Variation in element release rate from different mineral size fractions from the B horizon of a granitic podzol

Abstract

Far-from-equilibrium batch dissolution experiments were carried out on the 2000–500, 500–250, 250–53 and 53–2 μm size fractions of the mineral component of the B horizon of a granitic iron humus podzol after removal of organic matter and secondary precipitates. The different size fractions were mineralogically and chemically similar, the main minerals present being quartz, alkali and plagioclase feldspar, biotite and chlorite. Specific surface area increased with decreasing grain size. The measured element release rates decreased in the order 53–2⋙2000–500>500–250>250–53 μm. Surface area normalised element release rates from the 2000–500, 500–250 and 250–53 μm size fractions (0.6–77×10 −14 mol/m 2/s) were intermediate between literature reported surface area normalised dissolution rates for monomineralic powders of feldspar (0.1–0.01×10 −14 mol/m 2/s) and sheet silicates (100×10 −14 mol/m 2/s) dissolving under similar conditions. Element release rates from the 53–2 μm fraction (400–3000×10 −14 mol/m 2/s) were a factor of 4–30 larger than literature reported values for sheet silicates. The large element release rate of the 53–2 μm fraction means that, despite the small mass fraction of 53–2 μm sized particles present in the soil, dissolution of this fraction is the most important for element release into the soil. A theoretical model predicted similar (within a factor of <2) bulk element release rates for all the mineral powders if observed thicknesses of sheet silicate grains were used as input parameters. Decreasing element release rates with decreasing grain size were only predicted if the thickness of sheet silicates in the powders was held constant. A significantly larger release rate for the 53–2 μm fraction relative to the other size fractions was only predicted if either surface roughness was set several orders of magnitude higher for sheet silicates and several orders of magnitude lower for quartz and feldspars in the 53–2 μm fraction compared to the other size fractions or if the sheet silicate thickness input in the 53–2 μm fraction was set unrealistically low. It is therefore hypothesised that the reason for the unpredicted large release rate from the 52–3 μm size fraction is due to one or more of the following reasons: (1) the greater reactivity of the smaller particles due to surface free energy effects, (2) the lack of proportionality between the BET surface area used to normalise the release rates and the actual reactive surface area of the grains and, (3) the presence of traces quantities of reactive minerals which were undetected in the 53–2 μm fraction but were entirely absent in the coarser fractions.