Phosphorus sorption behaviour of some South African water treatment residues.

Abstract

Water treatment residues (WTRs), which are by-products from the production of potable water, are chemically benign, inorganic materials which are suitable for disposal by land application. Their high phosphorus (P) sorption capacities have, however, generated some concern in an agronomic context where P is recognised as a growth limiting plant nutrient. The extent to which labile P pools are reduced or enhanced by WTR amendments is, therefore, a central issue with respect to their disposal by land application. Therefore, the aim of this study was, through the use of empirical adsorption isotherm equations and chemical fractionation of P within the residues, to investigate the chemical processes responsible for the retention and release of P from 15 South African WTRs. Chemical characterisation revealed considerable variation in residue properties relevant to P sorption-desorption processes. pH, exchangeable Ca and organic carbon content ranged from 4.77 to 8.37, 238 to 8 980 mg kg-1 and 0.50 to 11.6 %, respectively. Dithionate, oxalate and pyrophosphate extractable Al fractions ranged from 741 to 96 375, 1 980 to 82 947 and 130 to 37 200 mg kg-1, respectively. Dithionate, oxalate and pyrophosphate extractable Fe ranged from 441 to 15 288, 3 865 to 140 569 and 230 to 90 000 mg kg-1 respectively. Therefore mechanisms of retention were hypothesised to be residue specific, being dependent on the unique chemical properties of the sorbent. Elevated Ca and amorphous Al and Fe concentrations did, nevertheless, suggest that all residues had the capacity to adsorb high amounts of P and to retain this P in forms unavailable for plant uptake. These arguments were confirmed by the sorption study where labile P was, for all residues, found to constitute a small fraction of total applied P even at high application concentrations (128 mg P L-1). Sequential P fractionation revealed that most of the inherent P (which ranged from 1 149 to 1 727 mg P kg-1) and applied P were retained in highly resistant mineral phases or fixed within the organic component. Thus P replenishment capacities were restricted even though residual P concentrations were often within adequate ranges for plant growth. Phosphorus adsorption data was described by four empirical adsorption isotherm equations in an effort to determine possible mechanisms of retention. Sorption data was, for most of the WTRs, described by the Temkin isotherm while the Freundlich and linear models fitted data for two residues each. A key finding was that the distribution coefficient (Kd) tended to increase with the quantity of P adsorbed (S) as opposed to decrease or remain constant in accordance with model assumptions. Therefore, the models could not be used for mechanistic interpretation, even though they provided excellent descriptions of the data. The direct relationship between Kd and S suggested a mechanism of retention involving the activation of sorption sites. This notion was supported by the fractionation study which showed that P addition results in the transfer of an increasing quantity of organically bound P to resistant residual forms. Model affinity parameters were strongly correlated to dithionate and pyrophosphate extractable Al and Fe which suggested that P was adsorbed primarily through ligand exchange mechanisms. The mobility of P bound to organic fractions did indicate that P was retained through weaker forces of attraction such as monodentate ligand exchange, charge neutralisation or proton transfer. Evidence to support the notion that P is immobilised through the formation of Ca phosphates was lacking. Based on P fractionation data, it was suggested that strong chemisorption mechanisms and the diffusion of P into WTR micropores were largely responsible for the minimal quantity of P desorbed by disequilibria desorption processes. A greater quantity of P was desorbed in the presence of oxalate and citrate which suggested that plants may increase bio-available pools through the release of organic ligands. Phosphorus desorbed in the presence of these ligands did, however, decline with P addition which confirmed that the affinity of the WTR surface for P increases with P application. Therefore, it was concluded that the application of P to WTRs is an uneconomical process unless sorption sites are already saturated or immobilisation processes are inhibited. In light of these findings, it was suggested that the absence of plant P deficiencies under the field application of WTRs is due primarily to inhibited sorption.