Occurrence and fate of selected surfactants in seawater at the outfall of the Marseille urban sewerage system
- First Online:
- Cite this article as:
- Robert-Peillard, F., Syakti, A.D., Coulomb, B. et al. Int. J. Environ. Sci. Technol. (2015) 12: 1527. doi:10.1007/s13762-014-0577-0
- 122 Downloads
This paper describes an investigation of linear alkylbenzene sulfonates (LAS) and nonylphenol ethoxylates (NPEO) and their metabolites in the vicinity of the Marseille sewage outfall (wastewater treatment plant with a capacity of 1.860 million inhabitant equivalents, Northwestern Mediterranean, southeast of France). This analytical survey describes their occurrence in the subsurface and sea surface layers and investigates their possible fates in this marine environment. The results indicated the presence of LAS in both layers and up to 3 km from the discharge point, whereas the concentration of sulfophenyl carboxylic acids, which are the main metabolites of LAS, was only significant near the sewer outfall and in the surface layer. The NPEO were present only in minor quantities, especially near the discharge point, and no other selected metabolites were detected. The fate of the surfactants in question was then assessed by two types of experiments according to their potential means of degradation under natural conditions. Biodegradation assays were conducted according to a protocol defined by the United States Environmental Protection Agency (“Biodegradability in sea water, 835.3160”), with variations in the substrate input frequencies. Photodegradation experiments were carried out in a solar simulator reactor. These results demonstrated the low photodegradability and rapid primary biodegradation of LAS (with half-life times between 10.3 and 11.5 days) in the coastal area under study, although some LAS metabolites were more recalcitrant to biodegradation in this specific environment, which was also validated by linear alkylbenzene analysis in the two selected sediment stations.
KeywordsSurfactants LAS NPEO LAB Seawater Sediment Biodegradation Photodegradation
Synthetic surfactants are organic compounds that are widely used in many domestic and industrial applications. They are used as ingredients not only in detergents, cosmetics, and personal care products but also in paints, paper and leather industries, pesticides, and more. The obvious consequences of this widespread application are the very high quantities of these compounds that are discharged into wastewater treatment facilities or directly in the environment (Mungray and Kumar 2009; Tubau et al. 2010).
Another consequence is the huge variety of surfactant molecules in use, although two main groups of molecules prevail; that is, the anionics and non-ionics, which represent more than 90 % of the European consumption (Lara-Martin et al. 2008a). Among these molecules, the linear alkylbenzene sulfonates (LAS) and alkylphenol polyethoxylates (APEO) are cause for major environmental concern.
The LAS are anionic surfactants, and they are the most widely used synthetic surfactants in Europe, with an annual production of more than 400,000 tons in 2008 (CESIO, European Committee of Surfactants and their Organic Intermediates). Concerns about LAS arise from their bioaccumulation potential, their ability to enhance the apparent aqueous solubility of hydrophobic organic contaminants and their toxicity to bacteria, plants and animals (Hampel et al. 2001; Tan et al. 2010). Many studies have shown that LAS can be readily biodegraded under aerobic conditions, resulting in the formation of sulfophenyl carboxylic acids (SPC) via ω-oxidation of the terminal carbon atom of the alkyl chain and successive β-oxidations (Boudenne et al. 2001; Swisher 1987). LAS and SPC concentrations in seawater vary greatly in the literature (from μg to mg L−1), depending on the presence of treated or untreated wastewater discharges close to the sampling points (Lara-Martin et al. 2008b; Munoz et al. 2009).
With regards to non-ionic surfactants, APEOs have been banned in household cleaning products in northern Europe since 1995 but are still extensively used for industrial purposes. APEOs have a low biodegradability and form undesirable biodegradation products in wastewater treatment plants. The main types of APEO are nonylphenol polyethoxylates (NPEO), which can be transformed into various degradation products such as carboxylated metabolites (NPEC), mono- and diethoxylate derivatives (NP1E, NP2E) and nonylphenol (NP). These metabolites are considered to be endocrine-disrupting compounds that can potentially alter the normal hormonal function and physiological status of animals (Jobling et al. 1996). NP and NPEO have been detected in seawater at concentrations up to 20 μg L−1 (Bester et al. 2001; Gonzalez et al. 2004).
