Efficiency of UV-Oxidation in Removal of Pharmaceuticals from Waster Water Samples and Toxicological Evaluation Before and After the Oxidative Treatment

Abstract

To gain biologically and chemically safe conditions in aquatic systems, as required by the EU Water Framework Directive, advanced oxidation processes are the only conceivable possibility to remove pharmaceuticals and their metabolites from waste water. Water treatment plants are not designed for the complete mineralization of these substances, so that oxidation by-products occur, whose chemical and toxicological properties are often unknown. In this study an UV-oxidation system with and without H2O2 was tested to optimize the elimination of pharmaceuticals, and replenished with toxicological investigations of untreated and oxidatively treated samples. The substances carbamazepine, metoprolol, amidotrizoic acid, diclofenac, and sulfamethoxazole were used as lead compounds. They were studied separately and as a substance mixture. The UV-system had a low-pressure lamp with an emission wavelength of 254 nm and a maximum UV output of 15 W. The sample volume was 1 L with a flow rate of 0.03 m³ h−1. Different matrices (waste water, ultrapure water) were used to detect possible matrix effects of real samples. The degradation rates of the substances were analyzed and quantified using LC-MS/MS with an API 3000 and a Q TRAP 3200. The comparison revealed that the additional use of H2O2 enhances the degradation significantly, especially in waste water samples. Cyto- and genotoxicity were tested with the MTT Assay and Alkaline Comet Assay, respectively, using Chinese Hamster Ovary cells. Generally, samples treated only with UV showed no toxic effects. Toxic results occurred only for samples with the substance mixture that were treated with UV+H2O2, indicating that combination effects might lead to the formation of toxic by-products.

Full text

EFFICIENCY OF UV-OXIDATION IN REMOVAL OF PHARMACEUTICALS
FROM WASTER WATER SAMPLES AND TOXICOLOGICAL EVALUATION
BEFORE AND AFTER THE OXIDATIVE TREATMENT
H. Bielak1, A. Boergers2, J. Raab, J. Tuerk2, E. Dopp1
1IWW Rheinisch-Westfälisches Institut für Wasserforschung gGmbH, Mülheim,
Germany
2Institut für Energie- und Umwelttechnik e. V., IUTA, (Institute of Energy and
Environmental Technology e. V.) Duisburg, Germany
1 INTRODUCTION
Conventional waste water treatment plants are not able to remove micropollutants
completely, resulting in the influent of those substances into surface waters. Especially
pharmaceuticals are very often resistant against the biological degradation because of
their polarity and persistence [1] [2]. As components in waste water treatment plant
(WWTP) effluents these substances can influence the aquatic environment (surface- and
ground water) and cause adverse effects in aquatic organisms [3]. Low concentrations of
pharmaceuticals can also be detected in drinking water which is of particular importance
for human exposure [4]. Moreover, according to the WHO it is expected that additive
toxicological effects of the different micropollutants may occur [5].
According to the German Federal Environment Agency [6], the yearly consumption of
pharmaceuticals, measured as active pharmaceutical ingredients (API), exceeds 25000
tons. Concentration ranges of API in surface water in German rivers are up to the μg L-1
range [6]. The highest influent of API into waste water is due to private households,
hospitals and retirement homes [7]. Hospital effluents are a main source for the
discharge of very polar and biological stable iodated contrast media (ICM) into the
sewage system [8] [9]. Because of their persistence and low sorption ability ICM can be
found in ground- and drinking water in concentrations up to 1 μg L-1 [10].
Concentrations of pharmaceuticals in waste water are related to the pharmacokinetic
properties of these substances [11]. Ternes et al. [12] have shown that concentrations of
pharmaceuticals in waste water from municipal WWTP effluents correlate with the
pharmaceutical concentrations that were detected in the receiving water from those
WWTP.
In the present study five substances were investigated: (1) amidotrizoic acid with a mean
concentration of 2.6 μg L-1 in WWTP effluents [13], (2) the anti-epileptic agent
carbamazepine which is persistent in the environment and is verified as one of the most
frequent pharmaceuticals in surface water with a maximum concentration of 1.8 μg L-1
[13] [14], (3) the analgetic diclofenac which is used as non-steroidal anti-rheumatic
agent [15], (4) the beta-blocker metoprolol which was detected in WWTP effluents at a
concentration of 0.48 μg L-1 and in surface water at a concentration of 0.017 μg L-1 [13]
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online
View Journal
| View Issue

