The removal of two blood lipid regulators, clofibric acid (CLA) and gemfibrozil (GFZ), was evaluated using two identical aerobic membrane bioreactors with 6.5 L effective volume each. Polysulfone ultrafiltration hollow fiber membranes were submerged in the reactors. Different operating conditions were tested varying the organic load (F/M), hydraulic residence time (HRT), biomass concentration measured as total suspended solids in the mixed liquor (MLTSS) and the sludge retention time (SRT). Complete GFZ removal was obtained with F/M of 0.21–0.48 kg COD kgTSS−1 d−1, HRT of 4–10 hours, SRT of 10–32 d and MLTSS of 6–10 g L−1. The GFZ removal can be attributed to biodegradation and there was no accumulation of the compound in the biomass. The CLA removals improved with the SRT and HRT increase and F/M decrease. Average removals of 78–79% were obtained with SRT 16–32 d, F/M of 0.21–0.34 kgCOD kgTSS−1 d−1, HRT of 7–10 hours and MLTSS of 6–10 g L−1. Biodegradation was found to be the main removal pathway.

INTRODUCTION

A large number of pharmaceutical compounds of different varieties are consumed annually worldwide. Some of the more relevant ones in terms of environmental contamination are anti-inflammatory drugs, antibiotics and blood lipid regulators (Santos et al. 2010). After their administration, these compounds are partially metabolized, excreted in urine and feces, and subsequently disposed to the wastewater treatment plants (WWTPs). However, conventional WWTPs are not designed to remove hardly biodegradable or highly polar micro-pollutants such as the pharmaceutical compounds (Bendz et al. 2005). Consequently, these compounds have been detected in numerous effluents from conventional WWTPs, as well as in surface and groundwater samples (Petrovic & Barcelo 2007). Lipid regulators are frequently used to decrease the concentration of blood-circulating cholesterol and triglycerides. Clofibric acid (CLA) was the first lipid regulator reported in the environment; it is the main metabolite and the pharmacologically active component of the blood lipid-regulating pharmaceuticals clofibrate, etofibrate and etophyllinclofibrate. CLA is a scarcely biodegradable, polar compound, with a high motility, very persistent in the environment, with a half-life of 21 years and water residence time of 1–2 years; it causes endocrine disruption activity through interference with cholesterol synthesis (Buser et al. 1998; Pfluger & Dietrich 2001). CLA concentrations of 0.01–15 μg L−1 have been detected in WWTP effluents and of 0.07–0.27 μg L−1 in surface and ground waters (Herberer et al. 1997; Tauxe-Wuersch et al. 2005; Barcelo & Lopez 2007; Gagnon et al. 2008). Gemfibrozil (GFZ) is a fibric acid derivate. GFZ concentrations of 0.01–5.2 μg L−1 have been found in WWTP effluent (Rosal et al. 2010; Kosma et al. 2010) and of 0.7–1.5 μg L−1 in surface waters (Sanderson et al. 2003). According to Quinn et al. (2008), GFZ could be classified as a toxic compound. Estrogenic and carcinogenic effects had been reported for both CLA and GFZ (Mimeault et al. 2005; Urase et al. 2005; Radjenovic et al. 2009; Gross et al. 2010). The CLA and GFZ removals determined in conventional WWTPs with activated sludge systems are relatively low, of 30–60% (Clara et al. 2005; Bernhard et al. 2006; Radjenovic et al. 2009; Jelic et al. 2011).

Membrane bioreactors (MBRs) are a relatively new technology currently used for municipal wastewater treatment. This technology offers several advantages over the conventional activated sludge plants, such as operation at high biomass concentration, reduced excess sludge production and superior effluent quality. Furthermore, long sludge retention time (SRT) positively affects the overall activity of slow growing micro-organisms acting in nitrification or degradation of specific refractory pollutants. The aim of this work was to assess the removal of CLA and GFZ using MBR with submerged ultrafiltration membranes. The effects of the organic load, SRT and hydraulic residence time (HRT) on the removal effectiveness were investigated.

