Synthetic hematites with Al substitutions between 0 and 18 mol % were synthesized at different temperatures and water activities. The cell-edge lengths a for different synthesis conditions decreased linearly with increasing Al substitution. The regression lines, however, had different slopes and intercepts: the series with the highest synthesis temperature (1270 K) had the most negative slope. With increasing Al substitution, the hematites contained increasing amounts of non-surface water. Significant correlations were found between these chemically determined water contents and the deviations of the unit-cell parameters a, c, and V relative to the corresponding 1270 K regression lines. To explain the measured X-ray peak intensities, structural OH had to be included into the theoretical calculations. From intensity ratios normalized to I113, it is possible to determine the structural OH separately from the Al substitution, which can be assessed by the shift of the cell-edge lengths relative to the 1270 K regression lines. The incorporation of Al and OH into the hematite structure induces strain, which was quantified by X-ray diffraction.

Seven Al-containing lepidocrocite samples, γ-Fe1−xAlxOOH, prepared from FeCl2/Al(N03)3 solutions with initial Al/(Al + Fe) mole ratios Ci of 0.0025, 0.01, 0.025, 0.05, 0.075, 0.10 and 0.15 mol/mol, were examined by means of Mössbauer spectroscopy at room temperature (RT) and at various temperatures in the range of 8 to 80 K. The spectra at RT and 80°K consist of broadened quadrupole doublets and were analyzed in terms of a single doublet and of a model-independent quadrupole-splitting distribution, the latter yielding the best fit. The observed variations of the quadrupole-splitting parameters with increasing Ci are inconclusive as to whether the Al cations are substituting into the structure. The temperature at which the onset of magnetic ordering is reflected in the spectra, was measured by the thermoscan method with zero source velocity. A gradual shift from 50 K for Ci = 0.0025 mol/mol to 44 K for Ci = 0.10 mol/mol was observed for that temperature. As compared to earlier studies of Al-free γ-FeOOH samples with similar morphological characteristics, the fractional doublet area in the mixed sextet-doublet spectra at 35 K is significantly higher for the present lepidocrocites. This observation is ascribed to the substitution of Al cations into the lepidocrocite structure. A similar conclusion is inferred from the variation with Ci of the maximum-probability hyperfine field derived from the spectra recorded at 8 K and fitted with a model-independent hyperfine-field distribution. The magnetic results suggest that for the sample corresponding to Ci = 0.15 mol/mol, not all of the initially present Al has been incorporated into the structure.

Synthetic aluminum-substituted maghemite samples, γ-(Fe1-xAlx)2O3, have been prepared by thermal decomposition of Al-lepidocrocite (γ-Fe1-xAlxOOH), with × = 0, 0.04, 0.06, 0.14 and 0.18. The particles are needle-shaped and the mean crystallite diameter along the [311] crystallographic direction was found to be between 2.0 and 5.0 nm. Mössbauer spectra were collected at 6 K and from 80 K up to 475 K at steps of 25 K. In a wide range of temperatures the spectra of the non-substituted sample consist of a superposition of a broad sextet and a superparamagnetic doublet, whereas for the Al-maghemites this range is much smaller. From the temperature variation of the fractional doublet area two different parameters were defined: the temperature corresponding to a 50/50 doublet-sextet spectrum (T1/2), and the temperature below which the doublet ceases to exist (T0). These two parameters (T1/2 and T0) decrease from 390 K and 92 K (Al-free sample), to 118 K and 64 K (4 mole % Al) and to 100 K and 48 K (18 mole % Al), respectively. The average hyperfine fields at 6 K undergo a steep drop in going from the Al-free sample (Hhf = 506 kOe) to the sample with 4 mole % Al (Hhf = 498 kOe), but for higher substitutions the effect is much smaller. The A- and B-site quadrupole splittings, obtained from the data between 220 K and 475 K, were found as: ΔEQ,A = 0.86 ± 0.04 mm/s and ΔEQ,B = 0.65 ± 0.04 mm/s for the 4 mole % Al sample. The characteristic Mössbauer temperature, determined from the temperature dependence of the average isomer shift, was found to be in the range of 500–600 K.

Several Al-substituted goethites were synthesized by hydrolysis of Fe3+ salt solutions. The kinetics of the reductive dissolution of these goethites in dithionite-ethylenediaminetetra acetic acid (D-EDTA) was studied at pH 5.5, at 303, 323 and 333 K. The initial dissolution rate (R) per unit of surface area decreases with Al substitution. In the sample with greater Al content (G″7), the kinetic profiles of the dissolved Fe fraction vs. time gave a small positive intercept. The kinetic profile of R as a function of EDTA initial concentration shows a significant weakening in the presence of Al. The maximum is flatter and wider in Al-substituted goethite than that of pure goethite. In sample G″7, where the Al content is 11.3 mol.% the maximum is obtained when the [D]:[EDTA] initial ratio is ∼4.5 vs. 2 in un-substituted goethite. These results can be attributed to the lesser density of the more active dimeric sites, the presence of more strongly bonded Al-O-Fe with regard to Fe-O and the small value for the ≡ Al-EDTA surface species constant. Activation energy (Ea) increases with Al substitution. Its value is doubled from GO (pure goethite) to G″7 (11.3 mol.% of Al). The frequency factor (A) acts in the opposite sense to Ea, but it is not sufficient to outweigh the effect of Ea.

Hematite (α-Fe2O3) possesses distinct spectral properties which facilitate its identification in mineral mixtures. This paper reports the relationships between the visible diffuse reflectance (DR) spectrum and the color and crystal properties of a group of 81 natural and synthetic hematites. The visible DR spectra for powdered hematite samples diluted to 4 wt.% with BaSO4 white standard were recorded and used to calculate the corresponding Munsell colors. The second derivative of the Kubelka-Munk function of the DR was used to estimate the position and intensity of the main absorption bands, which occurred at ∼435, ∼485 and ∼545 nm. The Munsell hue ranged from 9.5P to 5.3YR and was negatively correlated with the position of the ∼545 nm band, so it became yellower as the band shifted to shorter wavelengths (higher energies). The Munsell value, which ranged from 4.9 to 8.6, and chroma, which ranged from 1.4 to 8.3, were negatively and positively correlated, respectively, with the intensity of the ∼545 nm band, the position of which exhibited a weak negative correlation with the degree of Al substitution (x). The position and intensity of this band also exhibited a weak negative correlation with the specific surface area (SSA); both, however, were uncorrelated with domain shape as measured by the ratio between the X-ray mean coherence lengths (MCLs) in the [110] and [104] directions. The properties of the ∼435 and ∼485 nm bands were unrelated to x, SSA or MCL110/MCL104. The negative relationship between the position of the ∼545 nm band and x does not support the assignment of this band to the 2(6A1) → 2(4T1g(4G)) electron pair transition (EPT), which relates to the to Fe3+−Fe3+ magnetic coupling between face-sharing octahedra. The ∼485 nm band might reflect a ‘goethitic’ structural component in Fe-defective hematites as it appears at the same wavelength as the hypothetical EPT related to Fe3+−Fe3+ magnetic coupling between edge-sharing octahedra in goethite (α-FeOOH).