Disinfection By-products Prevention and Treatment 181
and (5) sulfamethoxazole, a bacteriostatic acting antibiotic that occurs in WWTP
effluents and surface waters at a mean concentration of 0.33 and 0.013 μg L-1,
respectively [13]. For example at conventional WWTPs the maximum elimination rate
for carbamazepine is 7% [14].
Further removal of these substances from waste water is possible e. g. by using advanced
oxidation processes (AOP) [16] [17]. However, several studies have shown that
formation of oxidation by-products is possible. Information about these transformation
products are rare, especially their toxicological properties are often unknown [1] [18]
[19].
AOP methods are able to reduce the concentration of pharmaceutical residues in
municipal WWTP effluents to the protective goal of 1 ng L-1 which is given by the
European Medicines Agency (EMA) [2]. The most applied processes are UV/H2O2,
O3/H2O2 and O3/UV [17].
Chemical analyses are able to detect transformation products but cannot give any
information about the toxicity of oxidation by-products. Therefore, toxicological testing
systems (bioassays) are a necessary supplement to the chemical analysis [13].
The aim of the present study was the investigation of the efficiency of an UV-oxidation
system using different matrices and the toxicological evaluation of the detected
substances.
2 METHODS
2.1 Sample preparation
The degradation of amidotrizoic acid, carbamazepine, diclofenac, metoprolol and
sulfamethoxazole (Sigma-Aldrich, Germany) by oxidation was measured for the
individual substances and for a substance mixture consisting of 100 mg L-1 of each
substance. In the present study toxicological evaluations were performed for the
substance mixture of amidotrizoic acid, carbamazepine and ciclofenac.
For the chemical measurement the samples were transferred to autosampler-vials using a
syringe filter. The solid phase extraction (SPE) of the samples with ENV+ (Biotage,
Germany) and Strata XL (Phenomenex, Germany) cartridges was performed by a
Gilson-System (Gilson International BV, Germany).
Before the toxicological examinations all samples were sterile filtrate (pore size 0.2 μm)
to remove microbial contaminations.
2.2 Determination of the degradation rate during UV-oxidation
In the laboratory system a low-pressure lamp with an emission wavelength of 254 nm
and a maximum UV output of 15 W (TNN 15/32, Heraeus, Germany) was used. The
sample volume was 1 L with a flow rate of 0.03 m³ h-1. The low-pressure UV lamp
needed 5 minutes to reach its maximum output. For the UV oxidation experiments a
single substance or the substance mixture was added to HPLC water or the WWTP
effluents. After 15 minutes of mixing, the UV lamp was activated. After 2-10 minutes
the maximal efficiency was reached. Samples were taken before oxidation and at regular
time intervals during the oxidation process. The oxidation by an UV/H2O2 system was
also evaluated. For these experiments 0.3 g L-1 H2O2 was added after the warm-up phase
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

