Phosphate reactions with natural allophane, ferrihydrite and goethite

The reactions of phosphate with natural samples of allophane, ferrihydrite, hematite and goethite were measured for up to 30 d. The amount of phosphate sorbed on allophane showed the biggest increase with time whereas the amount sorbed on goethite showed the least increase with time. The total amount of phosphate sorbed either at high levels of phosphate addition or after 10 d followed the order hematite < goethite < ferrihydrite < allophane and was probably related to the specific surface. Si was desorbed as phosphate was adsorbed on the minerals.
The reactions of phosphate on allophane involved rapid, strong adsorption, probably at defect sites, followed by weaker adsorption, followed, probably, by disruption of the allophane structure together with precipitation of aluminium phosphates. Previous suggestions either of diffusive penetration of phosphate into surfaces or about the formation of aluminium phosphate coatings, are unlikely to hold for allophane, if all the Al is at the surface and if the structure can be ruptured.
The reactions of phosphate with iron oxides involved a rapid, strong ligand exchange, followed by weaker ligand exchange, and, probably, by a relatively slow penetration at defect sites and pores. Highly crystalline goethite has virtually no slow reaction and therefore solid-state diffusion of phosphate does not readily occur. The extent of phosphate uptake during the slow penetration reactions probably depends on the degree of crystallinity or porosity of iron oxides.
The most reactive adsorbents, such as allophane, ferrihydrite and Al-humus complexes do not have planar surfaces, and this needs to be considered when modelling phosphate reactions.