Smectite (from South Dakota, Wyoming, and Mississippi) and Vermiculite (Transvaal) were treated with solutions of Al(OH)B(3-B)+, with B varying from 0 to 2.5. The average basicity (OH/Al = B) of the Al adsorbed differed very much from the basicity of the Al added. The average basicity of the Al adsorbed by smectite was always above the average basicity of the Al added. In contrast to smectite, Vermiculite adsorbed smaller hydroxy-Al complexes. One reason for the different selective behavior was the difference in expansion between smectite (about 18 Å) and vermiculite (about 14 Å). Because of the adsorption of the relatively more basic OH-Al by smectite, smectite adsorbed considerably more Al than vermiculite. The total amount of aluminum in the interlayer generally could not be calculated by the difference between Al added and that remaining in solution after the reaction because of possible protonation of the clay mineral and adsorption of structural Al and other cations, which is more pronounced for vermiculite. The results in the present study demonstrated that neither the quantitative nor the qualitative composition of an Al(OH)B-treated exchanger can be deduced from B of the Al salts added. These points are frequently overlooked when cation exchangers are pretreated with Al of variable basicity and are used for further investigations, such as studies of CEC, surface area, interlayer spacing, anion reactions, the formation of gibbsite, etc. Before these kinds of investigations are conducted employing the pretreated OH-Al-exchangers, their composition should be known precisely.

The presence of Al hydroxy species in solution during the synthesis of lepidocrocite had been previously found to influence the reaction towards goethite formation. However, under certain conditions, which are not unrealistic in terms of the natural soil environment, this influence does not occur, and Al appears to substitute for Fe(III) in the lepidocrocite structure. This substitution causes the unit-cell dimensions to decrease along the “a” direction and to increase along the “b.” From the differential line broadening of X-ray powder diffraction peaks, the incorporation of Al was found to inhibit crystal growth preferentially in the b-axis direction, the hkl peaks being more broadened the higher the value of k relative to h and l. Al-substituted lepidocrocites have been suggested to occur in soils, and although they can be synthesized under conditions approaching those expected in soils, it is considered that their formation in nature is unlikely or restricted to unusual environments.

Acidified suspensions of Al-saturated kaolinite, montmorillonite, mica, illite, and biotite in 10−3 M NaNO3 were potentiometrically titrated with 0.1 N NaOH and 0.1 N HNO3 in succession in a CO2-free nitrogen atmosphere. The resulting curves were compared with those for Al(NO3)3 solutions of similar Al concentration in the supernatant solution and corrected for Al in the entrained solution in the clay.

Base titrations of Al ions adsorbed on all the minerals, except montmorillonite, showed two pH inflections separated by a buffering range. With montmorillonite, there were three pH inflections similar to those for Al in solution. The first inflections in the titration of suspensions occurred at lower pHs and were less pronounced than for Al in solution. These represent the titration of H3O+ sorbed during the pretreatment. The buffering by adsorbed Al ions is also less than that by Al in solution.

The OH− used up by adsorbed Al ions between the first and last inflections was equal to, or slightly greater than, the CEC of the minerals, except for mica where it was more than twice the CEC, because new interlayer surfaces were formed during the acid pretreatment. Acid titration curves of Al ions in the adsorbed and solution states showed hysteresis when related to the base titration curves. The use of two titration speeds (3 and 0.3 pH units/hr) only slightly affected the titration curves of the minerals suggesting that the observed effects were not caused by lack of equilibrium with added base or acid.

A stable porous system consisting of montmorillonite cross-linked by Al-hydroxide oligomers was synthesized by reacting at room temperature an aqueous solution of such oligomers with a unit-layer dispersion of montmorillonite. The resulting cross-linked montmorillonite (Al-CLM) is a nonswelling material, showing basal spacings of 14.4 to 18.8 Å after air drying and between 14.2 to 18.0 Å after treatment at 110°C. The basal spacing is found to depend on the age and OH/Al ratio of the Al-hydroxide solution, as well as on the relative amounts of the two reactants. A specific surface area of 160 m2/g and a diffraction pattern with a dominant basal spacing of 17.5 Å is obtained by using Al-hydroxide with OH/Al = 1.85, aged for at least 5 days, and by applying an Al/montmorillonite ratio greater than 1.5 in the cross-linking process. The basal spacing of Al-CLM remains essentially unchanged after heating at 220°C, while the specific surface area is not affected by heat treatment up to 480°C.