MATERIALS AND METHODS

Feed wastewater composition

The experiments were performed using synthetic wastewater (Table 1). Methanol was used as a major carbon source, providing a chemical oxygen demand (COD) of 800–1,000 mg L−1 in the synthetic wastewater. CLA and GFZ were added at concentrations of 0.4 and 0.8 μg L−1, respectively, in accordance with reported concentrations in wastewaters for Mexico (Gibson et al. 2007; Siemens et al. 2008). The synthetic wastewater was complemented with nitrogen, phosphorus and micronutrients; the wastewater pH was adjusted at 7 ± 0.14.

Table 1

Composition of the synthetic wastewater

Compound Concentration Units Compound Concentration Units 
Methanol 533–667 mg L−1 (NH4)6Mo7·4H20.01 mg L−1 
CLA 0.4 μg L−1 CaCl2·2H24.4 mg L−1 
GFZ 0.8 μg L−1 MgSO4·7H212.2 mg L−1 
NH4Cl 90 mg L−1 ZnSO4·7H20.132 mg L−1 
K2HPO4 mg L−1 MnSO4·H20.04 mg L−1 
KH2PO4 8.4 mg L−1 CoCl2·6H20.03 mg L−1 
FeSO4.7H217.4 mg L−1    
Compound Concentration Units Compound Concentration Units 
Methanol 533–667 mg L−1 (NH4)6Mo7·4H20.01 mg L−1 
CLA 0.4 μg L−1 CaCl2·2H24.4 mg L−1 
GFZ 0.8 μg L−1 MgSO4·7H212.2 mg L−1 
NH4Cl 90 mg L−1 ZnSO4·7H20.132 mg L−1 
K2HPO4 mg L−1 MnSO4·H20.04 mg L−1 
KH2PO4 8.4 mg L−1 CoCl2·6H20.03 mg L−1 
FeSO4.7H217.4 mg L−1    

Experimental setup and procedure

Two aerobic MBRs with submerged membranes (6.5 L effective volume) were continuously fed with synthetic wastewater. The schematic diagram of the experimental setup is presented in Figure 1. The MBRs were equipped with ultra-filtration hollow fiber membranes made of polysulfone (General Electric). The membrane had a molecular weight cutoff 100 kDa and total surface area of 0.042 m2. The reactors were operated under 3 minutes suction and 45 seconds backwashing sequential cycles which were remotely controlled using a control panel and solenoid valves. Level sensors were installed to control the feeding pumps and prevent overflows. Continuous aeration was provided using a blower. The air flow entered the reactor through a stone diffuser placed in the bottom of each reactor, the oxygen concentration was kept between 1 and 2 mg L−1.

Figure 1

Experimental setup.

Figure 1

Experimental setup.

The inoculation of the reactors was performed using biomass from the activated sludge reactor of a WWTP. The biomass was acclimated to the substrates in the synthetic wastewater by feeding continuously the reactors during 172 days at an organic load of 0.48 kgCOD kgTSS−1 d−1 in the first reactor (MBR1) and 0.21 kgCOD kgTSS−1 d−1 in the second one (MBR2). The mixed liquor total suspended solids (MLTSS) was maintained at 10,000 mg L1 in both reactors. The determination of the trace organic compounds began once the process was completely stabilized (more than 95% COD and NH4-N removals and a constant biomass growth). The operational parameters are presented in Table 2 for each experimental phase. As can be seen the organic loads (F/M) were between 0.21 and 0.77 kgCOD kgTSS−1 d−1, the HRT between 4 and 10 hours, the SRTs between 5 and 32 d. Daily biomass extractions were performed in the reactors to keep the desired MLTSS and SRT in each experimental phase. Two biomass concentrations had been tested: 6,000 and 10,000 mg TSS L−1.

