Kinetic isotope and trace element partitioning during calcite precipitation from aqueous solution.

摘要

Precipitation of carbonate minerals is ubiquitous in the near-surface environment, and the isotopic and trace element composition of carbonates may be used to reconstruct the conditions of growth. Little is known about the mechanisms controlling isotope and trace element distribution into carbonates. Proposed growth mechanisms are typically inferred from the supersaturation dependence of CaCO3 precipitation rates. As different experimental techniques often generate different apparent reaction orders, numerous hypothesized mechanisms can be found in the literature. These descriptions of crystallization pathway cannot be used to identify processes controlling trace element and isotope partitioning.;Recent advances in experimental methodology allow us to observe microscopic to nanoscopic structures at the mineral surface during growth. The mechanisms controlling calcite growth have been directly determined by using fluid flow cells placed in an atomic force microscope (AFM; chapter 3). Calcite growth below the threshold oversaturation for amorphous calcium carbonate (ACC) formation---typical of seawater and most terrestrial fluids---occurs primarily through the attachment of ions to kink sites on the surface. The net flux of ions to kink sites governs both overall growth rate and mineral composition.;In this thesis, I present a self-consistent model based on observed calcite growth mechanisms that may be used to predict growth rate (chapter 1), and isotopic (chapter 2; Nielsen et al., 2012) and trace element (chapter 5; Nielsen et al., in prep.) partitioning as a function of solution composition. I apply this model to calcium isotope fractionation during the precipitation of synthetic calcite, which I grew and analyzed using novel secondary ion mass spectrometry (SIMS) methods (chapter 3). In chapter 4 (Nielsen and DePaolo, in review), I model calcium isotope fractionation in carbonate minerals that I collected from the highly alkaline Mono Lake, CA. The mechanistic framework developed here may be extended to multicomponent systems, and may be adapted for use in reactive transport models. When interpreted through the lens of this model, trace element and isotope signatures preserved in carbonates may eventually be used to reconstruct the chemistry of natural aqueous fluids.