panels 0.20 Nielsen & Frieden (1972) 0.16 T = 6 °C pH = 7.5 0.12 y = 0.19exp(-3.82*[NaCl]) 0.08 0.04 KI=170 mM 0.00 0.0 0.2 0.4 0.6 0.8 [NaCl] (M) Figure 6: The pH-stat assays of Nielsen and Frieden (1972) conducted at pH 7.5 and a temperature of 6°C suggest an exponential dependence of CO2 (de)hydration reaction velocity (V ) on [NaCl]. The data gives an exponent of -3.82 and an inhibition constant K I = 170 mM. 63 V 1000lnα (Fig. 7b and 7c) because the δ18O of DIC species under non-equilibrium conditions depends on the δ18O of bubbled CO2 (Appendix A, Section 2): (1) experiments with high δ 18OCO (g) (22.2-24.3‰,2 Fig. 7b) and (2) experiments with low δ18OCO (g) (12.27-13.45‰, Fig.7c). Model outputs shown2 in Fig. 7b and 7c involve no free parameters and yet can explain several first-order features of the dataset: (1) The range of measured 1000lnαc/w values in the high δ 18OCO (g) experiments is about2 3‰. (2) About 1 µM bCA is required to maintain an isotopically equilibrated the DIC pool up to [NaCl] ∼ 0.35 M. (3) Maintaining DIC at isotopic equilibrium with bCA becomes increasingly difficult above [NaCl] = 0.35 M even with 2-3 µM of bCA added. (4) There is a general decrease, for a given [bCA], in the 1000lnαc/w values with increasing [NaCl]. However, the model 1000lnαc/w values are systematically higher than the measured values for all experiments with bCA, implying that the modeled NaCl inhibition is too weak and that the modeled DIC pool is more equilibrated than suggested by the data (Fig. 7d). The data-model agreement is improved using A = -5.5 (K I = 120 mM, Fig. 7e to 7h). There are a couple of possible reasons why the inhibitory effect of NaCl on bCA may be stronger in our experiments than suggested by the data of Nielsen and Frieden (1972). First, the Nielsen and Frieden (1972) parameters were derived from a solution with a lower temperature and pH than our experiments and it is conceivable that NaCl inhibition varies with either or both of these parameters. Second, it is possible that they used a different isozyme of bCA, and different isozymes can have a substantially different K I values (De Simone and Supuran, 2012). Overall, these results suggest that the most up-to-date values for the kinetic fractionation factors (Table 3) and functional form of the salt effect on bCA from Nielsen and Frieden (1972) are an accurate quantitative description of our calcite growth experiments and should be useful for modeling the δ18O of CaCO3 in other experiments and natural environments. 7 Implications 7.1 Towards a general model for kinetic oxygen isotope effects The current work builds on a large number of studies, each contributing to the development of a general model of kinetic oxygen isotope fractionation in the CaCO3-DIC-H2O system (e.g. Clark et al., 1992; Usdowski and Hoefs, 1993; Zeebe and Wolf-Gladrow, 2001; Beck et al., 2005; Kim et al., 2006; DePaolo, 2011; Nielsen et al., 2012; Watkins et al., 2013, 2014, 2017; Zeebe, 2014; Devriendt et al., 2017b; Sade and Halevy, 2017; Chen et al., 2018; Yumol et al., 2020; Zeebe, 2020; Guo and Zhou, 2019; Guo, 2020; Christensen et al., 2021). This study presents the first attempt to model kinetic oxygen isotope effects recorded in laboratory grown calcite with a fully reversible isotopic box model (Chen et al., 2018; Christensen et al., 2021) and provides a functional form for the inhibitory effect of NaCl on the catalytic effect of bovine carbonic anhydrase. The success of the model applied to our experiments with variably equilibrated DIC pools (Fig. 7f to 7h) suggests that the community is converging towards an accurate set of equilibrium and kinetic fractionation factors in the CaCO3-DIC-H2O system that will be useful for improving models of 64 3.0 3.0 x107 A x107 E 2.5 2.5 2.0 2.0 y = 2.7x107 exp(-3.82*[NaCl]) y = 2.7x107 exp(-5.5*[NaCl]) 1.5 1.5 1.0 1.0 0.5 K =170 mM 0.5I KI=120 mM 0.0 0.0 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 [NaCl] (M) [NaCl] (M) 29 29 bCA ( M) B F 2.0 Kim and O’Neil (1997) 28 28 Kim and O’Neil (1997) 1.5 27 27 1.0 26 26 0.5 0.0 25 25 24 18O CO2(gas) = +24‰ 24 18O CO2(gas) = +24‰ 23 23 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 [NaCl] (M) [NaCl] (M) 29 29 bCA ( M) C G 2.0 Kim and O’Neil (1997) 28 28 Kim and O’Neil (1997) 1.5 27 27 1.0 26 26 0.5 0.0 25 25 18 24 O CO2(gas) = +13‰ 24 18O CO2(gas) = +13‰ 23 23 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 [NaCl] (M) [NaCl] (M) 31 31 bCA ( M) D H 2.0 30 line 30 ne 1:1 :1 l i 1.5 29 29 1 R2 = 0.15 R2 = 0.63 28 28 1.0 27 27 0.5 26 26 0.0 25 25 24 24 23 23 22 22 22 23 24 25 26 27 28 29 30 31 22 23 24 25 26 27 28 29 30 31 Measured 1000ln c/w Measured 1000ln c/w Figure 7: Data-model comparison for the calcite-water oxygen isotope fractionation (1000lnαc/w) in solutions with different NaCl concentrations (0-1.4 M). Model outputs were obtained using Eq. (19) and (20) for the NaCl dependence of the catalyzed CO2 (de)hydroxylation reaction kinetic (as in Figure 6). Left panels (A-D) show results using an exponent of -3.82 (Nielsen and Frieden, 1972). Right panels (E-H) show results using an exponent of -5.5. For panels (B) and (F), the input gas has high δ18O (22.2-24.3‰). For panels (C) and (G), the input gas has low δ18O (12.27-13.45‰). The NaCl dependence of Nielsen and Frieden (1972) is weaker than indicated by our experiments, leading to model 1000lnαc/w values that are systematically higher (i.e., more equilibrated DIC) than the measured values (panel D). A stronger inhibition by NaCl yields better overall agreement between model and data (panel H). 65 Model 1000ln c/w 1000ln c/w 1000ln c/w kcat/KM Model 1000ln c/w 1000ln c/w 1000ln c/w kcat/KM isotopic disequilibrium and biological ‘vital effects’ in other settings. 7.2 Application to marine calcifiers Carbonic anhydrase (CA) is present at or near the site of calcification of many (if not all) biological organisms, including urchins (e.g. Mitsunaga et al., 1986), crustaceans (e.g. Henry, 2001), corals (e.g. Furla et al., 2000; Moya et al., 2008; Tambutté et al., 2011; Bertucci et al., 2011b; Mass et al., 2014), coccolithophores (e.g. Zhang et al., 2021), and foraminifera (e.g. de Goeyse et al., 2021). Despite the presence of CA, all of these organisms produce CaCO3 skeletons or tests that exhibit deviations from isotopic equilibrium (e.g., McConnaughey, 1989; Spero et al., 1997; Adkins et al., 2003; Kimball et al., 2014; Hermoso et al., 2014; Devriendt et al., 2017a; Chen et al., 2018). The pervasiveness of KIEs in the marine carbonate record has motivated efforts to develop isotopic biomineralization models. Recent approaches for coral and coccolith calcification invoke a CO2-fed fluid and explicitly account for carbonic anhydrase activity on KIEs in the DIC-water system (Devriendt et al., 2017b; Chen et al., 2018; Zhang et al., 2021). For both organisms, the parameter k cat/KM has been treated as a constant. Given that the composition of the calcifying fluid is likely to have higher pH and lower salinity than ambient seawater due to proton pumping and cation selectivity, respectively, the treatment of constant k cat/KM may warrant relaxation in future refinements. An obvious caveat is that the bCA used herein may not be representative of CA found in marine organisms. For example, it is possible that marine varieties of CA are less sensitive to salt inhibition, either through modifications to the enzyme structure or by screening inhibitors from the calcifying fluid (e.g. Dionisio-Sese and Miyachi, 1992; Bertucci et al., 2009). 8 Summary Calcite crystals were precipitated at 25°C in solutions of variable NaCl and bovine carbonic an- hydrase (bCA) concentrations to investigate the effect of a solution ionic strength on the oxygen isotope fractionation between calcite and an isotopically equilibrated DIC pool (EIC). The experi- mental results indicate no significant ionic strength effects on calcite-EIC oxygen isotope fraction- ation but revealed the inhibitory effect of NaCl on the catalyzed CO2 (de)hydration reaction. This led to DIC pools that were not isotopically equilibrated and variable 1000lnαc/w values in high ionic strength solutions ([NaCl] > 0.35 M). Using an updated isotopic box model and accurate measure- ments of solution parameters ([TA], [DIC] and δ18O of input CO2(g)), we successfully modeled the measured variability in 1000lnαc/w values at high [NaCl] and quantified the inhibitory effect of NaCl on bCA. The new parameterization of the dependence of bCA activity on [NaCl] can be used in biochemical models of calcification with the caveat that other isozymes of CA may be less sensitive to inhibition by dissolved salts and solution pH. This is a subject that warrants further investigation for understanding vital effects in biogenic calcite and for finding an alternative to bCA for equilibrating the DIC pool in experiments that involve seawater-like compositions. 66 9 Bridge In the preceding Chapter (III), I investigated the effect of solution composition and ionic strength on DIC-CaCO3 oxygen isotopic fractionation and found no significant effect, but a significant NaCl inhibition of the enzyme carbonic anhydrase that is used to catalyze CO2 (de)hydration. Another factor that has been proposed to affect the oxygen isotope signature of carbonates during non- equilibrium growth is the pH of the growth solution. In Chapter IV, I perform calcite growth experiments to examine the oxygen and carbon isotopic fractionation between calcite and the aqueous solution over a range of solution pH common in many natural waters from which calcite precipitates, from pH 7.5 to 9.3. I utilize the enzyme carbonic anhydrase to help facilitate equi- libration of the DIC pool, but in these low ionic strength solutions any inhibitory effect of NaCl should be negligible. 67 CHAPTER IV EFFECT OF pH ON THE ISOTOPIC COMPOSITION OF CALCITE GROWN FROM CO2-FED SOLUTIONS: EXPERIMENTS AND MODELING This chapter is in preparation for submission to Geochemistry, Geophysics, Geosystems, co- authored with J.M. Watkins and L.S. Devriendt. The experiments were performed by E.K. Olsen. The model was a code from an earlier paper of J.M. Watkins, with parameters adjusted and run by E.K. Olsen. The writing is by E.K. Olsen, with editorial assistance from J.M. Watkins. 1 Introduction The oxygen isotope composition of calcite exhibits a strong temperature dependence, and is com- monly used in paleoenvironment reconstructions to estimate the temperature of the oceans (Elder- field & Ganssen, 2000; Strassen et al., 2009 Van Geldern et al., 2006; Zachos et al., 2001), lakes (Lacey et al., 2018; Leng & Marshall, 2004; Leng et al., 2001), or caves (Bar-Matthews et al., 1997; Feng et al., 2014; Gascoyne, 1992) in which it formed. This δ18O “thermometer” is based on principles of isotopic equilibrium between the mineral and its host solution (McCrea, 1950; Urey, 1947). When calcite grows too quickly to maintain isotopic equilibrium with the solution, how- ever, non-equilibrium kinetic isotope effects (KIEs) that depend on factors such as solution pH, precipitation rate, source(s) of dissolved inorganic carbon (DIC), and solution composition affect the calcite δ18O (Baker, 2015; Dietzel et al., 2009; Gabitov et al., 2012; Kim & O’Neil, 1997; Levitt et al., 2018; Watkins et al., 2013, 2014). This study focuses on the effect of solution pH on the oxygen isotope composition of low temperature inorganic calcite. Calcite grows in a variety of geologic settings that span a range in pH, from fresh water lakes and rivers (pH ∼ 6-8), to alkaline lakes (pH ∼ 9-11), and seawater systems. While the modern surface ocean pH generally falls within a narrow range (pH 8.1-8.3), the full oceanic water column displays considerably greater pH-variation, from pH 7.6-8.4 (Beck et al., 2005). The pH of the ocean has also been proposed to vary significantly over geologic time. Some studies argue for a high pH (9-11), “soda ocean” with (bi)carbonate present in greater abundance than even chloride during the Archean (Kempe & Degens, 1985), while other studies suggest the Archean ocean pH was lower than that of the modern ocean. Calculations using projected Archean atmospheric CO2 concentrations suggest an ocean pH of 6.5-7.0 (Grotzinger & Kasting, 1993; 68 Halevy & Bachan, 2017). In contrast to the inorganic calcic sediments of the Precambrian, Phanerozoic calcic seafloor sediments are dominated by biogenic CaCO3 (Grotzinger & Kasting, 1993; Zeebe & Wolf-Gladrow, 2001). Biocalcifiers do not typically create their skeletons directly from seawater, but from a calcifying fluid modified from seawater composition via preferential pumping of Ca2+ in and H+ out of the calcifying space, creating an environment of higher pH and higher CaCO3 saturation state (Chen et al., 2018). Biogenic CaCO3 is always depleted in 18O and 13C relative to inorganic CaCO3 from similar environments, which has been hypothesized to result from kinetic isotope effects during CO -HCO−2 3 interconversion, or due to pH-dependent DIC speciation within the calcifying fluid (Chen et al., 2018; McConnaughey, 1989). The δ18O of foraminiferal calcite has been observed to vary with seawater carbonate ion concentration, recording isotopically lower values with increasing [CO2−3 ] or pH (Spero et al., 1997; Zeebe, 1999). Increasing ocean pH by 0.2-0.3 would correspondingly result in a decrease of δ18O of calcite by ∼ 0.22-0.33‰, which would typically be interpreted as an increase in ocean temperature (Coplen, 2007; Zeebe, 1999). In addition to being observed in biogenic calcite (Zeebe, 1999), the pH effect on oxygen isotope partitioning between calcite and water has also been studied experimentally (Baker, 2015; Dietzel et al., 2009; Watkins et al., 2014) with consistent findings that as solution pH increases, the oxygen isotope composition of calcite decreases. This shift in calcite δ18O mimics the shift in the δ18O of the DIC pool over the same pH range, as the CO2−3 proportion of DIC increases, which is depleted in 18O relative to HCO−3 (Beck et al., 2005; Millero et al., 2006, 2007; Zeebe, 2007). Thus, researchers have postulated that the relative contribution to the mineral lattice from each DIC species is pH-dependent. To investigate the effect of solution pH on the isotopic composition of calcite, we perform well-controlled inorganic calcite growth experiments at 25°C over a range in solution pH from 7.5-9.3 that are characterized by relatively constant HCO−/CO2−3 3 , degree of supersaturation (Ω), growth rate, and with time series for both total alkalinity (TA) and DIC concentration. Kinetic isotope effects (KIEs) in our CaCO3 are attributed to attachment and detachment of ions at the mineral surface, since use of the enzyme carbonic anhydrase (bCA) to speed up isotopic equilibration of the DIC pool reduces or eliminates KIEs arising due to aqueous DIC disequilibrium. We then adapt the ion-by-ion model of Watkins et al. (2014) and Watkins and Devriendt (2022) and report modified kinetic fractionation factors (KFFs) for calcite precipitation and dissolution reactions due to the KIEs recorded in the δ18O of our experimental CaCO3. 2 Methods We use the same experimental set-up as Watkins et al. (2013, 2014), and more recently, Olsen et al. (2022) (Chapter III). Of note for comparison to previous studies of inorganic calcite precipitation, we measure the isotopic composition of the input CO2 gas, which becomes important if the DIC pool is not fully equilibrated, and utilize the enzyme carbonic anhydrase (bCA), which catalyzes CO2 hydration and dehydration reactions and thus promotes equilibration of the DIC pool. Bovine 69 carbonic anhydrase (MP Biomedicals #153879) dissolved in distilled, deionized (DDI) water was added immediately after the start of the experiment when the solution was brought to the pH of interest. For low ionic strength solutions of pH ≤ 8.3, [bCA] of ∼ 0.2 µM is sufficient to equilibrate the DIC pool (Olsen et al., 2022; Uchikawa & Zeebe, 2012; Watkins et al., 2013, 2014). At high pH, bCA becomes ineffective at establishing an equilibrated DIC pool due to low CO2(aq) activity. CO2 (de)hydroxylation reactions take over moderating isotopic exchange between DIC and water at high pH, which are also slow, rate-limiting reactions but which are not catalyzed by carbonic anhydrase. To mitigate the lower bCA efficacy and longer DIC equilibration time, our solutions from pH 8.65 - 9.3 were carried out at [bCA] ∼ 0.5-1.5 µM (Table 1). Experimental solutions were prepared by dissolving CaCl2·2H2O (30 mM) and NH4Cl (5 mM) in 1.7 L of DDI water. Solutions were held at 25°C (± 0.2) in a temperature-controlled water bath. CO2-in-N2 mixtures (200 - 2000 ppm CO2) were bubbled through our solutions for several hours prior to the start of an experiment until the headspace CO2 concentration leveled off at approximately the gas tank value. Rate of gas flow ranged from 0.1 - 0.5 SCF/H (standard cubic feet per hour), which corresponds to a range of 0.025 - 1.264 mmol CO2 per hour, with lower pH solutions requiring higher CO2 flow in order to precipitate CaCO3 (Table 1). The CO2 concentration of the headspace was continuously recorded (Fig. 1a) using a K-30 USB CO2 Probe Data Logger (CM0039 from CO2meter.com). Experiments were maintained at constant pH through use of a Titronic 3000 autotitrator dis- pensing 1 M NaOH when solution pH measured 0.02 below its set point. Total alkalinity was measured by titration of a solution sample by 0.01 M HCl, and then calculated using the Gran method. Solution samples for DIC and TA were taken 2-3 times per day. Samples for measuring the δ18O of the experimental solution were taken at the end of each experiment. Ionic activities for our solutions are calculated using the R-package of PHREEQC with the Minteq.v4 database (Charlton & Parkhurst, 2011; De Lucia & Kühn, 2013). The degree of super- aCa2+×aCO2− saturation with respect to calcite (Ω = 3K ) is calculated using those ionic activities andsp a solubility product of calcite that accounts for ion pairs in solution (K = 10−8.48sp , Charlton & Parkhurst, 2011; De Lucia & Kühn, 2013; Jacobsen & Langmuir, 1974). Imaging of precipitates was performed on an FEI Quanta 200 Environmental Scanning Electron Microscrope (ESEM) at the Center for Advanced Materials Characterization (CAMCOR) at the University of Oregon (Appendix D). Isotopic analysis methods are the same as detailed in Chapter III. 3 Experimental results In many of our experiments, both the TA and DIC (Fig. 1b, e) increase at the beginning of the experiment, peak at or near the onset of calcite precipitation, and then decrease to a steady state value for the remainder of the experiment. Some experiments record a very high initial value, which might reflect peaks near the onset of calcite precipitation since calcite began growing quickly 70 800 10 2000 1600 600 20 1200 800 400 400 30 0 0 20 40 60 80 100 120 time (hours) 200 40 0 a d 50 0 20 40 60 80 100 120 0 20 40 60 80 100 120 2.5 1.6 1.4 2.0 pH 1.2 9.5 1.5 1.0 9.0 0.8 8.5 1.0 0.6 8.0 0.4 7.5 0.5 0.2 b e 0.0 0.0 0 20 40 60 80 100 120 0 20 40 60 80 100 120 12 12 60 10 10 40 8 8 20 6 6 0 0 30 60 90 120 time (hours) 4 4 2 2 c f 0 0 0 20 40 60 80 100 120 0 20 40 60 80 100 120 time (hours) time (hours) Figure 1: (a) Experiment headspace CO2 concentration (ppm), (b) total alkalinity (TA), (c) NaOH added (mEq/L), (d) DIC δ13C (‰ VPDB), (e) DIC concentration (mM), and (f) calculated degree of supersaturation (Ω) for each experiment. The inset of panel (a) depicts the high CO2 concen- tration experiments carried out at pH 7.5. The inset of panel (f) depicts the high Ω values early on in some experiments carried out at pH 8.65-9.3, as well as two experiments (CA10, CA11) with anomalously high Ω that we interpret as experimental error and should be disregarded. 71 NaOH (mEq/L) TA (mEq/L) CO2 (ppm) CO2 (ppm) [DIC] (mM) 13CDIC (‰) Table 1: Experimental parameters Experiment pH bCA [bCA] CO2 flow Exp duration Precip R log10R (mg) (µM) (mmol/h) (h) (h) (mmol/h) (mol/m2/s) S6 8.30 10.6 0.208 0.126 72.77 52.55 0.0440 -6.45 S8 8.30 10.48 0.205 0.126 52.15 34.15 0.0533 -6.27 CA4 8.30 0 0.000 0.126 30.55 21.55 0.1010 -6.07 CA10 9.30 50 0.980 0.025 68.29 60.29 0.0984 -6.51 CA11 9.30 75 1.471 0.038 71.11 64.11 0.0735 -6.54 CA16 7.90 10 0.196 0.253 41.19 31.69 0.1858 -6.23 CA17 7.90 10 0.196 0.202 46.42 33.02 0.1694 -6.25 CA18 8.30 20 0.392 0.126 42.68 35.48 0.1521 -6.28 CA20 8.30 10.1 0.198 0.126 46.82 40.35 0.0893 -6.34 CA21 8.65 25 0.490 0.126 41.69 37.69 0.1206 -6.31 CA22 9.00 25 0.490 0.126 42.28 40.78 0.0932 -6.34 CA23 9.30 0 0.000 0.038 67.82 61.82 0.0176 -6.52 CA24 9.30 0 0.000 0.038 85.19 79.69 0.0299 -6.63 CA25 8.65 50 0.980 0.101 66.77 55.77 0.0925 -6.48 CA26 7.50 20 0.392 1.264 92.66 66.26 0.0429 -6.25 CA27 7.50 10 0.196 0.758 112.69 63.99 0.0326 -6.17 CA28 9.00 50 0.980 0.076 47.43 42.93 0.0810 -6.37 CA29 9.30 50 0.980 0.038 72.51 71.01 0.0419 -6.58 CA30 7.50 20 0.392 1.264 92.24 55.38 0.1440 -6.08 CA31 7.50 10 0.196 1.264 36.68 21.51 0.2109 -6.03 in many of our higher pH solutions. In other experiments, the TA and DIC stay fairly constant throughout. Two experiments at pH 9.3 (CA10, CA11) record anomalously high [DIC] for their pH value (and therefore calculated Ω) that continues to increase throughout the experiment duration, which we attribute to human error during sampling for DIC measurements (e.g. not filtering the solution properly, improper sample storage conditions) since none of the other experimental values (TA, isotopic composition of calcite or solution) seem to be outliers for those experiments. We therefore exclude the values from CA10 and CA11 when fitting trends to Ω and ion activity aCa2+/aCO2− ratios later in this study.3 Though there is a wide spread in δ13C of DIC (Figure 1d) that primarily reflects the range in δ13C of our CO2-in-N2 gas sources, within each experiment the carbon isotopic composition of the overall DIC pool remains fairly constant throughout the duration. The average DIC for an experiment is typically ∼ 3-8 ‰ isotopically heavier than its source gas. DIC speciation is primarily a function of pH. HCO−3 is the most abundant DIC species across our entire experimental range, with an increasing [CO2−3 ] from ∼ 0.3 to 16% of the total DIC pool (Charlton & Parkhurst, 2011; De Lucia & Kühn, 2013; Millero et al., 2007). We calculated DIC speciation using the PHREEQC Minteq.v4 database, which is approximately equivalent to the HCO−3 /CO 2− 3 ratio of Millero et al. (2007) for simple NaCl solutions over this pH range. Our solutions are supersaturated with respect to calcite at all times, with changes in Ω being governed by changes in [DIC] (Figure 1f). We precipitate calcite without the use of seed crystals, 72 so the peak in Ω values for many experiments at the onset of calcite precipitation reflects the critical degree of supersaturation necessary to initiate calcite growth. Both the [HCO−/CO2−3 3 ] and aHCO−/aCO2− ratios decrease exponentially with increasing pH, though the ion activity ratio3 3 is consistently ∼ 2.1 times greater than the concentration ratio due to ions of greater charge magnitude having lower effective activity in solution relative to ions of smaller charge or neutral species. Since the proportion of CO2−3 of total DIC increases with pH while our starting [Ca 2+] was constant, both the concentration and ion activity ratios for Ca2+/CO2−3 decrease with pH. In all solutions, aCa2+ >>> aCO2− , ranging from ∼ 6,000 to 14,000 (Fig. 2c). As a consequence of DIC3 speciation shifts, the degree of supersaturation increases with pH (Figure 2a) since it is calculated from the ion activities of Ca2+ and CO2−3 , and does not consider any contribution from HCO − 3 . In contrast to the stable {101̄4} rhombohedral calcite face, scalenohedral faces in calcite are polar and unstable in stoichiometric solutions. These faces may be stabilized by the adsorption of charged species, which is promoted at high pH and high aHCO−/aCO2− (Ruiz-Agudo et al.,3 3 2011). Our solutions are characterized by particularly high aHCO−/aCO2− (∼ 6,000-14,000), which3 3 could explain the range in crystal morphology present in our precipitates. Some experiments were dominated by rhombohedral crystals, while others had an abundance of more equant, rhombo- scalenohedral morphologies. The slope of NaOH addition serves as a rough visual proxy for growth rate, with faster growth rates resulting in steeper slopes (Figure 1c). We hold our solutions at constant pH by use of an autotitrator that dispenses NaOH in response to processes occurring that acidify the solution: dissolution of CO2 gas and calcite precipitation. The NaOH addition curves often have change in slope at the onset of calcite precipitation, which helps in determining when growth began while early calcite crystals are too small to be directly observed. Experiments from pH 7.9 to 9.3 have mostly linear NaOH slopes, sometimes with a slightly steeper initial slope that corresponds to a more rapid growth rate at precipitation onset before reaching a steady state. The experiments at pH 7.5 are clear outliers among the NaOH curves, having a slope of zero for up to multiple days, followed by sudden exponential increase. Direct observation of these experiments reveal a few calcite grains precipitate during the slope of zero, but that calcite was not abundant until a while after NaOH addition began again. For these experiments, a uniform growth rate over the entire precipitation period is not a reasonable approximation. Instead, we adjust the growth periods of these experiments to reflect when growth rate was substantial. However the true rates at pH 7.5 are likely faster than our calculated values. As pH increased from 7.5 to 9.3, the growth rate decreased from log10R ∼ -6.1 to ∼ -6.5 mol/m2/s (Figure 2b, Table 1). The NaOH added over experiment duration serves as a pseudo- growth rate proxy (Fig. 1c), as both CO2 dissolution and CaCO3 precipitation act to lower pH and therefore drive NaOH addition because the autotitrator and software hold the pH to set value of ± 0.02. The overall trend of decreasing growth rate with increasing pH is an unexpected result due to the fact that growth rate often co-varies with degree of supersaturation (Dietzel et al., 2009; Nielsen et al., 2012), but we observe an inverse relationship between growth rate and Ω instead 73 7 a 6 5 4 3 2 1 y = 1.0859x - 5.1103 R2=0.51 0 7.5 8.0 8.5 9.0 9.5 5.8 b 6.0 6.2 6.4 6.6 y = -0.2214x - 4.4633 R2=0.72 6.8 7.5 8.0 8.5 9.0 9.5 16000 c 14000 12000 10000 8000 6000 y = -2367.2x - 30072 R2=0.41 4000 7.5 8.0 8.5 9.0 9.5 pH Figure 2: Over the pH range 7.5-9.3, (a) Ω increases, (b) growth rate decreases, and (c) aCa2+/aCO2−3 decreases. 74 aCa2+/aCO32- log10R (mol/m2/s) (Fig. 2a). 3.1 Carbon isotopic fractionation It is difficult to distinguish between contributions to the calcite mineral from HCO−3 and CO 2− 3 using carbon isotopes over the pH range of many natural waters (pH 7-10) due to the very small (∼ 0.4‰) equilibrium isotope fractionation between HCO−-CO2−3 3 resulting in the sum of DIC δ 13C varying minimally (Fig. 3; Millero et al., 2007; Mook, 1986; Zeebe, 2007). This is despite the fact that DIC speciation shifts significantly within this pH range, from 95% HCO−3 and ∼0.3% CO 2− 3 to 84% HCO−3 and 16% CO 2− 3 at pH 7.5 and 9.3, respectively (Charlton & Parkhust, 2011; De Lucia & Kühn, 2013; Millero et al., 2007). The HCO−3 fraction of the DIC pool reaches its maximum of 97.6% at pH 8.1, while [CO2−3 ] increases continuously over the pH range studied. We expected 1000lnαc−DIC to either be similar to that of the sum of DIC, or a few permil isotopically heavier. Seeded calcite growth experiments conducted at 10-30°C and pH 6.3-7.0 under chemo-stat conditions found the carbon fractionation between calcite and HCO−3 to be a temperature-independent value of 1.6‰ (Levitt et al., 2018). They suggested this temperature- independence may be a reflection of the temperature-independent CO2−3 carbon isotope composi- tion, since calcite primarily forms from carbonate ions, or that “oxygen isotopes control isotope partitioning between DIC and calcite, and carbon isotopes have little to no influence” (Levitt et al., 2018). When calculated as fractionations from the DIC pool, the fractionations from Levitt et al. (2018) are larger and temperature-dependent; 4.51‰ at 10°C, 3.72‰ at 20°C, and 4.14‰ at 30°C. Baker (2015) found ∆13Cc−DIC to be between -0.2 and 0.9‰. Over half of our experiments yield 1000lnαc−DIC values between -1 and 1‰, but many experiments yield calcite that is isotopically lighter than DIC by 1.5 to 9‰ (Fig. 3, Table 2). These low negative carbon isotope fractionations are beyond even what is hypothesized as the kinetic limit for carbon isotope calcite-DIC fractionation. Our δ13CDIC values used in these calculations are averages of several DIC water samples taken over the course of the experiment. Often, δ13C and concentration of DIC reaches and maintains a steady state shortly after the onset of calcite precipitation (Fig. 1E). In some of our experiments, however, the δ13C of DIC continually increases or decreases throughout an experiment, but still not of a magnitude that would cause the very large negative isotope fractionation discrepancy we observe in the data. Due to the many replicable DIC water samples, we conclude that the calcite samples from these experiments may have experienced contamination from isotopically light organic carbon. Importantly, the oxygen isotope fractionations seem unaffected by the anomalous carbon isotope fractionations. For example, the two experiments at pH 9.0 record very different 1000lnαc−DIC values, one being within the expected range (-0.15‰) and the other puzzlingly low (-9.21‰) (Fig. 3). Their oxygen isotope fractionations between calcite and water, however, are approximately equal (27.46‰ and 27.54‰) (Fig. 4, Table 2). Therefore, we feel confident in our oxygen isotope data despite harboring concerns regarding the carbon isotope data. 75 2 1 HCO3- 0 DIC CO32- 1 2 3 4 5 6 bCA (µM) 1.5 1.0 7 0.5 0.0 8 CO2 9 10 7.0 7.5 8.0 8.5 9.0 9.5 10.0 pH Figure 3: Calcite-DIC carbon isotope fractionation expressed as 1000lnαc−DIC. Carbon isotope results from this study’s experiments are highly variable and inconsistent with expected frac- tionations due to the complementary oxygen isotope fractionations. Values for 1000lnαCO2−DIC, 1000lnαHCO3−DIC, and 1000lnαCO3−DIC are calculated using the Zeebe (2007) expression modified for carbon isotopes, Millero et al. (2007) DIC pK values for simple NaCl solutions, and equilibrium fractionation factors reported in Mook (1986) and Zhang et al. (1995). 