Table 2

Operational parameters

Experimental phase   F/M, kgCOD kgTSS−1 d−1
 
SRT, d
 
HRT, h
 
MLTSS, mg L−1
 
reactor Operation days MBR1 MBR2 MBR1 MBR2 MBR1 MBR2 MBR1 MBR2 
Acclimatization 1–172 0.48 0.21 11 32 10 10,000 10,000 
173–256 0.48 0.21 11 32 10 10,000 10,000 
257–356 0.77 0.34 20 10 6,000 6,000 
357–400 0.43 0.26 10 16 6,000 10,000 
Experimental phase   F/M, kgCOD kgTSS−1 d−1
 
SRT, d
 
HRT, h
 
MLTSS, mg L−1
 
reactor Operation days MBR1 MBR2 MBR1 MBR2 MBR1 MBR2 MBR1 MBR2 
Acclimatization 1–172 0.48 0.21 11 32 10 10,000 10,000 
173–256 0.48 0.21 11 32 10 10,000 10,000 
257–356 0.77 0.34 20 10 6,000 6,000 
357–400 0.43 0.26 10 16 6,000 10,000 

Analysis of the trace organics CLA and GFZ

The gas chromatography-mass spectrometry method (GC-MS) was developed and implemented for CLA and GFZ determination in aqueous and biomass samples, using approaches reported in Patterson et al. (2000), Reddersen & Heberer (2003) and Gibson et al. (2007). Solid-phase extraction (SPE) was used for extracting and concentrating the compounds from all samples. Chromabond C18 EC cartridges (3 mL, 500 mg octadecyl modified silica as sorbent material) connected to a manifold (Teknokroma) and pumps as vacuum source, were used for the solid phase extraction. Liquid samples of 1 L were collected in amber glass bottles from the MBR influent and effluent. Aqueous samples were filtered through 1.0 μm glass fiber filters and adjusted to pH 2. The samples were then passed through Chromabond C18 EC cartridges which were preconditioned using 6 mL of methanol and 6 mL of acid water. The cartridges were washed with acid water (3 mL) and the target compounds were eluted from the cartridges with organic solvent (4 mL of methanol). The extracts were taken to dryness by a gentle stream of nitrogen gas, and re-dissolved in 1 mL of methanol. The extracts were stored in refrigeration before analysis. Biomass samples were collected from both MBRs. The biomass sample was submitted to the lyophilization process (−70 °C) to obtain 1 g dry weight, afterwards the sample was submitted to three sequential extractions, two with methanol and one with acetone. For each step of said extraction the sample was ultrasonicated (15 minutes) and then centrifuged (4,000 rpm, 15 minutes). The supernatants were taken and the three solvent fractions were combined and then filtered through Teflon filters (Pall Germany) of 0.45 μm. Subsequently, the sample was concentrated with nitrogen gas up to a volume of 4 mL. Before being submitted to SPE, the sample was reconstituted with 500 mL of HPLC water at pH 2.

Because of their high polarity and low volatility, the analytes required chemical derivatization prior to GC-MS analysis. Trimethylsilyldiazomethane (TSDM) was used as the derivatizing agent to form methylester derivatives. The samples were derivatizated adding 50 μL of TSDM and 100 μL of methanol, at environmental temperature for 2 hours. Chemical analysis was carried out by GC-MS (Varian CP-3800) with an ion-trap tandem mass spectrometer (Varian Saturn 2200). Standards and sample concentrates were injected using a Varian CP-8400 (Switzerland) automatic sample injector. A 30 m × 0.25 mm ID column coated with 0.25 μm chemically bonded phase VF-5 ms, 5% phenyl +95% dimethylpolysiloxane was used (Variant Technologies). Injector temperature was kept at 280 °C, under split/splitless mode. For the analysis of all compounds the oven program was as follows: 65 °C for 2 minutes, 30 °C min−1 to 180 °C, 15 °C min−1 to 230 °C for 1 minutes, 15 °C min−1 to 300 °C for 2 minutes. The detector was used predominantly in selected ion mode. The molecular ions [M]+ with m/z 128 and 143 correspond to compound derivatives of CLA and GFZ, respectively. The electron impact ion source temperature was 220 °C with electron energy of 70 eV. All quantitative results were calculated by integration of the peak areas obtained by monitoring the respective ion fragments using selective ion monitoring mode. The analytical methods for CLA and GFZ determination were validated using standard solutions in liquid and solid feces. Calibration curves, response linearity, sensitivity, limit of detection and quantification, recovery and precision of the analytical procedure were calculated. The results summarized in Table 3 indicated that the methods allow accurate, precise and reliable determination of both compounds.