Investigations of surfactant and metabolite concentrations and degradation in seawater are essential to understanding the impacts of these compounds in the marine environment. Little research has been published on these compounds in seawater, other than in the region of Cadiz (Atlantic Ocean) (Gonzalez-Mazo et al. 1997; Leon et al. 2004; Perales et al. 2003). Therefore, the aim of this study was to set up a global investigation of the most worrisome surfactants (LAS and NPE) in a Mediterranean environment to assess contaminant occurrence and fate. The goal of this work was thus to determine the amounts of LAS and APEO in the seawater subsurface, along with their metabolites. The surfactant concentrations were determined in samples that were taken from two transects previously selected according the prevailing currents and winds of the study area. Finally, biodegradation and photodegradation tests were conducted to predict the fate of the detected surfactants.
Materials and methods
Materials and reagents
LAS were purchased as a linear alkylbenzene sulfonic acid mixture from Alfa Aesar (Schiltigheim, France) and simply purified by recrystallization in NaOH (2 M). HPLC analysis provided the following alkyl chain distribution for LAS: C9, 0.5 %; C10, 20.6 %; C11, 46.3 %; C12, 30.6 %; and C13, 2.0 %. The C2–C8 SPCs were prepared by sulfonating commercial phenylcarboxylic acids according to a procedure described elsewhere (Sarrazin et al. 1997). A 4-nonylphenol of PESTANAL grade (NP), Igepal CO-210 (containing nonylphenol mono-, di- and triethoxylates) and Tergitol NP-9 (a nonylphenol ethoxylate mixture with an average number of nine ethoxy units) were purchased from Sigma-Aldrich (Saint-Quentin Fallavier, France). Nonylphenoxyacetic acid (NP1EC), nonylphenoxyethoxyacetic acid (NP2EC), nonylphenol monoethoxylate (NP1E) and nonylphenol diethoxylate (NP2E) were synthesized according to a procedure described by Diaz et al. (2002) from technical grade 4-nonylphenol. All synthesized compounds were characterized by mass spectrometry and purities were checked by HPLC.
Six mL ENVI-Carb cartridges were used for the solid-phase extraction (SPE). Dichloromethane (DCM) of RPE-ACS grade was purchased from Carlo-Erba (Peypin, France), and HPLC grade tetrahydrofuran (THF) and trifluoroacetic acid (TFA) was sourced from Alfa Aesar (Schiltigheim, France). HPLC grade methanol (MeOH) was obtained from Sigma-Aldrich (Saint-Quentin Fallavier, France).
All the solutions were prepared by dissolving or diluting appropriate amounts of reagents in ultra-high quality deionized water (Millipore, resistivity >18 MΩ cm).
Study site: history and location
Sample collection and preparation
The marine samples were collected at ten different points in the Bay of Marseille (Fig. 2), following two transects: (the discharge is close to sampling point number 4). Two different samples were taken at each sampling point, one representing the surface layer (0–5 mm depth) using a surface sampler (Hydro-Bios, Kiel, Germany), and one at 2 m deep using a homemade Go-Flo-type sampling bottle. A 2 % formaldehyde solution was added to each sample immediately after sampling for conservation purposes. The samples were taken back to the laboratory and kept in the dark at 4 °C until extraction (always less than 72 h after sampling) and subsequent analysis.
The surfactants were concentrated and extracted from marine samples using a classical SPE procedure described by Di Corcia et al. (1994). In brief, the extraction was performed with graphitized carbon black (GCB) cartridges (ENVI-Carb). Before extraction the samples were acidified to a pH of approximately 3 with 1 M HCl and vigorously shaken for 20 min to ensure adequate mixing and suspension of solid particles. After adequate preconditioning of the cartridges, 100 mL of water sample were passed through the GCB cartridges, followed by a washing step with ultrapure water and methanol. The desorption of non-ionic surfactants was performed by passing 7 mL of 9/1 DCM/MeOH + 25 mM formic acid through the cartridges, and the anionic surfactants were eluted with 7 mL of 9/1 DCM/MeOH + 10 mM tetramethylammonium hydroxide. This extract was neutralized with 6 μL of HCl (12 M), and the eluates were evaporated under a gentle stream of nitrogen. Residues containing non-ionic and anionic surfactants were finally reconstituted in 1 mL of 55/45 THF/Water and 50/50 MeOH/water (concentration factor = 100), respectively, and 20 μL of this final solution was injected into the LC system. All sample analyses were performed in duplicate according to this protocol.
The surfactants were analyzed with a VWR-Hitachi L-2130 high-performance liquid chromatograph equipped with a Dionex RF2000 fluorescence detector set at λex = 229 nm/λem = 289 nm for anionic surfactants and λex = 232 nm/λem = 301 nm for non-ionic surfactants. A 25 cm × 4.6 mm i.d. column filled with 5-μm particle C8 reverse-phase packing (Supelco Discovery) was used. The column was operated at ambient temperature with a 1 mL/min flow rate.
