Does phase separation quantitatively affect Al, Fe, and Si extraction from soil using acid oxalate?

Abstract

Soil extraction with ammonium oxalate/oxalic acid is commonly used to quantify aluminium (Al), iron (Fe), and silicon (Si) in poorly crystalline oxides and silicates and other pedogenic forms. However, separation of the extract solution from the extracted soil is not specifically defined by the frequently used protocols or in most studies citing these protocols. Thus, we checked potential effects of the filter size (0.45 or 2 µm) and additional centrifugation on the amounts of Al, Fe, and Si extracted by oxalate/oxalic acid from six subsoil samples and three allophanic materials as a function of extraction time (1 to 4 h). Centrifugation prior to 2-µm filtration decreased the amounts of Fe and Si extracted with all seven variants with significant differences. With three of the four variants of Al extraction with significant differences, filtration at 0.45 µm yielded smaller amounts. Release of Al from allophanic minerals was slow, incomplete, and biased by Al re-adsorption, confirming the inappropriateness of oxalate/oxalic-acid extraction for allophane quantification. We thus recommend considering the incompleteness and time dependency of Al release from allophanic minerals for soil characterization/classification purposes, and generally using 0.45-µm filters, which correspond to the definition of separating dissolved species.