182 Disinfection By-products in Drinking Water
of the UV lamp of 5 minutes. The residual peroxides in the samples after oxidation were
removed by the enzyme catalase from Aspergillus niger.
The degradation of each substance was analyzed using LC-MS/MS with an API 3000
and a Q TRAP 3200 (AB Sciex, Germany).
To calculate the degradation rate for the reactions a kinetic of first order was assumed.
The half-life period is then calculated by means of the reaction speed, i.e. the positive
slope of the correlation curve.
2.3 Cytotoxicity and genotoxicity
Toxicological investigations were performed for samples from the laboratory scale UV-
system. Samples with the substance mixture were tested up to 60 min of oxidative
treatment time, while samples with a single substance (amidotrizoic acid, carbamazepine
and diclofenac) were tested up to 30 min after the oxidation.
All samples and controls were sterile-filtrated (hydrophilic, pore size 0.2 μm) before
they were applied in cell culture tests. Samples treated with H2O2 were checked for
residual peroxides with peroxide test sticks (semi-quantitative, detection limit 0.5 – 25
mg L-1) to exclude false positive results in the cyto- and genotoxicity testing.
Chinese Hamster Ovary cells (CHO-9) (provided by ECACC, UK) were exposed to the
test samples in a 1:10 dilution (c = 10 μg L-1) for 24 h. The culturing of the cells was
performed in accordance with standard procedures. CHO 9-cells were grown in HAM’s
F12 medium at 37 °C, 5 % CO2 and 95 % RH.
Cytotoxicity was tested using the MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-
diphenyltetrazoliumbromid) assay. This test detects cytotoxic effects depending on the
mitochondrial activity of the cell. A total of 104 cells were seeded in 200 μL growth
medium per well in a 96-well plate. After one day of cell growth and additional 24 h of
exposure (samples were tested in triplicate), MTT solution was added to the wells for
two hours. MTT is a soluble, yellow tetrazolium salt that can be transformed to an
insoluble, purple formazan in the mitochondria of intact cells. After lysis of the exposed
cells the viability can be determined by a photometrical measurement. Therefore the
medium was removed after the incubation time, cells were lysed with a lysis solution for
15 minutes and the absorption at 595 nm was then measured immediately. For the
determination of the cell viability relative values were calculated by correlating the
absorption values of the samples to the absorption value of the negative control
(unexposed cells, 100 % viability). MMA III (monomethylarsenic acid (III), 50 μM) was
used as positive control.
To test the genotoxicity of the samples the Alkaline Comet Assay (“single cell gel
electrophoresis”) was performed according to the protocols of Ostling and Johanson [20]
and Singh et al. [21]. In this test, DNA single and double strand breaks of eukaryotic
cells can be detected through the migration of DNA in an electrical field during gel
electrophoresis, whereby smaller (broken) fragments move faster leading to the
formation of a so called comet tail that can be analyzed under the microscope after
staining of the DNA. Briefly, 104 cells were seeded in 2 mL growth medium per well in
a 24-well plate. After one day of cell growth and additional 24 h of exposure (samples
tested in duplicate), 8000 cells were re-suspended in 20 μL phosphate buffered saline,
mixed with low melting agarose (0.75 %, 37 °C) and transferred to previously prepared
agarose mini gels on ice. After incubation of the gels in lysis solution at 4 °C overnight,
the gels were incubated in cooled alkaline electrophoresis solution to unwind the DNA
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

Disinfection By-products Prevention and Treatment 183
with subsequent electrophoresis for 20 min (300 mA/ 25 V) followed by neutralization
and fixation of the gels. The dried gels were stained with SYBR green solution (0.001
%), fixed on microscopic slides and manually evaluated under a fluorescence
microscope (Leitz GmbH & Co. KG, Germany). The positive control used in the
Alkaline Comet Assay was N-Ethyl-N-Nitrosourea (ENU, 100 μg L-1, 30 min).
3 RESULTS
3.1 Degradation of the substances
All individual substances in HPLC-water were completely eliminated during the
oxidation experiments. The degradation rates of amidotrizoic acid and sulfamethoxazole
could not be determined because the degradation of these substances was too fast. The
degradation rates differed depending on the use of H2O2. The degradation rate for
carbamazepine was 8.6 min L-1 without H2O2 and 5.0 min L-1 with H2O2, showing an
enhancement of the oxidation by adding H2O2. For diclofenac no enhancement of the
oxidation was observed after the addition of H2O2.
The treatment of the substance mixture in HPLC-water showed similar trends in the
degradation (Figure 1). The fastest degradation was observable for amidotrizoic acid,
diclofenac and sulfamethoxazole. The degradation rate for amidotrizoic acid without
H2O2 was 0.8 min L-1 and 0.7 min L-1 with H2O2. For sulfamethoxazole the rates without
and with H2O2 are 1.1 and 1.02 min L-1, respectively. Carbamazepine and metoprolol
were more stable and the elimination was slower. Here the degradation rate for
metoprolol was enhanced from 30.8 min L-1 without H2O2 to 3 min L-1 with H2O2 and
for carbamazepine from 34 min L-1 to 3.5 min L-1 (Figure 1).
Figure 1 Degradation curves of amidotrizoic acid, carbamazepine, metoprolol,
sulfamethoxazole and diclofenac as substance mixture (c0 = 100 μg L-1) in 1 L
HPLC-water with 0.3 g L-1 H2O2.
In WWTP effluents diclofenac and sulfamethoxazole (2.5 min L-1) were degraded faster
than the other pharmaceuticals. Only 47 % metoprolol and 20 % carbamazepine were
eliminated from the WWTP effluent. Half of the concentration of metoprolol was
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