Two possible configurations of Al-hydroxide oligomers, homogeneously distributed between parallel montmorillonite unit-layers, were considered in order to account for the basal spacing of 17.5–18.8 Å, viz. (a) stacking of two oligomeric ring units in parallel orientation relative to the clay lamellae and (b) perpendicular orientation of individual oligomeric units.

Aluminum-substituted hematites (Fe2−xAlxO3) were synthesized from Fe-Al coprecipitates at pH 5.5, 7.0, and in 10−1, 10−2, and 10−2 M KOH at 70°C. As little as 1 mole % Al suppressed goethite completely at pH 7 whereas in KOH higher Al concentrations were necessary. Al substitution as determined chemically and by XRD line shift was related to Al addition up to a maximum of 16–17 mole %. The relationship between the crystallographic a0 parameter and Al substitution deviated from the Vegard rule. At low substitution crystallinity of the hematites was improved whereas higher substitution impeded crystal growth in the crystallographic z-direction as indicated by differential XRD line broadening. At still higher Al addition crystal growth was strongly retarded. The initial Al-Fe coprecipitate behaved differently from a mechanical mixture of the respective “hydroxides” and was, therefore, considered an aluminous ferrihydrite.

Ferrihydrite was transformed to goethite and/or hematite at various temperatures, [OH], and [Al]. Increasing temperature and [Al] favored hematite, increasing [OH] favored goethite. A given [Al] induces hematite more effectively at lower [OH]. Al substitution in goethites increased linearly with log[Al], but was independent of temperature. At a given [Al], substitution increased with decreasing [OH]. In a plot of Al/Fe in the goethite against [Al]/[Fe(OH)4−] in solution a straight line was obtained for all preparations independent of [OH].

The equilibrium diagrams developed for Al-hydroxide and for kaolinite by Garrels and Christ (1965) have been modified by taking into account the existence of gels. From the stability zones obtained, the “appropriate” concentrations can be deduced and utilized for synthesizing these species, provided the requirements to insure good crystal growth are observed. Among procedures to promote these crystallizations, homogeneous precipitation processes (La Iglesia et al., 1974, 1976) appear to be particularly adequate.

The theoretical considerations provide an explanation for most of the processes observed until now, both successful and unsuccessful syntheses, and also give an explanation for many field observations. The crystallizations, however, remain poorly reproducible, indicating that many factors are still poorly known. Some points requiring further investigation include (i) better values for ΔGr0, (ii) the influence of organic complexes, (iii) the effect of preexisting crystalline phases, (iv) those involving dehydration processes in these systems.

The rates of release of Al by M NH4NO3 (pH 3) from minerals saturated with Al ions at pH 3 suggest that Al ions migrated from the surface layers and the matrix cores of kaolinite, montmorillonite, illite, and biotite, but only from the matrix core of muscovite mica. From the 0.25–0.5 μm kaolinite and montmorillonite, part of the ‘surface’ Al is released ‘instantaneously’ and the rest by first order kinetics, but the coarse 1.5–2.5 μm kaolinite has only the former component. From illite and biotite, ‘surface’ Al is released by ‘bulk diffusion’ kinetics suggesting the existence of disturbed peripheral layers of finite thickness. The diffusion coefficients, Dm, for the matrix core follow the trend: mica ≃ biotite > illite > montmorillonite > kaolinite.

Based on models proposed in previous work, the ionic composition of the ‘surface’ Al is calculated. This shows that (1) this composition varies according to the mineral from 3 to 100% Al3+, the remainder being in the hydrolyzed form, and (2) the apparent hydrolysis constants are different for each mineral and significantly different from those of Al ions in solution.

The influence of the activity of water on the phase composition of aluminum hydroxides obtained by reaction of amalgamated aluminum with water has been studied. The reaction was carried out in solutions of water and dioxane, and in sodium chloride solutions of different concentrations; amorphous aluminum hydroxide was precipitated initially. The aging of this primary product to bayerite, pseudoboehmite, or an amorphous gel was controlled by the water content of the system. The pseudoboehmite has shown a significant reactivity when used as a source of alumina in the hydrothermal syntheses of kaolinite at relatively low temperature and pressure.