76 1000ln x-DIC Table 2: Isotopic data Exp [DIC] Ω δ13C δ18O δ18O δ13c c w CDIC 1000ln 13α 18c−DIC 1000ln αc−w (mM) (VPDB) (VSMOW) (VSMOW) (VPDB) S6 - - -46.06 16.83 -11.31 - - 28.057 S8 0.1451 2.69 -45.37 16.65 -11.39 -45.33 -0.042 27.961 CA4 0.1638 3.03 -25.44 13.68 -11.68 -25.07 -0.385 25.334 CA10 0.4223 38.92 -43.02 15.62 -11.83 -41.52 -1.565 27.396 CA11 0.4158 38.37 -43.34 15.66 -11.80 -42.45 -0.923 27.412 CA16 0.4157 3.23 -18.98 16.31 -11.85 -19.95 0.992 28.097 CA17 0.4340 2.74 -18.28 16.23 -11.84 -18.18 -0.101 28.009 CA18 0.2402 4.11 -17.67 16.21 -11.88 -17.36 -0.309 28.030 CA20 0.1895 3.08 -34.77 16.12 -11.95 -33.12 -1.705 28.018 CA21 0.1440 5.25 -33.22 16.29 -11.52 -27.91 -5.476 27.740 CA22 0.0816 4.95 -31.79 15.64 -11.87 -22.83 -9.213 27.457 CA23 0.0661 6.13 -47.43 8.30 -10.65 -41.70 -6.003 18.975 CA24 0.0521 4.83 -40.85 3.24 -10.85 -37.23 -3.763 14.148 CA25 0.1010 3.68 -42.79 16.89 -10.91 -40.54 -2.344 27.717 CA26 1.0581 3.28 -18.52 17.08 -10.96 -18.24 -0.285 27.961 CA27 1.0374 3.22 -18.93 16.98 -11.15 -18.32 -0.626 28.050 CA28 0.0802 5.16 -22.36 16.32 -11.28 -22.22 -0.146 27.537 CA29 0.0490 4.54 -21.57 16.25 -11.36 -21.43 -0.137 27.540 CA30 1.2360 3.81 -20.21 17.11 -10.86 -19.11 -1.123 27.891 CA31 1.2330 3.81 -19.82 16.89 -11.23 -19.93 0.113 28.044 3.2 Oxygen isotopic fractionation Our oxygen isotope fractionations between calcite and water at lower pH (7.5 – 8.3) are in ap- proximate agreement with the δ18O-T calibration of Kim and O’Neil (1997), but at higher pH our fractionations become smaller and fall off this calibration to lower values (Fig. 4). As many previous studies have concluded, the Kim and O’Neil (1997) calibration seems to be a good rep- resentative of oxygen isotope fractionation in many natural and synthetic calcites, but likely does not reflect true oxygen isotope equilibrium (Coplen, 2007; Daëron et al., 2019; Dietzel et al., 2009; Levitt et al., 2018; Watkins et al., 2013). The very slowly-grown cave calcites studied at Devils Hole and Corsica Cave together provide a better δ18O-T equilibrium calibration (Coplen, 2007; Daëron et al., 2019). Equilibrium oxygen isotope partitioning between calcite and water is strongly temperature-dependent, while kinetic isotope effects depend on mineral growth rate, solution pH, and degree of supersaturation, rather than temperature. Several experimental studies have aimed to quantify the 1000lnαc−w dependence on mineral growth rate, solution pH, and degree of supersaturation. We compare our results to that of other experiments where calcite precipitated from an equilibrated DIC pool. The experiments of Watkins et al. (2014), Baker (2015), and this study used sufficient carbonic anhydrase (bCA) to ensure isotopic equilibration of the DIC pool (Uchikawa & Zeebe, 2012). While experiments of Levitt et al. (2018) did not utilize bCA, the low pH of their solutions and slower growth rates suggests that isotopic equilibration of the DIC pool was achieved. Other studies, including Kim et al. (2006), Dietzel et al. (2009), and Gabitov et al. (2012) suffer from a nonequilibrated DIC pool, which 77 complicates interpretations of KIEs occurring at the mineral surface. This study systematically controlled solution pH, and calculated growth rate and degree of supersaturation for all experiments. As pH increases from 7.5 to 9.3, 1000lnαc−w decreases from ∼28.0 to ∼27.4‰ (Figure 4). Similar trends but with slightly different slopes and intercepts were found in Baker (2015) (1000lnαc−w ≈ 28.3 to 27.7‰) and Watkins et al. (2014) (1000lnαc−w ≈ 28.7 to 27.0‰) (Fig. 4, see inset) using the same experimental methods and set-up. The decrease in 1000lnαc−w observed due to increasing pH has previously been attributed to shifts in the proportions of different DIC species contributing to the calcite mineral lattice (Zeebe, 1999). As pH increases from 7.5 to 9.3, the relative proportion of CO2−3 increases while HCO − 3 and CO2 decrease, leading to an isotopically lighter DIC pool because equilibrium CO 2− 3 is 6.8‰ depleted in 18O relative to HCO−3 . We predict that as pH increases, the HCO − 3 contribution to calcite decreases, as the observed decreasing 1000lnαc−w trend would be in agreement with a decreasing contribution from the isotopically heavier species, HCO−3 . We observe smaller isotopic fractionations at pH 7.5 than we would expect based on the ap- proximately linear trend of our 1000lnαc−w values at higher pH. These smaller fractionations at pH 7.5 could be due to the elevated growth rate we observed in these experiments (see steep NaOH addition curve - Fig. 1c). Fast growth has previously been found to result in isotopically lighter calcite (Dietzel et al., 2009; Gabitov et al., 2012). A contrasting explanation of the pH 7.5 data is that the trend in 1000lnαc−w is not expected to continue mimicking the slope of the DIC pool isotopic composition at lower pH values with larger proportions of CO2(aq) since CO2(aq) is not inferred to contribute to the calcite lattice, while both HCO− and CO2−3 3 do (Watkins & Devriendt, 2022). 4 Updating the ion-by-ion model We use our new experimental CaCO3 oxygen isotope fractionation data to update the ion-by-ion model from Watkins et al. (2013, 2014) and Watkins and Devriendt (2022). The Watkins model modified that of Wolthers et al. (2012), which combined the simple cubic ionic crystal growth model of Zhang and Nancollas (1998) with the calcite surface speciation and complexation model of Wolthers et al. (2008). The ion-by-ion model assumes calcite growth is proceeding via the spiral growth of hillocks. Many previous atomic force microscopy (AFM) studies suggest that spiral growth, in which calcite grows by step flow at surface defects including screw dislocations, is favored at lower degrees of supersaturation (Ω ≤ 2.2) and at increased Ω the 2D nucleation mechanism becomes increasingly important and then dominates (Larsen et al., 2010; Teng et al., 2000). Some studies are at odds with this, and observe spiral growth even in solutions of high supersaturation (Ω ≈ 6.9, Gratz et al., 1993; Ω ≈ 8-20, Paquette and Reeder, 1995). This discrepancy may be reconciled by a recent study modeling AFM experimental data from Teng et al. (2000) which found that spiral growth is favored for sufficiently large calcite crystals (≥ 1 µm) regardless of Ω (Darkins et al., 2022), as 78 HCO3- 31 DIC 30 Coplen (2007) 29 Kim & O’Neil (1997) 28 27 29.0 28.5 26 28.0 bCA (µM) 1.5 27.5 1.0 25 27.00.5 7.5 8.0 8.5 9.0 9.5 0.0 pH CO32- 24 7.5 8.0 8.5 9.0 9.5 pH Figure 4: Oxygen isotope fractionation between calcite and experimental solution expressed as 1000lnαc−w. At higher pH, greater [bCA] should be necessary to ensure isotopic equilibration of the DIC pool. Oxygen isotope results from this study’s experiments (circles) are consistent with that of similar experiments from past studies (triangles - Watkins et al., 2014; diamonds - Baker, 2015). Oxygen isotope fractionations of this study from pH 7.5-8.3 are in agreement with the Kim and O’Neil (1997) value for 25°C, but trend to lower fractionations above pH 8.3. Inset depicts the slope of 1000αc−w vs. pH for each set of experiments. The DIC isotopic composition is calculated using the Zeebe (2007) expression, Millero et al. (2007) DIC pK values for simple NaCl solutions, and equilibrium fractionation factors reported in Beck et al. (2005). 79 1000ln x-water 1000ln c-w the experiments at high Ω observed spiral growth on sufficiently large calcite crystals (Gratz et al., 1993; Paquette and Reeder, 1995). Darkins et al. (2022) noted the importance of the supersaturation at the calcite surface being lower than the supersaturation in the bulk solution, and that while growth kinetics are determined by surface supersaturation, it is difficult to measure and that growth can be reasonably characterized by a combination of bulk supersaturation and average crystal size. The majority of crystals from our experiments are sufficiently large, and therefore despite the high bulk supersaturation (average Ω > 2.5 in our experiments), the spiral growth mechanism should be reasonably applicable. We do note that newly nucleated crystals are likely dominated by the 2D island nucleation growth mechanism until the crystal has grown to the critical length scale at which spiral growth takes over (Darkins et al., 2022). DIC speciation at the calcite surface differs significantly from that of the bulk solution (An- dersson et al., 2016b; Wolthers et al., 2008; Wolthers et al., 2012). In all of our experimental solutions, [HCO− 2−3 ]>[CO3 ] in the bulk solution, while the same may not be true for calcite surface speciation. The pK for HCO−-CO2−a 3 3 interconversion is 10.35 - 10.5 in the bulk solution, while it is approximately 3 pH units lower for a pKa ∼ 7.5 when adsorbed onto the mineral surface (Andersson et al., 2016b; Plummer and Busenberg, 1982). Bicarbonate in the calcite lattice has even lower pKa values, meaning that bicarbonate readily deprotonates upon incorporation into the mineral (Andersson et al., 2016b). The ion-by-ion model considers separately the DIC speciation on the mineral surface compared to the bulk solution, and modifies the bulk solution speciation to ac- count for temperature and the effect of background electrolytes (Millero et al., 2007; Watkins et al., 2014; Wolthers et al., 2012). How DIC speciation at the calcite surface is affected by temperature, background electrolytes, and factors other than pH is not known. In updating the model, we modified some of the initial parameters and fractionation factors employed in past iterations. We calculate DIC speciation in the bulk solution (φ) using the pK values of Millero et al. (2007) for simple CaCl2 solutions (Table 3). We update the ratio of HCO−3 /CO 2− 3 adsorbed on the calcite surface (θ) to reflect the density functional theory modeling of HCO−3 and CO 2− 3 adsorption energies to relevant calcite surface site geometries (i.e. the {101̄4} cleavage surface and acute and obtuse steps) (Table 3; Andersson et al., 2016b). Model parameters in Table 4 are calculated using the initial parameters given in Table 3. Tables 3 and 4 use a shorthand notation for the constituent ions where A is Ca2+ and B is (bi)carbonate, with B1 as CO2−3 and B2 as HCO − 3 , and A 2+ 1 as Ca associated with CO 2− 3 and A2 as Ca 2+ associated with HCO−3 . The oxygen and carbon fractionation factors used in the model calculations are presented in Table 5. The calcite growth ion-by-ion model predicts the 1000ln18α 13c−w and 1000ln αc−DIC as functions of temperature, solution pH, and growth rate. Results at 25°C are shown over a range of pH and growth rate (log10R = -4 to -10 mol/m 2/s) (Fig. 5). When calcite grows from an isotopically equilibrated DIC pool, KIEs recorded in the mineral may be attributed to processes occurring at the mineral surface (i.e. attachment and detachment of ions). We compare the calcite isotopic 80 Table 3: Model input parameters Parameter Symbol1 Value Units Kink formation energy  6.7 · 10−21 J Edge work γ 1.2 · 10−19 J Adsorbed HCO− 2− 7.5−pH3 /CO3 ratio θ ≈ 10 no units Bulk HCO−/CO2−3 3 ratio φ [B2] [B ] no units1 Attachment frequencies k 3.0 · 106 s−1A1 k ≈ k s−1A2 A1 kB1 = 2k 1+θ −1 A1 · 1+φ s kB ≈ k s−12 B1 Detachment frequencies νA 2 · 103 s−11 νA ≈ ν s−12 A1 νB = Ksk̄Ak̄B −1 1 ν̄A(1+θ) s ν ≈ ν s−1B2 B1 1 B1 = CO 2− − 2+ 3 and B2 = HCO3 , while A1 = Ca associated with CO2− 2+3 and A2 = Ca associated with HCO − 3 . Table 4: Model parameters calculated from input parameters Parameter Symbol1 Value Units Fraction of kink sites suitable for growth χ 1 no units Closest spacing between A and B sites a 3.199 × 10−10 m Molar density of calcite d 27,100 moles/m3 Solubility product for calcite K ≈ 10−8.48sp no units Saturation ratio for calcite S ( [A][B1] )1/2K no unitssp Rate coefficient for A attachment k̄A kA + θkA s −1 1 2 Rate coefficient for B1 and B2 attachment k̄B kB + φkB s −1 1 2 Rate coefficient for A detachment ν̄A νA1 + νA2 s −1 Rate coefficient for B1 and B2 detachment ν̄B ν −1 B1 + θνB2 s Probability that a given site is a B1 site P k̄B[B1]+ν̄A B no units1 k̄A[A]+ν̄B+(1+θ)(k̄B[B1]+ν̄A Probability that a given site is an A site PA 1 - (1 + θ) PB1 no units Probability that a given site is a B2 site PB 1 - PA - PB no units2 1 Net incorporation rate of A ions uA k̄ −1 A [A] PB - ν̄A PA s1 Net incorporation rate of B ions uB k̄ −1 B [B1] PA - ν̄B PB s1 Kink propagation rate uc uA + u −1 B s Rate of kink formation on B sites i 2exp(−2A kT )(S 2 - 1)( ν̄Bk̄A[A] ) s−1 k̄A[A]+ν̄B Rate of kink formation on A sites i 2exp(−2 )(S2 - 1)( ν̄Ak̄B[B] ) s−1B kT k̄B[B]+ν̄A Net rate of kink formation i iA+iBc 2 s −1 Steady state kink density ρ ( 2ic 1/2c u +u ) no unitsA B Step spacing y 19aγ0 kT lnS m 2 Calcite growth rate R ρcuca dc y moles/m 2/s 0 1 A = Ca2+, B = CO2−1 3 , and B = HCO − 2 3 . 81 30 32 DIC HCO -3 log R 30 Calcite (eq) 29 10 (mol/m2/s) fast 4 28 28 6 26 CO 2- 27 83 24 slow 10 22 26 4 5 6 7 8 9 10 11 12 6 7 8 9 10 this study 6 10 Baker, 2015HCO -3 CO 2- 4 Watkins et al., 20143 5 Levitt et al., 2018 DIC 2 Romanek et al., 1992 0 5 0 10 2 4 5 6 7 8 9 10 11 12 6 7 8 9 10 pH pH Figure 5: Updated ion-by-ion model of calcite growth at 25°C in 30 mM CaCl2 solutions. The model predicts 1000ln18αc−w (upper panels) and 1000ln 13αc−DIC (lower panels) as functions of temperature, solution pH, and growth rate (curves for log10R = -4, -6, -8, and -10 mol/m 2/s). Experiments plotted from this study (circles) and other studies (Baker, 2015 - diamonds; Levitt et al., 2018 - inverted triangles; Romanek et al., 1992 - squares; Watkins et al., 2014 - upright triangles) in which calcite grew from an equilibrated DIC pool, so that KIEs may be attributed to attachment and detachment of (bi)carbonate ions at the calcite surface. fractionations of this study’s experiments with that of other studies that reasonably assume calcite grew from an isotopically equilibrated DIC pool (Baker, 2015; Levitt et al., 2018; Romanek et al., 1992; Watkins et al., 2014). Since KIEs in 1000ln18αc−w should be a direct result of ion attachment/detachment, we modify the forward and backward KFFs attending calcite precipitation and dissolution reactions (Table 5) to fit the experimental CaCO3 data of this study. Due to concerns about our carbon isotope data, we did not use our 1000ln13αc−DIC to adjust the model results (Fig. 5 - gray circles). We instead use the carbon KFFs postulated by Watkins and Hunt (2015), which are in agreement with carbon isotope fractionations from Romanek et al. (1992) and Levitt et al. (2018) (Table 5; Fig. 5). This ion-by-ion model is a process-based model that fits experimental isotopic data rather well, despite many necessary assumptions. The model assumes a cubic mineral lattice, which does not account for the differential step and kink geometries of the acute and obtuse sites, which propagate at different rates depending on a variety of factors (i.e. [Ca2+]/[CO2−3 ], Ω; Andersson et al., 2016a; Larsen et al., 2010; Perdikouri et al., 2009; Ruiz-Agudo & Putnis, 2012; Sand et al., 2016). Calcite growth is modeled by the attachment and detachment of free Ca2+, HCO−3 , and CO 2− 3 ions to the 82 1000ln13 x-DIC 1000ln18 x-w Table 5: Oxygen and carbon fractionation factors used to generate the ion-by-ion model curves Fractionation Equation α (25°C) Notes factor Oxygen αeqCO2(g)−w 17.611/TK+0.9821 1.0412 Zeebe (2007) αeqCO (aq)−w exp( 2520 2 +0.01212) 1.0413 Beck et al. (2005) 2 TK αeqHCO −w exp( 2590 2 +0.00189) 1.0315 Beck et al. (2005) 3 TK αeq 2390CO −w exp( 2 -0.00270) 1.0245 Beck et al. (2005)3 TK αeq 17747c−w exp(( T - 29.777)/1000) 1.0302 Coplen (2007); Watkins etK al. (2013) αfc−HCO - 0.9966 Model parameter3 αfc−CO - 0.9995 Model parameter3 αeq eq eqc−HCO αc−w/α3 HCO −w 0.9987 Beck et al. (2005); Copen3 (2007); Watkins et al. (2013) αeq eq eqc−CO αc−w/αCO 1.0056 Beck et al. (2005); Copen3 3−w (2007); Watkins et al. (2013) αb f eqc−HCO αc−HCO /αc−HCO 0.9979 Beck et al. (2005); Copen3 3 3 (2007); Watkins et al. (2013) αb αf eqc−CO3 c−CO /αc−CO 0.9940 Beck et al. (2005); Copen3 3 (2007); Watkins et al. (2013) Carbon αeqCO (g)−HCO 1/((-0.1141*TC + 10.78)/1000 + 1) 0.9921 Zhang et al. (1995)2 3 αeqCO (aq)−CO (g) (0.0049*TC - 1.31)/1000 + 1 0.9988 Zhang et al. (1995)2 2 αeq −867CO3−HCO ( T + 2.52)/1000 + 1 0.9996 Mook (1986)3 K αeq −9866CO (aq)−HCO ( T + 24.12)/1000 + 1 0.9910 Mook (1986)2 3 K αeq exp((-2.4612 + 7666.3 - 2988000CO (g)−c T 2 )/1000) 0.9897 Bottinga (1968)2 K TK αfc−HCO - 1.000 Model parameter3 αfc−CO - 1.000 Model parameter3 αeq 1/(αeq eqc−HCO3 CO2(g)−c/αCO2(g)−HCO ) 1.0025 Bottinga (1968); Zhang et3 al. (1995) αeq αeq eqc−CO c−HCO /αCO −HCO 1.0029 Bottinga (1968); Mook3 3 3 3 (1986); Zhang et al. (1995) αb f eqc−HCO αc−HCO /αc−HCO 0.9975 Bottinga (1968); Zhang et3 3 3 al. (1995) αb f eqc−CO αc−CO /αc−CO 0.9972 Bottinga (1968); Mook3 3 3 (1986); Zhang et al. (1995) 83 calcite surface, assuming constant attachment (k̄A, k̄B) and detachment (ν̄A, ν̄B) frequencies, as well as equal frequencies for HCO−3 and CO 2− 3 ions (kB2 = kB1 , νB2 = νB1) all of which might not be a realistic description of processes occurring at the mineral surface. The importance of ion pairs and polynuclear complexes in calcite growth processes is a continued avenue of research, as ion pairs tend to have faster dehydration rates than free ions and can therefore attach more readily at the mineral surface (Andersson et al., 2016a; Ruiz-Agudo & Putnis, 2012; Ruiz-Agudo et al., 2011). The model assumes that calcite growth proceeds via spiral growth mechanism, which until recently was understood to be unrealistic at our high supersaturations, but a new study highlighting the importance of the critical length scale of calcite crystals suggests spiral growth takes over for crystals ≥ 1 µm (Darkins et al., 2022). However, our experiments are unseeded and therefore processes such as 2D island nucleation are dominant for newly nucleated crystals until they grow sufficiently large. Despite these limitations, the model still constitutes a physics-based parameterization that describes 1000lnα18c−w and 1000lnα 13 c−DIC over a broad range of T, pH, and growth rate. 5 Summary Most natural CaCO3 grows at rates by which the solution pH may result in kinetic isotope effects, even when the DIC is isotopically equilibrated (Baker, 2015; Daëron et al., 2019; Watkins et al., 2013, 2014). We report a decrease in 1000ln18αc−w over the pH range 7.5-9.3, with a slope of ∼ 0.33‰ per pH unit that is in agreement with the experiments of Baker (2015), but which exhibit a higher y-intercept by ∼ 0.12‰ than experiments of this study (Fig. 4). Our use of the enzyme carbonic anhydrase to facilitate isotopic equilibration of the DIC pool allows us to isolate KIEs that result from attachment and detachment of ions at the calcite surface (Uchikawa & Zeebe, 2012; Watkins et al., 2013). We update the ion-by-ion model of Watkins et al. (2014) with improved constraints on equilibrium fractionation between DIC species and water, and DIC speciation on the calcite surface, and report updated oxygen KFFs attending CaCO3 precipitation and dissolution reactions. Quantifying the inorganic DIC-CaCO3 isotopic fractionations and better understanding the mechanisms and processes leading to these KIEs gives insight into separating vital effects from mass-dependent kinetic fractionations in biogenic CaCO3. Understanding biocalcification processes is of high priority since the vast majority of calcic seafloor sediments are biogenic and constitute an important global paleoenvironmental archive (Zachos, 2001; Zeebe & Wolf-Gladrow, 2001). 6 Bridge The prior two chapters examine the effect of ionic strength (Chapter III) and solution pH (Chap- ter IV) on kinetic isotopic effects recorded in nonequilibrium inorganic calcite, but under growth conditions that result in relatively small magnitude KIEs (∼ 2-5‰). In Chapter V, I conduct exper- 84 iments investigating extreme kinetic isotopic effects during far-from-isotopic-equilibrium carbonate precipitation. These experiments simulate processes occurring at hyperalkaline springs upwelling through ultramafic ophiolitic bodies, where CaCO3 travertines precipitate at the air-water interface as the high pH, DIC-free spring water interacts with atmospheric CO2. This work quantifies the kinetic fractionation factors attending CO2 hydroxylation, and contributes valuable insights into understanding the isotopic signatures of carbonates from these environments that helps lay the groundwork for their use as paleoenvironmental archives. 85 CHAPTER V EXTREME ISOTOPIC FRACTIONATIONS IN CACO3 FROM A HYPERALKALINE SPRING ANALOG: QUANTIFYING KINETIC FRACTIONATIONS ATTENDING CO2 HYDROXYLATION This chapter is in preparation for submission to Geochemistry, Geophysics, Geosystems, co- authored with J.M. Watkins and L.S. Devriendt. The experiments were performed by E.K. Olsen. J.M. Watkins modeled the results and compared them to a previous study. All other writing is by E.K. Olsen, with editorial assistance from J. M. Watkins. 1 Introduction Hyperalkaline springs upwelling through ultramafic ophiolite formations commonly result in traver- tine precipitation upon contact with atmospheric CO2. Carbonate precipitated in this environment records large kinetic isotope effects that are not fully understood, but likely reflect kinetic fractiona- tions during CO2 hydroxylation. The isotopic composition of carbonates formed in this environment is a potential paleoproxy to reconstruct the isotopic composition of atmospheric CO2 and mete- oric water from which it grew (Christensen et al., 2021). However, to utilize natural travertines as a paleoenvironmental archive, we must more fully understand and quantify the kinetic isotopic fractionations that occur during their precipitation. The spring waters typically originate as meteoric water seeping downward into the peridotite, reacting with olivine and pyroxene to form serpentine, brucite, iron hydroxides, and other hydrated secondary minerals (Chavagnac et al., 2013; Falk et al., 2016; Mervine et al., 2014; Palandri & Reed, 2004). During serpentinization, the groundwaters become isolated from the atmosphere and evolve with progressive reactions. Calcium silicate minerals are generally unstable in high Mg serpentine systems, so the breakdown of clinopyroxene results in increased Ca2+ in the groundwater (Morrill et al., 2013; Palandri & Reed, 2004). The dissolved carbon content is precipitated at depth in the form of carbonate minerals, commonly magnesite (MgCO3) (Falk et al., 2016). Overall, serpentinization of the peridotite shifts water composition from its Mg-HCO3 meteoric origin to increased pH, increased Ca2+, decreased Mg2+, and decreased dissolved inorganic carbon (DIC) (Chavagnac et al., 2013; Falk et al., 2016; Morrill et al., 2013). These high pH (up to pH 11.7, Chavagnac et al., 2013; Christensen et al., 2021; up to pH 12-12.1, Falk et al., 2016; Giampouras et al., 2019; Morrill et al., 2013) spring waters then upwell at the surface as Ca-OH waters, with atmospheric 86 CO2 dissolving into solution and being rapidly precipitated out as CaCO3. 1.1 CO2 captured as CaCO3 As CO2 gas diffuses from the atmosphere into the Ca-OH pools, it dissolves into solution by the reaction: K0 CO2(g)←−−−→− CO2(aq), (1) during which the isotopes fractionate according to mass-dependent processes. CO2 exchange at the gas-solution interface is rapid, so we assume isotopic equilibrium is achieved between gaseous and aqueous CO2 (Clark et al., 1992). During equilibrium isotopic fractionation, the heavier isotopes tend to be enriched in the phase in which they are more strongly bound, and in this case carbon and oxygen have opposite fractionation effects where 13C prefers CO2(g) and 18O prefers CO2(aq). At 25°C, the resulting CO2(aq) is depleted in 13C by ∼1.1-1.4‰ (Vogel et al., 1970; Yumol et al., 2020; Zhang et al., 1995), and enriched in 18O by ∼0.3‰ (Barkan and Luz, 2012; Beck et al., 2005). Reactions involving molecules containing 13C or 18O proceed more slowly than reactions in- volving molecules containing only the light isotopes, thereby enriching the product species in light carbon and oxygen if the reactions are unidirectional, or mostly unidirectional. Upon dissolving into solution, the CO2 undergoes primarily unidirectional hydroxylation: − −k+CO2(aq) + OH −→ 4 HCO −3 (2) producing HCO−3 that is depleted in both heavy carbon and oxygen relative to the reacting CO2(aq). CO2 hydration and hydroxylation are the slowest DIC exchange reactions, and are also the only reactions through which DIC exchanges oxygen with water (Sade & Halevy, 2017; Zeebe & Wolf- Gladrow, 2001). In disequilibrium systems, the kinetic isotope effects attending these reactions are more likely to be recorded in carbonate minerals compared to the other, more rapid DIC exchange reactions. The relative production of HCO−3 from CO2 hydration versus hydroxylation depends on solution pH. In low ionic strength solutions at 25°C, the contribution from CO2 hydration is negligible at pH 10.5 or higher (Bajnai and Herwartz, 2021; Devriendt et al., 2017). The isotopic composition of the HCO−3 should reflect the CO2 hydroxylation kinetic fraction- ation factors (KFFs) for both oxygen and carbon, provided that the amount of back-reaction is negligible as we reasonably expect for our solutions. The dissolved CO2 is the only source of carbon, while oxygen comes from both the dissolved CO2 and OH −. A single KFF can describe the carbon kinetic fractionation, while two KFFs are required to fully explain the oxygen isotope signature of the resulting precipitate: the KFF for CO2(aq)-HCO − 3 and the KFF for OH −-HCO−3 . However, previous studies have been unable to tease apart the KFFs for CO2 and OH − separately during CO2 hydroxylation, and instead report 18O/16O fractionation using a single bulk oxygen KFF that is weighted according to the number of oxygen atoms each source contributes to the 87 HCO−3 (Christensen et al., 2021; Dietzel et al., 1992). HCO−3 then converts to CO 2− 3 either via deprotonation: kH+ − −−+5HCO3 ←−−−→− CO 2− +3 + H (3) kH+−5 or by reacting with OH−: kOH−+5 HCO − −3 + OH −←−−−−−−→− CO 2−3 + H2O · (4) kOH−−5 HCO−3 - CO 2− 3 exchange is many orders of magnitude faster than hydroxylation over our pH range of interest, so therefore instantaneous isotopic equilibration of HCO3-CO3 may be assumed (Eigen, 1964; Christensen et al., 2021; Pinsent et al., 1956; Sade & Halevy, 2017; Zeebe & Wolf- Gladrow, 2001). Equilibrium CO2−3 is 0.4‰ and 6.8‰ isotopically lighter than HCO − 3 for carbon and oxygen, respectively (Beck et al., 2005; Christensen et al., 2021; Mook, 1986). However, no net isotopic fractionation would be expected during quantitative conversion of HCO− to CO2−3 3 , which is a reasonable assumption for pH > 12 where HCO−3 /DIC < 1% (Christensen et al., 2021). With unidirectional CO2 hydroxylation, the full kinetic fractionation is expressed in the EIC (equilibrated inorganic carbon = HCO−3 + CO 2− 3 ) prior to CaCO3 precipitation. Assuming near quantitative conversion of EIC to CaCO3, then the CaCO3 precipitates would reflect the CO2 hydroxylation KFFs. Previous studies have addressed the isotopic composition of travertines from natural hyperalkaline springs as well laboratory experiments in an effort to quantify the KFFs for CO2 hydroxylation. Many studies agree that the 13C KFF is quite large, approximately -17‰ with little to no temperature dependence, while the bulk 18O KFF appears to have a stronger temperature-dependence, and an approximate value at 25°C of -7.2‰ when expressed relative to the weighted sum of CO2(aq)+OH −, or -13.9‰ when expressed relative to the weighted sum of CO2(aq)+H2O (Böttcher et al., 2018; Christensen et al., 2021; Clark et al., 1992; Falk et al., 2016; Mervine et al., 2014; Zeebe, 2020). If the EIC is not quantitatively precipitated as CaCO3, then we would expect isotopic fraction- ation to occur based on the degree of EIC distillation, but also based on the extent the CaCO3 precipitation reactions are unidirectional or bidirectional. At low supersaturation, CaCO3 precip- itation reactions would be bi-directional. However, in our high supersaturation solutions (Ω > 5), negligible back-reaction is expected: CO 2− 2+ −k−+c3 + Ca → CaCO3 · (5) Unidirectional precipitation of CaCO3 from CO 2− 3 should result in calcite with δ 13C ∼ 0.5‰ heavier and of δ18O ∼ 0.5‰ lighter than the reacting CO2−3 (Devriendt et al., 2017; Kim et al., 2006; Sade et al., 2020). 88 1.2 OH−-H2O fractionation The dissociation of water: H2O − Kw ←−−→− H+ + OH− (6) is a relatively rapid reaction, with equilibration timescales akin to the bicarbonate-carbonate ex- change, and therefore is regarded as having attained equilibrium when dealing with the DIC system, which equilibrates much slower due to the rate-limiting CO2 hydration and hydroxylation reactions through which DIC exchanges oxygen with H2O. There is considerable disagreement as to the equilibrium oxygen fractionation between OH− and H2O. Experimental studies suggest that at 25°C, 1000lnαOH−−H O ∼-42.5‰ (Bajnai & Herwartz,2 2021; Green & Taube, 1963), while theoretical calculations suggest much smaller values of -19.1 to -23.5‰, depending on the number of water molecules involved in the reactions (Zeebe, 2020). This ∼20‰ offset between experimental and theoretical studies introduces a significant degree of uncertainty into our fractionation factor calculations. Therefore, we choose to present bulk oxygen KFFs from CO2+H2O, in addition to the KFFs from CO2+OH − which would be the true contributing source of the oxygen. 