184 Disinfection By-products in Drinking Water
eliminated after 81.5 min, for carbamazepine after 315 min. The additional use of H2O2
decreased the degradation of sulfamethoxazole to 1.9 min L-1, of metoprolol to 5.8
min L-1 and of carbamazepine to 5.9 min L-1. Diclofenac was eliminated immediately,
thus it was not possible to calculate the degradation rate. The same results were obtained
in measurements with WWTP effluent samples from another treatment plant (data not
shown).
Matrix effects were observed for the degradation of the substance mixture in waste water
effluent. The degradation rates were higher than those of the substance mixture in
HPLC-water. For amidotrizoic acid the degradation rate was increased to 1.5 min L-1, for
carbamazepine to 10.6 min L-1, for metoprolol to 4.7 min L-1 and for sulfamethoxazole
to 1.7 min L-1.
3.2 Cytotoxicity
To evaluate the cytotoxicity by using the MTT assay, the viability of the cells after
exposure was calculated in relation to the viability of the negative control which was set
to 100 %. According to DIN ISO 1993-5 there are three grades of cytotoxicity: not
cytotoxic (100-81 %, grade 0), weakly to moderately cytotoxic (80-61 %, grade 1-2) and
highly cytotoxic (60-0 %, grade 3). This DIN-ISO norm was used for the evaluation of
the results in the present study.
In the water effluent samples from both WWTPs enriched with substance mixture (c0 =
100 μg L-1) no cytotoxic effects were detected after treatment with UV-light. In one
WWTP only the sample which was treated with UV/ H2O2 showed very weak cytotoxic
effects (79.4 ± 7.5 % cell viability) after 24 h of exposure. No cytotoxicity was observed
before the oxidative treatment (data not shown).
Amidotrizoic acid, carbamazepine and diclofenac were tested as single compounds in
HPLC-water (c0 = 100 μg L-1). For all the substances the viability was above 90 %,
which means that no cytotoxicity occurred before or after the oxidative treatment.
However, HPLC-water with the substance mixture treated for 60 min with UV/ H2O2
showed weak cytotoxicity effects (81 ± 6.5 % cell viability) after 24 h of exposure
(Figure 2).
3.3 Genotoxicity
To detect the genotoxic potential of the tested samples the alkaline comet assay was
performed. The prepared gels were stained with SYBR-green staining solution and
analyzed manually using the program “Comet Assay IV” (Perspective Instruments, UK)
under a fluorescence microscope at a 400x magnification. For the analysis 50 randomly
distributed cells were counted and the olive tail moment (OTM) was analyzed [23] [24].
Results were statistically evaluated by using the Mann-Whitney-Test for two
independent groups. Then the results of the treated cells were compared to the results of
the negative control (cell culture medium). To express the significance, p-values were
FDOFXODWHGS ! QRWVLJQLILFDQWS  VLJQLILFDQWS  KLJKO\VLJQLILFDQW
SH[WUHPHO\VLJQLILFDQW
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