SPCs: elution gradient from 100 % A to 100 % B in 17 min, then 100 % B for 7 min (to remove LAS from the column).
LAS: 100 % B for 15 min.
NPEC: 100 % C for 20 min.
NP–NPE: 100 % D for 13 min.
GC quantification and recoveries
Recoveries were calculated by triplicate assays with the 20 μg L−1 calibration solution to assess the efficiency of the extraction procedure. Detection limits (LOD) were calculated by using a signal-to-noise ratio of 3:1.
The biodegradation of LAS was assessed with natural seawater from the Mediterranean Sea. The seawater that was used in the biodegradation assays was sampled at the start of the experiment at sampling point number 10 (map in Fig. 2, 3.5 km from the wastewater discharge, 1,000 m from the coasts) at a depth of 2 m. The physico-chemical parameters that were measured on-site were the temperature, 18.2 °C; salinity, 36.3 g L−1; and pH 8.2.
Two assays were conducted according to the Office of Prevention, Pesticides and Toxic Substances (OPPTS) guideline 835.3160 “Biodegradability in sea water” (US.EPA 1998). The shake-flask method, which is used to verify that the study compounds are biodegradable in a marine environment, was employed with no added inoculum. According to the US.EPA protocol, the seawater was pretreated to remove coarse particles by filtering through a coarse paper filter. Assays were conducted in 1 L borosilicate glass bottles covered with aluminum foil to protect the samples from sunlight. One half liter of filtered seawater was added to each bottle and spiked to an initial concentration of 1 mg L−1 LAS. Four nutrient stock solutions were prepared. The first stock solution was composed of 8.5 g L−1 KH2PO4, 21.75 g L−1 K2HPO4, 26.6 g L−1 Na2HPO4 and 0.5 g L−1 NH4Cl. The second, third and fourth stock solutions were composed of 27.5 g L−1 CaCl2, 22.5 g L−1 MgSO4 and 0.15 g L−1 FeCl3, respectively.
Experiment A: 0.5 mL of four nutrient stock solutions was added at the beginning of the experiment (procedure described in the USEPA guideline, which represents a large amount of added nutrients).
Experiment B: 17 μL of the four nutrient stock solutions were added every day in the morning for 5 days per week (which represents regular and low additions of nutrients).
Each bottle was covered with an aluminum foil sheet and agitated at room temperature on a shaking table at 100 rpm. The room temperature was measured every day and was stable between 17 and 19 °C (average: 18 °C).
The LAS and SPC concentrations were determined by HPLC (eluents described above) by direct injection of 1 mL seawater. Calibration curves were made by direct injection of spiked water samples. When analysis was not possible within 4 h, 2 % formaldehyde was added to preserve the samples.
The photodegradation of LAS was assessed on a 5 mg L−1 LAS solution that was prepared in natural seawater (same sample as biodegradation experiments), which was sterilized by adding 2 % formaldehyde. 50 mL of this solution was transferred into a 65 mL cylindrical glass reactor (16 cm2 surface for irradiation) equipped with a magnetic stirrer. The sample temperatures were maintained at 19 ± 1 °C during the experiments by using continuous water cooling. Photolysis was conducted in a Luzchem Solar Simulator equipped with a 300 W xenon lamp (Luzchem Research Inc., Ottawa, Canada). The xenon lamp was set to generate a light intensity of 44,000 lux, which represents the annual average solar irradiation in southern France (Suri et al. 2007). The LAS and SPC were analyzed as described for the biodegradation experiments.
Linear alkyl benzene analysis
Freeze-dried sediments (of approximately 10 g) were transferred to a pre-cleaned cellulose extraction thimble and extracted with a Soxhlet extractor apparatus for 16 h with a 200 ml mixture of dichloromethane and hexane (1:1, V/V). Prior to the extraction, 1-phenlynonane was added to the samples as an internal standard (IS) for LAB quantification. All or part of the extractable organic matter (EOM) was dissolved in n-hexane and applied to a 50 % alumina/50 % silica (8 g of each, both deactivated with 5 % H2O) chromatography column (30 × 1 cm). The saturated fraction was eluted with 30 mL of n-hexane, resulting in a fraction that contained LAB. The LAB were separated by capillary gas chromatography (GC) with the following equipment: GC Autosystem XL Perkin Elmer chromatograph with on-column injection and a Perkin Elmer Elite-XLB column (30 m × 0.25 mm ID × 0.25 μm film thickness). Helium was used as the carrier gas at a constant rate of 1 mL min−1. The temperature was programmed from 70 to 285 °C (5 °C min−1) and then held for 30 min. The mass spectrometer was operated in electron impact ionization (EI) mode (70 eV) and simultaneously scanned in both full scan and selected ion monitoring modes (SIFI mode). The identification of LAB relied on both the retention times and characteristic ions. The 91 m/z ion was used for quantification, and the 105 and 119 m/z ions were used for confirmation.
