Transformation of haematite and Al-poor goethite to Al-rich goethite and associated yellowing in a ferralitic clay soil profile of the middle Amazon Basin (Manaus, Brazil)

The red and yellow colours of ferralitic soils in the tropics have for long intrigued pedologists. We have investigated the upward yellowing in a 10-m thick profile representative of the Ferralsols of the plateaux of the Manaus region of Brazil. We determined changes in the nature and crystal chemistry of their Fe oxides by optical and Mossbauer spectroscopy as well as Rietveld refinement of X-ray diffraction patterns. We attribute the upward yellowing of the soil to a progressive transformation of the Fe oxides at nearly invariant iron contents. Aluminium in contrast is strongly mobilized in the uppermost clay-depleted topsoil where there is preferential dissolution of kaolinite and crystallization of gibbsite. Haematite decreases from 35 to 10% of the Fe oxides from the bottom to the top of the profile and the particles become smaller (75-10 nm). Its Al for Fe-substitution remains almost unchanged (10-15 mol %). The average Al-substitution rate of goethite increases from 25 to 33 mol %, and its mean crystal diameter remains in the range 20-40 nm. The proportion of Al-rich goethite (33 mol %) increases at the expense of less Al-substituted Fe oxides (haematite and goethite). This conversion with restricted transfer of iron means that the amount of Al stored in Fe oxides gradually increases. Kaolinite, haematite and Al-poor goethite are thus witnesses of earlier stages of ferralitization of the soil. In contrast, Al-rich goethite and gibbsite initiate the alitization (or bauxitization) of the soil. They correspond to the last generation of soil minerals, which most likely reflects the present-day weathering conditions. The progressive replacement of kaolinite, haematite and Al-poor goethite by new generations of Al-rich goethite and gibbsite attests to greater activities of water and aluminium and smaller activity of aqueous silica in the topsoil than in the subsoil. We interpret this as a consequence of longer periods of wetting in the topsoil that could result from soil aging, more humid climate or both.