Al-substituted goethites were prepared by rapid oxidation of mixed FeCl2-AlCl3 solutions at pH 6.8 in the presence of CO2 at 25°C. A combination of Al substitution and adsorption of CO2 reduced crystal size (except for an increase at small additions of Al) and produced unusual thin, porous particles. Product goethites had surface areas up to 283 m2/g and unit-cell expansions induced by hydration. Substitution of Al for Fe reduced the 111 spacing and increased infrared OH-bending vibrational frequencies. Al substitution split the goethite dehydroxylation endotherm during differential thermal analysis into a doublet and increased the temperature of all reactions. Both cold and hot alkali solutions dissolved Al from the goethite structure.

After drying the product in vacuo at 110°C. X-ray powder diffraction data indicated minimal deviation from Vegard's law for the goethite-diaspore solid solution up to about 30 mole % Al substitution. Goethite prepared in the presence of 40 mole % Al had a 111 spacing of 2.403 Å corresponding to 36 mole % structural Al if Vegard's law was obeyed. Rapid oxidation of mixed FeCl2-AlCl3 solutions appears to be conducive to a higher degree of Al substitution in goethite than alkaline aging of hydroxy-Fe(III)-Al coprecipitates.

The influence of Al on the products formed by aerial oxidation at pH 5.5-7 and 20°C of Fe(II) chloride, sulfate and carbonate solutions, was examined. In all cases Al at levels Al/Al + Fe = 0.09−0.30 inhibited the formation of y phases (lepidocrocite and maghemite) in favor of goethite under conditions where, in the absence of Al, these y phases formed. The influence of Al in these laboratory studies was supported by field observations.

At higher levels of Al, ferrihydrite formation was favored. This effect of Al was seen to be the result of a slowing down in the hydrolysis/oxidation rate of the Fe(II) system.

The presence of Al not only changed the direction of mineral formation, but also caused the formation of Al substitued goethites which resembled in particle size and morphology the natural aluminiferous goethite extracted from a soil.

Cross-linking of Li-montmorillonite by hydroxy-aluminum oligomers was performed in a specially constructed mixing apparatus. Observations on flocculation and solution composition were carried out during and after the cross-linking reaction; the dry product was studied by scanning electron microscopy.

Flocculation was most pronounced at Al/montmorillonite ratios between 0.98 and 2.45 mM/g; below and above this range, flocculation was much less intensive. These observations can be explained by heterocoagulation and protecting colloid action. A complete neutralization of the montmorillonite charge was estimated at 1.9 mM adsorbed Al per g clay, and in order to account for the electrical charge of the hydroxy-Al, polymers with an average charge of 0.5 per Al atom must be assumed on the montmorillonite surface. Assuming that the hydroxy-aluminum form in the unreacted solution is Al6(OH)3+12, the adsorbed polymer will be Al24(OH)12+60. Alternatively, assuming Al6(OH)3+15 in the unreacted solution, this form will remain unchanged upon adsorption onto the montmorillonite surface.

Differences in the microfabric of dry Al-CLM as a function of Al/montmorillonite ratio can be explained along the lines of the interpretation of the flocculation studies.

Polycations of limited molecular size were prepared from 0.1 M solutions of Fe(III)(NO3)3 and A1(NO3)3 by ultrafiltration. Various amounts of the polycations were added to Na-kaolinite, Na-mont-morillonite, and a Na-soil clay and the effect on the flocculation of the clay and electrophoretic mobility compared. Flocculation occurred just before zero net charge was obtained. Addition of further polycation resulted in the dispersion of clay with a net positive charge. The Al polycations possessed a high positive charge (0.49 per gram atom of Al), and their interaction with the clays indicated a planar shape. Adsorption of Al polycations decreased markedly the cation-exchange capacities of the kaolinite and the soil clay but had little effect on surface areas determined by low-temperature N2 adsorption. The Fe polycations were spheres 10–100 Å in diameter with a positive charge of 0.2 per gram atom. The surface areas of the kaolinite and the soil clay were substantially increased by the addition of the Fe polycations but their cation-exchange capacities were reduced by one fifth. Al polycations increased the surface areas of the montmorillonite (to 300 m2) presumably by propping open the interlamellar spaces and rendering the a-b planes accessible for nitrogen adsorption. The Al polycations in interlamellar spaces prevented collapse to 14 Å on heating to 150°C. There was no evidence of regular interlayer Fe as might be anticipated from the size of the spheres.