1.3 Motivation Past studies have addressed the isotopic composition of travertines from natural hyperalkaline springs and carbonate precipitation experiments conducted at high pH, with the aim of quantifying the kinetic isotope fractionations attending CO2 hydroxylation and/or the equilibrium oxygen isotope fractionation between H −2O and OH . However, separating the kinetic fractionation factors attending CO2 hydroxylation from equilibrium isotopic fractionations between OH − and H2O have proved difficult, and past studies have failed to do so definitively (Zeebe, 2020). The isotopic composition of the carbonate mineral may be described by minimal OH−-H2O fractionation and a large CO2 hydroxylation KFF (Böttcher et al., 2018), or by a large OH −-H2O fractionation and a small to negligible CO2 hydroxylation KFF (Clark et al., 1992), or anywhere in between those end-member scenarios. We conducted well-controlled laboratory experiments in simple CaCl2 solutions at high pH in an attempt to simulate what occurs naturally at hyperalkaline springs such as The Cedars, California, and the Samail Ophiolite, Oman, in order to better understand how these large isotopic fractionations arise. We sought to recreate the (1) mineralogy, (2) morphology, and (3) isotopic fractionations observed in natural travertines, with the goal of isolating and quantifying the KFFs attending CO 13 12 18 16 18 16 −2 hydroxylation ( C/ C on CO2(aq), O/ O on CO2(aq), and O/ O on OH ). 89 2 Methods Our initial experimental design (Figure 1) consists of a sealed beaker containing 1.3 L distilled, deionized (DDI) water with 10 millimolar (mM) or 30 mM CaCl2·2H2O sitting in a temperature- controlled water bath, held at 10°C or 25°C (Table 1). Initially, pure N2 gas is fluxed into a rapidly-stirring solution until the CO2 concentration of the headspace drops to < 30 ppm CO2, typically over 1-1.5 hours. N2 was bubbled for at least 1.5 hours to minimize the amount of remnant DIC in the experimental solution so that experimental precipitates can be reasonably expected to form from the CO2 introduced later. For experiments CH2 - CH10, a mixture of 3 M KCl and 1 M NaOH was added to the solution to raise the pH to ∼12.4, with the final solution containing 50 mM KCl and 12 mM NaOH. For experiments CH1, CH11 - CH27, 1 M NaOH was added to reach the desired pH (11 - 12.6), with the final solution containing 1 - 18 mM NaOH (Table 1). The gas bubbler was moved above the solution into the headspace, with N2 still flowing. Stirring was turned down to a very slow rate (∼ 6-12 rpm) or off entirely and the gas was then switched to a mixture of CO2 in N2 (200 ppm or 2000 ppm CO2). As CO2 diffuses into the high pH, DIC-free solution, calcium carbonate precipitates on the surface of the water, forming a mineral crust. Solution pH decreases slightly throughout an experiment (Figure 2, Table 1) as both CO2 diffusion and CaCO3 precipitation act to lower pH, but remains high enough in the majority of our experiments so that kinetic isotope effects from CO2 hydration reactions are assumed to be negligible. We found that in many of our early experiments, the CO2 concentration of the experimental headspace was lower than that of the gas tank, which can complicate interpretation of results. To mitigate this, we adjusted our experimental design (Figure 1b) to maximize the headspace volume to solution surface area where CO2 was being taken up into solution. Experiments CH14-15, 18-19, 21, and 23 were experiments in which a smaller beaker of experimental solution was positioned within the larger experimental beaker that sat in the temperature-controlled water bath. These experiments were entirely unstirred even during the N2 flow period because the smaller beaker had to sit elevated from the magnetic stir bar for the probe and bubbler to reach the solution. We also adjusted our experimental routine to carry out experiments that grew CaCO3 from laboratory air CO2 rather than from a gas tank. These experiments (CH13, 17, 20, 22, 24-27) still follow the original experimental design (Figure 1) and start enclosed with pure N2 gas fluxing through the experimental solution until the CO2 concentration of the headspace is minimized, but then the experiment lid is removed to expose the solution to the laboratory atmosphere. These particular experiments were carried out during weekends and other periods of time in which human presence in the lab air space was minimized to lessen the impact of human CO2 exhalation on the isotopic composition of the CO2 in the lab air. Solution samples for DIC and δ18O were collected at the end of each experiment (Table 2). Mid-experiment DIC samples were also collected for a few experiments (CH5, CH7, CH20) to assess whether the DIC concentration or isotopic composition was evolving over time. Experiments CH22, 24, and 25 were conducted as time series, with mid-experiment water sampling for DIC and δ18O. For later experiments, total alkalinity (TA) was also measured by Gran titration using an 90 91 Table 1: Experimental parameters, solution composition and pH for all experiments of this study. Exp Temp CO2 Volume I CaCl2·2H2O Ca2+ KCl NaOH N2 Exp dur. CO2 flow pH pH pH (°C) (ppm) (mL) (g) (mM) (mM) (mM) (h) (h) (SCF/H) max end unstirred CH1 25 200 1300 0.031 1.922 10.06 0 1.23 2.12 166.93 0.5, 0.25, 0.15 11.029 8.43 CH2 25 200 1300 0.152 5.735 30.01 50 12 2.08 96.27 0.5, 0.2 12.528 12.28 CH3 25 2000 1300 0.152 5.733 30.00 50 12 1.65 70.4 0.5, 0.2 11.93 11.82 CH4 25 2000 1300 0.152 5.735 30.01 50 12 3.6 44.2 0.5, 0.2 12.25 12.03 CH5 25 200 1300 0.152 5.734 30.00 50 12 1.6 117.53 0.5, 0.2 12.426 12.154 CH6 10 2000 1300 0.152 5.738 30.03 50 12 4.12 97.33 0.5, 0.2 12.225 11.87 11.8 CH7 25 2000 1300 0.152 5.734 30.00 50 12 1.68 46.07 0.5 12.444 12.016 8.89 CH8 10 2000 1300 0.152 5.737 30.02 50 12 3 48.67 0.5 12.769 12.355 CH9 25 200 1300 0.092 1.913 10.01 50 12 2.5 66.47 0.5 12.669 12.371 CH10 10 200 1300 0.152 5.735 30.01 50 12 4.83 70.73 0.5 12.238 12.051 CH11 25 2000 1300 0.031 1.913 10.01 0 1.08 5.35 14.47 0.5 11.404 9.443 CH12 25 2000 1300 0.039 1.916 10.03 0 9.23 5.42 17.93 0.5 12.535 12.147 CH13 25 lab air 1300 0.040 1.917 10.03 0 10 4.13 40.53 lab air 12.606 12.254 CH14 25 2000 130 0.042 0.196 10.27 0 10.77 3.48 21.07 0.5 12.528 11.91 11.55 CH15 25 2000 130 0.045 0.197 10.33 0 13.85 4.02 36.53 0.5 12.105 11.305 10.42 CH17 25 lab air 1300 0.034 1.918 10.03 0 3.85 2.07 41.47 lab air 11.735 10.615 CH18 25 2000 130 0.035 0.192 10.05 0 4.62 1.63 27.87 0.5 11.873 10.46 9.036 CH19 25 2000 130 0.034 0.194 10.17 0 3.08 1.67 18.67 0.5 11.606 10.52 8.737 CH20 25 lab air 1300 0.032 1.916 10.03 0 2 2.9 23.93 lab air 11.429 10.617 CH21 25 2000 130 0.034 0.193 10.07 0 3.85 2.62 25.40 0.5 11.813 11.027 10.117 CH22 25 lab air 1300 0.035 1.917 10.03 0 4.62 2.48 30.77 lab air 12.076 10.993 CH22 1 12.076 11.96 CH22 2 11.96 10.993 CH23 25 2000 130 0.033 0.191 10.02 0 3.08 2.67 19.77 0.5 11.868 10.973 9.253 CH24 25 lab air 1300 0.036 1.912 10.01 0 6.15 3.20 39.12 lab air 12.317 10.976 CH24 1 12.317 12.08 CH24 2 12.08 11.78 CH24 3 11.78 10.976 CH25 10 lab air 1300 0.039 1.914 10.01 0 8.46 2.38 72.50 lab air 12.323 11.267 CH25 1 12.323 12.169 CH25 2 12.169 11.853 CH25 3 11.853 11.267 CH26 10 lab air 1300 0.043 1.913 10.01 0 17.69 2.93 43.63 lab air 12.599 12.465 CH27 10 lab air 1300 0.039 1.914 10.01 0 10 2.75 37.73 lab air 12.367 12.143 pH probe Sampling syringe Gas Gas outlet inlet Chiller Heater Rotating disk Plastic “donut” Stir bar Magnetic Stirrer Control Figure 1: Experimental apparatus schematic (left) modified from Watkins et al. (2013) showing the set-up for our 1300 mL experiments where we pulled the gas inlet bubbler above the solution surface into the headspace before turning on the CO2-in-N2 gas, resulting in CaCO3 precipitating and remaining at the air-water interface. Open air experiments used the same design with the lid removed. A photo of our apparatus (right) in the early N2 bubbling stages of a 130 mL experiment with the gas inlet below the solution surface, and the water bath actively heating up. autotitrator dispensing 0.01 M HCl (Table 2). Solution ionic strength and free activities of Ca2+ and CO2−3 were calculated through the R-package of PHREEQC using the Minteq.v4 database (Charlton & Parkhurst, 2011; De Lucia & Kühn, 2013). The degree of supersaturation (Ω = (αCa2+ × αCO2−)/K sp) was determined by using these ionic activities and the solubility product of3 calcite at 25°C (K = 10−8.48sp , Charlton & Parkhurst, 2011; De Lucia & Kühn, 2013; Jacobson & Langmuir, 1974). The measured solution pH of unstirred experiments decreases rapidly upon water sampling at the end of the experiment (Fig. 2g). The pH probe measures solution pH below the air-water interface to ensure the probe remains fully submerged and functioning properly, and solutions of unstirred experiments developed evident pH stratification of the near-surface region, with the true surface pH commonly being much lower than was being measured by the probe. After all water samples were taken, with care to minimize disturbance of the floating precipitates, the solutions were removed from the water bath and precipitates were skimmed off the surface. Skimmed precipitates, as well as the remaining solution, were filtered through 2.5 µm filter paper and rinsed with DDI. 92 Table 2: Direct experimental solution measurements Experiment [DIC] δ13CDIC TA δ 18Ow (mM) (VPDB) (mEq/L) (VSMOW) CH1 0.179 -35.778 - -11.290 CH2 0.067 -33.317 - -11.289 CH3 0.036 -34.781 - -11.193 CH4 0.073 -36.665 - -11.003 CH5 1 0.125 -27.203 - - CH5 2 0.091 -32.954 - -10.484 CH6 0.136 -41.978 - -10.769 CH7 1 0.071 -29.039 - - CH7 2 0.101 -34.104 - -10.642 CH8 0.207 -42.305 - -10.524 CH9 0.113 -36.810 - -10.680 CH10 0.173 -38.061 - -10.806 CH11 0.138 -47.870 0.313 -11.260 CH12 0.077 -43.807 3.947 -11.187 CH13 0.070 -24.813 5.141 -10.365 CH14 0.133 -47.476 2.629 -10.823 CH15 0.337 -49.015 2.546 -10.548 CH17 0.064 -26.695 0.627 -10.544 CH18 0.274 -44.418 0.936 -10.955 CH19 0.254 -37.191 0.448 -10.913 CH20 1 0.041 -19.491 - - CH20 2 0.038 -24.674 0.466 -10.577 CH21 0.687 -33.210 0.700 -10.478 CH22 1 0.125 -26.660 2.906 -10.904 CH22 2 0.095 -15.891 0.646 -10.107 CH23 0.258 -35.351 0.602 -10.360 CH24 1 0.213 -27.925 4.107 -10.664 CH24 2 0.184 -23.007 2.156 -10.409 CH24 3 0.076 -27.272 0.660 -9.728 CH25 1 0.099 -27.945 5.318 -10.787 CH25 2 0.221 -31.791 2.823 -10.797 CH25 3 0.086 -28.783 0.989 -10.745 CH26 0.120 -27.199 11.975 -10.879 CH27 0.180 -26.294 5.991 -10.993 During the time series experiments (CH22, CH24, CH25), each time after the mid-experiment water sampling, all floating crystal precipitates were also skimmed, filtered, and stored separately. Most experiments also resulted in tiny crystals encrusted on the bottom of the experimental beaker, which were also rinsed with DDI and later gently scraped from the beaker. Imaging of precipitates was performed on an FEI Quanta 200 Environmental Scanning Electron Microscrope (ESEM) at the Center for Advanced Materials Characterization (CAMCOR) at the University of Oregon. Isotopic analyses methods are the same as detailed in Chapter III. In this contribution, all isotopic analyses are reported with respect to VPDB unless otherwise specified. 93 94 CH19 CH18 600CH3 CH15 2000 CH4 2000 CH23 CH14 CH21 CH1 CH6 CH24 CH13 200 CH8 CH17CH11 CH7 CH25 1600 1600 CH20 CH22 CH27 CH26 CH10 CH12 400 1200 1200 CH9 CH2 100 800 800 initial pH CH5 200 400 400 12.5 a b c d 0 0 0 0 12.0 13 13 13 13 11.5 CH9 CH12 CH8CH2 CH26 CH5 CH7 CH13 12 CH6CH10 12 12 12 CH27 11.0 CH4 CH14 CH3 CH15 CH22 CH24 CH25 11 11 11 11 CH20 CH18 CH17 CH19 CH11 10 10 10 CH21 10 CH1 CH23 9 9 9 9 e f g h 8 8 8 8 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 0 10 20 30 40 50 0 10 20 30 40 50 60 70 80 time (hours) time (hours) time (hours) time (hours) Figure 2: Experimental headspace CO2 concentration (ppm) (panels a-d) and solution pH (panels e-h) over the course of each experiment. The gray dashed lines (panels a-c) represent the gas tank concentration of CO2 (ppm). The left-most panels are the 200 ppm CO2 experiments, the left-middle panels are 2000 ppm CO2 experiments in which the headspace CO2 took a long time to reach the gas tank value, the right-middle panels are 2000 ppm CO2 experiments in which the headspace quickly reached the gas tank value, and the right-most plots are open, laboratory air experiments. C O ( p p m ) 2 p H C O ( p p m ) 2 p H C O ( p p m ) 2 p H C O ( p p m ) 2 p H 3 Results and Discussion 3.1 Experiment results 3.1.1 CaCO3 precipitates SEM images of precipitates from several experiments illustrate the range in both CaCO3 crystal morphology and polymorphs present, and indicate the experimental design achieved our aim of simulating CaCO3 crusts on the surface of hyperalkaline springs (Figure 3). Additional SEM images of precipitates are presented in Appendix E. Rhombohedral calcite, as well as both disaggregated and radiating acicular aragonite needles are common morphologies produced. Many crystals show a less well-defined habit. Early experiments that contained 30mM Ca2+ tended to produce cohesive, rigid carbonate crusts on the water surface characterized by a flat top side and a hummocky underside, with crystals stemming from the flat top and extending into the solution below. Our CaCO3 crust flakes look remarkably similar to those of natural carbonate crusts from hyperalkaline springs, which also consist of a mixture of calcite and aragonite (see Christensen et al., 2021: Figure 4). Travertines from natural hyperalkaline springs have a variety of mineralogies based on several different factors, including water composition, temperature, water turbulence, and degree of mixing with surface or other groundwaters. Atmospheric CO2 dissolution into the hyperalkaline Ca-OH type spring waters with little to no surface water mixing typically results in precipitation of calcite, aragonite, or a mixture of the two polymorphs. Increased mixing of hyperalkaline spring water with Mg-HCO3 type surface waters tends to result in precipitation of hydroxide minerals in addition to the much more abundant CaCO3 (Chavagnac et al., 2013). Rhombohedral calcite is more likely to form from slower flowing discharge, while more dynamic flows result in aragonite precipitation of a variety of morphologies including needle, dumbbell, and spheroidal (Chavagnac et al., 2013). In our simple CaCl2 solutions, we do not allow for mixing with additional water types (i.e. surface Mg-HCO−3 type waters) as might occur at natural hyperalkaline spring systems. We main- tained a low degree of water turbulence with a very slow stirring rate or even unstirred, in an effort to keep the CaCO3 floating at the air-water interface. We do not observe systematic differences in polymorph prevalence between our stirred and unstirred experiments. The dominant factors influencing CaCO3 polymorph formation are water temperature and aque- ous Mg/Ca ratio (Chavagnac et al., 2013; Folk, 1994). Findings from a combination of experiments and studies of natural spring samples suggest that only calcite forms at temperatures <25°C, a mixture of calcite and aragonite precipitates from 30-60°C, and only aragonite grows at elevated temperatures >70°C (Chavagnac et al., 2013; Folk, 1994; Kitano, 1962). However, another main control in addition to water temperature on CaCO3 polymorph formation is the Mg/Ca ratio of the springs. Aragonite dominates at Mg/Ca = 1, superseding the effect of water temperature and precipitating aragonite in cold springs that would otherwise be favored to grow calcite (Chavagnac et al., 2013; Folk, 1994). Some studies agree, finding calcite dominated springs at low temperatures and low Mg/Ca (<25°C and Mg/Ca ∼ 0, Chavagnac et al., 2013; <40°C and Mg/Ca < 0.5, Gi- 95 ampouras et al., 2019), while others are at odds. Despite the low temperature and Mg/Ca ratio of the springs, Christensen et al. (2021) found that aragonite was the dominant CaCO3 polymorph, with minor calcite, which suggests that while temperature and solution Mg/Ca are important, CaCO3 polymorph formation is influenced by additional environmental and/or geochemical fac- tors. Our experimental solutions contain no Mg, and CaCO3 precipitated in the temperature range where calcite is favored, and yet, all experiments yielded mixtures of calcite and aragonite. In many hyperalkaline springs, carbonate minerals precipitate as a crust on the surface of the water, with additional unconsolidated material either growing or collecting at the bottom of the pools. Different studies use different terms for each of these mineralizations: the surface material has been referred to as carbonate crusts (Chavagnac et al., 2013), surface films (Falk et al., 2016) or floes (Christensen et al., 2021) while the unconsolidated bottom material has been called bottom floc (Falk et al., 2016) or snow (Christensen et al., 2021). Falk et al. (2016) found calcite to be more common in surface crusts, and aragonite more common in unconsolidated deposits at the bottom of spring pools. Christensen et al. (2021) attributes the snow at the bottom of the pool to be sunken surface floe material, whereas other studies of natural springs do not speculate on where the bottom material originally precipitated. In this study, we report isotopic analyses of CaCO3 crusts precipitated at the air-water interface, and, for our open-air experiments, analyses of CaCO3 crystals that are reasonably interpreted as having precipitated at the bottom of the experimental beaker (“beaker crystals”: “ b”, Table 3) rather than be sunken surface crystals. 3.1.2 Impact of varying experimental parameters We varied key experimental parameters to determine the effect on the morphology, mineralogy, and isotopic composition of the resulting carbonate crusts. The key variables at play are temperature (10°C or 25°C), [Ca2+] (10 mM or 30 mM), CO2 gas source and isotopic composition (200 ppm or 2000 ppm CO2 gas tanks, and laboratory air ∼ 450 ppm CO2), the CO2 concentration of the experiment headspace relative to the CO2 concentration of the gas source, the presence or absence of a NaOH/KCl mixture to help control solution pH, the initial starting pH of the solution, the final solution pH of the solution, and whether the solution was stirred or not. The solution pH and headspace CO2 (ppm) of each experiment was continuously measured (Figure 2). In many early experiments, the concentration of CO2 in the experiment headspace was lower than the CO2 concentration of the gas tank source for the majority of the experiment duration (Fig. 2a). Each gas tank experiment began with an initial 0.5 SCF/H (standard cubic feet per hour) flux of the CO2-in-N2 mixture. For CH2-6 the gas flow rate was lowered to 0.2 SCF/H once the CO2 concentration of the headspace reached the gas tank value, which then resulted in a sudden drop in headspace CO2 early on in those experiments (Fig. 2a, b). The flow rate of CH1 was lowered ∼ 44 hours into the experiment, which seems to have had minimal impacts on the headspace CO2 and precipitates as solution pH was already fairly low. For all gas tank experiments CH7 and onward, the flow rate remained at 0.5 SCF/H throughout the entire experiment. The rate of CO2 flow at 0.5 SCF/H for 200 ppm and 2000 ppm CO2-in-N2 gas tanks corresponds to 96 CH4 hummocky CH7 CH6 crust underside radiating aragonite rhombohedral needles calcite rhombohedral calcite CH3 CH5 disaggregated CH10 acicular aragonite flat top of crust blocky, aragonite irregular needles calcite hummocky underside Figure 3: SEM images of CaCO3 precipitates from several experiments. Some experiments produced flaky, cohesive carbonate crusts with a flat top and hummocky underside (CH3, CH4, CH10) while others resulted in less cohesive, disaggregated crystals (CH5, CH6, CH7). Calcite is present ranging from rhombohedral crystals with distinct crystal faces to blocky, irregular masses with poorly defined to absent crystal faces. Aragonite is present as radiating acicular clusters to disaggregated needles. 0.126 and 1.264 mmol CO2 per hour. Even without manually lowering the CO2 flow rate, most early gas tank experiments did not reach their gas source CO2 concentration until far into the experiment. CH1 is the only 200 ppm CO2 experiment in which the headspace is approximately 200 ppm CO2 for the experiment duration, while the other experiments conducted on that gas tank have headspace values that range from ∼100-150 ppm CO2 (Figure 2a). We were concerned that the experimental headspace having a lower CO2 concentration than the source gas meant that CO2(g) distillation was occurring. We use the term CO2 distillation to describe the chemical and isotopic evolution of the CO2 reservoir that would occur when the rate of CO2 uptake by the solution outpaces the replenishment of the gas to the headspace. During this gaseous CO2 reservoir distillation, the concentration of CO2 in the headspace would be less than that of the gas tank source. Its isotopic composition would also be affected, though minimally so, due to small KFFs attending CO2 dissolution (Vogel et al., 1970; Zhang et al., 1995). Experiments with higher starting pH (12-12.5) consistently had headspace CO2 concentrations lower than their gas source, while CH1 (initial pH 11) did not, so to mitigate potential CO2(g) 97 reservoir distillation effects we carried out experiments with lower initial pH and tried to avoid solution pH dropping too low while still precipitating enough CaCO3 for analysis. CH1 started with 10 mM Ca2+ rather than the 30 mM Ca2+ used in most of our other early experiments, so we began using 10 mM Ca2+ instead. This is still much more concentrated than that of most natural hyperalkaline springs, which have a [Ca2+] ∼ 1-2 mM (Christensen et al., 2021; Falk et al., 2016; Morrill et al., 2013). While differences in experimental behavior or CaCO3 crust traits were observed, our experiments do not show significant differences in the carbon and oxygen isotopic composition of carbonate precipitates due to differences in temperature, [Ca2+], or presence or absence of the NaOH/KCl control mixture. Solutions with 30 mM Ca2+ resulted in precipitation of an interlocking, more cohesive carbonate crust on the surface of the experimental solution, in contrast to the smaller precipitate yields of the 10 mM Ca2+ solutions. Precipitates also grew more slowly in 10°C solutions compared to 25°C solutions. CH24 and CH25 had the same starting pH ∼12.3, but CH25 (at 10°C) took much longer for pH to decrease due to CO2 uptake and CaCO3 precipitation than CH24 (at 25°C). Experiments CH2-CH10 consisted of a mixture of NaOH and KCl to maintain a high solution pH throughout. Initial solutions were 12 mM NaOH and 50 mM KCl, with slightly variable starting pH values from 11.9-12.8. Experiments CH1 and CH11-CH27 had 1 M NaOH added until reaching the pH of interest, with starting solutions ranging from 1-18 mM NaOH. The solution pH of experiments with the NaOH + KCl control mixture tended to record slightly slower decreases in pH over experiment duration. For example, CH3 and CH4 are characterized by a slower pH decrease than other 25°C, slowly stirred, 2000 ppm CO2 experiments that did not contain KCl, such as CH12. Stirring or not stirring an experiment affects the pH recorded with the probe, as well as the pH at the gas-solution interface. Not stirring an experiment causes the solution to become pH-stratified, with lower pH nearer to the air-water interface and higher pH deeper into the solution because both CO2 dissolution and CaCO3 precipitation reactions act to lower pH. We discovered this pH stratification in unstirred experiments when the pH as measured by the probe drops significantly when water samples are taken due to sufficient lowering of the water level so that the probe tip is now approaching the gas-solution interface. By not stirring, the pH-stratified solution has a lower surface pH, which may mean that there is a greater contribution from CO2 hydration reactions than would be otherwise expected based on the higher solution pH values measured by the probe. We experimented with variably stirring or not stirring a few of our early gas tank experiments, in addition to all small beaker experiments (0.13 L solution; CH14, CH15, CH18, CH19, CH21, CH23) being unstirred due to the necessary alterations of the experimental design not allowing for a magnetic stir bar. CH6 and CH7 were unstirred throughout the entire experiment, whereas CH5 was stirred only for the first half. Experiments at 25°C that were unstirred for their entire duration had a significantly lower pH boundary layer at the water surface in contrast to the higher pH that the probe was recording a couple centimeters below the water line. The unstirred 10°C experiment, however, had a surface pH much closer to that of where the probe was measuring (fully below the 98 air-water interface) when compared to the steep pH decreases recorded in 25°C solutions, which could be due to the observed slower CaCO3 growth rate in the 10°C solutions. In our high pH solutions, CO2−3 is the dominant DIC species, with variable amounts of HCO − 3 and always negligible CO2. At pH 11, ∼10 % of the DIC is HCO−3 , whereas the HCO − 3 portion decreases to ∼ 0.2% of the DIC pool at pH 12. CO2−3 was >99% of the total DIC pool at the start of all experiments except CH1, CH11, CH20, CH21, and CH23, and it makes up >98% of the total DIC pool for the entire experiment duration for experiments CH2-CH10, CH12-CH14 and CH26-CH27. Several experiments dropped into the pH range in which we predict CO2 hydration would be contributing to HCO−3 production. At 25°C and low ionic strength (I = 0.05), the contribution from CO2 hydration is negligible at or above a solution pH of 10.5 (Bajnai & Herwartz, 2021; Devriendt et al., 2017). 3.2 Isotopic results 3.2.1 CO2, DIC, and CaCO3 This study includes experimental travertines grown from three isotopically distinct CO2 gas sources (Figure 4 - pentagons). All 200 ppm CO2 experiments were performed using a single gas tank, as well as for the 2000ppm CO2 experiments, and their measured isotopic compositions are reported in Table 3. We assume the isotopic composition of the laboratory air to be the same as estimates of modern atmospheric CO2 for our open-air experiments. One study of the isotopic composition of atmospheric CO2 that spanned several years found that seasonal variability for both oxygen and car- bon isotopes stayed within a range of a few permil, from -8.5 to -7.5‰ VPDB-CO 132 for δ C and -2 to +1‰ VPDB-CO2 for δ18O (Trolier et al., 1996). Atmospheric CO2 monitoring from NOAA agrees with those values (Global Monitoring Laboratory Data Viewer: https://gml.noaa.gov/dv/iadv/). For this study, we approximate the isotopic composition of modern atmospheric CO2 and the lab- oratory air that contributed CO2 for our open-air experiments to be δ 13C = -8 ‰ VPDB-CO2 and δ18O = -0.5‰ VPDB-CO2. For carbon, there is no offset between the VPDB-CO2 and VPDB scales, but for oxygen there is an approximately 10‰ offset and requires conversion using the equation (Srivastava and Verkouteren, 2018; Swart et al., 1991): δ18O 18VPDB−CO2 = (δ OVPDB − 10.25)/1.01025 (7) so the δ18O of modern atmospheric CO2 is 9.75‰ VPDB. Despite the δ18O of all our experimental solutions being tightly clustered at -10.8 ± 0.5 ‰ VSMOW (-40.5 ‰ VPDB), the drastically different isotopic compositions of the CO2 sources results in each suite of CaCO3 recording very different isotopic compositions (Figure 4). This supports the notion that these carbonates reflect extreme kinetic processes, and are quite far from isotopic equilibrium with the experimental solution, since equilibrium CaCO3 would be expected to record δ18O ∼ -11.6 ‰ VPDB according to the Coplen (2007) δ18O-T calibration at 25°C. Overall, 99 15 Mervine et al. (2014) atmospheric CO2 10 Falk et al. (2016) 5 Christensen et al. (2021) 0 Clark et al. (1992) Böttcher et al. (2018) 5 200ppm CO2 2/3 CO2 + 1/3 H2O 10 CO2 distillation 15 2/3 CO2 + 1/3 OH- 2000ppm CO2 20 25 30 35 40 H2O 45 60 55 50 45 40 35 30 25 20 15 10 5 0 13CVPDB (‰) Figure 4: Isotopic compositions of precipitated carbonates and gas sources. The data are grouped by gas source, with the 200 ppm CO2 experiments in purple, 2000 ppm CO2 experiments in red, and open air experiments using the modern CO2 atmosphere in orange. Pentagons are the gas sources, circles are 25°C gas tank experiments, diamonds are 10°C gas tank experiments, upright triangles are 25°C open air experiments, inverted triangles are 10°C experiments, and squares are open air experiment crystals from the bottom of the beaker. The isotopic outliers are CH21 and CH23 (red circles), and CH22 beaker crystals (orange square). The hexagons represent where the data would trend towards if experiencing CO2 distillation. The gray triangles are the isotopic compositions of natural and experimental travertines, in order of increasingly dark shade of gray: Mervine et al. (2014), Falk et al. (2016), Christensen et al. (2021), Clark et al. (1992), and Böttcher et al. (2018), with the gray star representing the most highly fractionated travertine from The Cedars (Christensen et al., 2021). 