Empirical modelling of the kinetics of phosphate sorption to macropore materials in aggregated subsoils

preferential transport of phosphate through macropores increases the significance of phosphate sorption to macropore wall materials compared with bulk soil materials. Therefore we studied the kinetics of phosphate sorption to soil bulk materials from the Ap and Btg horizons, from the iron oxide-poor (Albic) centres and from iron oxide-rich (Red) exteriors of the fractures in a clayey pseudogley in batch at initial phosphate concentrations (Po) up to a maximum of 650 mu M and at sampling times up to 7 days. Uptake of phosphate was least in the phosphate-rich Ap, and the Albic material also adsorbed little. By contrast the Red material sorbed phosphate strongly. By plotting the logarithm of the solution phosphate concentration against log time, two or three different kinetic regions could be distinguished: a fast reaction within the first minute of reaction; a lag period at intermediate to large initial values of phosphate concentration (P-0): and a steady slow reaction. The lag phase cannot be described satisfactorily by the well-known adsorption models. The Langmuir equilibrium expression produced rather good fits at any fixed time of reaction, but the estimated adsorption maximum increased steadily with time. Empirical models were fitted to the data, among them a 'lag-linear' model developed to include the lag phase. The relative contribution of the fast reaction to total adsorption during 7 days decreased from 50 to 804b at small P-0 to less than 10% at large P-0 values. The fraction of fast-adsorbed phosphate followed the order Red>Btg>Albic>Ap regardless of P-0; the slow reaction is, by a log-log rate, ranked Albic congruent to Btg>Red>Ap for P-0<20 mu M and Red>Btg>Albic>Ap at larger initial concentrations. The Albic material does not minimize the risk of phosphate leaching through macropores, as demonstrated by its small P sorption capacity, the relatively slow adsorption rate and its small actual P content.