Results and discussion
The separation of SPC and LAS has already been described in detail (Sarrazin et al. 1997; Leon et al. 2000; Marcomini et al. 1993) and was successfully performed with methanol and an aqueous phase containing trifluoroacetic acid (TFA) as a phase modifier (Fig. 3a, b).
NP and NPE analysis for various ethoxylate chain lengths is known to be problematic with conventional LC–FL because all the compounds co-elute into a single broad peak with methanol or acetonitrile as the organic eluent. Mass spectrometry is therefore necessary to quantify these compounds when using these types of mobile phases. However, less conventional eluents such as tetrahydrofurane (THF) provides much better NPE compound separation because of its more specific interactions with the organic eluent (NPE and THF belong to the ether compounds) (Zgola-Grzeskowiak et al. 2009). An isocratic elution using a mixture of water and THF (55/45) enabled the separation of NP, NP1E, NP2E and other NPE compounds, the latter of which were identified as a broad peak (Fig. 3C).
Finally, NP1EC and NP2EC (carboxylated biotransformation products of NPE) were analyzed using a water–methanol mixture with ammonium acetate (4 mM), which provided better separation and sensitivity, i.e., the same effects as those reported for LC–MS (Jahnke et al. 2004; Houde et al. 2002). The chromatograms of standard mixtures obtained under the reported experimental conditions are shown in Fig. 3.
Limits of detection and recoveries
Average recoveries and standard deviations from triplicate determination of 20 μg L−1 standard solutions
58 ± 4.5
85 ± 2
94 ± 4
75 ± 2.5
90 ± 2.5
95 ± 3.5
85 ± 4
85 ± 4
91 ± 5
42 ± 6
96 ± 3.5
65 ± 3
95 ± 4
77 ± 3.5
90 ± 3.5
92 ± 5
The detection limits were 0.2 μg L−1 for C2SPC, 0.1 μg L−1 for other anionic surfactants and metabolites, 0.1 μg L−1 for NP and 0.05 μg L−1 for all other non-ionic compounds. These results represent the LOD for LAS and SPC two times better than the previously reported results (Leon et al. 2000) and on the same order for non-ionic surfactants (Nunez et al. 2007). These LOD are relevant to commonly observed surfactant concentrations in coastal waters, and they enable trace analysis of these compounds.
Results of field sample analyses
The LAS were the most frequently detected surfactants in our marine samples, and they were present in the whole samples and at concentrations up to 100 μg L−1 at the closest location to the wastewater discharge (Fig. 4a). LAS contamination was also elevated at distances as far as 3 km away from this discharge point (between 20 and 30 μg L−1 for surface samples at sampling points 1 and 9), clearly demonstrating the persistence and ease of transport of these organic contaminants. On average, LAS surface concentrations are three to four times higher than those at 2 m deep. Concentrations and differences between the surface and subsurface are on the same order as those described by Leon et al. (2000) for estuarine waters in the Bay of Cadiz in Spain. With regards to the LAS homolog distributions, C11LAS and C12LAS were the most significant ones (between 60 and 70 % of the total LAS), followed by C13LAS (15–25 %) and C10LAS (10–15 %). This distribution was similar at all sampling points, regardless of the distance from the wastewater discharge area, which presumably indicates the transportation of discharged LAS mixtures before biodegradation takes place.
The SPC were only detected in surface layer samples close to the discharge point (Fig. 4B), with the highest value of 15 μg L−1 at sampling point number 4 (consistent with the above-mentioned study in the Bay of Cadiz). No more SPC could be identified at sampling point number 3 (600 m from sampling point number 4) and farther away. A dilution of SPC in the deepest seawater layers is likely to be responsible for this fast decrease in concentrations, SPC being less surface-active than LAS. The C4–C6 form accounted for more than 95 % of the total SPC.
