Fate of veterinary pharmaceuticals in soil: An experimental and numerical study on the mobility, sorption and transformation of sulfadiazine

Among other veterinary pharmaceuticals sulfadiazine (SDZ) is a widely used antimicrobial substance in intensive livestock production to prevent and treat diseases. Up to 40 % of the administered sulfonamides are eliminated as microbial active substances with the excretions. Antibiotics such as sulfadiazine reach agricultural soils directly through grazing livestock or indirectly through the spreading of manure or sewage sludge on the field. Knowledge about the fate of antibiotics in soil is crucial to assess the environmental risk of these compounds, including possible transport to groundwater. Sorption, transport and transformation of 4$^{14}$C-labelled SDZ in a silty loam were investigated using batch-type and column experiments. The batch sorption/ desorption experiments were conducted at various concentration levels (0.044 to 13 mg L$^{−1}$ initial solute concentration) and time-scales (0.75 to 272 days). Sorption of SDZ in the investigated soil was time-dependent and strongly non-linear with regard to the concentration. The time to reach the apparent sorption equilibrium was about 20 days. However, desorption was very slow and 41 days were insufficient to reach the desorption equilibrium. In annealed soil the sorption affinity was lower and the desorption was also very slow. Transport of $^{14}$C-labelled SDZ was investigated in disturbed soil columns at a constant flow rate of 0.26 cm h$^{−1}$ near saturation. $^{14}$C-SDZ was applied in different concentrations (5.7 or 0.57 mg L$^{−1}$) for either a short or a long pulse duration (7 or 70 hours). Breakthrough curves (BTCs) of $^{14}$C-SDZ and the non-reactive tracer chloride were measured. At the end of the leaching period the soil concentration profiles were determined. The peak maxima of the BTCs were delayed by a factor of 2 to 5 compared to chloride and the decreasing limbs are characterized by an extended tailing. The maximum relative concentrations differed as well as the eluted mass fractions, ranging from 18 to 83 % after 500 hours of leaching. Mineralization of SDZ during the batch and column experiments was neglectable. Inverse modelling techniques were used to identify relevant sorption processes of SDZ and its transformation products during the batch and column experiments. One-, two- and three-domain sorption models were tested, involving linear or non-linear, instantaneous equilibrium, rate-limited reversible and irreversible sorption. The various sorption concepts were included in a convective–dispersive transport model for the description of the column experiments. Models involving similar processes and complexity were necessary to describe the characteristic features of the batch and transport experiments. The non-linear sorption in the batch experiments was best described by the Freundlich sorption isotherm and may also be one reason for the pronounced tailing of the BTCs in the transport experiments. Rate-limited sorption accounts for the slow attainment of the apparent sorption equilibrium during the batch experiments and can also contribute to the tailing of the BTCs. However, sorption sites exhibiting instantaneous equilibrium sorption were required in addition to the kinetic sorption to describe both, the one-day adsorption and the BTCs of the column transport experiments with the long pulse duration. Only the consideration of irreversible sorption enabled [...