THE DISSOLUTION OF NATURALLY WEATHERED FELDSPAR AND QUARTZ

Surface area measurements and dissolution experiments were performed on subsamples from a naturally weathered mineral assemblage (100-1000 mu m) consisting of feldspar and quartz. The subsamples were obtained by splitting the assemblage into four different ranges of grain density, each of which was sieved to three different size fractions. BET-krypton and geometric surface areas, combined with mineralogical data and average grain diameters, showed that (1) surface roughness factors of the subsamples are generally much higher than those of freshly created surfaces by grinding and (2) for individual density ranges (i.e., at constant mineralogical composition), the surface roughness factor decreases linearly with decreasing grain diameter. Scanning electron microscopy and X-Ray diffraction showed that contributions to the surface roughness factors from secondary mineral coatings, macropores (diameters >50 nm), and etch pits are insignificant. In contrast, krypton adsorption data indicated that by far most surface roughness is due to the presence of micropores and mesopores (diameters 150 nm). These findings strongly suggest that, during natural weathering, micropores/mesopores develop at sites whose density (cm(-2) of geometric surface area) is approximately proportional to grain diameter. Multivariate linear regression showed that, at similar grain diameters, the micropore/mesopore density increases in the order: quartz < microcline < albite < oligoclase/andesine. This sequence is similar to the well-known sequence of relative weatherability of these minerals, suggesting a relationship between weatherability and micropore/mesopore density. At pH 3 HCl and ambient temperature, dissolution rates of Na, K, Ca, Al, and Si from the subsamples, normalized to the BET-krypton surface area, were essentially independent of the grain diameter. Due to effects from surface roughness, dissolution rates normalized to the geometric surface area were essentially proportional to grain diameter. Comparison with micropore/mesopore and etch pit densities showed that the dissolution rates are determined by the pores, rather than by the etch pits. Furthermore, theoretical arguments indicate that the pore area perpendicular to the mineral surface (i.e., the pore ''walls'') is essentially nonreactive, and that the dissolution rates are largely determined by the pore area parallel to the mineral surface (the pore ''bottoms''). The pore area parallel to the mineral surface is equivalent to (1) the leached layer/fresh mineral interface, if the micropores/mesopores develop in leached layers or (2) the dislocation outcrops where strained mineral material is in contact with the solution, if the micropores/mesopores develop at crystal defects. Tentative calculations suggest that (1) as in the laboratory, dissolution during the previous natural weathering of the sample occurred from the micropore ''bottoms'' rather than from the etch pits, and (2) dissolution from the etch pits becomes more important with increasing exposure time to weathering conditions.