The central control of mineral weathering rates on biogeochemical systems has motivated studies of dissolution for more than 50 years. A complete physical picture that explains widely observed variations in dissolution behavior is lacking, and some data show apparent serious inconsistencies that cannot be explained by the largely empirical kinetic ''laws.'' Here, we show that mineral dissolution can, in fact, be understood through the same mechanistic theory of nucleation developed for mineral growth. In principle, this theory should describe dissolution but has never been tested. By generalizing nucleation rate equations to include dissolution, we arrive at a model that predicts how quartz dissolution processes change with undersaturation from step retreat, to defect-driven and homogeneous etch pit formation. This finding reveals that the ''salt effect,'' recognized almost 100 years ago, arises from a crossover in dominant nucleation mechanism to greatly increase step density. The theory also explains the dissolution kinetics of major weathering aluminosilicates, kaolinite and K-feldspar. In doing so, it provides a sensible origin of discrepancies reported for the dependence of kaolinite dissolution and growth rates on saturation state by invoking a temperature-activated transition in the nucleation process. Although dissolution by nucleation processes was previously unknown for oxides or silicates, our mechanism-based findings are consistent with recent observations of dissolution (i.e., demineralization) in biological minerals. Nucleation theory may be the missing link to unifying mineral growth and dissolution into a mechanistic and quantitative framework across the continuum of driving force.silica ͉ kinetics ͉ mineralization O ver long time scales, the geochemistry of earth systems is, in large part, controlled by the kinetics of silicate mineral dissolution. Because waters contain a wide variety of solute types and concentrations, including significant levels of aqueous silica, there is considerable need to understand the dependence of silicate mineral dissolution rates on chemical driving force, as measured by the extent of undersaturation. This need has motivated intense investigations of both mineral weathering and the corrosion behavior of silica-based glasses.Basic thermodynamic principles predict that mineral dissolution rates should increase with increasing driving force or chemical potential; however, experimental studies of major silicate minerals show that this dependence is complex. A further complication is the so-called ''salt effect'' reported for quartz, SiO 2 , whereby the dissolution rate of this oxide end-member to all silicates is increased up to 100 times in the presence of the major cationic solutes found in natural waters (Na ϩ , K ϩ , Ca 2ϩ , Mg 2ϩ ) (1-3). In contrast, dissolution rates of silicate minerals have only a weak sensitivity to the introduction of electrolytes (4). To explain these behaviors, many of the widely used rate models are based on variants of transition state t...

Polyaspartate domains are a prominent feature of proteins associated with biogenic carbonates and have been implicated in modifying crystal morphology through specific interactions with step edges. Here, we show that the morphology and growth kinetics of calcite are modified in a systematic way when a series of poly-L-aspartates, Asp 1-6 , are introduced into solution. In-situ measurements of step propagation rates by atomic force microscopy reveal these effects are strongly chain-length dependent and specific to the crystallographically distinct, obtuse and acute step types. Direct observations of differential roughening and rounding of the step edges demonstrate that, while Asp 1 and Asp 2 have stronger effects on acute step edges, a crossover occurs for the longer Asp 4,5,6 peptides that preferentially affect obtuse steps. Independent analysis of Asp n -step edge interactions by semiempirical quantum mechanical modeling gives estimates of aspartate-step edge binding energies and predicts that the crossover should occur at n ) 2. The switch occurs because, upon Asp n binding, the energy required to dehydrate acute steps is greater than that at the obtuse steps when n ) 3-6.Step velocity measurements show that the concentration of Asp n needed to stop growth scales exponentially and inversely with the calculated binding energies. A simple model of Asp n adsorption to the steps is derived from these results. These findings suggest a process by which small fluctuations in primary structure in proteins can control calcite shape.