Disinfection By-products Prevention and Treatment 185
Figure 2 Viability of CHO-9 cells after 24 hours of exposure to a pharmaceutical mix
(amidotrizoic acid, carbamazepine, diclofenac, metoprolol and sulfamethoxazole)
dissolved in HPLC-water (100 μg L-1) before (-20, -5 min) and after (60 min)
oxidative treatment with UV or UV/H2O.
In the effluents with additional substance mixture of both tested WWTPs treated with
UV no genotoxicity was detected. The OTMs of the samples were not significantly
different from the OTM of the negative control, neither before nor after the oxidative
treatment (data not shown). Also the single substances (amidotrizoic acid,
carbamazepine, diclofenac) in HPLC-water treated with UV showed no significant
change in the OTM compared to the negative control in the cells. So, no genotoxicity
occurred in the cells after the exposure to the substances before and after oxidative
treatment.
In contrast, the substance mixture treated with UV-light in HPLC-water for 60 minutes
showed elevated genotoxic effects (OTM = 0.23) after 24 h of exposure, while the
untreated sample was not genotoxic (Figure 3).
The genotoxicity of the samples treated with additional H2O2 could not be assessed
because of high cell damage (cytotoxicity 80 %). To exclude genotoxic effects of
catalase in samples treated with additional H2O2, catalase in HPLC-water was also tested
and showed no genotoxic potential.
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

186 Disinfection By-products in Drinking Water
Figure 3 Genotoxic potential of a pharmaceutical mix (amidotrizoic acid, carbamazepine,
diclofenac, metoprolol and sulfamethoxazole) dissolved in HPLC-water (100 μg L-
1) before (-20, -5 min) and after (60 min) oxidative treatment with UV or UV/H2O
in CHO cells after 24 hours of exposure. ** = p  SQDQRW
assessable because of high cytotoxicity.
4 CONCLUSIONS
In the present study, the efficiency of an UV-oxidation system using different matrices
was investigated and toxicologically evaluated.
It was shown that the application of the substances as a mixture has only a slight effect
on the degradation of the individual compounds. Furthermore, the photo-oxidative
impact of the Hg-low pressure source was very low which is reflected in long
degradation times.
The addition of H2O2 accelerated the degradation by the formation of hydroxyl-radicals.
Those radicals are very reactive and facilitate the degradation of all substances in the
aqueous solution. Also, other matrix substances are degraded by UV-radiation and OH-
radicals which leads to competition reactions and to a slower degradation of the
substances of interest [22]. The additional radical source leads to a higher amount of
hydroxyl-radicals, thereby lowering the competition reaction by the matrix and
accelerating the degradation. The matrix effects are also shown in the comparison of
HPLC-water with effluent water. In HPLC-water no other substances are present, so that
the leading substances amidotrizoic acid, carbamazepine, diclofenac, metoprolol and
sulfamethoxazole were degraded faster than in the complex waste water effluent which
contains suspended matter and particles leading to turbidity and absorption of some of
the UV-light.
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

