- Email: [email protected]
Photocatalytic degradation of pharmaceuticals present in conventional treated wastewater by nanoparticle suspensions
Accepted Manuscript Title: Photocatalytic degradation of pharmaceuticals present in conventional treated wastewater by nanoparticle suspensions Author: Sara Teixeira Robert Gurke Hagen Eckert Klaus K¨uhn Joachim Fauler Gianaurelio Cuniberti PII: DOI: Reference:
S2213-3437(15)30045-2 http://dx.doi.org/doi:10.1016/j.jece.2015.10.045 JECE 838
To appear in: Received date: Revised date: Accepted date:
2-7-2015 26-10-2015 31-10-2015
Please cite this article as: Sara Teixeira, Robert Gurke, Hagen Eckert, Klaus K¨uhn, Joachim Fauler, Gianaurelio Cuniberti, Photocatalytic degradation of pharmaceuticals present in conventional treated wastewater by nanoparticle suspensions, Journal of Environmental Chemical Engineering (2015), http://dx.doi.org/10.1016/j.jece.2015.10.045 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M an
100
ed
80
TiO2
60
1.5 μm
ZnO
ce pt
Degradation ratio C(t)/C(0) (%)
us
cr
ip t
Graphical Abstract
40
Ac
20 0
0
10
20 30 40 Time t (min)
50
60
Page 1 of 12
*Manuscript
Photocatalytic degradation of pharmaceuticals present in conventional treated wastewater by nanoparticle suspensions Sara Teixeiraa , Robert Gurkeb,c , Hagen Eckerta,d,∗, Klaus K¨uhna , Joachim Faulerb , Gianaurelio Cunibertia,d,e a Institute
for Materials Science and Max Bergmann Center of Biomaterials, TU Dresden, 01062 Dresden, Germany of Clinical Pharmacology, Faculty of Medicine Carl Gustav Carus, TU Dresden, 01307 Dresden, Germany c Research Association Public Health Saxony and Saxony Anhalt, Faculty of Medicine Carl Gustav Carus, TU Dresden, 01307 Dresden, Germany d Dresden Center for Computational Materials Science (DCCMS), TU Dresden, 01062 Dresden, Germany e Center for Advancing Electronics Dresden, TU Dresden, 01062 Dresden, Germany
cr
ip t
b Institute
us
Abstract
Pharmaceuticals have become an important public health issue as environmental pollutants over the last years. After
an
ingestion, pharmaceuticals are partly excreted unchanged. They can reach the wastewater treatment plant (WWTP) via the sewer network. Because the conventional treatments are ineffective in their removal, new methods should be approached, for example semiconductor photocatalysis. Several of the hitherto published studies analyzed the
M
degradation of model pollutants but for the degradation of pharmaceuticals in unspiked real wastewater further investigations are required. Therefore, we want to focus on the removal of pharmaceuticals in an actual effluent from a WWTP and investigate the effluent background effect. This study shows the heterogeneous photocatalytic degrada-
d
tion of 14 pharmaceuticals with initial concentrations Ci > 0.3 µgL−1 present in a WWTP effluent. We found that
te
UVA (1.5 mWcm−2 , intensity peak at 365 nm) irradiation of TiO2 P25 (A s = 56 m2 g−1 ) or ZnO (A s = 5.23 m2 g−1 ) nanoparticles leads to considerable degradation of the analyzed pharmaceuticals. With ZnO nanoparticles, 40 min
ce p
UVA irradiation was sufficient to degrade over 95 % of these pharmaceuticals (kapp = 8.6 · 10−2 s−1 ). Using TiO2 P25 on the other hand, it would take more than six times longer to reach the same level (kapp = 1.4 · 10−2 s−1 ). Carbamazepine dissolved in millipore water served as a comparison model. Also in this system ZnO presents faster degradation.
Keywords: Photocatalysis, pharmaceuticals, ultraviolet radiation, wastewater
1
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
1. Introduction
5
ceuticals in drinking water the World Health Organiza-
6
tion already reviewed scientific evidence to address this
2
Pharmaceuticals emerging in the aquatic ecosystems
7
issue. They are mostly introduced in the sewage sys-
3
have become an important public health issue over the
8
tem through excretion of unmetabolized compounds af-
4
past few years. To evaluate the impact of those pharma-
9
ter medical use or inappropriate disposal [1, 2, 3, 4] and
10
then transported into the wastewater treatment plants
11
(WWTPs). However, conventional WWTPs are not
12
designed to treat water polluted with pharmaceuticals
author. Tel.: +49 351 463-31461 Email address: [email protected] (Hagen
∗ Corresponding
Eckert) Preprint submitted to Journal of Environmental Chemical Engineering
October 26, 2015
Page 2 of 12
carbamazepine.
present at trace levels and therefore, the applied treat-
48
14
ments are ineffective in their removal [5, 6].
