Hydration Reactions at the Mineral—Fluid Interface: Experimental and Computational Studies

Abstract

Reactions that occur at the mineral-fluid interface, such as mineral dissolution, trace Aelemeat adsorption, and secondary phase precipitation, are of particular interest to geochemists. Several new surface-sensitive analytical techniques have recently been applied to the study of these reactions. In addition, recent advances in ab initio quantum mechanical calculations have made it much easier to translate experimental results into models of the actual atomic-scale reactions that occur at surfaces. The penetration depth of protons into the feldspar near-surface environment during hydrolysis has been successfully measured using Elastic Recoil Detection Analysis (ERDA) (Casey et al., 1989). Using this method, the penetration profile of H § in minerals as a function of reaction time, temperature, pH, and crystallographic orientation can be determined. The H + distribution in the mineral near-surface can be used to: [1] constrain the thicknesses of surfaceillayers that have hydrated to form distinct new crystal structures, [2] define the depth to which cation-proton exchange reactions penetrate, and [3] construct a charge balance inventory. Surface sensitive X-ray techniques, such as Xray absorption spectroscopy (electron-yield geometry), X-ray reflectivity, and low-angle surface diffraction have also been applied to mineral surface experiments and can provide unique atomic-scale structural and physical data (Chiarello et al., 1993). Detailed information about changes in surface morphology, the nature and growth dynamics of altered or recrystallized layers, and the atomic arrangements of adsorbates can all be produced by these types of measuremeats. Unambiguous results with simple oxides such as periclase will provide a foundation to our understanding of how more complicated Mgbearing minerals react with aqueous fluids. X-ray results for periclase and ERDA results for olivine (Fogo) are presented below. Accurate periodic Hartree-Fock calculations (Scamehorn et aL, 1993) and experiments (Onishi et al., 1987) show that water physisorbs onto clean (100) periclase surfaces but that there is no dissociation or chemisorption. However a reaction takes place on exposure to water over a period of several hours to produce an opaque layer at the surface (presumably brucite). A reconstructive mechanism is necessary to explain the slow reaction rate. We observe that the atomic structure at a periclase (111) surface is isomorphic to brucite (0001) with a 5% discrepancy in lattice parameter. This suggests that protonation acts to stabilize the otherwise unstable (111) surface by forming a brucite-like layer and allows the hydration reaction to proceed by growth of brucite at (111) etch pits. Theoretical calculations will demonstrate this stabilization.