These concentration ranges for LAS and SPC are consistent with other studies, although some SPC could be measured at 2 m deep in the field study by Leon et al. (2000), which was contrary to our results. A faster dilution in our marine environment might be responsible for this discrepancy.
When compared to anionic surfactants, non-ionics (NPE) were present only in small amounts, even in the surface layer (maximum ~2 μg L−1), and with low concentrations in the wastewater discharge area (Fig. 4c). This finding is consistent with the fact that NPEs have been banned in household cleaning products in France since 2005. The NP3–13Es were the predominant NPE compounds, especially at sampling points 1–3, which have the most intense boat traffic and which exhibited the highest NPE contamination (possibly related to boat traffic or pollution from nearby creeks). The NP and NPEC were not detected in any samples at concentrations exceeding the LOD levels. The NPEC are usually detected when significant amounts of NPE enter the wastewater treatment system (Diaz et al. 2002; Petrovic et al. 2001), thereby supporting the idea that low NPE contamination is present in the wastewaters and seawater of the area surrounding Marseille compared to other areas in the world (Jonkers et al. 2005).
SPCs were detected starting on day 8 in the first two experiments, and they reached their maximum concentrations between days 18 and 20. In experiment A, the SPC degradation nearly stopped after day 30, while it continued until the end of experiment B (day 48) in which only 10 % remained in the medium. SPC can therefore not be considered to be totally biodegradable in our specific marine environment, especially when using reference protocol A. The C2–C7 SPC were detected throughout the time of the assays, with C4, C5 and C6 being the major variants at the beginning of the degradation. At the end of the experiments, C2 and C3 were the major metabolites.
In comparing the results of the two protocols, it can be noted that the biodegradation of anionic surfactants and their metabolites is faster and more complete when nutrients are added gradually (Protocol B). In this case, microorganism degradation activity was indeed still ongoing after 48 days, whereas with protocol A, which follows EPA guidelines, the large amount of nutrients that were added at the beginning were used inefficiently after a certain experimental duration (approximately 30 days). Protocol B thus seems better suited for reproducing real marine conditions in which the nutrients are brought into the seawater in relatively stable and small amounts. This protocol may thus be used to predict the fate of surfactants in coastal environments, which are subjected to constant sewage inputs.
These results were also compared with those obtained by Perales et al. (2003) who used seawater samples from the Bay of Cadiz (Atlantic Ocean), and where the experimental temperature was set at 20 °C (18 °C in our study). Their experimental conditions were similar to those of protocol A. In their assay, the LAS degraded faster (with a half-life time of 6.2 days compared to our 11.5 days) and the SPC had totally disappeared within 42 days, showing a higher microbial degradation activity in their samples. The choice of seawater sampling in their study most likely explains the differences between their experiments and ours. Indeed, our samples were taken in an area that was far from the coast and likely with only natural marine bacteria, whereas the Perales group took their seawater samples much closer to the urban areas. It is therefore possible that their samples contained not only natural marine bacteria but also microorganisms originating from wastewater discharges. Our study seems more relevant for investigating genuine natural biodegradation in seawater.
Linear alkyl benzenes (LAB)
These results show the presence and persistence of LAS (mainly C10–C13 compounds) in the sea surface layer and at greater depths up to 3 km from the discharge point, whereas SPC (C4–C6 compounds) were only detected in the surface layer near the discharge point. The LAS therefore seem to be slowly degrading in the study area. An important result is their presence at concentrations that are three to four times higher in the surface layer than at a depth of 2 m. NPEO were measured only in minor quantities, and no other metabolites (NP, NPEC) were detected.
The degradation of LAS under controlled conditions in natural seawater shows that photodegradation under annual average solar irradiation conditions in the south of France results in no appreciable drop in the LAS concentration. The results of our biodegradation experiments under natural marine conditions are in accordance with the results from other studies and show that LAS are quickly degraded to SPC and disappear after the first 20 days. The SPC are then slowly degraded, but no complete biodegradation was observed in our study. This finding was confirmed by the LAB analysis for stations 2 and 4. Moreover, this paper also suggests that laboratory biodegradation assays should be conducted by using regular substrate inputs to be closer to the natural conditions found in coastal environments and especially to better mimic the conditions found in marine sewage discharge areas.
This research was supported by a grant from the Rhone-Mediterranean and Corsica Water Agency (Grant 2009-1316). This work was partially supported by ANR project MARSECO (ANR-CESA-018-06). We acknowledge Ms. Valentini-Poirier and Mr. Boissery for sparking our interest in this study.