Potentiometric titration analysis was used to examine the hydrolysis behavior of Fe2+ Fe3+, and Al3+ in pure solution and in mixture, in order to evaluate the potential for coprecipitation and mixed solid-phase formation. Mixtures of Fe3+ and Al3+ did not interact during neutralization; base consumed in their respective buffer regions was equivalent to the total metal added. Fe2+-Al3+ solutions, however, showed excess base consumption in the Al3+ buffer region, indicating hydrolysis of Fe2+ at lower than normal pH. Ferric/ferrous iron analyses of systems at the Al endpoint (pH 5.5) showed amounts of oxidized Fe equivalent to the excess base consumption (∼10% of total Fe), with substantial amounts of Fe2+ sorbed to or occluded within Al polymers present. Increased electrolyte levels or the presence of SO42- inhibited oxidation and sorption of Fe2+ on Al surfaces, suggesting that Fe hydrolysis and oxidation was catalyzed at the surfaces. Increasing Al3+: Fe2+ ratios in the titrated solutions also increased the amount of Fe2+ coprecipitation, supporting a surface-mediated reaction mechanism. Ferrous iron oxidation was sensitive to O2 levels, which also affected the amount of coprecipitation. These findings suggest that surface-facilitated oxidation of Fe2+ may be important in the formation of mixed Fe-Al mineral phases in dilute soil solutions.

Lepidocrocite samples, γ-FeOOH, containing 0–10 mole % Al-for-Fe substitution were synthesized at 15°C and pH 8 by oxidizing mixed FeCl2-AlCl3 solutions. The unit-cell parameters a, b, and c were measured from step-counted X-ray powder diffractograms using seven lines and Si as an internal standard. With increasing Al substitution from 0 to 10 mole % the unit-cell edge lengths a, b, and c decreased regularly by 0.3, 0.8, and 0.6%, respectively. Furthermore, the crystals became smaller, but gained in thermal stability. Decrease in crystal size parallel to the y axis led to a significant increase of the OH-stretch vibration and a decrease of the out-of-plane OH-bending vibration due to a weakening of the hydrogen bond between the zig-zag layers in the structure.

Mössbauer spectra of 15 hematites with Al substitutions between 0 and 10 mole % were taken at room temperature. X-ray powder diffraction indicated dimensions of these hematites in the c-direction to range upwards from 27 nm to crystals large enough to show no line broadening. The Mössbauer spectra showed that magnetic hyperfine fields decreased both with increasing Al-for-Fe substitution and with decreasing crystal size. These relationships indicate that hyperfine field variations cannot, as has been done in the past, be unequivocally related to Al substitution alone. Hyperfine field reductions were paralleled by Mössbauer line broadening due to hyperfine field distributions. Only the hematites heated to 1000°C showed a significant variation of quadrupole splittings with Al substitution. No dependence of quadrupole splitting on crystal size was observed, indicating no detectable distortion of coordination polyhedra in the particle size range studied.

Manganiferous karst bauxites are rare on a worldwide scale. One such body, recently mined at Kincsesbánya, Hungary, has been studied by chemical, petrographic, X-ray powder diffraction, scanning electron microscopic, and energy dispersive X-ray analytical techniques. The bauxite deposits of Kinesesbánya are of Paleocene to Lower Eocene age; however, the enrichment of manganese in them was a much later, epigenetic process. Lithiophorite is the main Mn mineral in this bauxite and occurs chiefly in dusters of < 1-μm size crystallites. Well-developed crystallites, however, 5–10 μm in size, line the walls of many microfissures and voids.

The oxidation of pyritic bauxite and lignitic clays in the overlying beds apparently mobilized finely disseminated Mn and Fe. Downward-migrating acidic solutions were gradually neutralized, and Mn and Fe minerals precipitated. The manganiferous bauxite was found only along the eastern rim of heavily eroded Middle Eocene sedimentary rocks. Here, epigenetic oxidation and mobilization were optimum. Farther to the east, pyrite-rich overburden and bauxite were apparently eroded away before Fe and Mn could be mobilized.