Table 3

Detection limits, accuracy and precision of the analytical method

Compound Phase LD LQ RSD (%) Recovery (%) 
CLA Water 0.0026 μg L−1 0.0053 μg L−1 0.25 89 ± 2 
 Biomass 0.0014 μg g−1 0.0026 μg g−1 1.8 90 ± 2 
GFZ Water 0.0033 μg L−1 0.0071 μg L−1 0.21 106 ± 4 
 Biomass 0.0025 μg g−1 0.0033 μg g−1 1.5 86 ± 4 
Compound Phase LD LQ RSD (%) Recovery (%) 
CLA Water 0.0026 μg L−1 0.0053 μg L−1 0.25 89 ± 2 
 Biomass 0.0014 μg g−1 0.0026 μg g−1 1.8 90 ± 2 
GFZ Water 0.0033 μg L−1 0.0071 μg L−1 0.21 106 ± 4 
 Biomass 0.0025 μg g−1 0.0033 μg g−1 1.5 86 ± 4 

LD – limit of detection; LQ – limit of quantification; RSD – relative standard deviation = (standard deviation/mean value) 100.

All the target compounds as well as the internal standard 4,4′-dichlorobiphenyl and the derivatization agent TSDM (trimethylsilyl diazomethane, 2.0 mol/L solution in hexane) were purchased from Sigma–Aldrich (Gillingham, UK) with >97% purity grade. The solvents and water were HPLC grade, provided by Burdik & Jackson (Morristown NJ, USA).

Other analyses

To assess the process performance in the MBR, in addition to the CLA and GFZ analysis, COD, NH4-N and Ptotal had been determined twice a week in the influent and in the MBR effluents, as well as NO2-N, NO3-N in the effluents. TSS was determined once a week in the reactors. These parameters were determined according to Standard Methods (APHA 2012). Dissolved oxygen and temperature were measured daily.

RESULTS AND DISCUSSION

Process performance

During the acclimatization period, COD removals higher than 95% were reached in almost 1 week and a constant biomass growth was obtained in 30 days; however, the NH4-N removal increased slowly and 95% was reached at day 119 after the startup. The determination of the CLA and GFZ concentrations was started at day 174, the day when the Phase 1 evaluation began. The results obtained during the three experimental phases are presented in Figure 2. As can be seen, the CLA concentration decreased slightly over time and the average removals for days 209–256 were 65 ± 10% and 78 ± 2% in MBR1 and MBR2, respectively. The higher removal obtained in MBR2 can be attributed to the higher SRT and lower F/M applied in this reactor. Meanwhile the GFZ could be completely removed in both reactors despite the different operation conditions. During the first experimental stage the average COD and NH4-N removals were between 98 and 99%, average NO3-N concentrations were 6 and 8 mg L−1 in the effluents of MBR1 and MBR2, respectively. These results indicated a good process performance in both reactors.

Figure 2

CLA and GFZ concentrations during the MBR operation: (a) MBR1; (b) MBR2.

Figure 2

CLA and GFZ concentrations during the MBR operation: (a) MBR1; (b) MBR2.

The F/M increase performed on day 257 caused an increase of the CLA and GFZ concentrations in the effluents of both reactors. The CLA concentrations increased in MBR1 while they decreased gradually over time in MBR2. The F/M increase to 0.77 kgCOD kgTSS−1 d−1 and the SRT decrease to 5 d affected the CLA removal in MBR1 during the second experimental phase. The average CLA removal was 67 ± 8% in MBR1. Almost 50 days were required to reach again the CLA concentrations already obtained in MBR2 effluent during the first phase. The average CLA removal for days 300–356 was 79 ± 7% in MBR2, similar to the one obtained during the first phase. The highest CLA removal of 83% was reached on day 356. The results suggest that the F/M increase to 0.34 kgCOD kgTSS−1 d−1 and the SRT decrease to 20 d did not affect the CLA removal, as only a period of time was required for the acclimation to the new operational conditions. GFZ concentrations decreased gradually in both reactors over time and 100% removal was reached at day 300 and day 276 in MBR1 and MBR2, respectively. This means that 36 and 12 days were required to reach the removal capacity already obtained in MBR1 and MBR2. The average GFZ removals were 99 ± 1% (for days 292–256) in MBR1 and 100% (for days 276–356) in MBR2. The operation conditions slightly affected the GFZ removal in MBR1 during the second phase, while they did not affect the removal in MBR2. During the second experimental stage the average COD and NH4-N removals were between 99.3 and 99.9%, average NO3-N concentrations were 5 and 6 mg L−1 in the effluents of MBR1 and MBR2, respectively. The biological process was stable and a good performance was determined in both reactors.

