Oxidative and photochemical processes for the removal of galaxolide and tonalide from wastewater

Abstract

Synthetic musks have been reported in wastewaters at concentrations as high as tens of micrograms per litre. The two most significant polycyclic musk fragrance compounds are 1,3,4,6,7,8-hexahydro-4,6,6,7,8,8-hexamethylcyclopenta(g)-2-benzopyran (HHCB, trade name galaxolide^®) and 7-acetyl-1,1,3,4,4,6-hexamethyltetrahydronaphthalene (AHTN, trade name tonalide^®). We report the result of several irradiation and advanced oxidation processes carried out on samples of the effluent of a wastewater treatment plant located in Alcalá de Henares, Madrid. Wastewater samples were pre-ozonated and spiked with 500 ng/L of tonalide or galaxolide in order to obtain final concentrations in the same order as the raw effluent. The treatments assayed were ozonation with and without the addition of hydrogen peroxide (O₃, O₃/H₂O₂), ultraviolet (254 nm low pressure mercury lamp) and xenon-arc visible light irradiation alone and in combination with ozone (UV, O₃/UV, Xe, O₃/Xe) and visible light photocatalytic oxidation using a Ce-doped titanium dioxide photocatalyst performed under continuous oxygen or ozone gas bubbling (O₂/Xe/Ce–TiO₂, O₃/Xe/Ce–TiO₂). In all cases, samples taken at different contact times up to 15 min were analyzed. An analytical method based on stir bar sorptive extraction (SBSE), followed by comprehensive two-dimensional gas chromatography (SBSE-GC × GC–TOF-MS), was used for the automatic searching and evaluation of the synthetic musks and other nonpolar or semipolar contaminants in the wastewater samples. In all cases tonalide was more easily removed than galaxolide. The best results for the latter (more than 75% removal after 5 min on stream) were obtained from ozonation (O₃) and visible light photocatalytic ozonation (O₃/Xe/Ce–TiO₂). A significant removal of both pollutants (∼60% after 15 min) was also obtained during visible light photocatalysis (O₂/Xe/Ce–TiO₂). UV radiation was able to deplete tonalide (+90%) after 15 min but only reduced the concentration of galaxolide to about half of its initial concentration. The toxicity of treated samples decreased for O₃/UV and O₃/Ce–TiO₂, but increased during irradiation processes UV, Xe and Xe/Ce–TiO₂. Ozone treatments tend to decrease toxicity up to a certain dosage, from which point the presence of toxic transformation products has adverse effects on aquatic microorganisms.

Introduction

Synthetic musks are currently manufactured and used in large amounts for incorporation in a wide variety of personal care products. Their role is to fix fragrances by slowing down the release of volatiles (Reiner and Kannan, 2006). Synthetic musks comprise a wide variety of substances, of which only nitro, polycyclic and macrocyclic derivatives have achieved any commercial importance. They are artificial substances which do not occur in the nature and do not share chemical or structural relationship with natural musks. Currently, nitro musks, a family comprising musk xylene and musk ketone, are being gradually displaced by polycyclic derivatives, which account for about three-quarters of the worldwide market (Sommer, 2004). Of these, 1,3,4,6,7,8-hexahydro-4,6,6,7,8,8-hexamethylcyclopenta(g)-2-benzopyran (HHCB, galaxolide^®) and 7-acetyl-1,1,3,4,4,6-hexamethyltetrahydronaphthalene (AHTN, tonalide^®) are by far the two largest volume products, representing about 95% of the EU market and 90% of the US market for all polycyclic musks (Clara et al., 2011).

The widespread use of synthetic musks leads to their release to the environment in large amounts through the discharge of wastewater treatment plants (WWTP). Rosal et al. (2010a) measured values of up to 25 μg/L of HHCB and 1.9 μg/L of AHTN in the influent of a wastewater treatment plant, very high figures compared with the 30 ng/L found for celestolide and 6 ng/L for traseolide. As a consequence, HHCB is one of the most abundant xenobiotic in untreated wastewater together with caffeine, its metabolite paraxanthine and the non-steroidal anti-inflammatory drug acetaminophen. According to their non-polar character, the removal efficiency of these compounds is high in conventional activated sludge WWTP, but their high concentration still leads to occurrences in the microgram per litre range in treated effluents (Bester, 2004; Ternes et al., 2007; Lv et al., 2010). In comparison, the average concentration of musk xylene and musk ketone have been found in the 50–100 ng/L range (Rosal et al., 2010a). Synthetic musks have also been detected in similar concentrations in environmental samples (Gómez et al., 2012). Bester (2005) reported concentrations of up to 600 ng/L of polycyclic musks in the River Ruhr at the discharge point of a WWTP. In the part of the same river where drinking water is extracted, he found 60 ng/L HHCB and 10 ng/L AHTN. These are typical values for rivers in populated areas. In Korean surface waters Lee et al. (2010) reported concentrations in the 100–272 ng/L range for HHCB and 30–52 ng/L for AHTN. Synthetic musks are persistent compounds, a fact stressed by Schmid et al. (2007) who investigated their presence in lipid-based fish tissue from remote Alpine lakes. Although at levels far below those for fish from lakes and rivers receiving effluents of WWTP, HHCB and AHTN as well as other nitro and polycyclic musks were nonetheless still detectable on account of their entrance by atmospheric precipitation.