100 18OVPDB (‰) Table 3: Isotopic data for all experiments Experiment δ13C δ18O δ13C CO (g) δ18 18 18CaCO CaCO 2 O CO3 3 2(g) δ Ow δ Ow (VPDB) (VPDB) (VPDB) (VPDB) (VSMOW) (VPDB) CH1 -42.24 -28.18 -26.40 -6.61 -11.29 -40.93 CH2 -32.44 -29.93 -26.40 -6.61 -11.28 -40.92 CH3 -44.75 -38.84 -36.16 -19.86 -11.19 -40.84 CH4 -46.23 -39.17 -36.16 -19.86 -11.00 -40.65 CH5 -34.07 -29.62 -26.40 -6.61 -10.48 -40.15 CH6 -49.78 -38.50 -36.16 -19.86 -10.74 -40.40 CH7 -50.47 -38.88 -36.16 -19.86 -10.62 -40.29 CH8 -50.44 -39.44 -36.16 -19.86 -10.52 -40.19 CH9 -37.61 -30.28 -26.40 -6.61 -10.65 -40.31 CH10 -37.50 -30.42 -26.40 -6.61 -10.80 -40.46 CH11 -52.55 -38.15 -36.16 -19.86 -11.26 -40.91 CH12 -49.41 -39.39 -36.16 -19.86 -11.19 -40.83 CH13 -26.21 -20.42 -8 9.75 -10.37 -40.04 CH13 b -25.40 -21.17 -8 9.75 -10.37 -40.04 CH14 -51.89 -39.27 -36.16 -19.86 -10.82 -40.48 CH15 -50.54 -38.56 -36.16 -19.86 -10.55 -40.22 CH17 -28.24 -21.82 -8 9.75 -10.54 -40.21 CH17 b -27.72 -22.21 -8 9.75 -10.54 -40.21 CH18 -49.72 -36.18 -36.16 -19.86 -10.95 -40.61 CH19 -49.82 -35.98 -36.16 -19.86 -10.91 -40.57 CH20 -27.99 -20.42 -8 9.75 -10.58 -40.24 CH20 b -27.40 -20.53 -8 9.75 -10.58 -40.24 CH21 -29.61 -13.32 -36.16 -19.86 -10.48 -40.15 CH22 1 -27.84 -20.40 -8 9.75 -10.90 -40.56 CH22 2 -28.47 -20.64 -8 9.75 -10.11 -39.79 CH22 b -18.74 -8.05 -8 9.75 -10.11 -39.79 CH23 -35.51 -30.82 -36.16 -19.86 -10.36 -40.03 CH24 1 -28.94 -20.80 -8 9.75 -10.66 -40.33 CH24 2 -28.72 -20.78 -8 9.75 -10.41 -40.08 CH24 3 -29.69 -19.92 -8 9.75 -9.73 -39.42 CH24 b -28.27 -20.75 -8 9.75 -9.73 -39.42 CH25 1 -26.20 -21.20 -8 9.75 -10.79 -40.45 CH25 2 -27.08 -21.75 -8 9.75 -10.80 -40.46 CH25 3 -28.19 -20.94 -8 9.75 -10.74 -40.41 CH25 b -26.85 -21.65 -8 9.75 -10.74 -40.41 CH26 -25.99 -21.38 -8 9.75 -10.88 -40.54 CH26 b -19.07 -21.63 -8 9.75 -10.88 -40.54 CH27 -25.94 -21.31 -8 9.75 -10.99 -40.65 CH27 b -23.69 -21.71 -8 9.75 -10.99 -40.65 there is a relatively consistent offset in δ18O between CO2(g) and CaCO3 for all gas compositions, whereas we observe up to 10‰ variability in δ13C of CaCO3 for a given δ13C of CO2 gas. We plotted the isotopic compositions of natural (Clark et al., 1992; Mervine et al., 2014; Falk et al., 2016; Christensen et al., 2021) and experimental (Böttcher et al., 2018) travertines from past studies on Figure 4 (gray triangles). The natural travertines record a wide array of co-varying δ13C and δ18O compositions. Our open-air experiments overlap with the most isotopically depleted natural travertines. This suggests that human exhalation in the laboratory space had minimal 101 impact on the isotopic composition of the overall CO2 reservoir from which our carbonates grew because the isotopic composition of human CO2 exhalation is distinctively isotopically lighter (δ 13C: -18.7 to -23.5‰ VPDB, δ18O: 2.8 to 5.9‰ VPDB; Epstein & Zeiri, 1988) than atmospheric CO2. We conclude that our laboratory-air CaCO3 precipitates may be directly compared to natural hyperalkaline spring travertines and also may reflect the kinetic fractionation factors attending CO2 hydroxylation, as both natural and laboratory travertines precipitated in this manner have been interpreted to reflect (Böttcher et al., 2018; Christensen et al., 2021; Clark et al., 1992). The hexagons on Figure 4, color coded by gas source, represent where the CaCO3 isotopic compositions would trend towards if experiencing gaseous CO2 distillation, which notably, our CaCO3 does not seem to be trending toward for any of the gas sources. For carbon, the trend would go to the δ13C of CO2(aq) in equilibrium with the CO2 gas source. For oxygen, the trend would go towards the sum of 2/3 CO2(aq) + 1/3 OH −. The CaCO3 isotopic compositions would be expected to trend towards the isotopic compositions of its constituent reservoirs because there is no net isotopic fractionation during quantitative conversion, which would occur in the most extreme case of CO2 distillation where all CO2(g) is converted to CO2(aq) and subsequently CaCO3. For these calculations, equilibrium fractionations of CO2(g)-CO2(aq) of -1.1‰ (Vogel et al., 1970) and +0.3‰ (Barkan and Luz, 2012; Beck et al., 2005) for carbon and oxygen, respectively, were used, as well as the OH-H2O equilibrium fractionation of -21.5‰ (Zeebe, 2020). In the majority of our experiments, the δ13C of the CaCO3 is isotopically lighter than that of the DIC in the solutions from which they precipitated. This is unexpected, because CaCO3 is typically isotopically heavier than the DIC it grew from (Sade et al., 2020). With the exception of isotopic outliers described in the following section (CH21, CH23), the 2000 ppm CO2 experiments consistently produce DIC δ13C isotopically heavier (from around 1 - 10+ ‰) than its CaCO3 (Figure 5B). The 200 ppm CO2 and open air experiments, however, seem to mostly straddle the 1:1 line where δ13C 13CaCO3 ≈ δ CDIC. The 25°C open air experiments result in CaCO3 that is slightly isotopically lighter than the DIC, while the 10°C open air experiments produce CaCO3 that is slightly isotopically heavier (Figure 5C). When most of the CO2−3 is converted to CaCO3, then the CaCO3 records the KFF and the remaining CO2−3 is pushed to lighter values, while conversely, when only a small fraction of the CO2−3 is converted to CaCO3 then the KFF is recorded in the EIC and the CaCO3 is heavier (Christensen et al., 2021). However, this offset is expected to be small based on the proposed KFF of +0.5‰ where CaCO3 is just slightly isotopically heavier than the CO2−3 it grew from (Sade et al., 2020). 3.2.2 Outliers There are several clear isotopic outliers (Figure 4), as well as experiments in which solution pH dropped low enough that there is reason to expect that processes other than unidirectional CO2 hydroxylation (e.g. CO2 hydration, DIC equilibration) may be contributing to the isotopic signature of the CaCO3 minerals. 102 20 a 1:1 line 25 200ppm CO2 CaCO3 lighter than DIC 30 CH5 35 CH1 CH2CH9 40 CH10 CaCO3 heavier than DIC 45 45 40 35 30 25 20 13CCaCO3 (‰)3 25 b 30 CaCO3 lighter than DIC CH7 CH3 CH23 35 CH21 initial pH CH19 CH4 2000ppm CO2 12.5 40 CH6 CH8 12.0CH12 45 CH18 11.5 CH14 CH11 50 CH15 CaCO3 heavier than DIC 11.0 55 60 60 55 50 45 40 35 30 25 13CCaCO3 (‰)3 5 c atmospheric CO2 10 CaCO3 lighter than DIC 15 CH22_2 CH22_b 20 CH24_2 CH20 CH22_1 CH13 25 CH17 CH27 CH24 CH26 30 CH25 CaCO3 heavier than DIC 35 35 30 25 20 15 10 5 13CCaCO3 (‰) Figure 5: The δ13C of CaCO and δ133 C of DIC in the solutions in which the CaCO3 precipitated. (A) Experiments on the 200 ppm CO2 tank (δ 13C = -26.40‰ VPDB), (B) experiments on the 2000 ppm CO2 gas tank (δ 13C = -36.16‰ VPDB), and (C) open air experiments using atmospheric CO2 (δ13C = -8‰ VPDB). 103 13C (‰) 13 13DIC CDIC (‰) CDIC (‰) CaCO3 from CH21 and CH23 (red circles), and CH22 and CH26 beaker crystals (orange squares) record unusual isotopic fractionations relative to our other experiments (Figure 4). CH23 is depleted from its reservoirs with respect to oxygen but very slightly enriched with respect to carbon. CH21 is the only experiment that records enrichments in both of the heavy isotopes relative to its reservoirs. Both CH21 and CH23 are unstirred, resulting in a pH-stratified solution with a lower pH surface. The final pH values near the surface are 9.25 and 10.1 respectively, which are low enough that there might be some contribution from CO2 hydration. However, CH18 and CH19 are also unstirred experiments that have a lower surface pH of 9.0 and 8.7, respectively, and precipitates from those experiments have similar isotopic compositions as the earlier 2000 ppm CO2 experiments, with a slight caveat that the oxygen isotope compositions of CH18 and CH19 are noticeably isotopically heavier by about 2‰ compared to earlier 2000 ppm CO2 experiments. CH18, CH19, CH21, and CH23 are the last four experiments conducted on the 2000ppm CO2 gas tank, which had in total 13 experiments conducted on it. This suggests that something regarding the gas tank source itself might be confounding interpretation of the CaCO3 isotopic compositions, such as that the isotopic composition of the tank might be changing substantially as it gets emptier, or at least that the gas delivered to the experiment headspace might be getting fractionated as the tank approaches empty. The “beaker crystals” are the tiny crystals that grew at the bottom of the experimental beakers, which we analyzed for each of our open air experiments. These crystals were adhered to the beaker, in contrast to sunken surface floes, which did not adhere to the beaker surface. CH22 beaker crystals (CH22 b) record a depletion in 13C that is not as extreme as similar experiments, and an enrichment in 18O relative to its reservoirs, possibly reflecting partial equilibration of the DIC. CH22 b plots on an array between the most extremely fractionated precipitates and the isotopic composition of the atmospheric CO2 gas source, among CaCO3 from natural hyperalkaline springs (Figure 4). CH26 beaker crystals record the same 13C depletion as CH22 b, but notably the δ18O matches that of the rest of the open air experiments. Some of the experiments with lower initial starting pH (closer to pH 11-11.5 - CH1, CH11) record smaller oxygen fractionations, which may reflect partial equilibration of the DIC, and/or small contributions from CO2 hydration rather than only CO2 hydroxylation. 3.2.3 CaCO3 precipitated via CO2 hydroxylation Since this study includes experimental travertines grown from three isotopically distinct CO2 gas sources (Figure 4 - pentagons), we therefore must compare calculated isotopic fractionations be- tween phases rather than the standalone isotopic values of the CaCO3 in order to make meaningful comparisons across all of our experiments. We effectively removed any existing DIC from our ex- perimental solutions prior to the start of each experiment by bubbling pure N2 gas for > 1.5 hours, so we may conclude that the carbon atoms in CaCO3 come from one source: the dissolved CO2. For oxygen, we calculate a weighted sum because dissolved CO2 contributes two oxygen atoms and OH− contributes one oxygen atom to CaCO3 that precipitates as a result of CO2 hydroxylation: 104 δ18 2 1 OCO2(aq)+OH− = · δ18OCO2(aq) + · δ18OOH− · (8)3 3 However, due to uncertainty in the OH−-H2O oxygen isotope fractionation factor (Bajnai & Her- wartz, 2021; Green & Taube, 1963; Zeebe, 2020), we also present our fractionations using H2O rather than OH−: δ18 2 1 O 18 18CO2(aq)+H2O = · δ OCO2(aq) + · δ OH2O, (9)3 3 since δ18O of our experimental H2O is a measured value. Quantifying kinetic isotope fractionations for oxygen from CO2(aq)+OH − or CO2(aq)+H2O both constitute a “bulk oxygen” fractionation because at this point we are unable to quantify the fractionations independently for oxygen on CO2(aq) and OH −. We present isotopic fractionations as epsilon () values (Table 4, Figure 6), where: A/B = (αA/B − 1) · 1000 (10) and the fractionation between phases A and B (α) is equal to: δ18OA + 1000 αA/B = δ18 · (11) OB + 1000 From the isotopic fractionations between CaCO3 and CO2(aq) for carbon and CO2(aq)+H2O for oxygen (Figure 6), we found that in experiments with CO2 from a gas tank: 1. The carbon and oxygen isotope fractionations are smaller than at The Cedars. 2. The carbon isotope variability is much larger than the oxygen isotope variability. 3. The slope of the δ13C-δ18O relationship is lower than what would be expected from isotopic distillation of CO2(g). 4. In open-air experiments, the δ13C of carbonates is lower than in gas tank experiments. Hence, the possibility of variable amounts of air getting into the apparatus cannot explain the vari- ability and trend towards heavier δ13C in gas tank experiments. These observations led us to postulate that the hydroxylation KFF on oxygen isotopes may be near unity whereas the KFF for carbon isotopes is large. If so, the carbon isotope composition of CO2(aq) would get isotopically distilled during conversion to HCO−3 , leading to heavy compositions in δ 13C with little to no effect on the δ18O of CO2(aq). Such a scenario would imply that CO2 exchange at the air-water interface is not replenishing the CO2(aq) fast enough to keep it at the equilibrium concentration or equilibrium isotopic composition relative to the CO2(g) in the headspace. In the model described below, we test this hypothesis and show that the δ13C-δ18O variations can be used to separate the two oxygen isotope KFFs acting on the reactants CO −2(aq) and OH . 105 5 CH19 10 CH1 CH18 CH11 CH15 CH6 CH5 CH2 CH9 CH14 CH7 CH4 CH20 CH22 CH13 CH3 CH24 CH10Cedars CH8 CH12 CH25 15 CH17 CH27 CH26_b CH26 initial pH 12.5 12.0 11.5 20 11.0 20 15 10 5 ε13CaCO3/CO2(aq) (‰) Figure 6: Carbon and bulk oxygen kinetic isotope fractionations (KIFs) for all experiments of this study, calculated as  = (α - 1) · 1000. Gas tank experiments (25°C - circles; 10°C - diamonds) on 200 ppm CO2 (black outline) or 2000 ppm CO2 (no outline). Open air experiments (25°C - upright triangles; 10°C - inverted triangles) with crystals from the bottom of those beakers (squares). The star represents the KFFs determined from The Cedars, California (Christensen et al., 2021). 106 ε18CaCO3/CO2(aq)+H2O (‰) Table 4: Carbon and bulk oxygen kinetic isotope fractionations 13C 18O 18O Experiment CaCO3−CO2aq CaCO3−(CO2aq+OH) CaCO3−(CO2aq+H2O) CH1 -15.16 -3.46 -10.52 CH2 -5.09 -5.26 -12.30 CH3 -7.78 -5.42 -12.53 CH4 -9.32 -5.83 -12.93 CH5 -6.76 -5.21 -12.25 CH6 -13.01 -5.22 -12.32 CH7 -13.72 -5.65 -12.75 CH8 -13.69 -6.26 -13.36 CH9 -10.40 -5.83 -12.87 CH10 -10.28 -5.93 -12.96 CH11 -15.88 -4.68 -11.79 CH12 -12.62 -5.99 -13.08 CH13 -17.27 -6.92 -13.87 CH13 b -16.45 -7.67 -14.62 CH14 -15.19 -5.99 -13.09 CH15 -13.79 -5.35 -12.45 CH17 -19.32 -8.28 -15.22 CH17 b -18.79 -8.67 -15.61 CH18 -12.94 -2.75 -9.87 CH19 -13.04 -2.55 -9.68 CH20 -19.07 -6.84 -13.79 CH20 b -18.47 -6.96 -13.91 CH21 7.94 20.75 13.46 CH22 1 -18.91 -6.72 -13.68 CH22 2 -19.55 -7.22 -14.17 CH22 b -9.73 5.54 -1.49 CH23 1.82 2.60 -4.56 CH24 1 -20.02 -7.21 -14.16 CH24 2 -19.80 -7.27 -14.21 CH24 3 -20.78 -6.62 -13.57 CH24 b -19.34 -7.45 -14.40 CH25 1 -17.26 -7.57 -14.52 CH25 2 -18.14 -8.13 -15.07 CH25 3 -19.26 -7.32 -14.26 CH25 b -17.92 -8.04 -14.99 CH26 -17.05 -7.73 -14.67 CH26 b -10.07 -7.98 -14.92 CH27 -16.99 -7.61 -14.56 CH27 b -14.73 -8.02 -14.96 4 Box model approach We are going to get an expression for the steady state concentration and isotopic composition of CO2(aq) that depends on (a) the KFFs attending hydroxylation and (b) the forward and backward exchange rate at the air-water interface (Fig. 7). Parameters for the model are listed in Table 5. 107 CO2(g) αf αb Ff Fb CO2(aq) FH αH EIC Figure 7: Schematic denoting the carbon reservoirs in our box model, and the fluxes and frac- tionations between them. The flux between gaseous and aqueous CO2 is bidirectional, whereas we assume no back-reaction between EIC and aqueous CO2. Table 5: Model parameters Parameter Value Reference/Note Ff ≥ FH CO2 dissolution flux: CO2(g) → CO2(aq) Fb Ff - FH CO2 degassing flux: CO2(g) ← CO2(aq) FH k+4[CO2][OH −] CO2 hydroxylation flux: CO − 2(aq)+OH → HCO−3 kp 2000 Isotope results not sensitive to this (13.635− 2895 ) k 10 T+4 K Pinsent et al. (1956); Uchikawa and Zeebe (2012) 18αeq 1.0003 CO2(g) is lighter than CO2(aq) 18αf 1.000 Results not sensitive to this 18α 18α /18b f αeq Equilibrium constraint 18αH variable Depends on the KFF assigned to OH − 13αeq 0.9989 CO2(g) is heavier than CO2(aq) 13αf 1.000 Results not sensitive to this 13α 13α /13b f αeq Equilibrium constraint 13αH 0.9814 Christensen et al. (2021) 4.1 Model assumptions 1. CO2(g) does not get isotopically distilled. Although there is evidence of CO2(g) depletion in the headspace of many experiments, the influence on the isotopic composition of resid- ual CO (g) should be negligible because the 13C/12C and 182 O/ 16O isotopic fractionations attending the CO2(g) → CO2(aq) reaction are small (Vogel et al., 1970; Zhang et al., 1995). 2. The dehydroxylation reaction is negligible, implying that HCO−3 is isotopically fractionated from CO2(aq) by the forward hydroxylation reaction only. This is convenient because it means we can calculate the isotopic composition of EIC directly from that of CO2(aq). 108 3. The CO2(aq) is a well-mixed reservoir. This is a necessary simplification that ignores the chemical stratification of the fluid near the air-water interface. 4.2 Concentrations At steady state, the influxes must balance the outfluxes: F︸ f −︷︷Fb︸ = FH, (12) net flux in where Ff is the CO2 dissolution flux, Fb is the CO2 degassing flux, and FH is the CO2 hydroxylation flux. The value of FH is known as a function of pH and [DIC]. The values of Ff and Fb are not known, but since Fb cannot be negative, Ff must be greater than or equal to FH. Hence, we treat Ff as an adjustable parameter from which the value of Fb can be calculated. The hydroxylation flux is given by: FH = k+4[CO2][OH −], (13) Following Olsen et al. (2022), the net influx can be written as: Ff − Fb = kp ([CO2]eq − [CO2]) , (14) where kp is akin to a ‘piston velocity’ that describes the efficiency of CO2 exchange across the air-water interface. From Eqs. 12, 13, and 14, the steady state CO2 concentration is given by: kp[CO2]eq [CO2]ss = . (15) k+4[OH−] + kp For isotopes, it doesn’t really matter what value we use for kp, but as you might expect, as kp increases, [CO2]ss → [CO2]eq. 4.3 Isotopes The formula for a changing oxygen isotope ratio (RCO2) with time is: dRCO dN18 dN16 N 216 = −RCO2 , (16)dt dt dt which comes from taking the time derivative of the expression for the ratio, R = (18O/16CO2 O)CO2 . The number of moles of 16O in CO2 (N16) changes with time according to the equation: dN16 = Ff − Fb − FH. (17) dt 109 The number of moles of 18O in CO2 (N18) changes with time according to the equation dN18 = RgαfFf −RCO dt 2 αbFb −RCO2αHFH, (18) where the fluxes refer to 16O but are multiplied by the isotope ratio to get the relevant 18O fluxes. Combining Eqs. 17 and 18 with 16 leads to: dRCO N 216 = Ff (Rgαf −RCO2)− Fb (RCO2αb −RCO2)− FH (RCO2αH −RCO2) . (19)dt At steady state, we have: Ff (Rgαf −RCO2) = Fb (RCO2αb −RCO2) + FH (RCO2αH −RCO2) , (20) which leads to an equation for RCO2 : FfRgαf RCO2 = . (21)Ff + Fb(αb − 1) + FH(αH − 1) The box model tracks how the kinetic isotope effects (KIEs) recorded in the EIC change as the rate of CO2 dissolution (F f) shifts relative to a fixed CO2 hydroxylation rate (FH) (Figure 8). At log10(F f/FH) = 0, F f = FH and all dissolved CO2 undergoes hydroxylation, no longer preserving the isotopic depletion from unidirectional CO2 hydroxylation in the EIC, and therefore the CaCO3. At log10(F f/FH) = 2, F f >>FH, meaning the rate of CO2 dissolution far outpaces the rate of CO2 hydroxylation so there is no isotopic distillation of CO2(aq), so the EIC and therefore CaCO3 are expected to record large isotopic depletions due to CO2 hydroxylation, up to the full KFFs. Distillation of CO2(aq) has a larger effect on carbon KIEs than oxygen (Fig. 8), which is consistent with the large range in 13EIC/CO and relatively small range in 18 2 EIC/CO observed in many of our2 small headspace experiments (Figure 6). Keeping with the assumption that there is negligible back-reaction, we consider the following hydroxylation reactions that are occurring simultaneously but at different reaction rates: 13 − −k−’+→4CO2(aq) + OH H13CO −3 (22) a+4 CO2(aq) + 18OH− −−→ HC18OO −2 (23) and C18 b+4 OO(aq) + OH− −−→ HC18OO −2 (24) in addition to Eq. 2 (k+4), which involves only light carbon and oxygen isotopes. The same KFF for 13C is used for all model curves (k’+4). We test out different KFF values for oxygen between HCO−3 -CO2 (b+4/k+4), with corresponding shifts to HCO − 3 -OH − (a+4/k+4) (Table 6) to maintain the overall bulk oxygen KFF, to see if we can recreate the slope of 18 13CaCO3/CO2+H2O/ CaCO3/CO2 observed in the experiments that experienced CO2(aq) distillation (Figure 9). The CO2 gas tank 110 0 CO2(aq) in equil. with CO2(g) −5 −10 −15 hydroxylation of CO2(aq) −20 0 1 2 log10(Ff/FH) −10 2/3 CO2(aq) + 1/3 OH- = -7.33 −11 −12 −13 −14 hydroxylation of CO2(aq) −15 0 1 2 log10(Ff/FH) Figure 8: Model outputs showing how the KIEs change with increasing Ff relative to a fixed hydroxylation flux FH. At log10(Ff/FH) = 0, Ff=FH so all dissolved CO2 undergoes hydroxylation. When Ff >> FH, there is no distillation of the CO2(aq) reservoir. experiments seem to fall along the model curve of b+4/k+4 = 0.996 and a+4/k+4 = 0.986, which corresponds to KFFs of -4‰ for HCO−3 -CO2 and -14‰ for HCO − 3 -OH −. 4.4 Comparison to previous work There is one study in the literature that is relevant to the present discussion. Clark et al. (1992) carried out an experiment where they connected five beakers in series to a CO2-N2 tank with known isotopic composition. Each beaker contained a BaCl2 solution that was brought to high pH (12.8) by addition of NaOH. Between each beaker there was a sampling port that enabled them to measure the CO2 partial pressure and its isotopic composition. Their results show that the CO2(g) gets 111 18ε 13EIC/CO ε2 EIC/CO2 Table 6: Model curves varying KFFs for CO2 hydroxylation with 18O on CO2 (b+4) and OH − (a+4) b+4/k+4 a+4/k+4 0.992 0.994 0.994 0.990 0.996 0.986 0.998 0.982 1 0.978 distilled to heavier isotopic compositions as it exchanges with CO2(aq) in the fluid (Fig. 10; Clark et al., 1992, Table 3). Since the CO2(g)-CO2(aq) isotope effects are relatively small, the results can be used to get an estimate of the hydroxylation KFFs (Fig. 10). Clark et al. (1992) modeled their data using a simple Rayleigh model equation: Rr = f (α−1), (25) Ro −10 0.992 ) istillation of CO2(aq d increasing 0.994 initial pH −12 CH5 0.996 12.5 CH9 CH14 CH12 CH2 CH3 0.998 12.0 CH4 11.5 1 −14 11.0 Cedars −16 −18 −16 −14 −12 −10 −8 −6 −4 ε13CaCO3/CO2(aq) (‰) Figure 9: Box model results of 18EIC/CO +H O/ 13 EIC/CO slopes using different KFF values for 18O on 2 2 2 CO2, and correspondingly on OH −, during CO2 hydroxylation with increasing degrees of CO2(aq) distillation. Numerical line labels correspond to HCO−3 -CO2 KFFs in Table 6. We plot our 25°C gas tank experiments during which solution pH always remained higher than pH = 10.5, and find the slope of the experiments suggests there is a -4‰ KFF for oxygen on HCO−3 -CO2 (black dashed line). 112 ε18CaCO3/CO2(aq)+H2O (‰) where Rr is the isotopic ratio of the residual CO2, Ro is the initial isotopic ratio of CO2, f is the fraction of CO2 remaining, and α is the net fractionation factor. The results, presented as a set of curves in Figure 10, suggest a carbon KFF of  ≈ -17h, which is consistent with what we obtained in our experiments. Applying the same model to the oxygen isotope results indicates  ≈ -2h, which is somewhat lower than the value we obtained above (Fig. 9), but supports the overall conclusion that the hydroxylation KFF on CO2(aq) is much smaller for oxygen isotopes than for carbon isotopes (Fig. 10). Clark et al. (1992) closed system experiment 20 T = 22 °CpH = 12.8 0 α = 0.980 α = 0.983 α = 0.986 −20 −40 −60 1.0 0.8 0.6 0.4 0.2 0.0 Fraction of CO2(g) remaining −5 −10 α = 0.997 α = 0.998 α = 0.999 −15 −20 1.0 0.8 0.6 0.4 0.2 0.0 Fraction of CO2(g) remaining Figure 10: Results from Clark et al. (1992) showing isotopic distillation of headspace CO2 during hydroxylation. The effects are much larger for δ13C than for δ18O, consistent with a small KIE on CO2 as it undergoes hydroxylation to form HCO − 3 . Results suggest a carbon KFF of about -17h and an oxygen KFF of about 2h. These are approximate because (a) there is large uncertainty in f for the rightmost data points, (b) the Rayleigh calculations assume a single fractionation factor even though there are multiple steps involved, and (c) it’s not really a closed system experiment because gas was continually flowing through five beakers connected in series. 113 δ18O (PDB) δ13C (PDB) 5 CO2 hydroxylation KFFs The largest isotopic fractionations from this study were observed in our open-air experiments, which appear to recreate the extreme isotopic fractionations recorded in natural travertines such as the Cedars. Our largest carbon and oxygen fractionations are not from the same experimental CaCO3. CH24 records the largest carbon fractionation of CaCO −CO (aq) = -20.8 ‰, while the largest oxygen3 2 fractionation is recorded in CH17 as CaCO −(CO (aq)+OH−) = -8.3 ‰, or 3 2 CaCO3−(CO2(aq)+H =2O) -15.2 ‰ (Table 4). However, these values are highly sensitive to our selection of the δ13C and δ18O of atmospheric CO2, which has been shown to have both regional and seasonal variability (Trolier et al., 1996), in addition to the complication that our laboratory air likely has a slightly different isotopic composition than atmospheric CO2 anyways. We add our open-air experiment data to the CO2 hydroxylation KFF vs. temperature literature compilation from Christensen et al. (2021, 2023), presented in Table 7 and Figure 11. The carbon KFFs do not show a clear overall temperature dependence, while the bulk oxygen KFFs do, with larger kinetic fractionations occurring at lower temperature (Fig. 11). Of note, the isotopic compositions of our experimental CaCO3 support 1000lnαOH−−H2O = -21.5‰ of Zeebe (2020) rather than the larger 1000lnαOH−−H O = -42.53‰ determined by Bajnai2 and Herwartz (2021), because if ∼ -42.53‰ is used, then the bulk oxygen isotopic fractionations (CaCO −(CO (aq)+OH−)) of the overwhelming majority of experiments would be positive rather than3 2 negative. 5.1 Complications and possible causes of variability Here we briefly list some confounding factors and possible causes of variability in this study: • Variable contribution of CO2 hydration in some experiments where the solution pH dropped Table 7: Literature compilation1 of KFFs during CO2 hydroxylation Study Temp 13C/12C KFF 18O/16O KFF 18O/16O KFF pH (°C) from CO2(aq) from CO2(aq)+OH− from CO2(aq)+H2O Usdowski and Hoefs 1986 18 -17 - - 10 Clark et al. 1992 22 -13.9 -8.5 -15 11.5 Clark et al., 1992 22 -16.3 -7.9 -14.5 12.8 Dietzel et al., 2009 5 - -10 -16.8 10.5 Böttcher et al., 2018 4 -11.8 -9.7 -16.6 12.4 Böttcher et al., 2018 21 -15.8 -8.4 -14.9 12.4 Clark et al., 1992 28 -17 -6.5 -13 11.5 Clark et al., 1992 28 -16.9 -7.2 -13.6 11.5 Mervine et al., 2014 28 -17 -6.4 -12.9 11.0 Falk et al., 2016 27 -17.2 -6.9 -13.4 11.7 Christensen et al., 2021 17 -17.1 -7.3 -14 11.0 This study 10 -17.74 -7.67 -14.62 12.26 This study 25 -19.34 -7.14 -14.08 12.0 1 Data from Christensen et al. (2021, 2023) Table 7. 114 10 a 12 14 16 18 20 22 0 5 10 15 20 25 30 5 b 6 pH 7 13 8 12 9 11 10 10 y = 0.1093x - 9.958 r2 = 0.6953 11 0 5 10 15 20 25 30 12 c 13 14 15 16 17 y = 0.1253x - 16.925 r2 = 0.7822 18 0 5 10 15 20 25 30 Temperature (°C) Figure 11: Carbon and bulk oxygen KFFs vs. temperature, plotting the average KFFs from our open-air experiments (10°C - inverted triangle, 25°C - upright triangle) with data from past studies (circles) (Table 7, from Christensen et al., 2021, 2023, Table 7). 18O/16O KFF is expressed from CO − 18 162(aq)+OH , while O/ O KFF* is expressed from CO2(aq)+H2O. The bulk oxygen KFF shows a significant temperature-dependence, while the carbon KFF does not. below pH 10.5, which is difficult to quantify because we do not know the KFFs for CO2 hydration for each reactant independently • It is unclear why the CO2(aq) is not being replenished efficiently during gas tank experiments but is being replenished fine during our open air experiments, despite many of the gas tank experiments being conducted at higher CO2 concentrations • The gas tank experiments could be affected by the humidity of the experimental headspace, 115 18O/16O KFF* (‰) 18O/16O KFF (‰) 13C/12C KFF (‰) as opposed to the open air experiments which are not enclosed • There might be isotopic fractionation occurring during transfer of the CO2-in-N2 gas from the gas tank to the experimental headspace, so that the measured gas tank value is not an accurate reflection of the actual CO2 reaching the experimental solution • The last few experiments conducted on the 2000 ppm CO2 gas tank resulted in abnormal CaCO3 isotopic values, which could be due to the isotopic composition of the gas tank evolving as it becomes progressively emptier, or that the gas tank is isotopically stratified • Our modeled curves on Figure 9 represent EIC not CaCO3 • In every experiment we were able to obtain DIC analyses, which suggests that the DIC is not quantitatively precipitated and there should be some offset between our modeled EIC curves and precipitated CaCO3 • We precipitated multiple CaCO3 polymorphs in our experiments (as occurs at natural hy- peralkaline springs), but having both calcite and aragonite present in variable proportions might obscure some of the quantitative results if the polymorphs have different fractionations attending their precipitation reactions • At the beginning of very high pH experiments (>12), we observed a surface sheen not present for solutions at lower starting pH, and hypothesize this might be an amorphous calcium carbonate (ACC) precursor, which may affect the isotopic composition of the analyzed CaCO3 in ways that are not fully understood 6 Summary In summary: 1. We performed open air and small headspace experiments that successfully reproduce the nat- ural process of high pH travertine formation, including mineralogy, morphology, and isotopic composition. 2. The largest isotopic fractionations occurred during our open air experiments, which mimic the most isotopically depleted natural travertines. We used these experiments to determine the carbon and bulk oxygen kinetic fractionation factors (KFFs) attending CO2 hydroxylation. 3. Smaller and more variable fractionations occurred in small headspace experiments. We used these experiments to separate, for the first time, the KFFs for CO2-HCO − 3 and OH −-HCO−3 based on a box model of CO2(aq) distillation during hydroxylation. 4. We compared our results to Clark et al. (1992) and found them to be broadly consistent. 116 5. We add our KFFs to those of past studies and present equations for the temperature depen- dence of the bulk oxygen KFFs. CO2 sequestration in the form of stable carbonate minerals is a promising avenue through which significant quantities of CO2 may be removed from the atmosphere. Studies of carbon capture and storage assess carbonation rates in-situ within lithologic units that have high carbonation potential (e.g. ultramafic to mafic ophiolite bodies, Kelemen and Matter, 2008) and ex-situ in industrial set- tings (Kelly et al., 2011; Kemache et al., 2016), and weight those against the costs, required energy, space, materials, and waste generated. Understanding and quantifying the isotopic evolution of CO2 gas injected for sequestration is being developed as a technique to assess the degree of success of the CO2 capture processes (Flude et al., 2017). In a similar vein, understanding and quantifying the isotopic composition of the product CaCO3 may be able to give a comparable evaluation since its isotopic composition reflects the CO2 hydroxylation KFFs at low uptake fractions, trending towards that of the initial injected CO2 at high levels of reservoir distillation. 