The organic loads were reduced at the beginning of the third experimental phase. The SRT was increased in MBR1 and decreased in MBR2. HRT was increased in MBR1 and it was decreased in MBR2. The new operational conditions enhanced the CLA removal in MBR1, while a slight concentration increase was observed in the effluent from MBR2, which can be attributed to the SRT and HRT decreases. After 30 days of operation an 84% removal was reached in MBR2. The average CLA removals were calculated of 71 ± 6% and 78 ± 6% (days 377390) in MBR1 and MBR2, respectively. The GFZ was completely removed in both reactors despite the different operation conditions. During the third experimental stage the average COD and NH4-N removals were between 99.5 and 99.9%, average NO3-N concentrations were 5 ± 1 mg L1 in the effluents of both reactors; so the process performance was also good during the third experimental stage.

The effects of the operation parameters on the CLA and GFZ removals are summarized in Table 4. The CLA removal decreased with the organic load increase from 0.34 to values in the range of 0.43–0.77 kgCOD kgTSS−1 d−1. The average CLA removals were 65–71% with SRT of 511 d and 78–79% with SRT of 16–32 d. Thus, the SRT increase enhanced the CLA removal. These results were consistent with Kimura et al. (2005), as they reported 80% of CLA removal in a MBR with SRT of 65 d and lower removal (50%) when the STR decreased to 15 d. Other authors such as Quinn et al. (2008), Radjenovic et al. (2009) and Sipma et al. (2010) had also observed that better removals of pharmaceutical compounds could be obtained with higher SRT. The HRT increase favored the CLA removal; removals of 78–79% were obtained with HRT of 10 hours. The lowest removals of CLA and GFZ (65% and 99%, respectively) were obtained with HRT of 4 hours. Kosjek et al. (2007) indicated that the increase of HRT enhanced the biodegradation, results that agree with this study. Nevertheless they obtained CLA removal of only 29% using pilot scale MBR with an HRT of 48 hours and SRT of 20 d.

Table 4

Effects of the operation parameters on CLA and GFZ removals in MBR

SRT, d F/M, kgCOD kgTSS−1 d−1 HRT, h MLTSS, mg L−1 CLA removal (%) GFZ removal (%) 
0.77 6,000 67 ± 8 99 ± 1 
10 0.48 10,000 65 ± 10 100 
11 0.43 6,000 71 ± 6 100 
16 0.26 10,000 78 ± 6 100 
20 0.34 10 6,000 79 ± 7 100 
32 0.21 10 10,000 78 ± 2 100 
SRT, d F/M, kgCOD kgTSS−1 d−1 HRT, h MLTSS, mg L−1 CLA removal (%) GFZ removal (%) 
0.77 6,000 67 ± 8 99 ± 1 
10 0.48 10,000 65 ± 10 100 
11 0.43 6,000 71 ± 6 100 
16 0.26 10,000 78 ± 6 100 
20 0.34 10 6,000 79 ± 7 100 
32 0.21 10 10,000 78 ± 2 100 

