Allophane is a very fine-grained clay mineral which is especially common in Andosols. Its importance in soils derives from its large reactive surface area. Owing to its short-range order, allophane cannot be quantified by powder X-ray diffraction (XRD) directly. It is commonly dissolved from the soil by applying extraction methods. In the present study the standard extraction method (oxalate) was judged to be unsuitable for the quantification of allophane in a soil/clay deposit from Ecuador, probably because of the large allophane content (>60 wt.%). This standard extraction method systematically underestimated the allophane content but the weakness was less pronounced in samples with small allophane contents. In the case of allophane-rich materials, the Rietveld XRD technique, using an internal standard to determine the sum of X-ray amorphous phases, is recommended if appropriate structural models are available for the other phases present in the sample. The allophane (+imogolite) content is measured by subtracting the amount of oxalate-soluble phases (e.g. ferrihydrite). No correction would be required if oxalate-soluble Fe were incorporated in the allophane structure. The present study, however, provides no evidence for this hypothesis. Mössbauer and scanning electron microscopy investigations indicate that goethite and poorly ordered hematite are the dominant Fe minerals and occur as very fine grains (or coatings) being dispersed in the cloud-like allophane aggregates.
Allophane is known to adsorb appreciable amounts of water, depending on ambient conditions. The mass fraction of the sample attributed to this mineral thus changes accordingly; the choice of a reference hydration state is, therefore, a fundamental factor in the quantification of allophane in a sample. Results from the present study revealed that (1) drying at 105°C produced a suitable reference state, and (2) water adsorption has no effect on quantification by XRD analysis.