The Retention of Phosphate By the Cation Exchange Surface of Muscovite

Abstract

Many soils have the ability to remove phosphate from soil solution and it is known that although the phosphate retained by the solid phase may initially be available for plant growth, with time its availability decreases and it is said to be 'fixed'. As phosphate plays an important role in plant nutrition the 'fixation' of phosphate makes necessary the repeated application of phosphatic fertilizers, even though relatively large amounts of phosphate may already be present in the soil. In the hope of finding ways of making more efficient use of soil phosphate the interaction of phosphate with solid phases present in soils has often been studied. Results obtained using soils have been difficult to interpret because of the complex character of soils and for this reason clay minerals, which are known to be largely responsible for phosphate retention and fixation, have been used as model systems. Because clay minerals are not pure, usually containing other clay minerals and decomposition products, and because they have more than one type of surface, interpretation of experimental results is still difficult. Because of the similarity of the 001 cleavage surface of muscovite mica to many surfaces existing in soils certain aspects of the chemistry of this surface were determined in the hope of obtaining an understanding of some of the fundamental processes responsible for phosphate retention and fixation by soils. The low specific area of the muscovite sheets used made isotope dilution techniques necessary for the quantitative investigation of adsorption by this surface and special methods of cleaning and manipulation were necessary to prevent contamination. The distribution of adsorbed phosphorus-32 could be determined using autoradiographic techniques. Preliminary experiments using methods developed by previous workers showed that treatment of mica strips with 1M solutions of AlCl3 or FeCl3 followed by thorough rinsing produced a phosphate adsorbing surface. Autoradiography showed that whereas the FeCl3 treatment produced a surface which adsorbed phosphate uniformly, the adsorption occurred in streaks on the AlCl3, treated surface. The nature of the streaks and evidence from other experiments suggested that AlCl3 solution washed out of the edges during rinsing undergoes hydrolysis on dilution, the hydroxo—aluminium species produced being adsorbed by, and conferring their phosphate retaining properties to, the 001 surface. Direct electron microscopic observation of such AlCl3 and FeCl3, treated surfaces showed the presence of small particles, apparently hydrolysis products, in both cases, suggesting that pre-treatment with solutions containing other hydrolyzable cations could also cause phosphate retention by the same mechanism, differences in uniformity of adsorption such as streaking arising from differences in the rate of hydrolysis. Phosphate retention by freshly cleaved muscovite. Freshly cleaved surfaces are found to retain only small amounts of phosphate at short treatment times but the amount retained increases with time up to at least 3000 minutes without any apparent equilibrium being reached. The characteristics of the sorption of phosphate by freshly cleaved surfaces suggest that decomposition of the muscovite may be involved in the retention. A mechanism for the observed retention is suggested involving decomposition of the mineral and sorption of positively charged decomposition products, having phosphate retaining properties, by the negatively charged surface. Decomposition will occur at edges of the mica sheet and at 'steps' present on the basal surface and may also occur at the 001 face itself. If decomposition does occur at the 001 face the resulting disordered surface layer would probably also have phosphate retaining properties. Treatments leading to increased phosphate retention. Phosphate retention significantly greater than that observed for freshly cleaved mica could be obtained if certain cations were present in the phosphate solution or if the surface had been pretreated with certain solutions. Enhanced phosphate retention was found to occur in the following cases: (a) when Pb2+ was present in solution at a sufficient concentration for the solubility product of Pb3(PO4)2 to be exceeded. As the phosphate concentration was much smaller than the Pb2+ concentration the Pb3(PO4)2 would have been present as positively charged colloidal particles which could be retained by the negatively charged surface; (b) when one of the hydrolyzable cations Al(III), Fe(III), Th(IV), or Zr(IV) was present in the phosphate solution. It is suggested that basic phosphates (e.g. Al(OH)3-x (H2PO4)x) are formed in these solutions and that these exist as positively charged colloidal particles which can be retained by the 001 surface; (c) when the surface has been pretreated with solutions containing one of the hydrolyzable cations Fe(III), Cr(III), Th(IV), or Zr(IV), or with partially neutralized solutions containing one of the cations Co(III), Mn(III), Cr(III) Al(III), Fe(III), Th(IV), or Zr(IV). It is suggested that colloidal hydrous oxides or polymeric hydroxo-complexes, present in the pretreatment solutions or formed during rinsinq, are adsorbed by, and confer their phosphate retaining properties to, the 001 surface. The presence of phosphate as a solid phase in solutions from which significant retention occurred was shown in cases (a) and (b) by spotting the solutions on to chromatography paper and chromatogramming with distilled water, the solid phase phosphate remaining at the origin. Electrostatic adsorption of colloids by muscovite. By using a solution containing BaCl2 and S-35 labelled K2SO4 at various relative concentrations it was shown that colloidal particles of BaSO4, the charge of which depend on the relative Ba2+ and SO2-4 concentrations in the suspension, were adsorbed by the negatively charged 001 surface of muscovite only when positively charged. Similarly Fe-59 labelled colloidal hydrous ferric oxide was retained only at a pH below its isoelectric point. These results suggest that the adsorption of colloids by muscovite is due to electrostatic interaction and probably occurs for any surface and colloid with the right electrostatic charges. Desorption of colloidal BaSO4 from the muscovite surface in a solution with a high concentration of sulphate was found not to occur even though the free colloid would have been negatively charged in this solution. It is suggested that desorption by a mechanism of colloid charge reversal does not occur because anion exclusion by the muscovite surface and inhibited access to the surface of the colloid adjacent to the muscovite prevent the positive charges of the colloid nearest to the muscovite from being changed. The desorption of adsorbed Fe-59 from the surface of muscovite when placed