GFZ removal of 99% was only obtained with SRT of 5 d and F/M of 0.77 kgCOD kgTSS−1 d−1. The rest of the SRT values (10–32 d) with F/M of 0.21–0.48 kgCOD kgTSS−1 d−1 allowed for a complete GFZ removal. Regarding the effect of HRT it was observed that complete GFZ removal was obtained with HRT of 7 and 10 hours. Complete GFZ removal was also obtained with HRT of 4 hours when the SRT was 10 d, F/M of 0.48 kgCOD kgTSS−1 d−1 and MLTSS of 10,000 mg L−1. The GFZ removal was 99% applying HRT of 4 hours when the SRT was 5 d, with F/M of 0.77 kgCOD kgTSS−1 d−1 and MLTSS of 6,000 mg L−1. These results are better than the ones reported by Radjenovic et al. (2009) for MBR, of 30–40% with HRT of 7–15 hours and mixed liquor volatile suspended solids (MLVSS) of 1,4008,400 mg L1. GFZ removals of 20–60% had been determined by Gross et al. (2010) and Jelic et al. (2011) in conventional activated sludge systems.

CLA and GFZ concentrations in biomass samples

Because of the low values of the Henry coefficients of the studied compounds, the major removal mechanisms during the treatment are biodegradation and sorption onto the biomass. The determinations of the CLA and GFZ concentrations in biomass samples from both MBRs were, performed to elucidate the sorption contribution to the removal, started at day 129. The results obtained are presented in Figure 3. As can be seen, the GFZ content in the biomass decreased over time, reaching values lower than the detection limits at day 209 after the start up. The slow decrease (209 days) can be explained by the adjustment of the biomass consortia and metabolism to the biodegradation of the micro-pollutants, which is limited by their low concentrations and difficult biodegradability.

Figure 3

CLA and GFZ concentrations in the biomass of MBR1 (a) and MBR2 (b).

Figure 3

CLA and GFZ concentrations in the biomass of MBR1 (a) and MBR2 (b).

The decrease of the GFZ concentration was slightly faster in the biomass of MBR2, which can be attributed to the higher SRT applied in this reactor during the acclimatization and the first evaluation phase. At day 209 the GFZ concentrations in the effluents from both MBR were also lower than the detection limit. During the experimental phases 2 and 3, the GFZ content in the biomass of both MBRs remained below the detection limits, which indicated that after the acclimatization, the major mechanism of the GFZ removal was the biodegradation.

The CLA content in the biomass also decreased over time (slightly faster in MBR2), however concentrations lower than the detection limit were reached at day 256. During the first evaluation stage, the average CLA concentration in the biomass was 0.05 mg kg−1 in both MBRs. During the next experimental phases, the CLA concentrations were lower than the detection limit most of the time, except days 296 and 385 when concentrations of 0.03 and 0.15 mg kg−1 had been found in MBR2 and day 342 when a concentration of 0.02 had been determined in MBR1. The mass balance performed for CLA indicated that the removal due to the extractions of biomass with sorbed CLA was only 1.0–2.5% when the best operational conditions were applied (HRT of 10 hours and STR of 20–32 d) and the CLA removals were 78–79%. The contribution of the sorption onto the biomass was 1.5–3.6% when low HRT and SRT were applied (4 hours and 5–11 d, respectively) and CLA removals were 65–71%. This means that the major mechanism of the CLA removal was also the biodegradation. The CLA was however harder to biodegrade compared with the GFZ and small CLA can remain sorbed onto the biomass even operating with HRT of 10 hours and SRT of 32 days.

CONCLUSIONS

The MBR with submerged ultrafiltration membranes are able to remove the blood lipid regulator compound GFZ present in wastewaters. SRT of 10 d and HRT of 4 hours would be sufficient for complete GFZ removal, under organic load of 0.48 kg COD kgTSS−1 d−1 and MLTSS of 10,000 mg L−1. Slightly greater SRT and HRT, of 11 d and 7 hours, respectively, would be required when MLTSS is 6,000 mg L−1. The gemfibrozil removal can be attributed to biodegradation and there is no accumulation of the compound in the biomass. The scarcely biodegradable pharmaceutical compound CLA could be removed up to 84% using MBR with submerged ultrafiltration membranes. The SRT and HRT increases enhance the CLA removal. Average removals of 78–79% were obtained with SRT 16–32 d, F/M of 0.21–0.34 kgCOD kgTSS−1 d−1, HRT of 7–10 hours and MLTSS of 6,000–10,000 mg L−1. Based on the CLA content determined in the biomass, biodegradation was found to be the main removal pathway, as the contribution of the sorption was estimated to be of only 1.0–2.5%.