Disinfection By-products Prevention and Treatment 187
In the present study, the toxic effect of oxidation by-products after the UV-treatment of
waste water effluents and enriched ultrapure water with and without additional H2O2 was
compared. Samples treated only with UV showed no cyto- or genotoxic effects. Only in
one case the UV treated substance mixture was significantly genotoxic. Slightly
cytotoxic results (61-80% cell viability) were obtained for samples with the substance
mixture treated with UV/H2O2, indicating that combination effects have led to the
formation of some toxic by-products. No difference in toxicity was observed between
enriched and original WWTP effluent samples and enriched HPLC-water samples. No
cyto- or genotoxic effects were detected during treatment of single substances.
The results indicate that the treatment of a mixture of the test substances leads to
oxidation by-products which have more toxic potential than those of the individual
substances. Comparing the degradation of single substances with the elimination of the
substances from the substance mixture in HPLC-water, no significant difference was
observed; in both cases the substances were degraded within 60 min. So, the assumption
is supported that different by-products are formed during the treatment process.
However, toxic results occurred only for samples in ultrapure water, in waste water
effluents no genotoxicity was observed after treatment with UV. In this case, matrix
effects may lead to the formation of different by-products and also to a slower
degradation of the trace substances. These results show that toxicological tests are
necessary not only for single substances but especially for substance mixtures.
In summary, the results of the present study show that UV-oxidation can be used for the
elimination of pharmaceuticals from WWTP effluents. The additional use of H2O2
resulted in higher elimination rates. However, the use of H2O2 also led to weak
toxicological reactions in the cell culture tests.
References
1. IGF: Industrielle Gemeinschaftsforschung und –entwicklung, AiF-FV , 2006, 15852.
2. O. A. H. Jones, N. Voulvoulis and J. N. Lester, Water Research, 2002, 36, 5013.
3. M. Pedrouzo, F. Borrull, E. Pocurull and R. M. Marcé, Water Air Soil Pollut, 2010,
217, 267.
4. J. Fick, H. Söderström, R. H. Lindberg, C. Phan, M. Tsklind and D. G. J. Larsson,
Environmental Toxicology and Chemistry, 2009, 28, 2522.
5. WHO: World Health Organization, Guidelines for Drinking-Water Quality, 1996, 2.
6. H. H. Dieter, K. Götz, K. Kümmerer, B. Rechenberg and F. Keil, Umweltbundesamt,
Berlin, 2010.
7. A. Wick, G. Fink, A. Joss, H. Siegrist and T. A. Ternes, Water Research, 2009, 43,
1074.
8. A. Haiss, P. Hubner, J. Zipfl and K. Kümmerer, vom Wasser, 1998, 91, 315.
9. H. Schröder, S. Osterhorn and V. Flöser, gwf Wasser/Abwasser, 1999, 140, 20.
10. R. Sattelberger, Umweltbundesamt GmbH, Wien, 1999.
11. T. A. Ternes, Water Research, 1998, 32, 3245.
12. T. A. Ternes, H. Siegrist and G. Berthold in Umweltbundesamt Berlin, 2005, 29,
101.
13. BLAC: Bund/Länderausschuss für Chemikaliensicherheit, Arzneimittel in der
Umwelt – Bewertung der Untersuchungsergebnisse, 2003.
14. B. Ferrari, R. Mons, B. Vollat, B. Fraysse, N. Paxeus, R. L. Giudice, A. Pollio and J.
Garric, Environmental Toxicology and Chemistry, 2004, 23, 1344.
15. K. Aktories, U. Förstermann, F. Hofmann and K. Starke, Elsevier GmbH, Urban-
Fischer-Verlag, 2009.
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online

188 Disinfection By-products in Drinking Water
16. T. Oppenländer, WILEY-VCH Verlag, 2003.
17. C. Gottschalk, J. A. Libra, A. Saupe, WILEY-VCH Verlag, 2010.
18. J. Tuerk, B. Sayder, A. Boergers, H. Vitz, T. K. Kiffmeyer and S. Kabasci, Water
Science & Technology – WST, 2010, 61, 985.
19. M. B. Heringa, D. J. H. Harmsen, E. F. Beerendonk, A. A. Reus, C. A. M. Krul, D.
H. Metz and G. F. IJpelaar, Water Research, 2011, 45, 366.
20. O. Ostling and K. J. Johanson, Biochemical and Biophysical Research
Communications, 1984, 123, 291.
21. N. P. Singh, M. T. McCoy, R. R. Tice and E. L. Schneider, Experimental Cell Res.,
1988, 175, 184.
22. K. Höll, Wasser; de Gruyter, 2002, 161.
23. P. L. Olive, J. P. Banath and R. E. Durand, Radiation Research, 1990, 122, 86.
24. M. de Boeck, N. Touil, G. de Visscher, P. A. Vande and M. Kirsch-Volders, Mutat.
Res., 2000, 469, 181.
Published on 29 September 2015. Purchased by h.bielak@iww-online.de on 06 July 2016.
View Article Online