in a high concentration phosphate solution observed by a previous worker is therefore probably due to dissolution of the hydrous ferric oxide and formation of phosphato-iron complexes rather than the reversal of charge on the adsorbed colloid. Artificial weathering of muscovite. Muscovite, artificially weathered by long term treatment with water adjusted to various pH values, was found in some cases, depending on the pH and time of treatment, to retain phosphate at the basal surface. Such retention is thought to arise from adsorption by weathering products present on the 001 surface or by the modified surface resulting from weathering. Electron micrographs of such surfaces showed a range of features which could be attributed to the presence of both crystalline and apparently amorphous particles. The electron micrographs of surfaces treated with acid solutions (pH 1) also showed features which could be attributed to the presence of relics of colloidal silica particles remaining after dissolution of aluminium from the surface layers. Properties of hydroxo-aluminium species in muscovite surfaces. Because soils usually contain a large amount of aluminium and because hydroxo-aluminium species are probably formed during weathering, and are therefore likely to be electrostatically shorbed by negatively charged soil surfaces, more detailed investigations were made of the phosphate retention properties of muscovite surfaces treated with hydrolyzed AlCl3 solutions. Basicity of hydroxo-aluminium species. Pretreatment of muscovite surfaces with hydrolyzed AlCl3 solutions is found to cause enhanced phosphate retention providing the OH/Al ratio of the pretreatment solution is in the range 0 to 3. Phosphate retention thus varies with the basicity of the pretreatment solution in much the same way as does the retention of hydroxo-aluminium species by montmorillonite in the formation of interlayer complexes. The size of the hydroxo-aluminium species adsorbed by the surface during pretreatment appears to increase with the basicity of the AlCl3 solution. Where the OH/Al ratio of the pretreatment solution is in the range 2 to 3 the hydroxo-aluminium species adsorbed by the surface are sufficiently large (~50A) to be seen in the electron microscope. Aging of surface. Surfaces prepared from the more basic AlCl3 solutions (Oh/Al = 3.0) were found to lose much of their ability to retain phosphate when stored in distilled water for some time before determining phosphate retention. This aging effect was not noticeably present when the surfaces were prepared from the less basic solutions and is easily understood in terms of the crystallization of Al(OH)3 which is known to occur more rapidly at higher pH values or in more basic precipitates. Adsorption kinetics. The shape of the phosphate adsorbed vs time curve was found to depend on the basicity of the hydroxo-aluminium species on the surface. When the surface was prepared using AlCl3 solutions with an OH/Al ratio of 2, equilibrium was reached in about 30 minutes and no significant changes in the amount adsorbed were found for times of up to a day. However, when the pretreatment solution had an OH/Al ratio of 3 the adsorption curve reached a peak in about 30 to 60 minutes and then dropped slowly. Loss of phosphate from the surface at adsorption times longer than about 60 minutes is conveniently explained by the aging processes occurring in the hydroxo-aluminium species on the surface which are accompanied by crystallization and decrease in surface area. Desorption kinetics. The desorption kinetics were found to depend on the adsorption time used. Phosphate retained during the longer adsorption times was found to desorb more slowly than that retained during short adsorption times. Although this could be attributed to aging phenomena in the case where the basicity of the surface hydroxo-aluminium species was 3.0, no such aging phenomena were evident for the surface with a basicity of 2.0 and hence this fixation process is attributed to a mechanism involving the phosphate such as local crystallization of aluminium hydroxy phosphate or the formation of more stable aluminium phosphate bonds with time. The basicity of the hydroxo-aluminium species present on the surface did not appear to significantly affect the desorption kinetics. Effect of high phosphate concentrations. Electron micrographs of surfaces containing hydroxo-aluminium species with an OH/Al ratio of 3.0 show that treatment of this surface with a 10ˉ²M KH2PO4 solution removes the larger particies present to produce a surface with particles of size about 100A. This indicates that at higher concentrations phosphate is retained not by the adsorption mechanism operating at low concentrations but by a dissolution-precipitation mechanism. Relevance of results to the chemistry of phosphate in soils. The applicability of the 001 muscovite surface as a model soil surface and the significance of the results obtained in this thesis to the understanding of the retention and fixation of phosphate by soils is discussed. It is concluded that as clay mineral surfaces in soils are often “contaminated” by weathering products the 001 muscovite surface may not be directly applicable as a model surface. However, pretreatment of the muscovite with basic AlCl3 solutions produces a suitable model surface. As the adsorption of positively charge hydroxo-aluminium species by negatively charged surfaces results from non-specific electrostatic forces such species, which are weathering products, are probably adsorbed by other negatively charged soil surfaces besides those of the clay minerals. The muscovite surface can therefore be used as a model for all negatively charged soil surfaces. Results in this thesis indicate that cation exchange surfaces in soils will retain phosphate by adsorption of positively charged colloidal hydrous oxides or polymeric hydroxo-complexes containing sorbed phosphate, or by adsorption of phosphate by such hydroxo-species already present on the surface. At high phosphate concentrations, as exist near a fertilizer particle, retention of phosphate can occur by phosphate induced decomposition of clay minerals or hydrous oxides and precipitation of clay minerals or hydrous oxides and precipitation of phosphate containing products. These products may also be adsorbed on soil surfaces. Liming of a soil may induce aging of hydroxo-aluminium species and therefore reduce the phosphate retention capacity of the soil. Such a process may, in part, be responsible for the beneficial effect of liming on the utilization of fertilizer phosphate by plants. It is suggested that the fixation process observed probably occurs, not only with systems similar to the hydroxo—aluminium pretreated muscovite surface investigated, but also with other phosphate sorbing surfaces where sorption occurs by formation of aluminlum-phosphate or iron—phosphate bonds.