REFERENCES

REFERENCES
American Public Health Association (APHA)
2012
Standard Methods for the Examination of Water and Wastewater
,
22nd edn
.
APHA-AWWA-WEF
,
Washington, DC
,
USA
.
Barceló
D.
López
A.
2007
Contaminación y calidad química del agua: el problema de los contaminantes emergentes. Fundación Nueva Cultura del Agua. Panel científico-técnico de seguimiento de la política de aguas. Universidad de Sevilla-Ministerio de Medio Ambiente. 28 pp
.
Bendz
D.
Paxeus
N. A.
Ginn
T. R.
Loge
F. J.
2005
Occurrence and fate of pharmaceutically active compounds in the environment, a case study: Höje River in Sweden
.
Journal of Hazardous Materials
122
(
3
),
195
204
.
Gagnon
C.
Lajeunesse
A.
Cejka
P.
Gagné
F.
Hausler
R.
2008
Degradation of selected acid and neutral pharmaceutical products in a primary-treated wastewater by disinfection processes
.
Ozone: Science and Engineering
30
,
387
392
.
Herberer
T.
Dunnbier
U.
Reilich
C.
Stan
H. J.
1997
Detection of drugs and drug metabolites in ground water samples of a drinking water treatment plant
.
Fresenius Environmental Bulletin
6
(
7–8
),
438
443
.
Jelic
A.
Gross
M.
Ginebrada
A.
Cespedes-Sánchez
V.
Petrovic
M.
Barcelo
D.
2011
Occurrence, partition and removal of pharmaceuticals in sewage water and sludge during wastewater treatment
.
Water Research
45
,
1165
1176
.
Kosjek
T.
Heat
E.
Kompare
B.
2007
Removal of pharmaceutical residues in a pilot plant wastewater treatment plant
.
Analytical and Bioanalytical Chemistry
387
,
1379
1387
.
Kosma
C.
Lambropoulou
D.
Albanis
T.
2010
Occurrence and removal of PPCPs in municipal and hospital wastewaters in Greece
.
Journal of Hazardous Materials
179
,
804
817
.
Patterson
D.
Brumley
W.
Kellier
V.
Ferguson
L.
2000
Determination of clofibric acid in sewage effluent by GC/MS: conversion to methyl ester with trimethylsilildiazomethane
.
American Laboratory
34
,
20
28
.
Petrovic
M.
Barceló
D.
2007
Analysis, Fate and Removal of Pharmaceuticals in the Water Cycle
.
Elsevier
,
Amsterdam
,
The Netherlands
.
Pfluger
P.
Dietrich
D.
2001
Effects on pharmaceuticals in the environment an over view and principle considerations
. In:
Pharmaceuticals in the Environment Sources, Fate, Effects and Risks
(
Kümmerer
K.
, ed.).
Springer
,
Berlin
, pp.
11
17
.
Rosal
R.
Rodea
I.
Boltes
K.
Fernandez
P.
Leganés
F.
Gonzalo
S.
Petre
A.
2010
Ecotoxicity assessment of lipid regulators in water and biologically treated wastewater using three aquatic organisms
.
Environmental Science and Pollution Research
17
,
135
144
.
Santos
L.
Araújo
A. N.
Fachini
A.
Pena
A.
DelerueMatos
C.
Montenegro
M.
2010
Ecotoxicological aspects related to the presence of pharmaceuticals in the aquatic environment
.
Journal of Hazardous Materials
175
,
45
95
.
Sipma
J.
Osuna
B.
Collado
N.
Monclús
H.
Ferrero
G.
Comas
J.
Rodriguez-Roda
I.
2010
Comparison of removal of pharmaceuticals in MBR and activated sludge systems
.
Desalination
250
,
653
659
.
Tauxe-Wuersch
A.
De Alencastro
L.
Grandjean
D.
Tarradellas
J.
2005
Occurrence of several acidic drugs in sewage treatment plants in Switzerland and risk assessment
.
Water Research
39
,
1761
1772
.