It has been shown that the biotransformation of HHCB into HHCB-lactone takes place in WWTP operating with activated sludge (Bester, 2004). The same author reported 20–30 ng/L HHCB-lactone in the River Ruhr at considerable distances from WWTP discharge points (Bester, 2005). HHCB-lactone was found in all samples taken from the River Henares (Madrid, Spain), with a profile along the river course similar to its parent compound HHCB. HHCB-lactone was found in the water samples at a comparable concentration or even higher than those for HHCB (Gómez et al., 2012). This metabolite has been identified in fish and water samples by other groups (Franke et al., 1999; Kallenborn et al., 2001). HHCB-lactone was found in all wastewater samples in the United States at concentrations in the 146–4000 ng/L range and in sludge at 3.16–22.0 mg/kg dry weight (Reiner et al., 2007). Besides its biological origin, some HHCB-lactone is included in the technical galaxolide product, so it can also be found at the plant entrance. While HHCB and AHTN concentrations decreased during treatment, the concentrations of HHCB-lactone increased in water after treatment from an average of 505 ng/L (897 ng/L) in the influent to 1620 ng/L (1740 ng/L) in the effluent of two WWTP receiving domestic (domestic and industrial) discharges in New York State (Reiner et al., 2007). Recently, Janzen et al. (2011) identified HHCB-lactone as transformation product from the ozonation of HHCB in pure water.

In spite of concern due to their persistence and potential to bioaccumulate, the toxicity and environmental risks of these chemicals have been usually regarded as low (Salvito, 2005). However, it has been demonstrated that nitro musks can behave as co-genotoxicants (Mersch-Sundermann et al., 2001) and musk xylene has been included in the Candidate List of Substances of Very High Concern of the European Chemicals Agency. Luckenbach and Epel (2005) showed that nitro and polycyclic musks can inhibit the activity of multidrug efflux transporters responsible for multixenobiotic resistance in the gills of a marine mussel (Mytilus californianus). Nitro musks were more effective inhibitors than polycyclic musks, with combined IC₅₀ values of 0.82 ± 0.53 μM for nitro musks and 2.34 ± 0.82 μM for polycyclic musks. Moreover, a long term inhibition of efflux transporters results in the continued accumulation of normally excluded toxicants even after direct exposure to the musk has ended.

In this work, an analytical method based on stir bar sorptive extraction (SBSE), followed by comprehensive two-dimensional gas chromatography (SBSE-GC × GC–TOF-MS) was used for the evaluation of synthetic musks (Baltussen et al., 1999). This technique has recently been proposed as a tool for the multiresidue analysis of priority and emerging contaminants in waters with excellent results in terms of separation efficiency, analysis time and detection limits (Prieto et al., 2007; Gómez et al., 2011). The depletion of HHCB and AHTN in biologically treated wastewaters spiked with environmentally relevant concentrations of both compounds was tracked. The treatments assayed were ozonation with and without the addition of hydrogen peroxide (O₃, O₃/H₂O₂), ultraviolet (254 nm low pressure mercury lamp) and xenon-arc visible light irradiation alone and in combination with ozone (UV, O₃/UV, Xe, O₃/Xe) and visible light photocatalytic oxidation using a cerium-doped titanium dioxide photocatalyst performed under continuous oxygen or ozone gas bubbling (O₂/Xe/Ce–TiO₂, O₃/Xe/Ce–TiO₂). In all cases, we analyzed samples taken at different contact times up to 15 min and monitored their toxicity for the microcrustacean Daphnia magna, the green alga Pseudokirchneriella subcapitata and the marine bacterium Vibrio fischeri.