117 CHAPTER VI SUMMARY This dissertation explores how non-equilibrium isotopic partitioning between calcium carbonate minerals and water is affected by environmental parameters through a combination of experiments and modeling. Equilibrium oxygen isotope fractionation between calcite and water exhibits a strong temperature dependence, and as a result is used in paleoclimate and paleoenvironment reconstruc- tions to estimate the water temperatures of caves, lakes, surface oceans, and hydrothermal systems in which the calcite grew. Commonly, however, isotopic equilibrium is not achieved and the car- bonate minerals record kinetic isotope effects (KIEs) that may depend on growth rate, source(s) of DIC, pH and solution composition. Non-equilibrium isotopic fractionation may be characterized by small departures from equilib- rium, or, in contrast, reflect the kinetic fractionation factors (KFFs). I investigate the effect of ionic strength and pH on isotopic partitioning in calcite under relatively near-isotopic equilibrium growth conditions characterized by reasonably small KIEs in Chapters III and IV, respectively. Experi- ments from Chapter IV and low ionic strength (up to [NaCl] = 0.35 M) experiments from Chapter III result in calcite grown from an isotopically equilibrated DIC pool, so that KIEs recorded in the CaCO3 may be attributed to processes occurring at the mineral surface during mass-dependent attachment and detachment of ions. We ensure isotopic equilibration of the DIC pool by utilizing sufficient concentrations of the enzyme carbonic anhydrase (bCA). Due to salt inhibition of bCA, calcite from experiments of Chapter III precipitated in solutions of higher ionic strength ([NaCl] ≥ 0.52 M) grew from a not fully isotopically equilibrated DIC pool, so resulting KIEs are due to a combination of DIC-H2O and DIC-CaCO3 disequilibrium. Experiments from Chapter V are char- acterized by a DIC pool that is far-from-isotopically equilibrated, as our experimental conditions promote unidirectional CO2 hydroxylation with negligible back-reaction, thus hindering isotopic exchange between DIC-H2O. In Chapter II, I provide background information regarding the chemical reactions and isotopic fractionations of the CaCO3-DIC-H2O system, factors affecting isotopic partitioning in carbonates, and an overview of several proxies based on trace element uptake and/or isotopic partitioning and their corresponding paleoenvironment applications. This chapter outlines core concepts of the stable isotope systematics of calcite which may prove beneficial to understanding the experimental and modeling studies of the subsequent chapters. In Chapter III, I investigated the effect of ionic strength on the isotopic composition of inorganic calcite, finding no significant effect on CaCO3-DIC fractionation, but a significant inhibition of the 118 enzyme carbonic anhydrase (bCA) by NaCl. This enzyme inhibition prevented isotopic equilibra- tion of the DIC pool above [NaCl] = 0.35 M, and as a result lower and more variable 1000lnαc/w. This finding is important because biocalcifiers use carbonic anhydrase during their calcification processes, and while the composition of the calcifying fluid is poorly constrained and likely species- dependent, the inhibition of bCA may be applied to biochemical models to help understand vital effects in biogenic calcite. In Chapter IV, I performed inorganic calcite experiments designed to isolate the effect of solution pH and subsequently HCO−3 /CO 2− 3 ratio of solution on the resulting calcite isotopic composition, finding a decrease in 1000lnαc/w with increasing pH consistent with past studies. Using my ex- perimental data, I then updated the ion-by-ion model of calcite growth, which is a generalizable model with a wide range of applications to natural and experimental carbonates. This study helps to further constrain DIC-CaCO3 isotopic fractionations, which in turn gives insight into separating vital effects from mass-dependent kinetic fractionations in biogenic CaCO3. In Chapter V, I investigate far-from-isotopic equilibrium carbonate growth proceeding via near unidirectional CO2 hydroxylation and characterized by large KIEs and probing the KFFs attending CO2 hydroxylation. These experiments simulate the direct air capture of CO2 via CO2 hydrox- ylation and subsequent travertine formation at hyperalkaline springs, which are characterized by large isotopic depletions in both carbon and oxygen. Our open-air experiments mimic the isotopic compositions of the most depleted natural travertines, likely reflecting the KFFs attending CO2 hydroxylation. The small headspace experiments resulted in overall smaller isotopic fractionations, as well as more variable carbon isotopic fractionations, leading us to model the effect of CO2(aq) distillation on the isotopic composition of the EIC (Equilibrated Inorganic Carbon). Modeling 13 18EIC/CO and EIC/CO +H O over increasing degrees of CO2(aq) distillation using different oxygen2 2 2 KFF values for HCO−3 -CO2, we found the slope of our experiments suggests oxygen KFFs of -4‰ for HCO−3 -CO2 and -14‰ for HCO − 3 -OH −. This is the first experimental study to report separate oxy- gen KFFs attending CO2 hydroxylation, which is a valuable contribution to the CaCO3-DIC-H2O stable isotope systematics and may aid in the development of isotopic paleoproxies for carbonates precipitated in this geologic setting. 119 APPENDIX A CHAPTER III SUPPLEMENTARY MATERIAL Supplementary material of Olsen, E.K., Watkins, J.M., and Devriendt, L.S. (2022). Oxygen isotopes of calcite precipitated at high ionic strength: CaCO3-DIC fractionation and carbonic anhydrase inhibition. Geochimica et Cosmochimica Acta 325, 170-186. 1 DIC speciation Although DIC speciation is primarily a function of pH, factors such as temperature, pressure, and solution composition are also important. DIC speciation is well known for seawater solu- tions (Millero et al., 2006), but our solutions (CaCl2 + NH4Cl + NaCl) are substantially simpler in composition. The stoichiometric equilibrium constants (pK) have been determined for similar NaCl solutions (Millero et al., 2007), and while this may provide an accurate description of DIC speciation in our system, the stoichiometric solubility product (Ksp) for simple NaCl solutions is unknown. Hence, even with an accurate description of solution speciation, it is not straightfor- ward to relate speciation to the degree of supersaturation with respect to calcite in non-seawater solutions. Consequently, we chose to use PHREEQC to model our solution speciation and degree of supersaturation in a self-consistent way using the Pitzer ionic activity database and our exact solution compositions as inputs (Charlton and Parkhurst, 2011; De Lucia and Kühn, 2013). Differences in speciation between seawater solutions and NaCl solutions are shown in Fig. S1.1 for three different [NaCl] (0, 0.52, 1.37 M). For both seawater and NaCl solutions, increasing salinity shifts pK1 and pK2 to lower pH so that at a fixed pH of 8.3 the [CO2−3 ] increases while [HCO−3 ] and [CO2] decrease. Increasing salinity has a greater effect on pK2 than on pK1 for both solution types. Overall DIC speciation differs between seawater and NaCl solutions, as increasing salinity in seawater dramatically shifts speciation so that CO2−3 becomes the dominant DIC species at approximately 1 M NaCl (salinity ∼75 g/kg) at a pH of 8.3 (Fig. S1.1b). Speciation shifts in NaCl solutions are more moderate, with CO2−3 increasing from 1.8% of total DIC in freshwater solutions to 6.7% of total DIC in highly saline solutions. It is important to note that the PHREEQC speciation for our solution closely matches the DIC speciation based on pure NaCl solutions (Fig S1.1e, Millero et al., 2007). 120 121 0 100 60 A B C 90 HCO3- 50 CO2 HCO3- CO32- 80 70 40 60 2 50 30 40 20 30 20 CO32- 10 10 CO2 4 0 0 5 6 7 8 9 10 11 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 pH [NaCl] (M) [NaCl] (M) 0 100 60 D F 90 E HCO3- 50 CO2 HCO3- CO32- 80 70 40 60 2 50 30 [NaCl] (M) 40 1.5 20 30 1.0 20 0.5 10 10 CO32- 0.0 CO2 4 0 0 5 6 7 8 9 10 11 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 pH [NaCl] (M) [NaCl] (M) Figure 1: Solution speciation in seawater solutions (a-c; Millero et al., 2006) versus NaCl solutions (d-f; solid lines: Millero et al., 2007; dashed lines: PHREEQC, R-package Pitzer database) at 25°C. [NaCl] 0, 0.52, and 1.37 M span the range of this study (a, d). Proportion of each DIC species and HCO−3 /CO 2− 3 for solutions at pH 8.3 (b-c, e-f). l o g c o n c e n t r a t i o n l o g c o n c e n t r a t i o n % o f t o t a l D I C % o f t o t a l D I C - 2 - - 2 - H C O / C O H C O / C O 3 3 3 3 60 50 40 equil. CO (aq) 30 2 20 Range of tank values in this study 10‰ 10 0 2‰ −10 water −20 −30 −40 −50 −16 −14 −12 −10 −8 −6 −4 −2 0 2 4 δ18Oweighted − δ18OeqEIC EIC Figure 2: Modeled sensitivity of the steady state EIC composition to the δ18O of CO2 gas. 2 Sensitivity of the kinetic isotope effects to the δ18O of CO2 gas To assess the sensitivity of 1000lnαc/w to the isotopic composition of gas used, we simulated the uncatalyzed case ([bCA] = 0) for different δ18O values of input CO2. The results shown in Fig. S2.1 reveal a simple linear relationship between the δ18O of CO2 and δ 18O of EIC at steady state (all other parameters fixed) and show that a large 10‰ shift in the δ18O of CO2 translates to a much smaller 2‰ shift in the δ18O of EIC (and calcite). This is attributed to the hydration reaction being bi-directional (Rb/Rf ∼ 0.8; Fig. 5g of Chapter II). The sensitivity of 1000lnα 18c/w to δ O of CO2 decreases as more bCA is added, because bCA pushes Rb/Rf closer to unity. 122 δ18O tank 3 Research data 123 Table 1: Additional parameters and isotopic data for experiments of this study Experiment Salinity Molality Ia Time (total) Time (precip) δ13CCaCO3 1000lnαCaCO3−DIC Ω a a peak Ωss (g/kg) (h) (h) (VPDB) (VPDB) S2 35 0.5542 0.633 51.9 28.4 -44.00 -0.16 5.59 4.20 S3 35 0.5542 0.633 44.3 27.9 -45.56 -0.42 8.46 4.02 S4 15 0.2134 0.274 62.4 42.9 -46.26 -0.97 5.86 2.38 S5 25 0.3843 0.452 41.6 23.3 -45.08 -0.47 7.78 2.52 S6b 3.5 - - 72.8 52.6 -46.06 - - - S7 15 0.2132 0.274 47.7 29.4 -45.29 -0.40 6.85 3.61 S8 3.5 0.0365 0.094 52.2 34.2 -45.37 -0.59 8.46 3.34 S9 25 0.3854 0.453 47.6 32.0 -45.22 -0.40 8.01 3.71 S10 45 0.7275 0.820 66.0 40.7 -45.28 -0.13 8.09 3.70 S11 25 0.3852 0.452 66.9 34.8 -44.32 0.31 10.89 3.68 S12 85 1.4121 1.603 45.9 27.7 -24.00 0.61 7.52 4.09 S13 65 1.0692 1.203 93.2 64.2 -25.13 -0.02 5.14 2.22 S14 55 0.8992 1.010 52.4 36.6 -24.25 -0.33 7.24 2.99 S15 75 1.2414 1.401 61.3 37.5 -23.35 0.41 8.66 5.34 CA1 35 0.5566 0.634 65.2 37.6 -23.78 0.06 8.34 3.48 CA2 35 0.5565 0.634 48.5 33.9 -23.79 -0.05 7.76 3.93 CA3 35 0.5566 0.634 47.5 34.5 -24.29 -0.19 5.66 2.73 CA4 3.5 0.0368 0.094 30.6 21.5 -25.44 -0.54 7.19 3.77 CA5 35 0.5565 0.634 62.2 37.5 -22.86 0.62 8.74 2.79 CA6 35 0.5570 0.634 61.2 40.5 -22.38 0.55 8.12 3.55 CA7 20 0.2999 0.363 61.5 37.1 -21.59 0.61 10.65 3.89 CA9 60 0.9847 1.106 60.3 43.3 -43.55 0.41 8.55 3.62 CA12 45 0.7280 0.820 66.9 51.0 -43.52 -0.38 7.49 2.46 CA13 65 1.0697 1.203 66.0 41.0 -43.35 0.09 7.06 2.25 CA14 15 0.2143 0.274 50.2 36.1 -43.24 -0.44 5.23 2.54 CA15 25 0.3856 0.453 52.0 37.5 -42.59 0.05 6.33 2.64 CA18 3.5 0.0387 0.095 42.7 35.5 -17.67 -0.30 9.03 5.11 CA20 3.5 0.0376 0.094 46.8 40.3 -34.77 -1.27 7.80 3.82 a Ionic strength (I), Ωpeak, and Ωss were calculated using the PHREEQC R-package Pitzer database. Ωpeak represents the highest degree of supersaturation during an experiments, which occurs at the onset of calcite precipitation. Ωss represents the average degree of supersaturation for an experiment during the “steady state” calcite growth period. b S6 lacks DIC measurements Table 2: Experiment measurements Experiment Time TA [DIC] δ13C 18DIC δ Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) S2 0.00 - 0.07 -35.52 - 206 - 7.40 0.65 0.36 -43.65 - 141 - 22.65 0.92 0.45 -43.49 - 192 - 26.40 0.73 0.32 -43.16 -11.65 170 - 31.97 0.68 0.38 -43.61 - 148 - 44.98 0.65 0.32 -44.52 - 140 - 51.45 0.65 0.33 -44.64 - 142 S3 3.05 0.54 0.24 -44.45 - 70 - 16.65 0.94 0.68 -44.93 - 129 - 21.78 0.73 0.48 -44.78 - 102 - 26.78 0.70 0.40 -45.13 - 97 - 40.47 0.62 0.27 -45.86 - 92 - 43.97 0.69 0.31 -45.81 -11.52 93 S4 4.18 - 0.10 -44.77 - 60 - 12.33 - 0.22 -45.37 - 76 - 18.35 0.62 0.36 -45.73 - 91 - 20.50 - 0.34 -45.37 - 90 - 24.28 0.63 0.16 -45.02 - 70 - 28.60 0.61 0.13 -44.90 - 66 - 39.02 0.63 0.15 -44.54 - 67 - 44.02 0.63 0.15 -46.07 - 68 - 50.77 0.64 0.13 -45.85 - 65 - 62.33 0.80 0.15 -45.80 -11.42 60 S5 4.38 0.53 0.24 -43.30 - 79 - 15.55 0.80 0.51 -44.66 - 131 - 18.98 0.87 0.57 -44.81 - 141 - 25.05 0.69 0.31 -44.04 - 165 - 30.97 0.48 0.23 -45.28 - 194 - 39.35 0.52 0.15 -45.50 - 210 - 41.42 0.54 0.17 -44.81 -11.32 209 S6 3.85 0.46 - - - - - 18.82 0.69 - - - - - 26.83 0.61 - - - - - 41.45 0.51 - - - - - 45.78 0.53 - - - - - 52.75 0.52 - - - - 124 Table 2 continued Experiment Time TA [DIC] δ13C 18DIC δ Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 63.75 0.50 - - - - - 66.63 0.50 - - - - - 72.78 0.50 - - -11.31 - S7 3.83 0.48 0.22 -44.33 - 67 - 13.92 0.71 0.37 -45.16 - 105 - 20.65 0.71 0.42 -44.65 - 117 - 27.57 0.55 0.28 -44.80 - 83 - 38.83 0.55 0.21 -44.47 - 82 - 45.00 0.55 0.24 -45.50 - 89 - 48.00 0.52 0.15 -45.47 -11.34 184 S8 3.65 0.47 0.14 -43.66 - 97 - 11.80 0.65 0.37 -44.47 - 142 - 19.07 0.66 0.26 -44.18 - 143 - 26.07 0.49 0.15 -44.60 - 103 - 34.97 0.50 0.17 -45.47 - 104 - 47.22 0.49 0.13 -45.70 - 102 - 52.15 0.60 0.13 -45.55 -11.39 91 S9 3.22 0.55 0.17 -44.63 - 72 - 14.05 0.85 0.48 -44.68 - 172 - 16.22 0.85 0.59 -44.25 - 179 - 20.40 0.63 0.30 -44.18 - 123 - 25.95 0.62 0.25 -45.22 - 118 - 37.83 0.64 0.25 -45.47 - 124 - 47.38 0.60 0.29 -45.48 -11.40 118 S10 3.08 0.35 - - - 78 - 11.67 0.63 - - - 142 - 16.90 0.84 - - - 171 - 24.42 0.97 - - - 198 - 26.50 0.99 0.69 -44.22 - 192 - 34.65 0.63 0.29 -44.45 - 136 - 39.30 0.60 0.27 -45.24 - 129 - 46.10 0.60 0.36 -45.53 - 128 - 59.03 0.59 0.34 -45.76 - 136 - 66.13 0.61 0.31 -45.74 -11.39 132 S11 4.83 0.61 - - - 107 - 14.50 0.90 0.59 -44.36 - 189 125 Table 2 continued Experiment Time TA [DIC] δ13C δ18DIC Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 22.10 1.08 0.61 -43.72 - 218 - 28.07 1.15 0.80 -43.03 - 235 - 38.32 0.68 0.34 -43.34 - 139 - 44.27 0.65 0.22 -45.10 - 130 - 50.52 0.63 0.27 -45.51 - 130 - 60.68 0.66 0.25 -45.85 - 131 - 66.98 0.63 0.28 -46.02 -11.62 127 S12 4.55 0.57 - - - 131 - 14.03 0.85 0.56 -24.53 - 171 - 20.20 0.92 0.62 -23.87 - 176 - 28.00 0.64 0.34 -24.38 - 127 - 37.50 0.61 0.35 -24.94 - 127 - 45.50 0.62 0.31 -25.23 -11.57 124 S13 4.05 0.46 0.14 -24.15 - 194 - 12.20 0.63 0.27 -24.89 - 221 - 20.13 0.80 0.37 -24.76 - 226 - 28.58 0.88 0.44 -24.53 - 237 - 37.10 0.61 0.22 -24.18 - 219 - 44.32 0.58 0.20 -24.71 - 216 - 51.98 0.58 0.19 -25.08 - 217 - 60.97 0.57 0.20 -25.35 - 210 - 68.32 0.59 0.21 -25.74 - 208 - 76.02 0.58 0.16 -25.83 - 209 - 86.78 0.58 0.18 -26.05 - 204 - 92.95 0.57 0.20 -26.00 -11.72 206 S14 3.52 0.61 0.21 -23.69 - 89 - 10.62 0.90 0.63 -23.82 - 138 - 21.80 0.67 0.25 -23.74 - 98 - 26.72 0.61 0.25 -24.53 - 94 - 34.52 0.64 0.25 -24.39 - 96 - 45.13 0.62 0.22 -24.83 - 95 - 52.00 0.63 0.32 -22.45 -11.82 98 S15 3.37 0.59 0.19 -24.27 - 85 - 13.27 0.95 0.60 -24.32 - 142 - 19.47 1.24 0.71 -23.73 - 172 - 26.03 1.26 0.73 -23.00 - 178 126 Table 2 continued Experiment Time TA [DIC] δ13C δ18DIC Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 36.60 0.96 0.50 -22.96 - 136 - 44.47 0.82 0.55 -23.34 - 122 - 50.45 0.74 0.43 -23.84 - 111 - 61.12 0.67 0.34 -24.53 -11.82 97 CA1 5.15 0.43 0.10 -23.32 - 54 - 14.25 0.56 0.31 -23.96 - 83 - 21.65 0.90 0.67 -23.73 - 148 - 31.08 0.89 0.50 -22.91 - 149 - 39.38 0.68 0.26 -23.68 - 110 - 47.20 0.65 0.33 -24.26 - 108 - 55.07 0.65 0.30 -24.36 - 109 - 64.80 0.66 0.24 -24.49 -11.83 106 CA2 7.63 0.72 0.42 -23.58 - 127 - 14.78 0.95 0.63 -22.89 - 162 - 22.12 0.67 0.34 -23.30 - 105 - 29.22 0.65 0.33 -23.96 - 101 - 38.82 0.64 0.35 -24.20 - 93 - 48.33 0.63 0.24 -24.49 -11.86 97 CA3 2.40 0.48 0.17 -24.53 - 76 - 11.70 0.95 0.46 -23.61 - 146 - 23.03 0.72 0.24 -23.30 - 103 - 26.47 0.65 0.25 -24.24 - 102 - 35.83 0.64 0.21 -24.38 - 99 - 47.05 0.65 0.18 -24.54 -11.91 96 CA4 3.22 0.74 0.14 -24.50 - 134 - 8.38 0.95 0.31 -24.88 - 165 - 21.25 0.70 0.14 -25.12 - 121 - 25.38 0.72 0.19 -25.04 - 138 - 30.22 0.71 0.16 -25.05 -11.68 124 CA5 4.07 0.69 0.24 -23.50 - 94 - 13.15 - 0.48 -23.32 - 159 - 21.82 1.23 0.64 -22.53 - 195 - 24.68 1.29 0.71 -22.26 - 200 - 37.20 0.73 0.25 -23.27 - 131 - 44.22 0.70 0.23 -24.03 - 132 - 51.25 0.69 0.22 -24.25 - 130 127 Table 2 continued Experiment Time TA [DIC] δ13C δ18DIC Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 61.87 0.67 0.20 -24.58 -11.17 128 CA6 4.00 0.60 0.27 -23.19 - 85 - 13.68 - 0.54 -22.54 - 169 - 18.03 1.09 0.66 -21.94 - 191 - 23.15 0.99 0.56 -21.49 - 167 - 29.03 0.79 0.33 -22.30 - 118 - 39.12 0.78 0.29 -23.36 - 107 - 45.95 0.77 0.28 -23.63 - 103 - 51.73 0.73 0.27 -23.75 - 105 - 60.92 0.74 0.27 -24.00 -11.18 102 CA7 3.17 0.58 0.23 -23.19 - 81 - 13.12 0.92 0.55 -22.00 - 173 - 19.50 1.04 0.73 -21.09 - 203 - 26.85 1.02 0.56 -20.41 - 191 - 37.43 0.68 0.27 -22.09 - 113 - 43.67 0.64 0.27 -22.66 - 112 - 50.35 0.65 0.29 -22.87 - 109 - 61.13 0.65 0.23 -23.20 -11.33 104 CA9 3.52 0.55 0.22 -44.09 - 34 - 12.45 0.91 0.57 -43.99 - 126 - 16.98 1.04 0.74 -43.55 - 154 - 25.58 0.68 0.29 -43.29 - 97 - 35.37 0.67 0.29 -44.15 - 95 - 42.58 0.68 0.31 -44.13 - 94 - 49.20 0.68 0.40 -44.03 - 92 - 59.87 0.65 0.28 -44.27 -11.57 90 CA12 3.83 0.63 0.32 -42.02 - 102 - 16.93 0.96 0.64 -42.92 - 184 - 22.37 0.97 0.63 -42.54 - 168 - 29.78 0.79 0.18 -42.25 - 128 - 40.15 0.63 0.24 -43.45 - 96 - 47.15 0.60 0.21 -43.67 - 87 - 53.40 0.60 0.21 -43.93 - 84 - 66.82 0.63 0.21 -44.46 -11.76 83 CA13 2.57 0.63 0.34 -43.70 - 67 - 14.40 0.93 0.54 -43.22 - 171 128 Table 2 continued Experiment Time TA [DIC] δ13C 18DIC δ Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 20.68 1.05 0.61 -42.61 - 194 - 26.35 1.00 0.51 -42.13 - 184 - 37.92 0.61 0.22 -43.38 - 103 - 44.25 0.63 0.20 -43.82 - 103 - 50.88 0.62 0.16 -43.87 - 102 - 60.17 0.60 0.18 -43.89 - 105 - 65.75 0.60 0.22 -44.36 -11.65 106 CA14 3.67 0.59 0.17 -42.28 - 87 - 16.05 0.84 0.32 -41.79 - 155 - 23.38 0.63 0.17 -42.22 - 96 - 30.77 0.59 0.15 -43.02 - 94 - 38.28 0.60 0.16 -43.29 - 98 - 44.03 0.62 0.16 -43.66 - 100 - 49.97 0.64 0.15 -43.43 -11.85 97 CA15 4.40 0.63 0.19 -42.63 - 98 - 14.92 0.95 0.46 -42.75 - 179 - 21.23 1.00 0.46 -41.89 - 169 - 29.10 0.79 0.24 -41.90 - 112 - 38.80 0.75 0.18 -42.91 - 98 - 45.10 0.72 0.18 -43.19 - 95 - 51.80 0.68 0.18 -43.19 -11.86 88 CA18 3.20 0.77 0.39 -17.44 - 220 - 13.93 0.71 0.31 -16.77 - 214 - 19.78 0.64 0.24 -17.40 - 192 - 26.93 0.64 0.22 -17.53 - 186 - 36.45 0.65 0.20 -17.66 - 188 - 42.28 0.64 0.22 -17.45 -11.88 191 CA20 6.22 0.74 0.34 -35.65 - 161 - 15.95 0.63 0.23 -35.59 - 127 - 22.50 0.63 0.22 -34.74 - 124 - 29.92 0.61 0.19 -32.34 - 115 - 39.62 0.60 0.16 -31.50 - 102 - 46.50 0.59 0.15 -31.45 -11.95 99 129 APPENDIX B CHAPTER III SUPPLEMENTARY MATERIAL: EXPERIMENT DATA AND SEM IMAGES OF PRECIPITATES 1 Experiment data For each experiment, we periodically took water samples that were sent for [DIC] and δ13C analysis, which are depicted by black circles, with curves interpolated between the data measurements (dark green and light green, respectively). We continuously monitored the concentration of CO2 (ppm) of the experimental headspace (light blue line). The CO2 concentrations of early experiments (S2- S8) were measured with a less reliable software and are not reported. We continuously measured solution pH (purple line) and amount of added NaOH (dark blue line), in addition to the periodic water samples we titrated to calculate the total alkalinity of the solution (red squares). The vertical gray line represents the onset of calcite precipitation for each experiment. 130 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −42 −44 −46 0 10 20 30 40 50 4 10.0 9.5 3 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 1: S2, [NaCl] = 0.52 M, salinity = 35 g/kg 131 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 −44 −46 0 10 20 30 40 4 10.0 9.5 3 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 time (hours) Figure 2: S3, [NaCl] = 0.52 M, salinity = 35 g/kg 132 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.2 0.0 0 10 20 30 40 50 60 −44 −46 0 10 20 30 40 50 60 4 10.0 9.5 3 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 time (hours) Figure 3: S4, [NaCl] = 0.18 M, salinity = 15 g/kg 133 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.6 0.4 0.2 0.0 0 10 20 30 40 −44 −46 0 10 20 30 40 10.0 9.5 2 9.0 8.5 1 8.0 7.5 0 7.0 0 10 20 30 40 time (hours) Figure 4: S5, [NaCl] = 0.35 M, salinity = 25 g/kg 134 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 10.0 3 9.5 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 5: S6, [NaCl] = ∼0 M, salinity = 3.5 g/kg 135 Alkalinity (mEq/L) pH 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −44 −46 0 10 20 30 40 50 3 10.0 9.5 2 9.0 8.5 1 8.0 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 6: S7, [NaCl] = 0.18 M, salinity = 15 g/kg 136 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.2 0.0 0 10 20 30 40 50 −44 −46 0 10 20 30 40 50 3 10.0 9.5 2 9.0 8.5 1 8.0 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 7: S8, [NaCl] = ∼0 M, salinity = 3.5 g/kg 137 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −44 −46 0 10 20 30 40 50 300 200 100 0 0 10 20 30 40 50 4 10.0 9.5 3 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 8: S9, [NaCl] = 0.35 M, salinity = 25 g/kg 138 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −44 −46 0 10 20 30 40 50 60 70 300 200 100 0 0 10 20 30 40 50 60 70 10.0 4 9.5 3 9.0 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 9: S10, [NaCl] = 0.69 M, salinity = 45 g/kg 139 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −42 −44 −46 −48 0 10 20 30 40 50 60 70 300 200 100 0 0 10 20 30 40 50 60 70 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 10: S11, [NaCl] = 0.35 M, salinity = 25 g/kg 140 pCO 132 (ppm) δ C (VPDB) [DIC] (mM)Alkalinity (mEq/L) pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −22 −24 −26 0 10 20 30 40 50 300 200 100 0 0 10 20 30 40 50 4 10.0 9.5 3 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 11: S12, [NaCl] = 1.37 M, salinity = 85 g/kg 141 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 80 90 100 −24 −26 0 10 20 30 40 50 60 70 80 90 100 300 200 100 0 0 10 20 30 40 50 60 70 80 90 100 4 10.0 9.5 3 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 80 90 100 time (hours) Figure 12: S13, [NaCl] = 1.03 M, salinity = 65 g/kg 142 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −20 −22 −24 −26 0 10 20 30 40 50 300 200 100 0 0 10 20 30 40 50 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 13: S14, [NaCl] = 0.86 M, salinity = 55 g/kg 143 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 −22 −24 −26 0 10 20 30 40 50 60 300 200 100 0 0 10 20 30 40 50 60 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 time (hours) Figure 14: S15, [NaCl] = 1.20 M, salinity = 75 g/kg 144 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −22 −24 −26 0 10 20 30 40 50 60 70 300 200 100 0 0 10 20 30 40 50 60 70 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 15: CA1, [NaCl] = 0.52 M, salinity = 35 g/kg 145 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −22 −24 −26 0 10 20 30 40 50 300 200 100 0 0 10 20 30 40 50 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 16: CA2, [NaCl] = 0.52 M, salinity = 35 g/kg 146 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −22 −24 −26 0 10 20 30 40 50 300 200 100 0 0 10 20 30 40 50 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 17: CA3, [NaCl] = 0.52 M, salinity = 35 g/kg 147 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.4 0.2 0.0 0 10 20 30 −24 −26 0 10 20 30 300 200 100 0 0 10 20 30 10.0 3 9.5 9.0 2 8.5 8.0 1 7.5 0 7.0 0 10 20 30 time (hours) Figure 18: CA4, [NaCl] = ∼0 M, salinity = 3.5 g/kg 148 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 −22 −24 0 10 20 30 40 50 60 300 200 100 0 0 10 20 30 40 50 60 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 time (hours) Figure 19: CA5, [NaCl] = 0.52 M, salinity = 35 g/kg 149 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 −22 −24 0 10 20 30 40 50 60 300 200 100 0 0 10 20 30 40 50 60 6 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 time (hours) Figure 20: CA6, [NaCl] = 0.52 M, salinity = 35 g/kg 150 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 −20 −22 −24 0 10 20 30 40 50 60 300 200 100 0 0 10 20 30 40 50 60 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 time (hours) Figure 21: CA7, [NaCl] = 0.26 M, salinity = 20 g/kg 151 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 −42 −44 −46 0 10 20 30 40 50 60 200 100 0 0 10 20 30 40 50 60 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 time (hours) Figure 22: CA9, [NaCl] = 0.95 M, salinity = 60 g/kg 152 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −40 −42 −44 −46 0 10 20 30 40 50 60 70 200 100 0 0 10 20 30 40 50 60 70 6 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 23: CA12, [NaCl] = 0.69 M, salinity = 45 g/kg 153 pCO (ppm) δ132 C (VPDB) [DIC] (mM)Alkalinity (mEq/L) pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −40 −42 −44 −46 0 10 20 30 40 50 60 70 200 100 0 0 10 20 30 40 50 60 70 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 24: CA13, [NaCl] = 1.03 M, salinity = 65 g/kg 154 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.2 0.0 0 10 20 30 40 50 −42 −44 0 10 20 30 40 50 200 100 0 0 10 20 30 40 50 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 25: CA14, [NaCl] = 0.18 M, salinity = 15 g/kg 155 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −42 −44 0 10 20 30 40 50 200 100 0 0 10 20 30 40 50 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 26: CA15, [NaCl] = 0.35 M, salinity = 25 g/kg 156 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.6 0.4 0.2 0.0 0 10 20 30 40 −16 −18 0 10 20 30 40 400 300 200 100 0 0 10 20 30 40 10.0 7 9.5 6 5 9.0 4 8.5 3 8.0 2 7.5 1 0 7.0 0 10 20 30 40 time (hours) Figure 27: CA18, [NaCl] = ∼0 M, salinity = 3.5 g/kg 157 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 0.4 0.2 0.0 0 10 20 30 40 50 −30 −32 −34 −36 −38 0 10 20 30 40 50 200 100 0 0 10 20 30 40 50 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 28: CA20, [NaCl] = ∼0 M, salinity = 3.5 g/kg 158 pCO (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) 2 pH 2 SEM images Figure 29: S2, [NaCl] = 0.52 M, salinity = 35 g/kg 159 Figure 30: S3, [NaCl] = 0.52 M, salinity = 35 g/kg 160 Figure 31: S4, [NaCl] = 0.18 M, salinity = 15 g/kg 161 Figure 32: S5, [NaCl] = 0.35 M, salinity = 25 g/kg 162 Figure 33: S6, [NaCl] = ∼0 M, salinity = 3.5 g/kg 163 Figure 34: S7, [NaCl] = 0.18 M, salinity = 15 g/kg 164 Figure 35: S8, [NaCl] = ∼0 M, salinity = 3.5 g/kg 165 Figure 36: S9, [NaCl] = 0.35 M, salinity = 25 g/kg 166 Figure 37: S10, [NaCl] = 0.69 M, salinity = 45 g/kg 167 Figure 38: S11, [NaCl] = 0.35 M, salinity = 25 g/kg 168 Figure 39: S12, [NaCl] = 1.37 M, salinity = 85 g/kg 169 Figure 40: S13, [NaCl] = 1.03 M, salinity = 65 g/kg 170 Figure 41: S14, [NaCl] = 0.86 M, salinity = 55 g/kg 171 Figure 42: S15, [NaCl] = 1.20 M, salinity = 75 g/kg 172 Figure 43: CA1, [NaCl] = 0.52 M, salinity = 35 g/kg 173 Figure 44: CA2, [NaCl] = 0.52 M, salinity = 35 g/kg 174 Figure 45: CA4, [NaCl] = ∼0 M, salinity = 3.5 g/kg 175 Figure 46: CA5, [NaCl] = 0.52 M, salinity = 35 g/kg 176 Figure 47: CA6, [NaCl] = 0.52 M, salinity = 35 g/kg 177 Figure 48: CA7, [NaCl] = 0.26 M, salinity = 20 g/kg 178 Figure 49: CA9, [NaCl] = 0.95 M, salinity = 60 g/kg 179 Figure 50: CA12, [NaCl] = 0.69 M, salinity = 45 g/kg 180 Figure 51: CA13, [NaCl] = 1.03 M, salinity = 65 g/kg 181 Figure 52: CA15, [NaCl] = 0.35 M, salinity = 25 g/kg 182 Figure 53: CA18, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 1st disc removed 183 Figure 54: CA18, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 2nd disc removed 184 Figure 55: CA18, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 3rd disc removed 185 Figure 56: CA18, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - disc in solution for the full experiment 186 Figure 57: CA20, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 1st disc removed 187 Figure 58: CA20, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 2nd disc removed 188 Figure 59: CA20, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 3rd disc removed 189 Figure 60: CA20, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - 4th disc removed 190 Figure 61: CA20, [NaCl] = ∼0 M, salinity = 3.5 g/kg. Time series experiment - disc in solution for the full experiment 191 3 Additional figures 15 20 25 30 35 [NaCl] (M) 1.5 40 1.0 0.5 45 0.0 50 0 10 20 30 40 50 60 70 80 90 100 time (hours) Figure 62: δ13C of DIC (‰ VPDB) for all experiments. 192 13CDIC 28.5 28.0 27.5 bCA (µM) 27.0 1.5 26.5 1.0 0.5 26.0 0.0 25.5 25.0 24.5 0 5 10 15 20 25 30 35 40 HCO-3/CO2-3 Figure 63: Oxygen isotope fractionation between calcite and experimental solution expressed as 1000lnαc−w, over the range of experimental solution HCO − 3 /CO 2− 3 . 193 1000ln c-w APPENDIX C CHAPTER III SUPPLEMENTARY MATERIAL: CARBONIC ANHYDRASE ENZYME ASSAYS 1 Carbonic anhydrase enzyme Carbonic anhydrase is an enzyme that catalyzes CO2 (de)hydration reactions, which are some of the slowest reactions in the DIC-H2O system, in addition to being the only pathway (other than hydroxylation) through which DIC may exchange with H2O (Zeebe and Wolf-Gladrow, 2001). At the pH of this study (pH 8.3), CO2 (de)hydration reactions are several orders of magnitude slower than HCO−3 -CO 2− 3 exchange, which may be assumed to achieve equilibrium instantaneously (Sade and Halevy, 2017). Catalyzing CO2 (de)hydration therefore helps in establishing an isotopically equilibrated DIC pool, which is important when trying to separate kinetic isotope effects (KIEs) due to disequilibrium of CaCO3-DIC during attachment and detachment of ions at the mineral surface, from KIEs due to disequilibrium of DIC-H2O as a result of precipitating calcite from a not fully equilibrated DIC pool. This enzyme has been used in a few prior studies of inorganic calcite precipitation (Uchikawa and Zeebe, 2012; Watkins et al., 2013, 2014; Baker, 2015), all of which were performed in solutions of low ionic strength. Since we use carbonic anhydrase in solutions of varying ionic strength, it was important at the outset of our study to assess whether the activity of carbonic anhydrase is affected by other ions in solution. 