Examination of crystal dissolution in 3D: A way to reconcile dissolution rates in the laboratory?

Abstract

Surface reactivity is a major parameter controlling mineral reactivity, and microscopic techniques investigating surface retreat with time have pointed at the heterogeneous and/or anisotropic reactivity of minerals, in relation with the diversity and stochastic distribution of energetic sites. However, in view of the discrepancies between rates determined in the laboratory, a thorough 3D approach of crystal reactivity might be particularly attractive to evaluate the respective contributions of single faces and crystal edges to the dissolution flux, and to fill the gap between the rates derived from face-specific, topography observations at micro-scale (i.e., with no contribution of the edges to dissolution) and those determined on crystal powders in continuously stirred reactors (with an overcontribution of the edges and surface defects to dissolution). Here, we provide a detailed 3D characterization of the geometry evolution and dissolution rate of a single crystal of calcite at pH 4.5 and 4.0 using X-ray micro-tomography (XMT) with a pixel size of 0.325 µm. Evaluation of the retreat and mapping of the reaction rates at the 3D crystal surface reveal a large range of dissolution rates reflecting the specific contributions of the different regions of the crystal. During dissolution and against all expectation, etch pits forming at the crystal surface progressively annihilate, primarily by intersecting with trains of steps coming from the near edge regions. The global rate determined at the crystal scale integrates the contribution of the local rates of all the crystal features, with r¯corner′ > r¯edge′ > r¯cleavage′ > r¯macrostep′ ∼ r¯pit′ > r¯macrostepbase′. Crystal rounding reveals that contribution from the crystal edges progressively dominates the dissolution process over pit formation at the {101¯4} surfaces. The contribution of the edges to dissolution increases the crystal dissolution rate by at least 1.6 to what would be a face-specific dissolution, and will be size- and time-dependent, as suggested by a simple geometric model based on uniform or non-uniform dissolution of the faces of a model crystal. Finally, comparison of the method to vertical scanning interferometry measurements and scanning electron microscopy observations on surface portions shows that XMT imaging is robust, suggesting that its application to the dissolution/precipitation of other minerals would be highly beneficial to determine reliable rates that can be further used to model mineral reactivity.