Incorporation of cobalt into synthetic iron oxyhydroxide systems: implications for natural systems

Phase pure synthetic reference samples of goethite and ferrihydrite and their cobalt-doped analogues have been prepared to enable comparison of their analytical data with natural goethite samples collected across the CoG3 consortium of Universities.

A systematic study of the incorporation of cobalt into goethite through the aging of ferrihydrite has been carried out for the first time and the degradation of these phases as a function of cobalt concentration by reductive dissolution investigated. Goethite samples were prepared in potassium hydroxide solutions (pH 12-13) by aging precipitates formed from a mixture of Fe(NO3)3·9H2O and Co(NO3)2·6H2O with 0, 2, 4, 6, 8, 10, 12, 15 and 20 mol% cobalt at room temperature, 40 and 70 °C. Precipitates were aged for 1, 4, 7, 11 and 50 days to investigate the effect of time on the different phases formed and their crystallinity. Furthermore, undoped and Co-doped ferrihydrite samples were prepared at both room temperature and for the first time at 70 °C by optimising the synthetic conditions to enable full characterisation of the analogous goethite samples to be achieved; since ferrihydrite is a goethite precursor phase, it can be retained in the product as an impurity. Solutions containing EDTA, oxalic acid and citric acid were all tested for the first time as milder alternatives to mineral acids for removing poorly crystalline phases, such as ferrihydrite, from synthetic goethite and Co-goethite with a continuous contact period of 15 days or a batch wash procedure where the solution was refreshed every 2 days for a total period of 6 days. Characterisation of the products was carried out before and after washing using a multi-technique approach (wet chemical methods and instrument analysis) to determine the phases present and the amount of cobalt incorporated and its effect on physical properties such as dissolution behaviour.

A multi technique characterisation of the different undoped and Co-doped goethite and ferrihydrite samples was necessary for understanding the complexity and the impact of Co on the formation of these phases due to variation in the crystallinity and hydration of the phases prepared at different temperatures.

Inductively Coupled Plasma Mass Spectroscopy and redox titration with potassium permanganate showed that the amount of cobalt incorporated into goethite varied with aging time and temperature to a maximum of 6.7 mol% for 11 days at 70 °C. Characterisation of the model ferrihydrite samples using thermogravimetric analysis (TGA) allowed the presence of ferrihydrite in goethite to be identified through a characteristic weight loss below 100 °C and allowed the efficacy of the washing techniques for the removal of ferrihydrite to be determined. Imaging by electron microscopy (scanning and transmission) indicated that the batch process using 0.1 M EDTA was the most effective washing technique, successfully removing ferrihydrite completely without damaging the crystallinity of the goethite phase. Phase identification using powder X-ray diffraction data by comparison with the ICDD and the appearance of bands at 292, 413, 1321 cm⁻1 and 470, 690 cm⁻1 in the Raman spectra showed that long aging times promoted the production of haematite and cobalt ferrite (CoFe2O4) respectively which were not removed by the EDTA wash. The Mössbauer hyperfine parameters were strongly affected by the incorporation of cobalt, suggesting Mössbauer is a powerful tool for determining cobalt doping levels in goethite. Le Bail analysis was used to show a linear decrease in cell parameters with increasing cobalt incorporation according to Vegard’s rule for the 11 days at 70 °C batch of samples in good agreement with the substitution of the smaller low spin Co(III) ion on the high spin iron(III) site. SEM and TEM indicated that the amount of poorly crystalline phase(s) increases on decreasing the synthesis temperature and the shape of the goethite crystals becomes whisker-like (or lath-shaped) rather than acicular with increasing cobalt concentration. Samples with a targeted dopant level of 10 mol% cobalt showed evidence by TGA and Raman-spectroscopy for the presence of ferrihydrite in the system irrespective of preparation temperature or washing time. The samples with dopant levels between 10 to 20 mol% were very poorly crystalline and dissolved completely in 0.1 M EDTA solution.

Reductive dissolution experiments examining the total dissolution of the sample over time have shown a retardation of the reaction of citrate-bicarbonate-dithionite (CBD) solution with the samples when cobalt is present. Thermogravimetric Analysis, Raman spectroscopy and Mössbauer-spectroscopy are the most effective techniques for the detection of impurities in the goethite-ferrihydrite system.