2 Assay methods The enzyme carbonic anhydrase from bovine erythrocytes (bCA) used in our work was purchased from MP Biomedicals (#153879), with a reported molecular weight of 30,000 g/mol. The standard method for determining bCA activity involves measuring the time required for a CO2-saturated solution to lower the pH of a veronal buffer solution from 8.3 to 6.3 at 0-4°C, both in the presence and absence of bCA (Wilbur and Anderson, 1948; Worthington, 1993; Uchikawa and Zeebe, 2012). The enzyme activity is reported in Wilbur-Anderson enzyme units (EU): · tblank − tCAEU = 2 (1) tCA 194 where tCA and tblank are the measured time interval for the pH decline with and without bCA, respectively. According to Eq. ??, two enzyme units corresponds to an uncatalyzed reaction time that is twice as long as the catalyzed reaction time. For blank assays, a 30 mL beaker containing 12 mL of ice-cold Tris-HCl (0.02 M; pH = 8.6) was set in an ice bath on a magnetic stir plate. The temperature of the Tris-HCl was monitored to ensure that it stayed within 0-4°C during the assay. The reaction was initiated by pipetting 8 mL of (nearly) CO2-saturated distilled, deionized H2O (DDI), which was prepared by briefly bubbling pure CO2 through 200 mL of ice-cold deionized water. The pH of this solution was found to be 3.96, in agreement with literature values (Peng et al., 2013). Upon addition of the CO2 solution to the Tris-HCl, there is a relatively sharp decrease in pH to about 8.3, followed by a more gradual decrease to the pH endpoint. For assays at high salinity, we prepared (nearly) CO2-saturated solutions using DDI mixed with 87.5 g/kg and 175 g/kg NaCl, such that the final mixtures of Tris-HCl and CO2-saturated solutions were 35 g/kg and 70 g/kg. For CA assays, we added 0.04 mL of 0.1 mg/mL bCA (4 x 10−3 mg of bCA) to the Tris-HCl buffer immediately prior to addition of the (nearly) CO2-saturated solution. This bCA solution was stored in a refrigerator and we noticed no significant drop in bCA activity over the course of several months. 3 Assay results Results from freshwater and saltwater assays with essentially the same tblank (∼135-150 s) show a resolvable influence of NaCl on bCA activity (Fig. ??). The catalyzed time for freshwater is about 25 s, and about 40 s and 45 s for saltwater of 35 g/kg and 70 g/kg, respectively. From these results we calculate a bCA activity of ∼2350 EU/mg for freshwater, which is consistent with the manufacturer’s quote as well as the values determined by Uchikawa and Zeebe (2012). From the saltwater results we calculate a bCA activity of about 1340 EU/mg in 35 g/kg salinity solutions and 1070 EU/mg in 70 g/kg solutions, indicating that 35-70 g/kg of background NaCl inhibits the activity of bCA by a factor of 1.8-2.2. This should be regarded with caution because the assays are carried out at variable pH and ∼0°C whereas our experiments are at fixed pH and 25°C. 4 Assay complications and challenges We encountered a number of issues with the assay protocol that have not been adequately addressed in the literature. In the discussion that follows, we attempt to lay bare these issues so that future researchers can avoid the pitfalls that we encountered. The standard method for determining the activity of CA involves measuring the time required for a CO2-saturated solution to lower the pH of a veronal buffer solution from 8.3 to 6.3 at 0-4°C, both in the presence of, and absence of, CA (Wilbur and Anderson, 1948). A clear recipe for 195 9.0 8.5 pH 8.3 8.0 Salinity (g/kg) blanks 80 7.5 60 w/ bCA 40 7.0 20 0 6.5 pH 6.3 6.0 5.5 5.0 0 50 100 150 200 time (s) 3000 2500 2000 1500 1000 500 0 0 10 20 30 40 50 60 70 salinity (g/kg) Figure 1: The activity of bovine CA is inhibited by NaCl. Freshwater assays yield 2150-2520 EU/mg. Saltwater assays yield 1240-1450 EU/mg at 35 g/kg and 1000-1100 EU/mg at 70 g/kg. Error bars represent one standard deviation from the mean. 196 EU/mg bCA pH carrying out the assay can be found in Worthington (1993), which was modified by Uchikawa and Zeebe (2012) who used different concentrations of reagents. To prepare our veronal buffer solution, Tris-base (121.14 g/mol) was added to DDI and placed in an ice bath. The initial pH of the ice-cold 0.02 M Tris solution was 11-11.5. Between 1-2 mL concentrated HCl was added to reach the target pH of 8.6 for the Tris-HCl solution. An alternative approach is to use Tris-HCl base, which contains equal molar quantities of Tris and HCl, and requires addition of a base such as NaOH to reach the target pH. Uchikawa and Zeebe (2012) bubbled pure CO2 gas through DDI water for 30-60 minutes to ensure CO2-saturation (Uchikawa and Zeebe, 2012; Worthington, 1993). When we bubbled CO2 through DDI for this amount of time or longer, the resulting tblank was much lower (∼30 to 80 s) than the tblank of ∼160 s reported by Uchikawa and Zeebe (2012). Additionally, our blank runs had poor reproducibility whereas Fig. 2 in Uchikawa and Zeebe (2012) shows excellent reproducibility. The bCA activities that we obtained from these early runs were generally about a factor of two lower than those obtained by Uchikawa and Zeebe (2012), as well as the manufacturer’s quote. Setting aside the disagreement between our blanks and theirs, we investigated several pos- sible causes for the lack of reproducibility, including changing the stirring rate, calibrating the pH probe at 0°C instead of room temperature, storing the probe in ice versus a room temper- ature solution between assays, using different batches of Tris, transferring the CO2-saturated solution in a large pipette to minimize interaction with the air, and performing the assay in a round bottom flask instead of a beaker to minimize interaction with the air. None of these approaches improved the reproducibility. Ultimately, we concluded that the variable blank re- sults were due to variably under- or over-saturated CO2 solutions. This conclusion aligns with a careful reading of the Sigma-Aldrich assay protocol (https://www.sigmaaldrich.com/technical- documents/protocols/biology/enzymatic-assay-of-carbonic- anhydrase.html): The general blank time after first opening [the CO2 solution] is approximately 40 sec- onds. Record this time as “Blank-1”. If this occurs, transfer [the CO2 solution] back fourth between a 1 L beaker and the “Vess Seltzer” bottle a couple of times. Place [the CO2 solution] back in ice. Repeat [the blank assay]. Record all blank times in seconds. The blank times tend to increase with each opening of the bottle containing [the CO2 solution]. Once a blank time of 65 seconds is reached, proceed with [the CA assay]. After all test runs, the final blank average must be in the range of 70 to 100 seconds. This text implies use of a partially degassed CO2 solution, which is not mentioned elsewhere in the literature (Worthington, 1993; Uchikawa and Zeebe, 2012). By bubbling CO2 and allowing the solution to partially degas, we managed to obtain blanks between 70-100 s but this still led to bCA activities that were a factor of two lower than the manufacturer’s quote (Fig. ??). An important caveat is that we followed the Uchikawa and Zeebe (2012) protocol, and for direct comparison to their results, we tried bubbling CO2 for shorter durations (<2 minutes) to obtain blank runs of ∼140-150 seconds to match their blank time. This led to a bCA activity in excellent agreement with their results in NaCl-free solutions (2325±175 versus 2358 EU/mg). 197 9.0 Fresh Salty 8.5 pH 8.3 8.0 Blanks 7.5 7.0 w/CA 6.5 pH 6.3 6.0 5.5 −20 0 20 40 60 80 100 120 time (s) Figure 2: Assay results for a shorter tblank of ∼75 seconds. The freshwater assay yields a bCA activity of ∼1375 EU/mg whereas the saltwater assay yields ∼450 EU/mg. These values are significantly lower than the manufacturer’s quote (∼2000 EU/mg), which is why we accept the assays with longer tblank in the main text. The choice of tblank does not affect the main conclusion that there is a salt inhibition effect of about a factor of 2-3. For saltwater assays, we adjusted the CO2 bubbling time in order to obtain the same tblank of ∼140-150 seconds to facilitate direct comparison to freshwater results. Importantly, regardless of tblank, whether it be 40, 80, or 150 seconds, we find that NaCl lowers bCA activity by about a factor of 2-3 in going from [NaCl] = 0 g/kg to 35 g/kg (Fig. ?? and Fig. ??). Additional References Peng C., Crawshaw J. P., Maitland G. C., Martin Trusler J. P. and Vega-Maza D. (2013) The pH of CO2-saturated water at temperatures between 308 K and 423 K at pressures up to 15 MPa. J. Supercrit. Fluids 82, 129–137. Wilbur K. M. and Anderson N. G. (1948) Electrometric and Colorimetric Determination of Car- bonic Anhydrase. J. Biol. Chem. 176, 147–154. Worthington V. (1993) Worthington Enzyme Manual. Worthington Biomedical Corporation, Lake- wood, New Jersey. 198 pH APPENDIX D CHAPTER IV SUPPLEMENTARY MATERIAL: EXPERIMENT DATA AND SEM IMAGES OF PRECIPITATES 1 Experiment data table Table 1: Experimental measurements Experiment Time TA [DIC] δ13C δ18DIC Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) S6 3.85 0.46 - - - 223 - 18.82 0.69 - - - 295 - 26.83 0.61 - - - 242 - 41.45 0.51 - - - 294 - 45.78 0.53 - - - 238 - 52.75 0.52 - - - 238 - 63.75 0.50 - - - 346 - 66.63 0.50 - - - 275 - 72.78 0.50 - - -11.31 S8 3.65 0.47 0.14 -43.66 - 97 - 11.80 0.65 0.37 -44.47 - 142 - 19.07 0.66 0.26 -44.18 - 143 - 26.07 0.49 0.15 -44.60 - 103 - 34.97 0.50 0.17 -45.47 - 104 - 47.22 0.49 0.13 -45.70 - 102 - 52.15 0.60 0.13 -45.55 -11.39 91 CA4 3.22 0.74 0.14 -24.50 - 134 - 8.38 0.95 0.31 -24.88 - 165 - 21.25 0.70 0.14 -25.12 - 121 - 25.38 0.72 0.19 -25.04 - 138 - 30.22 0.71 0.16 -25.05 -11.68 124 CA10 0.40 1.78 - - - 130 - 19.47 1.70 0.25 -38.84 - 70 199 Table 1 continued Experiment Time TA [DIC] δ13C δ18DIC Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 31.58 1.69 0.32 -40.45 - 71 - 44.33 1.70 0.45 -42.03 - 71 - 48.10 1.73 0.42 -42.16 - 69 - 55.23 1.75 0.52 -42.63 - 68 - 67.72 1.74 0.58 -43.02 -11.825 67 CA11 0.20 2.28 - - - 151 - 18.97 2.07 0.21 -40.40 - 64 - 23.77 2.00 0.26 -41.16 - 64 - 30.25 2.12 0.30 -41.97 - 64 - 43.42 2.08 0.39 -42.84 - 64 - 49.37 1.98 0.52 -43.21 - 64 - 66.67 1.96 0.63 -43.79 - 59 - 70.50 2.01 0.61 -43.82 -11.8 58 CA16 3.30 0.69 0.63 -18.22 - 469 - 12.53 0.85 0.72 -17.77 - 576 - 19.55 0.65 0.44 -19.52 - 418 - 27.12 0.63 0.42 -20.00 - 411 - 35.77 0.64 0.42 -20.11 - 404 - 40.75 0.68 0.39 -20.17 -11.85 398 CA17 3.47 0.57 0.36 -20.54 - 374 - 13.48 0.99 0.87 -19.73 - 673 - 19.97 0.90 0.68 -17.80 - 531 - 27.62 0.71 0.41 -18.03 - 383 - 37.42 0.67 0.32 -18.44 - 341 - 46.18 0.59 0.33 -18.45 -11.84 339 CA18 3.20 0.77 0.39 -17.44 - 220 - 13.93 0.71 0.31 -16.77 - 214 - 19.78 0.64 0.24 -17.40 - 192 - 26.93 0.64 0.22 -17.53 - 186 - 36.45 0.65 0.20 -17.66 - 188 - 42.28 0.64 0.22 -17.45 -11.88 191 CA20 6.22 0.74 0.34 -35.65 - 161 - 15.95 0.63 0.23 -35.59 - 127 - 22.50 0.63 0.22 -34.74 - 124 - 29.92 0.61 0.19 -32.34 - 115 - 39.62 0.60 0.16 -31.50 - 102 200 Table 1 continued Experiment Time TA [DIC] δ13C δ18DIC Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 46.50 0.59 0.15 -31.45 -11.95 99 CA21 4.38 1.03 0.38 -29.62 - 121 - 14.57 0.81 0.13 -27.80 - 86 - 22.38 0.81 0.14 -27.46 - 71 - 29.20 0.80 0.15 -28.02 - 65 - 41.30 0.82 0.16 -28.35 -11.52 54 CA22 3.45 1.40 0.10 -22.25 - 100 - 17.07 1.30 0.08 -23.69 - 66 - 24.40 1.40 0.07 -22.68 - 36 - 30.57 1.34 0.08 -22.56 - 29 - 41.92 1.37 0.07 -22.97 -11.87 18 CA23 3.62 1.30 0.08 -29.56 - 127 - 21.55 1.23 0.10 -40.87 - 150 - 31.75 1.24 0.06 -40.94 - 141 - 42.52 1.22 0.06 -42.19 - 144 - 55.65 1.25 0.05 -42.35 - 133 - 67.55 1.29 0.05 -42.14 -10.65 122 CA24 1.57 2.29 0.19 -18.86 - 56 - 11.32 2.06 0.08 -35.10 - 67 - 16.85 2.14 0.07 -37.74 - 63 - 23.72 2.06 0.06 -37.77 - 64 - 33.43 1.96 0.04 -36.58 - 64 - 42.25 2.05 0.05 -38.57 - 65 - 60.78 2.03 0.04 -37.82 - 58 - 71.80 1.93 0.04 -36.91 - 56 - 84.62 2.05 0.04 -37.36 -10.85 39 CA25 3.07 0.95 0.17 -39.61 - 56 - 13.25 1.10 0.25 -41.53 - 41 - 21.83 0.93 0.14 -39.77 - -4 - 29.60 0.90 0.10 -40.73 - -20 - 40.18 0.92 0.08 -40.69 - -38 - 52.60 0.86 0.08 -40.55 - -64 - 66.25 0.87 0.09 -40.96 -10.91 -85 CA26 2.65 0.94 0.71 -21.27 - 1638 - 13.55 1.25 1.04 -18.10 - 2042 - 19.73 1.22 1.06 -17.89 - 1976 201 Table 1 continued Experiment Time TA [DIC] δ13CDIC δ 18Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) - 26.38 1.27 1.05 -17.93 - 1983 - 39.55 1.23 1.03 -17.88 - 1962 - 50.77 1.26 1.14 -17.82 - 1934 - 61.87 1.21 1.09 -18.07 - 1888 - 74.55 1.34 1.15 -18.37 - 1787 - 92.33 1.25 0.89 -19.88 -10.96 1458 CA27 1.97 0.70 0.56 -22.30 - 1074 - 14.68 1.17 0.98 -18.35 - 2008 - 24.20 1.17 1.10 -17.88 - 1981 - 38.07 1.19 1.13 -17.59 - 1963 - 48.73 1.15 0.99 -17.92 - 1946 - 61.62 1.18 1.02 -17.87 - 1942 - 72.22 1.20 0.98 -17.97 - 1939 - 86.05 1.25 1.14 -18.04 - 1852 - 96.22 1.29 1.02 -18.59 - 1745 - 112.32 1.22 0.98 -20.68 -11.15 1434 CA28 5.03 1.58 0.35 -19.01 - 58 - 16.45 1.28 0.09 -22.35 - 63 - 24.67 1.28 0.09 -22.21 - 34 - 37.90 1.25 0.08 -22.31 - -9 - 47.15 1.32 0.07 -22.00 -11.28 -26 CA29 1.77 2.27 0.29 -14.10 - 74 - 12.60 1.78 0.08 -20.98 - 75 - 20.87 1.76 0.06 -21.58 - 72 - 31.38 1.72 0.05 -21.70 - 66 - 41.02 1.74 0.05 -20.85 - 40 - 51.58 1.71 0.04 -21.30 - 31 - 61.80 1.69 0.04 -21.86 - 17 - 72.15 1.75 0.04 -21.76 -11.36 3 CA30 1.58 0.68 0.64 -23.33 - 1168 - 19.4 1.37 1.32 -18.18 - 1933 - 36.87 1.51 1.28 -18.17 - 1896 - 49.57 1.35 1.23 -18.22 - 1878 - 62.93 1.44 1.27 -18.35 - 1707 - 74.65 1.47 1.35 -19.23 - 1652 - 91.98 1.06 0.96 -22.48 -10.86 1069 202 Table 1 continued Experiment Time TA [DIC] δ13C 18DIC δ Ow pCO2 (h) (mEq/L) (mM) (VPDB) (VSMOW) (ppm) CA31 1.82 0.77 0.69 -22.66 - 1190 - 15.17 1.30 1.34 -18.15 - 1988 - 26.12 1.44 1.42 -19.46 - 1719 - 36.3 1.16 0.94 -22.18 -11.23 1286 2 Experiment data figures For each experiment, we periodically took water samples that were sent for [DIC] and δ13C analysis, which are depicted by black circles, with curves interpolated between the data measurements (dark green and light green, respectively). We continuously monitored the concentration of CO2 (ppm) of the experimental headspace (light blue line). We continuously measured solution pH (purple line) and amount of added NaOH (dark blue line), in addition to the periodic water samples we titrated to calculate the total alkalinity of the solution (red squares). The vertical gray line represents the onset of calcite precipitation for each experiment. Data figures for experiments S6, S8, CA4, CA18, and CA20 were included as part of the experimental suite for Chapter III and may be found in Appendix B. 203 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −38 −40 −42 −44 0 10 20 30 40 50 60 70 9 10.0 8 9.5 7 6 9.0 5 8.5 4 3 8.0 2 7.5 1 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 1: CA10, pH = 9.3 204 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −40 −42 −44 0 10 20 30 40 50 60 70 8 10.0 7 9.5 6 9.0 5 4 8.5 3 8.0 2 7.5 1 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 2: CA11, pH = 9.3 205 δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 −16 −18 −20 −22 0 10 20 30 40 1000 800 600 400 200 0 0 10 20 30 40 8 10.0 7 9.5 6 9.0 5 4 8.5 3 8.0 2 7.5 1 0 7.0 0 10 20 30 40 time (hours) Figure 3: CA16, pH = 7.9 206 pCO2 (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 1.0 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 −16 −18 −20 −22 0 10 20 30 40 50 1000 800 600 400 200 0 0 10 20 30 40 50 10.0 7 9.5 6 5 9.0 4 8.5 3 8.0 2 7.5 1 0 7.0 0 10 20 30 40 50 time (hours) Figure 4: CA17, pH = 7.9 207 pCO2 (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.2 0.0 0 10 20 30 40 −26 −28 −30 −32 0 10 20 30 40 200 100 0 0 10 20 30 40 10.0 6 9.5 5 9.0 4 8.5 3 2 8.0 1 7.5 0 7.0 0 10 20 30 40 time (hours) Figure 5: CA21, pH = 8.65 208 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.2 0.1 0.0 0 10 20 30 40 −22 −24 0 10 20 30 40 200 100 0 0 10 20 30 40 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 time (hours) Figure 6: CA22, pH = 9.0 209 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.2 0.1 0.0 0 10 20 30 40 50 60 70 −28 −32 −36 −40 −44 0 10 20 30 40 50 60 70 200 100 0 0 10 20 30 40 50 60 70 3 10.0 9.5 2 9.0 8.5 1 8.0 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 7: CA23, pH = 9.3 210 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.2 0.1 0.0 0 10 20 30 40 50 60 70 80 90 −16 −20 −24 −28 −32 −36 −40 0 10 20 30 40 50 60 70 80 90 100 0 0 10 20 30 40 50 60 70 80 90 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 80 90 time (hours) Figure 8: CA24, pH = 9.3 211 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.3 0.2 0.1 0.0 0 10 20 30 40 50 60 70 −38 −40 −42 0 10 20 30 40 50 60 70 200 100 0 −100 0 10 20 30 40 50 60 70 10.0 7 9.5 6 5 9.0 4 8.5 3 8.0 2 7.5 1 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 9: CA25, pH = 8.65 212 pCO2 (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 1.6 1.2 0.8 0.4 0.0 0 10 20 30 40 50 60 70 80 90 100 −16 −18 −20 −22 −24 0 10 20 30 40 50 60 70 80 90 100 2400 2000 1600 1200 800 400 0 0 10 20 30 40 50 60 70 80 90 100 10.0 6 9.5 5 9.0 4 8.5 3 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 80 90 100 time (hours) Figure 10: CA26, pH = 7.5 213 pCO2 (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 1.6 1.2 0.8 0.4 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 −16 −18 −20 −22 −24 0 10 20 30 40 50 60 70 80 90 100 110 120 2400 2000 1600 1200 800 400 0 0 10 20 30 40 50 60 70 80 90 100 110 120 5 10.0 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 80 90 100 110 120 time (hours) Figure 11: CA27, pH = 7.5 214 pCO2 (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.2 0.0 0 10 20 30 40 50 −18 −20 −22 −24 0 10 20 30 40 50 200 100 0 0 10 20 30 40 50 6 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 time (hours) Figure 12: CA28, 9.0 215 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 0.4 0.2 0.0 0 10 20 30 40 50 60 70 −12 −16 −20 −24 0 10 20 30 40 50 60 70 200 100 0 0 10 20 30 40 50 60 70 10.0 5 9.5 4 9.0 3 8.5 2 8.0 1 7.5 0 7.0 0 10 20 30 40 50 60 70 time (hours) Figure 13: CA29, pH = 9.3 216 pCO2 (ppm) δ 13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 1.6 1.2 0.8 0.4 0.0 0 10 20 30 40 50 60 70 80 90 −16 −18 −20 −22 −24 −26 0 10 20 30 40 50 60 70 80 90 2400 2000 1600 1200 800 400 0 0 10 20 30 40 50 60 70 80 90 11 10.0 10 9 9.5 8 9.0 7 6 8.5 5 4 8.0 3 2 7.5 1 0 7.0 0 10 20 30 40 50 60 70 80 90 time (hours) Figure 14: CA30, pH = 7.5 217 pCO (ppm) 13 Alkalinity (mEq/L) 2 δ C (VPDB) [DIC] (mM) pH 1.6 1.2 0.8 0.4 0.0 0 10 20 30 40 −16 −18 −20 −22 −24 −26 0 10 20 30 40 2400 2000 1600 1200 800 400 0 0 10 20 30 40 7 10.0 6 9.5 5 9.0 4 8.5 3 8.0 2 1 7.5 0 7.0 0 10 20 30 40 time (hours) Figure 15: CA31, pH = 7.5 218 pCO2 (ppm) δ13C (VPDB) [DIC] (mM) Alkalinity (mEq/L) pH 3 SEM images SEM images for experiments S6, S8, CA4, CA18, and CA20 were included as part of the experi- mental suite for Chapter III and may be found in Appendix B. 219 Figure 16: CA10, pH = 9.3 220 Figure 17: CA11, pH = 9.3 221 Figure 18: CA16, pH = 7.9 222 Figure 19: CA17, pH = 7.9. Time series experiment - 1st disc removed 223 Figure 20: CA17, pH = 7.9. Time series experiment - 2nd disc removed 224 Figure 21: CA17, pH = 7.9. Time series experiment - disc in solution for the full experiment 225 Figure 22: CA21, pH = 8.65 226 Figure 23: CA22, pH = 9.0 227 4 Additional figures 29.0 28.5 28.0 bCA (µM) 27.5 1.5 1.0 27.0 0.5 0.0 26.5 0 50 100 150 200 250 300 350 HCO- 2-3/CO3 Figure 24: Oxygen isotope fractionation between calcite and experimental solution expressed as 1000lnαc−w, over the range of experimental solution HCO − 2− 3 /CO3 . Included are experiments from this study (circles), from Baker (2015) (diamonds), and from Watkins et al. (2014) (triangles). 228 1000ln c-w 28 26 24 22 bCA (µM) 20 1.5 1.0 18 0.5 16 0.0 14 0 50 100 150 200 250 300 350 HCO-3/CO2-3 Figure 25: Oxygen isotope fractionation between calcite and experimental solution expressed as 1000lnαc−w, over the range of experimental solution HCO − 3 /CO 2− 3 , including experiments that did not utilize the enzyme bCA. Included are experiments from this study (circles), from Baker (2015) (diamonds), and from Watkins et al. (2014) (triangles). 229 1000ln c-w APPENDIX E CHAPTER V SUPPLEMENTARY MATERIAL: EXPERIMENT DATA AND SEM IMAGES OF PRECIPITATES 1 Experiment data For each experiment, we continuously monitored the experimental solution pH and concentration of CO2 (ppm) of the experiment headspace. For open-air experiments, the concentration of CO2 reflects that of the laboratory air. 230 200 100 0 0 20 40 60 80 100 120 140 160 13 12 11 10 9 8 0 20 40 60 80 100 120 140 160 time (hours) Figure 1: CH1. 200ppm CO 2+2, 25°C, 10mM Ca 231 pH pCO2 (ppm) 200 100 0 0 20 40 60 80 100 13 12 11 10 9 8 0 20 40 60 80 100 time (hours) Figure 2: CH2. 200ppm CO2, 25°C, 30mM Ca2+ 232 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 20 40 60 80 13 12 11 10 9 8 0 20 40 60 80 time (hours) Figure 3: CH3. 2000ppm CO2, 25°C, 30mM Ca2+ 233 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 20 40 13 12 11 10 9 8 0 20 40 time (hours) Figure 4: CH4. 2000ppm CO2, 25°C, 30mM Ca2+ 234 pH pCO2 (ppm) 200 100 0 0 20 40 60 80 100 120 13 12 11 10 9 8 0 20 40 60 80 100 120 time (hours) Figure 5: CH5. 200ppm CO2, 25°C, 30mM Ca2+ 235 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 20 40 60 80 100 13 12 11 10 9 8 0 20 40 60 80 100 time (hours) Figure 6: CH6. 2000ppm CO2, 10°C, 30mM Ca2+ 236 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 20 40 13 12 11 10 9 8 0 20 40 time (hours) Figure 7: CH7. 2000ppm CO2, 25°C, 30mM Ca2+. Unstirred 237 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 20 40 13 12 11 10 9 8 0 20 40 time (hours) Figure 8: CH8. 2000ppm CO2, 10°C, 30mM Ca2+ 238 pH pCO2 (ppm) 200 100 0 0 20 40 60 13 12 11 10 9 8 0 20 40 60 time (hours) Figure 9: CH9. 200ppm CO2, 25°C, 10mM Ca2+ 239 pH pCO2 (ppm) 200 100 0 0 20 40 60 80 13 12 11 10 9 8 0 20 40 60 80 time (hours) Figure 10: CH10. 200ppm CO2, 10°C, 30mM Ca2+ 240 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 5 10 15 13 12 11 10 9 8 0 5 10 15 time (hours) Figure 11: CH11. 2000ppm CO2, 25°C, 10mM Ca2+ 241 pCO2 (ppm)pH 2000 1600 1200 800 400 0 0 10 20 13 12 11 10 9 8 0 10 20 time (hours) Figure 12: CH12. 2000ppm CO2, 25°C, 10mM Ca2+ 242 pCO2 (ppm)pH 600 400 200 0 0 20 40 13 12 11 10 9 8 0 20 40 time (hours) Figure 13: CH13. Lab air ∼450ppm CO2, 25°C, 10mM Ca2+ 243 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 10 20 13 12 11 10 9 8 0 10 20 time (hours) Figure 14: CH14. 2000ppm CO2, 25°C, 10mM Ca2+. Mini-beaker, unstirred 244 pCO (ppm) pH 2 2000 1600 1200 800 400 0 0 20 40 13 12 11 10 9 8 0 20 40 time (hours) Figure 15: CH15. 2000ppm CO2, 25°C, 10mM Ca2+. Mini-beaker, unstirred 245 pH pCO2 (ppm) 400 0 0 20 40 13 12 11 10 9 8 0 20 40 time (hours) Figure 16: CH17. Lab air ∼450ppm CO2, 25°C, 10mM Ca2+ 246 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 10 20 30 13 12 11 10 9 8 0 10 20 30 time (hours) Figure 17: CH18. 2000ppm CO2, 25°C, 10mM Ca2+. Mini-beaker, unstirred 247 pCO (ppm) pH 2 2000 1600 1200 800 400 0 0 10 20 13 12 11 10 9 8 0 10 20 time (hours) Figure 18: CH19. 2000ppm CO2, 25°C, 10mM Ca2+. Mini-beaker, unstirred 248 pCO (ppm) pH 2 600 400 200 0 0 10 20 30 13 12 11 10 9 8 0 10 20 30 time (hours) Figure 19: CH20. Lab air ∼450ppm CO2, 25°C, 10mM Ca2+ 249 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 10 20 30 13 12 11 10 9 8 0 10 20 30 time (hours) Figure 20: CH21. 2000ppm CO2, 25°C, 10mM Ca2+. Mini-beaker, unstirred 250 pCO (ppm) pH 2 600 400 200 0 0 10 20 30 13 12 11 10 9 8 0 10 20 30 time (hours) Figure 21: CH22. Lab air ∼450ppm CO2, 25°C, 10mM Ca2+. Time series (2 crystal skims) 251 pH pCO2 (ppm) 2000 1600 1200 800 400 0 0 10 20 13 12 11 10 9 8 0 10 20 time (hours) Figure 22: CH23. 2000ppm CO2, 25°C, 10mM Ca2+. Mini-beaker, unstirred 252 pCO (ppm) pH 2 600 400 200 0 0 10 20 30 40 13 12 11 10 9 8 0 10 20 30 40 time (hours) Figure 23: CH24. Lab air ∼450ppm CO2, 25°C, 10mM Ca2+. Time series (3 crystal skims) 253 pH pCO2 (ppm) 600 400 200 0 0 10 20 30 40 50 60 70 80 13 12 11 10 9 8 0 10 20 30 40 50 60 70 80 time (hours) Figure 24: CH25. Lab air ∼450ppm CO 2+2, 10°C, 10mM Ca . Time series (3 crystal skims) 254 pH pCO2 (ppm) 600 400 200 0 0 10 20 30 40 50 13 12 11 10 9 8 0 10 20 30 40 50 time (hours) Figure 25: CH26. Lab air ∼450ppm CO2, 10°C, 10mM Ca2+. 255 pH pCO2 (ppm) 600 400 200 0 0 10 20 30 40 13 12 11 10 9 8 0 10 20 30 40 time (hours) Figure 26: CH27. Lab air ∼450ppm CO2, 10°C, 10mM Ca2+. 256 pH pCO2 (ppm) 2 SEM images Figure 27: CH1. 200ppm CO2, 25°C, 10mM Ca2+ 257 Figure 28: CH2. 200ppm CO2, 25°C, 30mM Ca2+ 258 Figure 29: CH3. 2000ppm CO2, 25°C, 30mM Ca2+ 259 Figure 30: CH4. 2000ppm CO2, 25°C, 30mM Ca2+ 260 Figure 31: CH5. 200ppm CO2, 25°C, 30mM Ca2+ 261 Figure 32: CH6. 2000ppm CO2, 10°C, 30mM Ca2+ 262 Figure 33: CH7. 2000ppm CO2, 25°C, 30mM Ca2+. Unstirred 263 Figure 34: CH8. 2000ppm CO2, 10°C, 30mM Ca2+ 264 Figure 35: CH9. 200ppm CO2, 25°C, 10mM Ca2+ 265 Figure 36: CH10. 200ppm CO2, 10°C, 30mM Ca2+ 266 3 Additional figures 15 atmospheric CO2 10 5 0 5 200 ppm CO2 10 15 2000 ppm CO2 20 25 30 35 40 measured H2O OH- (Böttcher et al., 2018) 45 50 55 60 OH- (Zeebe, 2020) 65 70 75 80 OH- (Bajnai and Herwartz, 2021) 85 60 55 50 45 40 35 30 25 20 15 10 5 0 13CVPDB (‰) Figure 37: Isotopic compositions of precipitated carbonates and gas sources. The data are grouped by gas source, with the 200 ppm CO2 experiments in purple, 2000 ppm CO2 experiments in red, and open air experiments using the modern CO2 atmosphere in orange. Pentagons are the gas sources, circles are 25°C gas tank experiments, diamonds are 10°C gas tank experiments, upright triangles are 25°C open air experiments, inverted triangles are 10°C experiments, and squares are open air experiment crystals from the bottom of the beaker. The isotopic outliers are CH21 and CH23 (red circles), and CH22 beaker crystals (orange square). The orange star represents the most highly fractionated travertine from The Cedars (Christensen et al., 2021). The average measured δ18O of our experimental solutions is ≈ -40.4‰ VPDB, which would subsequently result in OH− that is ≈ -61.3‰, -81.6‰, or -40.4‰ VPDB using 1000lnαOH−−H O from Zeebe2 (2020), Bajnai and Herwartz (2021), or that proposed by Böttcher et al. (2018), respectively. 267 18OVPDB (‰) 5 CH19 10 CH1 CH18 CH11 CH15 CH6 CH5 CH2 CH9 CH14 CH7 CH4 CH20 CH22 CH13 CH3 CH10 CH24 Cedars CH8 CH12 CH25 15 CH17 CH27 CH26_b % pCO2CH26 100 80 20 60 20 15 10 5 ε13CaCO3/CO2(aq) (‰) Figure 38: Carbon and bulk oxygen kinetic isotope fractionations (KIFs) for all experiments of this study, calculated as  = (α - 1) · 1000. Gas tank experiments (25°C - circles; 10°C - diamonds) on 200 ppm CO2 (black outline) or 2000 ppm CO2 (no outline). Open air experiments (25°C - upright triangles; 10°C - inverted triangles) with crystals from the bottom of those beakers (squares). The star represents the KFFs determined from The Cedars, California (Christensen et al., 2021). Experiments are color coded according to the average experiment pCO2 as a percent of the pCO2 of its gas source (i.e. 200 ppm or 2000 ppm CO2 gas tanks). All open-air experiments are plotted as 100%. 268 ε18CaCO3/CO2(aq)+H2O (‰) 0 CH19 CH1 CH18 CH11 5 CH15 CH6 CH5 CH2 CH7 CH9 CH4 CH14 CH20 CH22 CH13 CH3 CH12 CH10 CH24 Cedars CH8 CH25 CH17 CH27 CH26_b CH26 10 initial pH 12.5 12.0 11.5 15 11.0 20 15 10 5 ε13CaCO3/CO2(aq) (‰) Figure 39: Carbon and bulk oxygen kinetic isotope fractionations (KIFs) for all experiments of this study, calculated as  = (α - 1) · 1000. Gas tank experiments (25°C - circles; 10°C - diamonds) on 200 ppm CO2 (black outline) or 2000 ppm CO2 (no outline). Open air experiments (25°C - upright triangles; 10°C - inverted triangles) with crystals from the bottom of those beakers (squares). The star represents the KFFs determined from The Cedars, California (Christensen et al., 2021). 1000lnαOH−−H O ≈ -21.5‰ from Zeebe2 (2020) is used when calculating the KIFs. 269 ε18CaCO3/CO2(aq)+OH- (‰) 10 5 CH19 CH1 CH18 CH11 CH15 CH6 CH5 CH2 CH7 CH9 CH4 CH14 CH20 CH22 CH13 CH3 0 CH12 CH10CH24 Cedars CH8 CH25 CH17 CH27 CH26_b initial pH CH26 12.5 12.0 5 11.5 11.0 20 15 10 5 ε13CaCO3/CO2(aq) (‰) Figure 40: Carbon and bulk oxygen kinetic isotope fractionations (KIFs) for all experiments of this study, calculated as  = (α - 1) · 1000. Gas tank experiments (25°C - circles; 10°C - diamonds) on 200 ppm CO2 (black outline) or 2000 ppm CO2 (no outline). Open air experiments (25°C - upright triangles; 10°C - inverted triangles) with crystals from the bottom of those beakers (squares). The star represents the KFFs determined from The Cedars, California (Christensen et al., 2021). 1000lnαOH−−H O ≈ -42.53‰ from2 Bajnai and Herwartz (2021) is used when calculating the KIFs. 270 ε18CaCO3/CO2(aq)+OH- (‰) REFERENCES CITED Chapter II Andersson M. P., Dobberschütz S., Sand K. K., Tobler D. J., De Yoreo J. J. and Stipp S. L. S. (2016) A Microkinetic Model of Calcite Step Growth. Angew. Chemie 128, 11252–11256. Bajnai D. and Herwartz D. (2021) Kinetic Oxygen Isotope Fractionation between Water and Aque- ous OH− during Hydroxylation of CO2. ACS Earth Sp. Chem. 5, 3375–3384. Bar-Matthews M., Marean C. W., Jacobs Z., Karkanas P., Fisher E. C., Herries A. I. R., Brown K., Williams H. M., Bernatchez J., Ayalon A. and Nilssen P. J. (2010) A high resolution and continuous isotopic speleothem record of paleoclimate and paleoenvironment from 90 to 53 ka from Pinnacle Point on the south coast of South Africa. Quat. Sci. Rev. 29, 2131–2145. Barkan E. and Luz B. (2012) High-precision measurements of 17O/16O and 18O/16O ratios in CO2. Rapid Commun. Mass Spectrom. 