Selective removal of 17β-estradiol at trace concentration using a molecularly imprinted polymer

Abstract

A molecularly imprinted polymer (MIP) was synthesized with 17β-estradiol (E2) as template. It was then capable to recover this compound by 100±0.6% from a 2 μg/L aqueous solution. By comparison, E2 recoveries of 77±5.2%, 87.1±2.3% and 19.1±7.8%, were achieved using a non-imprinted polymer (NIP) synthesized under the same conditions (but without template), a commercial C18 extraction phase and granular-activated carbon (GAC), respectively. When fluoxetine hydrochloride and acenaphthene were added as interferences to the aqueous solution at 2 μg/L each, E2 was recovered by 95.5±4.0% from the MIP, compared to 54.5±9.4%, 76.0±2% and 14.3±0.1% from the NIP, C18 and GAC phases, respectively. Estrogenic activity equivalent to the effect caused by 22.4 ng E2/L was recorded in the MIP extract from a wastewater sample whereas no activity was detected in the NIP extract. This suggested the imprinted polymers removed estrogenic compounds. This study therefore demonstrates the potential of MIPs for the selective removal of endocrine-disrupting compounds. By using a synthetic analogue to natural hormone receptors, adsorption is based on the same property that makes the contaminants harmful. Biological treatment of enriched E2 was also demonstrated.

Introduction

Endocrine-disrupting contaminants (EDCs) can interfere with the regulatory network in humans and wildlife even when present at trace concentration of ng/L–μg/L (Daughton, 2002; Daughton and Ternes, 1999; Groshart and Okkerman, 2000; Jobling et al., 2003). They have been, for instance, linked to drastic problems such as the feminization of male fish in receiving waters (Diniz et al., 2005) and even the drop of human male fertility observed over the last decades (Golden et al., 1998). Many EDCs are natural hormones or consumer products such as pharmaceuticals and personal care products (Daughton, 2002; Daughton and Ternes, 1999). They are therefore continuously introduced in the environment from multiple diffuse sources and are ubiquitously found, at low concentration, in water resources (Cargouet et al., 2004; Fawell et al., 2001; Kolpin et al., 2002). It is precisely their potency at extremely low level combined with the large number at which they can be simultaneously found that make EDCs so worrisome, and requires the development of highly efficient water treatment methods.

Conventional wastewater and drinking water treatment processes are partially inefficient in removing EDCs (Jones et al., 2005), explaining why contaminants such as drug metabolites or human estrogens are frequently found in water resources (Stackelberg et al., 2004; Ternes et al., 2003). As typical example, adsorption to activated carbon is only efficient for hydrophobic contaminants (Ternes et al., 2002) and considerably impacted by the presence of interfering substances such as humic acids and surfactants (Fukuhara et al., 2006; Zhang and Zhou, 2005). Likewise, biological treatment of trace contaminants is often difficult because microorganisms preferentially metabolize other substances present at higher concentrations (Auriol et al., 2006; Petrovic et al., 2003). Although ozonation and advanced oxidation processes are efficient in removing many pollutants (Nakagawa et al., 2002; Ternes et al., 2002), they are also limited by the competitive removal of interfering organic matter as one study, for instance, reported that 5 mg/L of O₃ were necessary to remove 2 μg/L of nonylphenol in river water (Zwiener and Frimmel, 2000), which more than 800 times the amount theoretically required to fully mineralize this compound. Finally, the efficiency of microfiltration and ultrafiltration is limited to hydrophobic pollutants adsorbed to particles whereas nanofiltration and reverse osmosis are efficient for most compounds but are not specific and highly energy demanding (Duin et al., 2000). Hence, the lack of specificity of current methods considerably reduces their efficiency at trace concentration as most of the adsorption or oxidation capacity is wasted in the removal of other, often harmless, compounds.

With this perspective, this study presents a new method to remove EDCs by selective extraction using artificial molecular receptors synthesized by guided polymerization of functional monomers around a target molecule (the template), which leaves a specific recognition site after removal of the template (Makoto et al., 2003). By using an EDC as template, the adsorbing material synthesized becomes a synthetic analogue to the natural receptors targeted by the contaminant. Thus, pollutant removal ingeniously relies on the same property that makes the pollutants so harmful: their capacity to bind to natural receptors, which should permit to remove any molecules having the capacity to bind natural receptors (i.e. all endocrine disrupters). This would be extremely advantageous as it is impossible to identify, determine the environmental fate and understand the effects of all anthropogenic molecules (and their metabolites) entering the environment. 17β-Estradiol (E2) was chosen as model contaminant for being considered as the most active estrogen (Jobling et al., 1998) and for being frequently detected in wastewater (Ying et al., 2002).