Sorption of Pharmaceuticals in Natural Waters

Pharmaceuticals are one class of emerging contaminants frequently detected in drinking water supplies and even in finished drinking waters. Because these chemicals can exhibit adverse effects on aquatic ecosystems and human health, a growing number of studies have been conducted to evaluate their fate in natural water bodies. However, a quantitative tool is lacking to evaluate the relative significance of natural attenuation mechanisms and provide insight into the environmental behavior of pharmaceuticals in natural waters. The purpose of this research is to evaluate the significance of natural attenuation processes on the fate and behaviors of pharmaceuticals through modeling analyses. Special emphasis is placed on sorption processes and sorption kinetics. As part of this research, a sorption kinetics model was developed and incorporated into a water quality model for the Patuxent Estuary to evaluate the effect of sorption kinetics. Although most current models are based on the assumption of instantaneous sorption equilibrium, pharmaceutical compounds may take long times (in days) to reach sorption equilibrium, strongly suggesting that this assumption is invalid. Four hypothetical pharmaceuticals representing four possible combinations of sorption coefficients and times to reach sorption equilibrium were used as target compounds. Model results reveal that assumption of instantaneous sorption equilibrium results in significant under-prediction of water column concentrations for some pharmaceuticals: up to 150% at upstream locations. Further, sorption kinetics affects a model’s ability to capture accumulation of pharmaceuticals into riverbeds and the transport of pharmaceuticals in estuaries. For the second part of this research, experiments were conducted to examine the sorption behavior, especially sorption kinetics, of two selected pharmaceuticals: triclosan and enrofloxacin. Both chemicals exhibit slow sorption; further verifying that the assumption of instantaneous sorption equilibrium is inadequate. The slow sorption of enrofloxacin may result from reduced molecular mobility due to local sorption. Enrofloxacin shows much higher sorption capacity than triclosan due to ionic interactions. For this reason, the sorption of enrofloxacin is strongly but adversely dependent on pH and/or ionic strength. Desorption kinetics experiments reveal a 23-28% increase in sorption coefficient for both chemicals, indicating a sorption-desorption hysteresis. Enrofloxacin exhibits nonlinear sorption due to limited sorption sites. A two-compartment Langmuir model, assuming a linear sorption component of sediments, generates adequate fits to nonlinear experimental data. Experimental results were then incorporated into the fate and transport model for the Patuxent Estuary to evaluate the significance of attenuation processes. Results verify that triclosan and enrofloxacin exhibit different environmental fate and behavior than each other due to their properties. For a 20-km area immediately downstream of the upper boundary of the study area, sorption causes a 7.9% - 51.5% decrease in dissolved-phase concentrations for enrofloxacin and a 1.0%-11.2% decrease for triclosan. Photolysis results in further decrease in dissolved-phase concentrations: 7.6% - 42.4% for enrofloxacin and 6.3% - 49.9% for triclosan. The role of sorption and photolysis depends on chemical properties, total suspended solid concentrations, flow conditions and environmental parameters such as light extinction coefficient. Both chemicals exhibit significant accumulation onto riverbeds with TCS levels exceeding its no effect concentrations for algae.