26, 2733–2738. Beck W. C., Grossman E. L. and Morse J. W. (2005) Experimental studies of oxygen isotope frac- tionation in the carbonic acid system at 15°, 25°, and 40°C. Geochim. Cosmochim. Acta 69, 3493–3503. Böhm F., Eisenhauer A., Tang J., Dietzel M., Krabbenhöft A., Kisakürek B. and Horn C. (2012) Strontium isotope fractionation of planktic foraminifera and inorganic calcite. Geochim. Cos- mochim. Acta 93, 300–314. Böttcher M. E., Neubert N., Escher P., von Allmen K., Samankassou E. and Nägler T. F. (2018) Multi-isotope (Ba, C, O) partitioning during experimental carbonatization of a hyper-alkaline solution. Chemie der Erde 78, 241–247. Bottinga Y. (1968) Calculation of Fractionation Factors for Carbon and Oxygen Isotopic Exchange in the System Calcite-Carbon Dioxide-Water. J. Phys. Chem. 72, 800–807. Bowen G. J. and Wilkinson B. (2002) Spatial distribution of δ18O in meteoric precipitation. Geol- ogy 30, 315–318. Brennan S. T., Lowenstein T. K. and Horita J. (2004) Seawater chemistry and the advent of bio- calcification. Geology 32, 473–476. 271 Budd D. A., Hammes U. and Ward W. B. (2000) Cathodoluminescence in Calcite Cements: New Insights on Pb and Zn Sensitizing, Mn Activation, and Fe Quenching at Low Trace-Element Concentrations. SEPM J. Sediment. Res. 70, 217–226. Carpenter S. J. and Lohmann K. C. (1992) Sr/Mg ratios of modern marine calcite: Empirical indi- cators of ocean chemistry and precipitation rate. Geochim. Cosmochim. Acta 56, 1837–1849. Charlton S. R. and Parkhurst D. L. (2011) Modules based on the geochemical model PHREEQC for use in scripting and programming languages. Comput. Geosci. 37, 1653–1663. Chen S., Gagnon A. C. and Adkins J. F. (2018) Carbonic anhydrase, coral calcification and a new model of stable isotope vital effects. Geochim. Cosmochim. Acta 236, 179–197. Christensen J. N., Watkins J. M., Devriendt L. S., DePaolo D. J., Conrad M. E., Voltolini M., Yang W. and Dong W. (2021) Isotopic fractionation accompanying CO2 hydroxylation and carbonate precipitation from high pH waters at The Cedars, California, USA. Geochim. Cos- mochim. Acta 301, 91–115. Christensen J. N., Watkins J. M., Devriendt L. S., DePaolo D. J., Conrad M. E., Voltolini M., Yang W. and Dong W. (2023) Corrigendum to “Isotopic fractionation accompanying CO2 hydroxylation and carbonate precipitation from high pH waters at The Cedars, California, USA” [Geochim. Cosmochim. Acta 301 (2021) 91-115]. Geochim. Cosmochim. Acta 343, 416-419. Clark I. D., Fontes J. C. and Fritz P. (1992) Stable isotope disequilibria in travertine from high pH waters: Laboratory investigations and field observations from Oman. Geochim. Cosmochim. Acta 56, 2041–2050. Clark I. D. and Lauriol B. (1992) Kinetic enrichment of stable isotopes in cryogenic calcites. Chem. Geol. 102, 217–228. Cléroux C., Cortijo E., Anand P., Labeyrie L., Bassinot F., Caillon N. and Duplessy J. C. (2008) Mg/Ca and Sr/Ca ratios in planktonic foraminifera: Proxies for upper water column temper- ature reconstruction. Paleoceanography 23, 1–16. Coggon R. M., Teagle D. A. H., Smith-Duque C. E., Alt J. C. and Cooper M. J. (2010) Recon- structing Past Seawater Mg/Ca and Sr/Ca from Mid-Ocean Ridge Flank Calcium Carbonate Veins. Science 327, 1114–1117. Coplen T. B. (2007) Calibration of the calcite-water oxygen-isotope geothermometer at Devils Hole, Nevada, a natural laboratory. Geochim. Cosmochim. Acta 71, 3948–3957. Coplen T. B., Kendall C. and Hopple J. (1983) Comparison of stable isotope reference samples. Nature 302, 236–238. Daëron M., Drysdale R. N., Peral M., Huyghe D., Blamart D., Coplen T. B., Lartaud F. and Zanchetta G. (2019) Most Earth-surface calcites precipitate out of isotopic equilibrium. Nat. Commun. 10, 1–7. 272 Davis K. J., Dove P. M., Wasylenki L. E. and De Yoreo J. J. (2004) Morphological consequences of differential Mg2+ incorporation at structurally distinct steps on calcite. Am. Mineral. 89, 714–720. De Lucia M. and Kühn M. (2013) Coupling R and PHREEQC: efficient programming of geochem- ical models. Energy Procedia 40, 464–471. Dennis K. J. and Schrag D. P. (2010) Clumped isotope thermometry of carbonatites as an indicator of diagenetic alteration. Geochim. Cosmochim. Acta 74, 4110–4122. DePaolo D. J. (2011) Surface kinetic model for isotopic and trace element fractionation during precipitation of calcite from aqueous solutions. Geochim. Cosmochim. Acta 75, 1039–1056. Devriendt L. S., Watkins J. M. and McGregor H. V. (2017) Oxygen isotope fractionation in the CaCO3-DIC-H2O system. Geochim. Cosmochim. Acta 214, 115–142. Dietzel M., Tang J., Leis A. and Köhler S. J. (2009) Oxygen isotopic fractionation during inorganic calcite precipitation - Effects of temperature, precipitation rate and pH. Chem. Geol. 268, 107–115. Dietzel M., Usdowski E. and Hoefs J. (1992) Chemical and 13C/12C-and 18O/16O-isotope evolution of alkaline drainage waters and the precipitation of calcite. Appl. Geochemistry 7, 177–184. Edgar K. M., Anagnostou E., Pearson P. N. and Foster G. L. (2015) Assessing the impact of di- agenesis on δ11B, δ13C, δ18O, Sr/Ca and B/Ca values in fossil planktic foraminiferal calcite. Geochim. Cosmochim. Acta 166, 189–209. Eigen M. (1964) Proton Transfer, Acid-Base Catalysis, and Enzymatic Hydrolysis. Part I: ELE- MENTARY PROCESSES. Angew. Chemie Int. Ed. English 3, 1–19. Eiler J. M. (2007) “Clumped-isotope” geochemistry—The study of naturally-occurring, multiply- substituted isotopologues. Earth Planet. Sci. Lett. 262, 309–327. Escobar J., Hodell D. A., Brenner M., Curtis J. H., Gilli A., Mueller A. D., Anselmetti F. S., Ariztegui D., Grzesik D. A., Pérez L., Schwalb A. and Guilderson T. P. (2012) A ∼43-ka record of paleoenvironmental change in the Central American lowlands inferred from stable isotopes of lacustrine ostracods. Quat. Sci. Rev. 37, 92–104. Falk E. S., Guo W., Paukert A. N., Matter J. M., Mervine E. M. and Kelemen P. B. (2016) Controls on the stable isotope compositions of travertine from hyperalkaline springs in Oman: Insights from clumped isotope measurements. Geochim. Cosmochim. Acta 192, 1–28. Fantle M. S. (2015) Calcium isotopic evidence for rapid recrystallization of bulk marine carbonates and implications for geochemical proxies. Geochim. Cosmochim. Acta 148, 378–401. Folk R. L. (1994) Interaction between bacteria, nannobacteria, and mineral precipitation in hot springs of Central Italy. Géographie Phys. Quat. 48, 233–246. 273 Gabitov R. I., Watson E. B. and Sadekov A. (2012) Oxygen isotope fractionation between calcite and fluid as a function of growth rate and temperature: An in situ study. Chem. Geol. 306–307, 92–102. Ghosh P., Adkins J., Affek H., Balta B., Guo W., Schauble E. A., Schrag D. and Eiler J. M. (2006) 13C-18O bonds in carbonate minerals: A new kind of paleothermometer. Geochim. Cosmochim. Acta 70, 1439–1456. Ghosh P., Eiler J., Campana S. E. and Feeney R. F. (2007) Calibration of the carbonate “clumped isotope” paleothermometer for otoliths. Geochim. Cosmochim. Acta 71, 2736–2744. Green M. and Taube H. (1963) Isotopic fractionation in the OH−-H2O exchange reaction. J. Phys. Chem. 67, 1565–1566. Grotzinger J. P. and Kasting J. F. (1993) New constraints on Precambrian ocean composition. J. Geol. 101, 235–243. Guo W. (2009) Carbonate Clumped Isotope Thermometry: Application to Carbonaceous Chondrites and Effects of Kinetic Isotope Fractionation. PhD thesis. California Institute of Technology. Guo W. (2020) Kinetic clumped isotope fractionation in the DIC-H2O-CO2 system: Patterns, con- trols, and implications. Geochim. Cosmochim. Acta 268, 230–257. Guo W. and Zhou C. (2019) Triple oxygen isotope fractionation in the DIC-H2O-CO2 system: A numerical framework and its implications. Geochim. Cosmochim. Acta 246, 541–564. Habermann D. (2002) Quantitative cathodoluminescence (CL) spectroscopy of minerals: Possibil- ities and limitations. Mineral. Petrol. 76, 247–259. Halevy I. and Bachan A. (2017) The geologic history of seawater pH. Science 355, 1069–1071. Hardie L. A. (1996) Secular variation in seawater chemistry: An explanation for the coupled secular variation in the mineralogies of marine limestones and potash evaporites over the past 600 m.y. Geology 24, 279. Helgeson H. C. and Kirkham D. H. (1976) Theoretical Prediction of the Thermodynamic Proper- ties of Aqueous Electrolytes At High Pressures and Temperatures - 3. Equation of State for Aqueous Species At Infinite Dilution. Am J Sci 276, 97–240. Hill P. S., Tripati A. K. and Schauble E. A. (2014) Theoretical constraints on the effects of pH, salinity, and temperature on clumped isotope signatures of dissolved inorganic carbon species and precipitating carbonate minerals. Geochim. Cosmochim. Acta 125, 610–652. Ichikuni M. (1973) Partition of strontium between calcite and solution: effect of substitution by manganese. Chem. Geol. 11, 315–319. 274 Jiang, L. Q., Carter, B. R., Feely, R. A., Lauvset, S. K. and Olsen, A. (2019) Surface ocean pH and buffer capacity: past, present and future. Sci Rep 9, 18624. Kelson J. R., Huntington K. W., Schauer A. J., Saenger C. and Lechler A. R. (2017) Toward a universal carbonate clumped isotope calibration: Diverse synthesis and preparatory methods suggest a single temperature relationship. Geochim. Cosmochim. Acta 197, 104–131. Kim S.-T., Hillaire-Marcel C. and Mucci A. (2006) Mechanisms of equilibrium and kinetic oxygen isotope effects in synthetic aragonite at 25 °C. Geochim. Cosmochim. Acta 70, 5790–5801. Kim S.-T. and O’Neil J. R. (1997) Equilibrium and nonequilibrium oxygen isotope effects in syn- thetic carbonates. Geochim. Cosmochim. Acta 61, 3461–3475. Kluge T., John C. M., Jourdan A. L., Davis S. and Crawshaw J. (2015) Laboratory calibration of the calcium carbonate clumped isotope thermometer in the 25-250°C temperature range. Geochim. Cosmochim. Acta 157, 213–227. Kupriyanova E. V and Pronina N. A. (2011) Carbonic Anhydrase: Enzyme That Has Transformed the Biosphere. Russ. J. Plant Physiol. 58, 197–209. Larsen K., Bechgaard K. and Stipp S. L. S. (2010) The effect of the Ca2+ to CO2−3 activity ratio on spiral growth at the calcite 1014 surface. Geochim. Cosmochim. Acta 74, 2099–2109. Lemarchand D., Wasserburg G. J. and Papanastassiou D. A. (2004) Rate-controlled calcium iso- tope fractionation in synthetic calcite. Geochim. Cosmochim. Acta 68, 4665–4678. Levitt N. P., Eiler J. M., Romanek C. S., Beard B. L., Xu H. and Johnson C. M. (2018) Near Equilib- rium 13C-18O Bonding During Inorganic Calcite Precipitation Under Chemo-Stat Conditions. Geochemistry, Geophys. Geosystems 19, 901–920. Lopez O., Zuddas P. and Faivre D. (2009) The influence of temperature and seawater composition on calcite crystal growth mechanisms and kinetics: Implications for Mg incorporation in cal- cite lattice. Geochim. Cosmochim. Acta 73, 337–347. Lowenstein T. K., Timofeeff M. N., Brennan S. T., Hardie L. A. and Demicco R. V. (2001) Os- cillations in Phanerozoic seawater chemistry: Evidence from fluid Inclusions. Science 294, 1086–1088. Machel H.-G. (1985) Cathodoluminescence in Calcite and Dolomite and its Chemical Interpreta- tion. Geosci. Canada 12, 139–147. Mackensen A. and Schmiedl G. (2019) Stable carbon isotopes in paleoceanography: atmosphere, oceans, and sediments. Earth-Science Rev. 197, 102893. Major R. P. and Wilber R. J. (1991) Crystal habit, geochemistry, and cathodoluminescence of magnesian calcite marine cements from the lower slope of Little Bahama Bank. Geol. Soc. Am. Bull. 103, 461–471. 275 Mason R. A. (1987) Ion microprobe analysis of trace elements in calcite with an application to the cathodoluminescence zonation of limestone cements from the Lower Carboniferous of South Wales, U.K. Chem. Geol. 64, 209–224. Mavromatis V., Gautier Q., Bosc O. and Schott J. (2013) Kinetics of Mg partition and Mg stable isotope fractionation during its incorporation in calcite. Geochim. Cosmochim. Acta 114, 188–203. Mavromatis V., Montouillout V., Noireaux J., Gaillardet J. and Schott J. (2015) Characterization of boron incorporation and speciation in calcite and aragonite from co-precipitation experi- ments under controlled pH, temperature and precipitation rate. Geochim. Cosmochim. Acta 150, 299–313. Mavromatis V., Schmidt M., Botz R., Comas-Bru L. and Oelkers E. H. (2012) Experimental quan- tification of the effect of Mg on calcite-aqueous fluid oxygen isotope fractionation. Chem. Geol. 310–311, 97–105. McCaffrey M. A., Lazar B. and Holland H. D. (1987) The Evaporation Path of Seawater and the Coprecipitation of Br− and K+ with Halite. J. Sediment. Petrol. 57, 928–937. McConnaughey T. (1989) 13C and 18O isotopic disequilibrium in biological carbonates: I. Patterns. Geochim. Cosmochim. Acta 53, 151–162. McCrea J. M. (1950) On the Isotopic Chemistry of Carbonates and a Paleotemperature Scale. J. Chem. Phys. 18, 849–857. Millero F. J., Graham T. B., Huang F., Bustos-Serrano H. and Pierrot D. (2006) Dissociation constants of carbonic acid in seawater as a function of salinity and temperature. Mar. Chem. 100, 80–94. Millero F., Huang F., Graham T. and Pierrot D. (2007) The dissociation of carbonic acid in NaCl solutions as a function of concentration and temperature. Geochim. Cosmochim. Acta 71, 46–55. Mook W. G. (1986) 13C in atmospheric CO2. Netherlands J. Sea Res. 20, 211–223. Morse J. W., Wang Q. and Tsio M. Y. (1997) Influences of temperature and Mg:Ca ratio on CaCO3 precipitates from seawater. Geology 25, 85–87. Nielsen L. C., DePaolo D. J. and De Yoreo J. J. (2012) Self-consistent ion-by-ion growth model for kinetic isotopic fractionation during calcite precipitation. Geochim. Cosmochim. Acta 86, 166–181. O’Leary M. H. (1988) Carbon Isotopes in Photosynthesis. Bioscience 38, 328–336. Paquette J. and Reeder R. J. (1995) Relationship between surface structure, growth mechanism, and trace element incorporation in calcite. Geochim. Cosmochim. Acta 59, 735–749. 276 Passey B. H. and Henkes G. A. (2012) Carbonate clumped isotope bond reordering and geospeedom- etry. Earth Planet. Sci. Lett. 351–352, 223–236. Peral M., Daëron M., Blamart D., Bassinot F., Dewilde F., Smialkowski N., Isguder G., Bonnin J., Jorissen F., Kissel C., Michel E., Vázquez Riveiros N. and Waelbroeck C. (2018) Updated cali- bration of the clumped isotope thermometer in planktonic and benthic foraminifera. Geochim. Cosmochim. Acta 239, 1–16. Pingitore N. E., Eastman M. P., Sandidge M., Oden K. and Freiha B. (1988) The coprecipitation of manganese(II) with calcite: an experimental study. Mar. Chem. 25, 107–120. Pinsent B. R. W., Pearson L. and Roughton F. J. W. (1956) The kinetics of combination of carbon dioxide with ammonia. Trans. Faraday Soc. 52, 1594–1598. Richter D. K., Götte T., Götze J. and Neuser R. D. (2003) Progress in application of cathodolu- minescence (CL) in sedimentary petrology. Mineral. Petrol. 79, 127–166. Sade Z. and Halevy I. (2017) New constraints on kinetic isotope effects during CO2(aq) hydration and hydroxylation: Revisiting theoretical and experimental data. Geochim. Cosmochim. Acta 214, 246–265. Sade Z., Yam R., Shemesh A. and Halevy I. (2020) Kinetic fractionation of carbon and oxygen isotopes during BaCO3 precipitation. Geochim. Cosmochim. Acta 280, 395–422. Srivastava A. and Verkouteren R. M. (2018) Metrology for stable isotope reference materials: 13C/12C and 18O/16O isotope ratio value assignment of pure carbon dioxide gas samples on the Vienna PeeDee Belmnite-CO2 scale using dual-inlet mass spectrometry. Anal. Bioanal. Chem. 410, 4153–4163. Sun W., Jayaraman S., Chen W., Persson K. A. and Ceder G. (2015) Nucleation of metastable aragonite CaCO3 in seawater. Proc. Natl. Acad. Sci. U. S. A. 112, 3199–3204. Swart P. K., Burns S. J. and Leder J. J. (1991) Fractionation of the stable isotopes of oxygen and carbon in carbon dioxide during the reaction of calcite with phosphoric acid as a function of temperature and technique. Chem. Geol. Isot. Geosci. Sect. 86, 89–96. Tang J., Dietzel M., Fernandez A., Tripati A. K. and Rosenheim B. E. (2014) Evaluation of kinetic effects on clumped isotope fractionation (∆47) during inorganic calcite precipitation. Geochim. Cosmochim. Acta 134, 120–136. Tang J., Köhler S. J. and Dietzel M. (2008) Sr2+/Ca2+ and 44Ca/40Ca fractionation during inor- ganic calcite formation: I. Sr incorporation. Geochim. Cosmochim. Acta 72, 3718–3732. Tang J., Niedermayr A., Köhler S. J., Böhm F., Kisakürek B., Eisenhauer A. and Dietzel M. (2012) Sr2+/Ca2+ and 44Ca/40Ca fractionation during inorganic calcite formation: III. Im- pact of salinity/ionic strength. Geochim. Cosmochim. Acta 77, 432–443. 277 Teng H. H., Dove P. M. and De Yoreo J. J. (2000) Kinetics of calcite growth: Surface processes and relationships to macroscopic rate laws. Geochim. Cosmochim. Acta 64, 2255–2266. Tripati A. K., Hill P. S., Eagle R. A., Mosenfelder J. L., Tang J., Schauble E. A., Eiler J. M., Zeebe R. E., Uchikawa J., Coplen T. B., Ries J. B. and Henry D. (2015) Beyond temperature: Clumped isotope signatures in dissolved inorganic carbon species and the influence of solution chemistry on carbonate mineral composition. Geochim. Cosmochim. Acta 166, 344–371. Trolier M., White J. W. C., Tans P. P., Masarie K. A. and Gemery P. A. (1996) Monitoring the isotopic composition of atmospheric CO2: Measurements from the NOAA global air sampling network. J. Geophys. Res. Atmos. 101, 25897–25916. Turchyn A. V. and DePaolo D. J. (2019) Seawater Chemistry Through Phanerozoic Time. Annu. Rev. Earth Planet. Sci. 47, 197–224. Uchikawa J., Penman D. E., Zachos J. C. and Zeebe R. E. (2015) Experimental evidence for kinetic effects on B/Ca in synthetic calcite: Implications for potential B(OH)−4 and B(OH)3 incorpo- ration. Geochim. Cosmochim. Acta 150, 171–191. Uchikawa J. and Zeebe R. E. (2012) The effect of carbonic anhydrase on the kinetics and equi- librium of the oxygen isotope exchange in the CO2-H2O system: Implications for δ 18O vital effects in biogenic carbonates. Geochim. Cosmochim. Acta 95, 15–34. Urey H. C. (1947) The Thermodynamic Properties of Isotopic Substances. J. Chem. Soc., 562–581. van Dijk J., Fernandez A., Storck J. C., White T. S., Lever M., Müller I. A., Bishop S., Seifert R. F., Driese S. G., Krylov A., Ludvigson G. A., Turchyn A. V., Lin C. Y., Wittkop C. and Bernasconi S. M. (2019) Experimental calibration of clumped isotopes in siderite between 8.5 and 62 °C and its application as paleo-thermometer in paleosols. Geochim. Cosmochim. Acta 254, 1–20. Vogel J. C., Grootes P. M. and Mook W. G. (1970) Isotopic fractionation between gaseous and dissolved carbon dioxide. Zeitschrift für Phys. 230, 225–238. Wacker U., Fiebig J., Tödter J., Schöne B. R., Bahr A., Friedrich O., Tütken T., Gischler E. and Joachimski M. M. (2014) Empirical calibration of the clumped isotope paleothermometer us- ing calcites of various origins. Geochim. Cosmochim. Acta 141, 127–144. Wang L., Ruiz-Agudo E., Putnis C. V. and Putnis A. (2011) Direct observations of the modifi- cation of calcite growth morphology by Li+ through selectively stabilizing an energetically unfavourable face. CrystEngComm 13, 3962–3966. Wang Z., Gaetani G., Liu C. and Cohen A. (2013) Oxygen isotope fractionation between aragonite and seawater: Developing a novel kinetic oxygen isotope fractionation model. Geochim. Cos- mochim. Acta 117, 232–251. Watkins J. M. and Hunt J. D. (2015) A process-based model for non-equilibrium clumped isotope effects in carbonates. Earth Planet. Sci. Lett. 432, 152–165. 278 Watkins J. M., Hunt J. D., Ryerson F. J. and DePaolo D. J. (2014) The influence of temperature, pH, and growth rate on the δ18O composition of inorganically precipitated calcite. Earth Planet. Sci. Lett. 404, 332–343. Watkins J. M., Nielsen L. C., Ryerson F. J. and Depaolo D. J. (2013) The influence of kinetics on the oxygen isotope composition of calcium carbonate. Earth Planet. Sci. Lett. 375, 349–360. Watson E. B. (2004) A conceptual model for near-surface kinetic controls on the trace- element and stable isotope composition of abiogenic calcite crystals. Geochim. Cosmochim. Acta 68, 1473–1488. Wolthers M., Nehrke G., Gustafsson J. P. and Van Cappellen P. (2012) Calcite growth kinetics: Modeling the effect of solution stoichiometry. Geochim. Cosmochim. Acta 77, 121–134. Yumol L. M., Uchikawa J. and Zeebe R. E. (2020) Kinetic isotope effects during CO2 hydration: Experimental results for carbon and oxygen fractionation. Geochim. Cosmochim. Acta 279, 189–203. Zaarur S., Affek H. P. and Brandon M. T. (2013) A revised calibration of the clumped isotope thermometer. Earth Planet. Sci. Lett. 382, 47–57. Zachos J., Pagani H., Sloan L., Thomas E. and Billups K. (2001) Trends, rhythms, and aberrations in global climate 65 Ma to present. Science 292, 686–693. Zeebe R. E. (1999) An explanation of the effect of seawater carbonate concentration on foraminiferal oxygen isotopes. Geochim. Cosmochim. Acta 63, 2001–2007. Zeebe R. E. (2007) An expression for the overall oxygen isotope fractionation between the sum of dissolved inorganic carbon and water. Geochemistry, Geophys. Geosystems 8, 1–7. Zeebe R. E. (2014) Kinetic fractionation of carbon and oxygen isotopes during hydration of carbon dioxide. Geochim. Cosmochim. Acta 139, 540–552. Zeebe R. E. (2020) Oxygen isotope fractionation between water and the aqueous hydroxide ion. Geochim. Cosmochim. Acta 289, 182–195. Zeebe R. E. and Wolf-Gladrow D. (2001) CO2 in seawater: equilibrium, kinetics, isotopes., Elsevier, Amsterdam. Zeebe R. E., Wolf-Gladrow D. A. and Jansen H. (1999) On the time required to establish chemical and isotopic equilibrium in the carbon dioxide system in seawater. Mar. Chem. 65, 135–153. Zhang J., Quay P. D. and Wilbur D. O. (1995) Carbon isotope fractionation during gas-water exchange and dissolution of CO2. Geochim. Cosmochim. Acta 59, 107–114. Zuddas P. and Mucci A. (1998) Kinetics of calcite precipitation from seawater: II. The influence of the ionic strength. Geochim. Cosmochim. Acta 62, 757–766. 279 Chapter III Adkins J. F., Boyle E. A., Curry W. B. and Lutringer A. (2003) Stable isotopes in deep-sea corals and a new mechanism for “vital effects.” Geochim. Cosmochim. Acta 67, 1129–1143. Baker E. B. (2015) Carbon and Oxygen Isotope Fractionation in Laboratory-Precipitated, Inor- ganic Calcite. University of Oregon. Beck W. C., Grossman E. L. and Morse J. W. (2005) Experimental studies of oxygen isotope frac- tionation in the carbonic acid system at 15°, 25°, and 40°C. Geochim. Cosmochim. Acta 69, 3493–3503. Berner R. A. and Morse J. W. (1974) Dissolution Kinetics of Calcium Carbonate in Sea Water IV. Theory of Calcite Dissolution. Am. J. Sci. 274, 108–134. Bertucci A., Innocenti A., Zoccola D., Scozzafava A., Allemand D., Tambutté S. and Supuran C. T. (2009) Carbonic anhydrase inhibitors: Inhibition studies of a coral secretory isoform with inorganic anions. Bioorganic Med. Chem. Lett. 19, 650–653. Bertucci A., Innocenti A., Scozzafava A., Tambutté S., Zoccola D. and Supuran C. T. (2011a) Car- bonic anhydrase inhibitors. Inhibition studies with anions and sulfonamides of a new cytosolic enzyme from the scleractinian coral Stylophora pistillata. Bioorganic Med. Chem. Lett. 21, 710–714. Bertucci A., Tambutté S., Supuran C. T., Allemand D. and Zoccola D. (2011b) A New Coral Car- bonic Anhydrase in Stylophora pistillata. Mar. Biotechnol. 13, 992–1002. Bigeleisen J. and Mayer M. G. (1947) Calculation of Equilibrium Constants for Isotopic Exchange Reactions. J. Chem. Phys. 15, 261–267. Brenninkmeijer C. A. M., Kraft P. and Mook W. G. (1983) Oxygen isotope fractionation between CO2 and H2O. Chem. Geol. 41, 181–190. Charlton S. R. and Parkhurst D. L. (2011) Modules based on the geochemical model PHREEQC for use in scripting and programming languages. Comput. Geosci. 37, 1653–1663. Chen S., Gagnon A. C. and Adkins J. F. (2018) Carbonic anhydrase, coral calcification and a new model of stable isotope vital effects. Geochim. Cosmochim. Acta 236, 179–197. Christensen J. N., Watkins J. M., Devriendt L. S., DePaolo D. J., Conrad M. E., Voltolini M., Yang W. and Dong W. (2021) Isotopic fractionation accompanying CO2 hydroxylation and carbonate precipitation from high pH waters at The Cedars, California, USA. Geochim. Cos- mochim. Acta 301, 91–115. Clark I. D., Fontes J. C. and Fritz P. (1992) Stable isotope disequilibria in travertine from high pH waters: Laboratory investigations and field observations from Oman. Geochim. Cosmochim. Acta 56, 2041–2050. 280 Coplen T. B. (2007) Calibration of the calcite-water oxygen-isotope geothermometer at Devils Hole, Nevada, a natural laboratory. Geochim. Cosmochim. Acta 71, 3948–3957. Coplen T. B., Kendall C. and Hopple J. (1983) Comparison of stable isotope reference samples. Na- ture 302, 236–238. De Goeyse S., Webb A. E., Reichart G.-J. and De Nooijer L. J. (2021) Car- bonic anhydrase is involved in calcification by the benthic foraminifer Amphistegina lessonii. Biogeosciences 18, 393–401. De Lucia M. and Kühn M. (2013) Coupling R and PHREEQC: efficient programming of geochem- ical models. Energy Procedia 40, 464–471. De Simone G. and Supuran C. T. (2012) (In)organic anions as carbonic anhydrase inhibitors. J. Inorg. Biochem. 111, 117–129. Del Prete S., Vullo D., Scozzafava A., Capasso C. and Supuran C. T. (2014) Cloning, characteriza- tion and anion inhibition study of the δ-class carbonic anhydrase (TweCA) from the marine diatom Thalassiosira weissflogii. Bioorganic Med. Chem. 22, 531–537. DePaolo D. J. (2011) Surface kinetic model for isotopic and trace element fractionation during precipitation of calcite from aqueous solutions. Geochim. Cosmochim. Acta 75, 1039–1056. Devriendt L. S., McGregor H. V. and Chivas A. R. (2017a) Ostracod calcite records the 18O/16O ratio of the bicarbonate and carbonate ions in water. Geochim. Cosmochim. Acta 214, 30–50. Devriendt L. S., Watkins J. M. and McGregor H. V. (2017b) Oxygen isotope fractionation in the CaCO3-DIC-H2O system. Geochim. Cosmochim. Acta 214, 115–142. Dionisio-Sese M. L. and Miyachi S. (1992) The Effect of Sodium Chloride on Carbonic Anhydrase Activity in Marine Microalgae. J. Phycol. 28, 619–624. DOE (1994) Handbook of Methods for the Analysis of the Various Parameters of the Carbon Diox- ide System in Sea water. Version 2, Dickson A. G. and Goyet C., Eds., ORNL/CDIAC-74, Oak Ridge, Tennessee. Furla P., Allemand D. and Orsenigo M. N. (2000) Involvement of H+-ATPase and carbonic anhy- drase in inorganic carbon uptake for endosymbiont photosynthesis. Am. J. Physiol. - Regul. Integr. Comp. Physiol. 278, R870–R881. Guo W. (2020) Kinetic clumped isotope fractionation in the DIC-H2O-CO2 system: Patterns, con- trols, and implications. Geochim. Cosmochim. Acta 268, 230–257. Guo W. and Zhou C. (2019) Triple oxygen isotope fractionation in the DIC-H2O-CO2 system: A numerical framework and its implications. Geochim. Cosmochim. Acta 246, 541–564. Henry R. P. (2001) Environmentally mediated carbonic anhydrase induction in the gills of euryha- line crustaceans. J. Exp. Biol. 204, 991–1002. 281 Hermoso M., Horner T. J., Minoletti F. and Rickaby R. E. M. (2014) Constraints on the vital ef- fect in coccolithophore and dinoflagellate calcite by oxygen isotopic modification of seawater. Geochim. Cosmochim. Acta 141, 612–627. Hong M. and Teng H. H. (2014) Implications of solution chemistry effects: Direction-specific re- straints on the step kinetics of calcite growth. Geochim. Cosmochim. Acta 141, 228–239. Jacobson R. L. and Langmuir D. (1974) Dissociation constants of calcite and CaHCO+3 from 0 to 50°C. Geochim. Cosmochim. Acta 38, 301–318. Kim S.-T., Gebbinck C. K., Mucci A. and Coplen T. B. (2014) Oxygen isotope systematics in the aragonite–CO2–H2O–NaCl system up to 0.7 mol/kg ionic strength at 25 °C. Geochim. Cos- mochim. Acta 137, 147–158. Kim S.-T., Hillaire-Marcel C. and Mucci A. (2006) Mechanisms of equilibrium and kinetic oxygen isotope effects in synthetic aragonite at 25 °C. Geochim. Cosmochim. Acta 70, 5790–5801. Kim S.-T. and O’Neil J. R. (1997) Equilibrium and nonequilibrium oxygen isotope effects in syn- thetic carbonates. Geochim. Cosmochim. Acta 61, 3461–3475. Kimball J. B., Dunbar R. B. and Guilderson T. P. (2014) Oxygen and carbon isotope fractionation in calcitic deep-sea corals: Implications for paleotemperature reconstruction. Chem. Geol. 381, 223–233. Mass T., Drake J. L., Peters E. C., Jiang W. and Falkowski P. G. (2014) Immunolocalization of skeletal matrix proteins in tissue and mineral of the coral Stylophora pistillata. Proc. Natl. Acad. Sci. U. S. A. 111, 12728–12733. McConnaughey T. (1989) 13C and 18O isotopic disequilibrium in biological carbonates: I. Patterns. Geochim. Cosmochim. Acta 53, 151–162. Millero F. J., Graham T. B., Huang F., Bustos-Serrano H. and Pierrot D. (2006) Dissociation constants of carbonic acid in seawater as a function of salinity and temperature. Mar. Chem. 100, 80–94. Millero F., Huang F., Graham T. and Pierrot D. (2007) The dissociation of carbonic acid in NaCl solutions as a function of concentration and temperature. Geochim. Cosmochim. Acta 71, 46–55. Mitsunaga K., Akasaka K., Shimada H., Fujino Y., Yasumasu I. and Numanoi H. (1986) Carbonic anhydrase activity in developing sea urchin embryos with special reference to calcification of spicules. Cell Differ. 18, 257–262. Morse J. W. (1978) Dissolution kinetics of calcium carbonate in sea water: VI. The near-equilibrium dissolution kinetics of calcium carbonate-rich deep sea sediments. Am. J. Sci. 278, 344–353. 282 Moya A., Tambutté S., Bertucci A., Tambutté E., Lotto S., Vullo D., Supuran C. T., Allemand D. and Zoccola D. (2008) Carbonic anhydrase in the scleractinian coral Stylophora pistil- lata: Characterization, localization, and role in biomineralization. J. Biol. Chem. 283, 25475–25484. Mucci A. (1983) The solubility of calcite and aragonite in seawater at various salinities, tempera- tures, and one atmosphere total pressure. Am. J. Sci. 283, 780–799. Nancollas G. H. and Reddy M. M. (1971) The crystallization of calcium carbonate. II. Calcite growth mechanism. J. Colloid Interface Sci. 37, 824–830. Nielsen L. C., DePaolo D. J. and De Yoreo J. J. (2012) Self-consistent ion-by-ion growth model for kinetic isotopic fractionation during calcite precipitation. Geochim. Cosmochim. Acta 86, 166–181. Nielsen S. A. and Frieden E. (1972) Some chemical and kinetic properties of oyster carbonic anhy- drase. Comp. Biochem. Physiol. Part B Comp. Biochem. 41, 875–889. O’Leary M. H. (1984) Measurement of the isotope fractionation associated with diffusion of carbon dioxide in aqueous solution. J. Phys. Chem. 88, 823–825. Pinsent B. R. W., Pearson L. and Roughton F. J. W. (1956) The kinetics of combination of carbon dioxide with ammonia. Trans. Faraday Soc. 52, 1594–1598. Ruiz-Agudo E., Kowacz M., Putnis C. V. and Putnis A. (2010) The role of background elec- trolytes on the kinetics and mechanism of calcite dissolution. Geochim. Cosmochim. Acta 74, 1256–1267. Ruiz-Agudo E., Putnis C. V., Wang L. and Putnis A. (2011) Specific effects of background elec- trolytes on the kinetics of step propagation during calcite growth. Geochim. Cosmochim. Acta 75, 3803–3814. Sade Z. and Halevy I. (2017) New constraints on kinetic isotope effects during CO2(aq) hydration and hydroxylation: Revisiting theoretical and experimental data. Geochim. Cosmochim. Acta 214, 246–265. Sofer Z. and Gat J. R. (1972) Activities and Concentrations of Oxygen-18 in Concentrated Aque- ous Salt Solutions: Analytical and Geophysical Implications. Earth Planet. Sci. Lett. 15, 232–238. Spero H. J., Bijma J., Lea D. W. and Bernis B. E. (1997) Effect of seawater carbonate concentration on foraminiferal carbon and oxygen isotopes. Nature 390, 497–500. Tambutté S., Holcomb M., Ferrier-Pagès C., Reynaud S., Tambutté É., Zoccola D. and Allemand D. (2011) Coral biomineralization: From the gene to the environment. J. Exp. Mar. Bio. Ecol. 408, 58–78. 283 Tang J., Köhler S. J. and Dietzel M. (2008) Sr2+/Ca2+ and 44Ca/40Ca fractionation during inor- ganic calcite formation: I. Sr incorporation. Geochim. Cosmochim. Acta 72, 3718–3732. Taube H. (1954) Use of Oxygen Isotope Effects in the Study of Hydration of Ions. J. Phys. Chem. 58, 523–528. Uchikawa J., Chen S., Eiler J. M., Adkins J. F. and Zeebe R. E. (2021) Trajectory and timescale of oxygen and clumped isotope equilibration in the dissolved carbonate system under normal and enzymatically-catalyzed conditions at 25 °C. Geochim. Cosmochim. Acta 314, 313–333. Uchikawa J. and Zeebe R. E. (2012) The effect of carbonic anhydrase on the kinetics and equi- librium of the oxygen isotope exchange in the CO2-H2O system: Implications for δ 18O vital effects in biogenic carbonates. Geochim. Cosmochim. Acta 95, 15–34. Uchikawa J. and Zeebe R. E. (2013) No discernible effect of Mg2+ ions on the equilibrium oxygen isotope fractionation in the CO2-H2O system. Chem. Geol. 343, 1–11. Urey H. C. (1947) The Thermodynamic Properties of Isotopic Substances. J. Chem. Soc., 562–581. Usdowski E. and Hoefs J. (1993) Oxygen isotope exchange between carbonic acid, bicarbonate, carbonate, and water: A re-examination of the data of McCrea (1950) and an expression for the overall partitioning of oxygen isotopes between the carbonate species and water. Geochim. Cosmochim. Acta 57, 3815–3818. Usdowski E., Michaelis J., Bottcher M. E. and Hoefs J. (1991) Factors for the oxygen isotope equilibrium fractionation between aqueous and gaseous CO2, carbonic acid, bicarbonate, car- bonate, and water (19 °C). Z. Phys. Chem. 170, 237–249. Wang L., Ruiz-Agudo E., Putnis C. V. and Putnis A. (2011) Direct observations of the modifi- cation of calcite growth morphology by Li+ through selectively stabilizing an energetically unfavourable face. CrystEngComm 13, 3962–3966. Watkins J. M., DePaolo D. J. and Watson E. B. (2017) Kinetic Fractionation of Non-Traditional Stable Isotopes by Diffusion and Crystal Growth Reactions. Rev. Mineral. Geochemistry 82, 85–125. Watkins J. M., Hunt J. D., Ryerson F. J. and DePaolo D. J. (2014) The influence of temperature, pH, and growth rate on the δ18O composition of inorganically precipitated calcite. Earth Planet. Sci. Lett. 404, 332–343. Watkins J. M., Nielsen L. C., Ryerson F. J. and DePaolo D. J. (2013) The influence of kinetics on the oxygen isotope composition of calcium carbonate. Earth Planet. Sci. Lett. 375, 349–360. Wolthers M., Nehrke G., Gustafsson J. P. and Van Cappellen P. (2012) Calcite growth kinetics: Modeling the effect of solution stoichiometry. Geochim. Cosmochim. Acta 77, 121–134. 284 Yumol L. M., Uchikawa J. and Zeebe R. E. (2020) Kinetic isotope effects during CO2 hydration: Experimental results for carbon and oxygen fractionation. Geochim. Cosmochim. Acta 279, 189–203. Zeebe R. E. (2007) An expression for the overall oxygen isotope fractionation between the sum of dissolved inorganic carbon and water. Geochemistry, Geophys. Geosystems 8, 1–7. Zeebe R. E. (2014) Kinetic fractionation of carbon and oxygen isotopes during hydration of carbon dioxide. Geochim. Cosmochim. Acta 139, 540–552. Zeebe R. E. (2020) Oxygen isotope fractionation between water and the aqueous hydroxide ion. Geochim. Cosmochim. Acta 289, 182–195. Zeebe R. E. and Wolf-Gladrow D. (2001) CO2 in seawater: equilibrium, kinetics, isotopes. Elsevier Oceanography Series, Amsterdam. Zhang H., Blanco-Ameijeiras S., Hopkinson B. M., Bernasconi S. M., Mejia L. M., Liu C. and Stoll H. (2021) An isotope label method for empirical detection of carbonic anhydrase in the calcification pathway of the coccolithophore Emiliania huxleyi. Geochim. Cosmochim. Acta 292, 78–93. Zuddas P. and Mucci A. (1998) Kinetics of calcite precipitation from seawater: II. The influence of the ionic strength. Geochim. Cosmochim. Acta 62, 757–766. Chapter IV Andersson, M. P., Dobberschütz, S., Sand, K. K., Tobler, D. J., De Yoreo, J. J., & Stipp, S. L. S. (2016a). A Microkinetic Model of Calcite Step Growth. Angewandte Chemie, 128 (37), 11252–11256. https://doi.org/10.1002/ange.201604357 Andersson, M. P., Rodriguez-Blanco, J. D., & Stipp, S. L. S. (2016b). Is bicarbonate stable in and on the calcite surface? Geochimica et Cosmochimica Acta, 176, 198–205. https://doi.org/10.1016/j.gca.2015.12.016 Baker E. B. (2015). Carbon and Oxygen Isotope Fractionation in Laboratory-Precipitated, Inor- ganic Calcite. (Master’s thesis). Eugene, OR: University of Oregon. Bar-Matthews, M., Ayalon, A., & Kaufman, A. (1997). Late Quaternary Paleoclimate in the East- ern Mediterranean Region from Stable Isotope Analysis of Speleothems at Soreq Cave, Israel. Quaternary Research, 47 (2), 155-168. https://doi.org/10.1006/qres.1997.1883 Beck, W. C., Grossman, E. L., & Morse, J. W. (2005). Experimental studies of oxygen isotope fractionation in the carbonic acid system at 15°, 25°, and 40°C. Geochimica et Cosmochimica Acta, 69 (14), 3493–3503. https://doi.org/10.1016/j.gca.2005.02.003 285 Bottinga, Y. (1968). Calculation of Fractionation Factors for Carbon and Oxygen Isotopic Ex- change in the System Calcite-Carbon Dioxide-Water. Journal of Physical Chemistry, 72 (3), 800–807. https://doi.org/10.1021/j100849a008 Charlton, S. R., & Parkhurst, D. L. (2011). Modules based on the geochemical model PHREEQC for use in scripting and programming languages. Computers and Geosciences, 37 (10), 1653–1663. https://doi.org/10.1016/j.cageo.2011.02.005 Chen, S., Gagnon, A. C., & Adkins, J. F. (2018). Carbonic anhydrase, coral calcification and a new model of stable isotope vital effects. Geochimica et Cosmochimica Acta, 236, 179–197. https://doi.org/10.1016/j.gca.2018.02.032 Coplen, T. B. (2007). Calibration of the calcite-water oxygen-isotope geothermometer at Devils Hole, Nevada, a natural laboratory. Geochimica et Cosmochimica Acta, 71 (16), 3948–3957. https://doi.org/10.1016/j.gca.2007.05.028 Daëron, M., Drysdale, R. N., Peral, M., Huyghe, D., Blamart, D., Coplen, T. B., Lartaud, F., & Zanchetta, G. (2019). Most Earth-surface calcites precipitate out of isotopic equilibrium. Nature Communications, 10 (1), 1–7. https://doi.org/10.1038/s41467-019-08336-5 Darkins, R., Kim, Y. Y., Green, D. C., Broad, A., Duffy, D. M., Meldrum, F. C., & Ford, I. J. (2022). Calcite Kinetics for Spiral Growth and Two-Dimensional Nucleation. Crystal Growth and Design, 22 (7), 4431–4436. https://doi.org/10.1021/acs.cgd.2c00378 De Lucia, M., & Kühn, M. (2013). Coupling R and PHREEQC: efficient programming of geochem- ical models. Energy Procedia 40, 464–471. https://doi.org/10.1016/j.egypro.2013.08.053 Devriendt, L. S., Watkins, J. M., & McGregor, H. V. (2017). Oxygen isotope fractionation in the CaCO3-DIC-H2O system. Geochimica et Cosmochimica Acta, 214, 115–142. https://doi.org/10.1016/j.gca.2017.06.022 Dietzel, M., Tang, J., Leis, A., & Köhler, S. J. (2009). Oxygen isotopic fractionation during in- organic calcite precipitation - Effects of temperature, precipitation rate and pH. Chemical Geology, 268 (1-2), 107–115. https://doi.org/10.1016/j.chemgeo.2009.07.015 Elderfield, H., & Ganssen, G. (2000). Past temperature and δ18O of surface ocean waters inferred from foraminiferal Mg/Ca ratios. Nature, 405 (6785), 442-445. https://doi.org/10.1038/35013033 Feng, W., Casteel, R. C., Banner, J. L., & Heinze-Fry, A. (2014). Oxygen isotope variations in rainfall, drip-water and speleothem calcite from a well-ventilated cave in Texas, USA: Assess- ing a new speleothem temperature proxy. Geochimica et Cosmochimica Acta, 127, 233-250. https://doi.org/10.1016/j.gca.2013.11.039 Gabitov, R. I., Watson, E. B., & Sadekov, A. (2012). Oxygen isotope fractionation between calcite and fluid as a function of growth rate and temperature: An in situ study. Chemical Geology, 306–307, 92–102. https://doi.org/10.1016/j.chemgeo.2012.02.021 286 Gascoyne, M. (1992). Palaeoclimate determination from cave calcite deposits. Quaternary Science Reviews, 11 (6), 609-632. https://doi.org/10.1016/0277-3791(92)90074-I Gratz, A. J., Hillner, P. E., & Hansma, P. K. (1993). Step dynamics and spiral growth on calcite. Geochimica et Cosmochimica Acta, 57 (2), 491–495. https://doi.org/10.1016/0016- 7037(93)90449-7 Grotzinger, J. P., & Kasting, J. F. (1993). New constraints on Precambrian ocean composition. Journal of Geology, 101 (2), 235–243. https://doi.org/10.1086/648218 Halevy, I., & Bachan, A. (2017). The geologic history of seawater pH. Science, 355 (6329), 1069–1071. https://doi.org/10.1126/science.aal4151 Hong, M., & Teng, H. H. (2014). Implications of solution chemistry effects: Direction-specific restraints on the step kinetics of calcite growth. Geochimica et Cosmochimica Acta, 141, 228–239. https://doi.org/10.1016/j.gca.2014.06.023 Jacobson, R. L., & Langmuir, D. (1974). Dissociation constants of calcite and CaHCO+3 from 0 to 50°C. Geochimica et Cosmochimica Acta, 38 (2), 301–318. https://doi.org/10.1016/0016- 7037(74)90112-4 Kempe, S., & Degens, E. T. (1985). An early soda ocean? Chemical Geology, 53 (1–2), 95–108. https://doi.org/10.1016/0009-2541(85)90023-3 Kim, S.-T., & O’Neil, J. R. (1997). Equilibrium and nonequilibrium oxygen isotope effects in syn- thetic carbonates. Geochimica et Cosmochimica Acta, 61 (16), 3461–3475. https://doi.org/10.1016/S0016-7037(97)00169-5 Lacey, J. H., Leng, M. J., Peckover, E. N., Dean, J. R., Wilke, T., Francke, A., Zhang, X., Masi, A., & Wagner, B. (2018). Investigating the environmental interpretation of oxygen and carbon isotope data from whole and fragmented bivalve shells. Quaternary Science Reviews, 194, 55-61. https://doi.org/10.1016/j.quascirev.2018.06.025 Larsen, K., Bechgaard, K., & Stipp, S. L. S. (2010). The effect of the Ca2+ to CO2−3 activity ratio on spiral growth at the calcite 1014 surface. Geochimica et Cosmochimica Acta, 74, 2099–2109. https://doi.org/10.1016/j.gca.2009.12.028 Leng, M. J., & Marshall, J. D. (2004). Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quaternary Science Reviews, 23 (7-8), 811-831. https://doi.org/10.1016/j.quascirev.2003.06.012 Leng, M., Barnker, P., Greenwood, P., Roberts, N., & Reed, J. (2001). Oxygen isotope analysis of diatom silica and authigenic calcite from Lake Pinarbasi, Turkey. Journal of Paleolimnology, 25, 343-349. https://doi.org/10.1023/A:1011169832093 287 Levitt, N. P., Eiler, J. M., Romanek, C. S., Beard, B. L., Xu, H., & Johnson, C. M. (2018). Near Equilibrium 13C-18O Bonding During Inorganic Calcite Precipitation Under Chemo-Stat Con- ditions. Geochemistry, Geophysics, Geosystems, 19 (3), 901–920. https://doi.org/10.1002/2017GC007089 McCrea, J. M. (1950). On the Isotopic Chemistry of Carbonates and a Paleotemperature Scale. Journal of Chemical Physics 18 (6), 849–857. https://doi.org/10.1063/1.1747785 Millero, F., Huang, F., Graham, T., & Pierrot, D. (2007). The dissociation of carbonic acid in NaCl solutions as a function of concentration and temperature. Geochimica et Cosmochimica Acta 71 (1), 46–55. https://doi.org/10.1016/j.gca.2006.08.041 Nielsen, L. C., De Yoreo, J. J., & DePaolo, D. J. (2013). General model for calcite growth ki- netics in the presence of impurity ions. Geochimica et Cosmochimica Acta, 115, 100–114. https://doi.org/10.1016/j.gca.2013.04.001 Olsen, E. K., Watkins, J. M., & Devriendt, L. S. (2022). Oxygen isotopes of calcite precipitated at high ionic strength: CaCO3-DIC fractionation and carbonic anhydrase inhibition. Geochimica et Cosmochimica Acta, 325, 170–186. https://doi.org/10.1016/j.gca.2022.01.028 Paquette, J., & Reeder, R. J. (1995). Relationship between surface structure, growth mechanism, and trace element incorporation in calcite. Geochimica et Cosmochimica Acta, 59 (4), 735–749. https://doi.org/10.1016/0016-7037(95)00004-J Perdikouri, C., Putnis, C. V., Kasioptas, A., & Putnis, A. (2009). An atomic force microscopy study of the growth of a calcite surface as a function of calcium/total carbonate concentration ratio in solution at constant supersaturation. Crystal Growth and Design 9 (10), 4344–4350. https://doi.org/10.1021/cg900200s Plummer, L. N., & Busenberg, E. (1982). The solubilities of calcite, aragonite and vaterite in CO2- H2O solutions between 0 and 90°C, and an evaluation of the aqueous model for the system CaCO3-CO2-H2O. Geochimica et Cosmochimica Acta, 46 (6), 1011-1040. https://doi.org/10.1016/0016-7037(82)90056-4 Romanek, C. S., Grossman, E. L., & Morse, J. W. (1992). Carbon isotopic fractionation in syn- thetic aragonite and calcite: Effects of temperature and precipitation rate. Geochimica et Cosmochimica Acta, 56 (1), 419–430. https://doi.org/10.1016/0016-7037(92)90142-6 Ruiz-Agudo, E., & Putnis, C. V. (2012). Direct observations of mineral fluid reactions using atomic force microscopy: the specific example of calcite. Mineralogical Magazine, 76 (1), 227–253. https://doi.org/10.1180/minmag.2012.076.1.227 Ruiz-Agudo, E., Putnis, C. V., Rodriguez-Navarro, C., & Putnis, A. (2011). Effect of pH on cal- cite growth at constant aCa2+/aCO2− ratio and supersaturation. Geochimica et Cosmochimica3 Acta 75 (1), 284–296. https://doi.org/10.1016/j.gca.2010.09.034 288 Sade, Z., Yam, R., Shemesh, A., & Halevy, I. (2020). Kinetic fractionation of carbon and oxy- gen isotopes during BaCO3 precipitation. Geochimica et Cosmochimica Acta, 280, 395–422. https://doi.org/10.1016/j.gca.2020.04.025 Sand, K. K., Tobler, D. J., Dobberschütz, S., Larsen, K. K., Makovicky, E., Andersson, M. P., Wolthers, M., & Stipp, S. L. S. (2016). Calcite Growth Kinetics: Dependence on Saturation Index, Ca2+:CO2−3 Activity Ratio, and Surface Atomic Structure. Crystal Growth and Design, 16 (7), 3602–3612. https://doi.org/10.1021/acs.cgd.5b01792 Spero, H. J., Bijma, J., Lea, D. W., & Bernis, B. E. (1997). Effect of seawater carbonate concen- tration on foraminiferal carbon and oxygen isotopes. Nature 390 (6659), 497–500. https://doi.org/10.1038/37333 Strassen, P., Dupuis, C., Morsi, A. M., Steurbaut, E., & Speijer, R. P. (2009). Reconstruction of a latest Paleocene shallow-marine eutrophic paleoenvironment at Sidi Nasseur (Central Tunisia) based on foraminifera, ostracoda, calcareous nannofossils and stable isotopes (δ13C, δ18O). Ge- ologica Acta, 7 (1), 93-112. https://doi.org/10.1344/105.000000273 Teng, H. H., Dove, P. M., & De Yoreo, J. J. (2000). Kinetics of calcite growth: Surface pro- cesses and relationships to macroscopic rate laws. Geochimica et Cosmochimica Acta, 64 (13), 2255–2266. https://doi.org/10.1016/S0016-7037(00)00341-0 Uchikawa, J., Chen, S., Eiler, J. M., Adkins, J. F., & Zeebe, R. E. (2021). Trajectory and timescale of oxygen and clumped isotope equilibration in the dissolved carbonate system under nor- mal and enzymatically-catalyzed conditions at 25 °C. Geochimica et Cosmochimica Acta, 314, 313–333. https://doi.org/10.1016/j.gca.2021.08.014 Uchikawa, J., & Zeebe, R. E. (2012). The effect of carbonic anhydrase on the kinetics and equi- librium of the oxygen isotope exchange in the CO2-H2O system: Implications for δ 18O vital effects in biogenic carbonates. Geochimica et Cosmochimica Acta, 95, 15–34. https://doi.org/10.1016/j.gca.2012.07.022 Urey, H. C. (1947). The Thermodynamic Properties of Isotopic Substances. Journal of the Chem- ical Society, 562–581. https://doi.org/10.1039/JR9470000562 van Geldern, R., Joachimski, M. M., Day, J., Jansen, U., Alvarez, F., Yolkin, E. A., & Ma, X. P. (2006). Carbon, oxygen and strontium isotope records of Devonian brachiopod shell calcite. Palaeogeography, Palaeoclimatology, Palaeoecology, 240 (1-2), 47-67. https://doi.org/10.1016/j.palaeo.2006.03.045 Watkins, J. M., & Devriendt, L. S. (2022). A combined model for kinetic clumped isotope effects in the CaCO3-DIC-H2O system. Geochemistry, Geophysics, Geosystems, 23 (8), e2021GC010200. https://doi.org/10.1029/2021GC010200 Watkins, J. M., Hunt, J. D., Ryerson, F. J., & DePaolo, D. J. (2014). The influence of temperature, pH, and growth rate on the δ18O composition of inorganically precipitated calcite. Earth and Planetary Science Letters, 404, 332–343. https://doi.org/10.1016/j.epsl.2014.07.036 289 Watkins, J. M., Nielsen, L. C., Ryerson, F. J., & DePaolo, D. J. (2013). The influence of kinetics on the oxygen isotope composition of calcium carbonate. Earth and Planetary Science Letters, 375, 1–12. https://doi.org/10.1016/j.epsl.2013.05.054 Wolthers, M., Charlet, L., & Van Cappellen, P. (2008). The surface chemistry of divalent metal carbonate minerals: A critical assessment of surface charge and potential data using the charge distribution multi-site ion complexation model. American Journal of Science, 308 (8), 905–941. https://doi.org/10.2475/08.2008.02 Wolthers, M., Nehrke, G., Gustafsson, J. P., & Van Cappellen, P. (2012). Calcite growth kinetics: Modeling the effect of solution stoichiometry. Geochimica et Cosmochimica Acta, 77, 121–134. https://doi.org/10.1016/j.gca.2011.11.003 Zachos, J., Pagani, H., Sloan, L., Thomas, E., & Billups, K. (2001). Trends, rhythms, and aberra- tions in global climate 65 Ma to present. Science, 292 (5517), 686–693. https://doi.org/10.1126/science.1059412 Zeebe, R. E. (1999). An explanation of the effect of seawater carbonate concentration on foraminiferal oxygen isotopes. Geochimica et Cosmochimica Acta, 63 (13-14), 2001–2007. https://doi.org/10.1016/S0016-7037(99)00091-5 Zeebe, R. E. (2007). An expression for the overall oxygen isotope fractionation between the sum of dissolved inorganic carbon and water. Geochemistry, Geophysics, Geosystems, 8 (9), 1–7. https://doi.org/10.1029/2007GC001663 Zhang, J., & Nancollas, G. H. (1998). Kink density and rate of step movement during growth and dissolution of an AB crystal in a nonstoichiometric solution. Journal of Colloid and Interface Science, 200 (1), 131–145. https://doi.org/10.1006/jcis.1997.5357 Zhang, J., Quay, P. D., & Wilbur, D. O. (1995). Carbon isotope fractionation during gas-water exchange and dissolution of CO2. Geochimica et Cosmochimica Acta, 59 (1), 107–114. https://doi.org/10.1016/0016-7037(95)91550-D Chapter V Bajnai, D., & Herwartz, D. (2021). Kinetic Oxygen Isotope Fractionation between Water and Aque- ous OH− during Hydroxylation of CO2. ACS Earth and Space Chemistry, 5 (12), 3375–3384. https://doi.org/10.1021/acsearthspacechem.1c00194 Barkan, E., & Luz, B. (2012). High-precision measurements of 17O/16O and 18O/16O ratios in CO2. Rapid Communications in Mass Spectrometry, 26 (23), 2733–2738. https://doi.org/10.1002/rcm.6400 Beck, W. C., Grossman, E. L., & Morse, J. W. (2005). Experimental studies of oxygen isotope fractionation in the carbonic acid system at 15°, 25°, and 40°C. Geochimica et Cosmochimica Acta, 69 (14), 3493–3503. https://doi.org/10.1016/j.gca.2005.02.003 290 Böttcher, M. E., Neubert, N., Escher, P., von Allmen, K., Samankassou, E., & Nägler, T. F. (2018). Multi-isotope (Ba, C, O) partitioning during experimental carbonatization of a hyper-alkaline solution. Chemie der Erde, 78 (2), 241–247. https://doi.org/10.1016/j.chemer.2018.01.001 Charlton, S. R., & Parkhurst, D. L. (2011). Modules based on the geochemical model PHREEQC for use in scripting and programming languages. Computers and Geosciences, 37 (10), 1653–1663. https://doi.org/10.1016/j.cageo.2011.02.005 Chavagnac, V., Ceuleneer, G., Monnin, C., Lansac, B., Hoareau, G., & Boulart, C. (2013). Min- eralogical assemblages forming at hyperalkaline warm springs hosted on ultramafic rocks: A case study of Oman and Ligurian ophiolites. Geochemistry, Geophysics, Geosystems, 14 (7), 2474–2495. https://doi.org/10.1002/ggge.20146 Christensen, J. N., Watkins, J. M., Devriendt, L. S., DePaolo, D. J., Conrad, M. E., Voltolini, M., Yang, W., & Dong, W. (2021). Isotopic fractionation accompanying CO2 hydroxylation and carbonate precipitation from high pH waters at The Cedars, California, USA. Geochimica et Cosmochimica Acta, 301, 91–115. https://doi.org/10.1016/j.gca.2021.01.003 Christensen, J. N., Watkins, J. M., Devriendt, L. S., DePaolo, D. J., Conrad, M. E., Voltolini, M., Yang, W., & Dong, W. (2023). Corrigendum to “Isotopic fractionation accompanying CO2 hydroxylation and carbonate precipitation from high pH waters at the Cedars, California, USA” [Geochim. Cosmochim. Acta 301 (2021) 91–115], Geochimica et Cosmochimica Acta, 343, 416-419. https://doi.org/10.1016/j.gca.2022.09.022 Clark, I. D., Fontes, J. C., & Fritz, P. (1992). Stable isotope disequilibria in travertine from high pH waters: Laboratory investigations and field observations from Oman. Geochimica et Cos- mochimica Acta, 56 (5), 2041–2050. https://doi.org/10.1016/0016-7037(92)90328-G Coplen, T. B. (2007). Calibration of the calcite-water oxygen-isotope geothermometer at Devils Hole, Nevada, a natural laboratory. Geochimica et Cosmochimica Acta, 71 (16), 3948–3957. https://doi.org/10.1016/j.gca.2007.05.028 De Lucia, M., & Kühn, M. (2013). Coupling R and PHREEQC: efficient programming of geochem- ical models. Energy Procedia 40, 464–471. https://doi.org/10.1016/j.egypro.2013.08.053 Devriendt, L. S., Watkins, J. M., & McGregor, H. V. (2017). Oxygen isotope fractionation in the CaCO3-DIC-H2O system. Geochimica et Cosmochimica Acta, 214, 115–142. https://doi.org/10.1016/j.gca.2017.06.022 Dietzel, M., Usdowski, E., & Hoefs, J. (1992). Chemical and 13C/12C-and 18O/16O-isotope evolu- tion of alkaline drainage waters and the precipitation of calcite. Applied Geochemistry, 7 (2), 177–184. https://doi.org/10.1016/0883-2927(92)90035-2 Eigen, M. (1964). Proton Transfer, Acid-Base Catalysis, and Enzymatic Hydrolysis. Part I: ELE- MENTARY PROCESSES. Angewandte Chemie International Edition in English, 3 (1), 1–19. https://doi.org/10.1002/anie.196400011 291 Epstein, S., & Zeiri, L. (2023). Oxygen and Carbon Isotopic Compositions of Gases Respired by Humans. Proceedings of the National Academy of Sciences of the United States of America, 85 (6), 1727–1731. https://doi.org/https://doi.org/10.1073/pnas.85.6.1727 Falk, E. S., Guo, W., Paukert, A. N., Matter, J. M., Mervine, E. M., & Kelemen, P. B. (2016). Controls on the stable isotope compositions of travertine from hyperalkaline springs in Oman: Insights from clumped isotope measurements. Geochimica et Cosmochimica Acta, 192, 1–28. https://doi.org/10.1016/j.gca.2016.06.026 Flude, S., Györe, D., Stuart, F. M., Zurakowska, M., Boyce, A. J., Haszeldine, R. S., Chalaturnyk, R., & Gilfillan, S. M. V. (2017). The inherent tracer fingerprint of captured CO2. Interna- tional Journal of Greenhouse Gas Control, 65 (August), 40–54. https://doi.org/10.1016/j.ijggc.2017.08.010 Folk, R. L. (1994). Interaction between bacteria, nannobacteria, and mineral precipitation in hot springs of Central Italy. Géographie Physique et Quaternaire, 48 (3), 233–246. Giampouras, M., Garrido, C. J., Zwicker, J., Vadillo, I., Smrzka, D., Bach, W., Peckmann, J., Jiménez, P., Benavente, J., & Garćıa-Ruiz, J. M. (2019). Geochemistry and mineralogy of serpentinization-driven hyperalkaline springs in the Ronda peridotites. Lithos, 350–351, 105215. https://doi.org/10.1016/j.lithos.2019.105215 Green, M., & Taube, H. (1963). Isotopic fractionation in the OH−-H2O exchange reaction. Journal of Physical Chemistry, 67 (7), 1565–1566. https://doi.org/10.1021/j100801a507 Jacobson, R. L., & Langmuir, D. (1974). Dissociation constants of calcite and CaHCO+3 from 0 to 50°C. Geochimica et Cosmochimica Acta, 38 (2), 301–318. https://doi.org/10.1016/0016- 7037(74)90112-4 Kelemen, P. B., & Matter, J. (2008). In situ carbonation of peridotite for CO2 storage. Proceedings of the National Academy of Sciences of the United States of America, 105 (45), 17295–17300. https://doi.org/10.1073/pnas.0805794105 Kelly, K. E., Silcox, G. D., Sarofim, A. F., & Pershing, D. W. (2011). An evaluation of ex situ, industrial-scale, aqueous CO2 mineralization. International Journal of Greenhouse Gas Con- trol, 5 (6), 1587–1595. https://doi.org/10.1016/j.ijggc.2011.09.005 Kemache, N., Pasquier, L. C., Mouedhen, I., Cecchi, E., Blais, J. F., & Mercier, G. (2016). Aqueous mineral carbonation of serpentinite on a pilot scale: The effect of liquid recircu- lation on CO2 sequestration and carbonate precipitation. Applied Geochemistry, 67, 21–29. https://doi.org/10.1016/j.apgeochem.2016.02.003 Kim, S.-T., Hillaire-Marcel, C., & Mucci, A. (2006). Mechanisms of equilibrium and kinetic oxygen isotope effects in synthetic aragonite at 25 °C. Geochimica et Cosmochimica Acta, 70 (23), 5790–5801. https://doi.org/10.1016/j.gca.2006.08.003 292 Kitano, Y. (1962). The behavior of various inorganic ions in the separation of calcium carbonate from a bicarbonate solution, Bulletin of the Chemical Society of Japan 35 (12), 1973–1980. https://doi.org/10.1246/bcsj.35.1973 Mervine, E. M., Humphris, S. E., Sims, K. W. W., Kelemen, P. B., & Jenkins, W. J. (2014). Carbonation rates of peridotite in the Samail Ophiolite, Sultanate of Oman, constrained through 14C dating and stable isotopes. Geochimica et Cosmochimica Acta, 126, 371–397. https://doi.org/10.1016/j.gca.2013.11.007 Mook, W. (1986). 13C in atmospheric CO2. Netherlands Journal of Sea Research, 20 (2-3), 211-223. https://doi.org/10.1016/0077-7579(86)90043-8 Morrill, P. L., Kuenen, J. G., Johnson, O. J., Suzuki, S., Rietze, A., Sessions, A. L., Fogel, M. L., & Nealson, K. H. (2013). Geochemistry and geobiology of a present-day serpentinization site in California: The Cedars. Geochimica et Cosmochimica Acta, 109, 222–240. https://doi.org/10.1016/j.gca.2013.01.043 Olsen, E. K., Watkins, J. M., & Devriendt, L. S. (2022). Oxygen isotopes of calcite precipitated at high ionic strength: CaCO3-DIC fractionation and carbonic anhydrase inhibition. Geochimica et Cosmochimica Acta, 325, 170–186. https://doi.org/10.1016/j.gca.2022.01.028 Palandri, J. L., & Reed, M. H. (2004). Geochemical models of metasomatism in ultramafic systems: Serpentinization, rodingitization, and sea floor carbonate chimney precipitation. Geochimica et Cosmochimica Acta, 68 (5), 1115–1133. https://doi.org/10.1016/j.gca.2003.08.006 Pinsent, B. R. W., Pearson, L., & Roughton, F. J. W. (1956). The kinetics of combination of carbon dioxide with ammonia. Transactions of the Faraday Society, 52, 1594–1598. https://doi.org/10.1039/tf9565201512 Sade, Z., & Halevy, I. (2017). New constraints on kinetic isotope effects during CO2(aq) hydration and hydroxylation: Revisiting theoretical and experimental data. Geochimica et Cosmochim- ica Acta, 214, 246–265. https://doi.org/10.1016/j.gca.2017.07.035 Sade, Z., Yam, R., Shemesh, A., & Halevy, I. (2020). Kinetic fractionation of carbon and oxy- gen isotopes during BaCO3 precipitation. Geochimica et Cosmochimica Acta, 280, 395–422. https://doi.org/10.1016/j.gca.2020.04.025 Srivastava, A., & Verkouteren, R. M. (2018). Metrology for stable isotope reference materials: 13C/12C and 18O/16O isotope ratio value assignment of pure carbon dioxide gas samples on the Vienna PeeDee Belmnite-CO2 scale using dual-inlet mass spectrometry. Analytical and Bioanalytical Chemistry, 410, 4153–4163. https://doi.org/10.1007/s00216-019-01740-2 Swart, P. K., Burns, S. J., & Leder, J. J. (1991). Fractionation of the stable isotopes of oxygen and carbon in carbon dioxide during the reaction of calcite with phosphoric acid as a function of temperature and technique. Chemical Geology: Isotope Geoscience Section, 86 (2), 89–96. https://doi.org/10.1016/0168-9622(91)90055-2 293 Trolier, M., White, J. W. C., Tans, P. P., Masarie, K. A., & Gemery, P. A. (1996). Monitor- ing the isotopic composition of atmospheric CO2: Measurements from the NOAA global air sampling network. Journal of Geophysical Research Atmospheres 101 (20), 25897–25916. https://doi.org/10.1029/96jd02363 Usdowski, E., & Hoefs, J. (1986). 13C/12C partitioning and kinetics of CO2 absorption by hydrox- ide buffer solutions. Earth and Planetary Science Letters, 80 (1-2), 130–134. https://doi.org/10.1016/0012-821X(86)90025-7 Vogel, J. C., Grootes, P. M., & Mook, W. G. (1970). Isotopic fractionation between gaseous and dissolved carbon dioxide. Zeitschrift Für Physik, 230 (3), 225–238. https://doi.org/10.1007/BF01394688 Yumol, L. M., Uchikawa, J., & Zeebe, R. E. (2020). Kinetic isotope effects during CO2 hydration: Experimental results for carbon and oxygen fractionation. Geochimica et Cosmochimica Acta, 279, 189–203. https://doi.org/10.1016/j.gca.2020.03.041 Zeebe, R. E. (2020). Oxygen isotope fractionation between water and the aqueous hydroxide ion. Geochimica et Cosmochimica Acta, 289, 182–195. https://doi.org/10.1016/j.gca.2020.08.025 Zhang, J., Quay, P. D., & Wilbur, D. O. (1995). Carbon isotope fractionation during gas-water exchange and dissolution of CO2. Geochimica et Cosmochimica Acta, 59 (1), 107–114. https://doi.org/10.1016/0016-7037(95)91550-D 294