Con-
49
As photocatalysts we chose TiO2 and ZnO and com-
15
sequently, they reach the aquatic system and can be
50
pared the degradation efficiencies of both photocata-
16
found in surface and ground water [7, 8], soil and sed-
51
lysts. Despite several semiconductors have been studied
17
iments [8, 9] and even in drinking [10, 11] and tap wa-
52
for applications in wastewater decontamination, ZnO
18
ter [8, 12]. Although, normally pharmaceuticals do not
53
and TiO2 are frequently the most studied photocata-
19
present acute toxic effects on aquatic organisms due to
54
lysts because of their interesting optical properties, low
20
their low concentrations, in the range of ng to µg per
55
cost, and availability [22]. Although ZnO is usually
21
liter, concerns have been raised for chronic exposure,
56
described as the most active semiconductor [23], TiO2
22
due to their continuous input into the environment, act-
57
is used more frequently because it is more stable than
23
ing as slightly persistent pollutants [2, 4, 13].
58
ZnO in aqueous solution [24]. We used the photocat-
us
cr
ip t
13
For these reasons, diverse efforts have been made
59
alysts as nanoparticles in a slurry mixture to maximise
25
to remove pharmaceuticals from wastewater, such as
60
the surface area of the system. The upscaling of such
26
membrane filtration, activated carbon adsorption and
61
27
advanced oxidation processes (AOPs). AOPs are re-
62
28
commended when water pollutants have a high chem-
63
29
ical stability, allowing to achieve almost the total miner-
64
think that this can be solved in the near future, for ex-
30
alization of contaminants to carbon dioxide, water and
65
ample through the use of magnetic core nanoparticles
31
inorganic compounds or, at least, allow their partial oxi-
66
[25, 26]. Whenever photocatalytic systems are applied
32
dation to become more biodegradable and/or less harm-
67
in an actual wastewater treatment plant, a risk assess-
33
ful [3, 14].
68
ment regarding the material output into the environment
69
is necessary, due to their photo activity, size distribution
an
24
a setup provides a challenge regarding the separation of the nanoparticles from water after the treatment. In the
te
d
M
light of an active research regarding this problem, we
Different techniques involve the generation of hy-
35
droxyl radicals, which are nonselective and have twice
70
and potential toxicity for aquatic organisms in the case
36
the oxidizing power of chlorine [4, 6, 15, 16, 17]. Het-
71
of ZnO [27].
37
erogeneous semiconductor photocatalysis has become
38
an attractive method to remediate environmental con-
72
2. Experimental part
39
tamination due to its high photocatalytic activity, nontoxicity and photostability [3, 15, 18, 19, 20, 21]. How-
73
40
2.1. Chemicals and Materials
41
ever, most of the studies do not use unspiked wastew-
74
For the degradation experiments, TiO2 P25 (kindly
42
ater from sewage treatment plants but aqueous solu-
75
provided by Evonik), ZnO (IOLITEC Ionic Liq-
43
tions of model compounds or surface waters. Therefore,
76
uids Technologies GmbH), and carbamazepine (Sigma
44
we want to degrade pharmaceuticals in effluent samples
77
Aldrich) were used in this work.
45
from a WWTP and investigate the effluent background
78
tained from a Millipore Milli-Q System (Water, Milli-
46
effect. To do so, we additionally investigated the degra-
79
pore). For the SPE-HPLC-MS/MS analysis, acetoni-
47
dation process in millipore water artificially spiked with
80
trile, methanol (HiperSolv, HPLC-grade), and ammo-
ce p
34
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Water was ob-
2
Page 3 of 12
Table 1: Drugs (LLoQ - Lower Limit of Quantification)
Analyte
Drug class
Provider
Internal standard
LLoQ
Sigma
Carbamazepine D10
50 ngL−1
Pfizer
Gabapentin D10
200 ngL−1
Lamotrigine
Sigma
Lamotrigine 13C, 15N4
50 ngL−1
Oxcarbazepine
Cerilliant
Carbamazepine D10
Wyeth
Venlafaxine D6
Merck
Oxprenolol
LGC Standards
Oxprenolol
LGC Standards
Venlafaxine D6
Carbamazepine Gabapentin
Bisoprolol Celiprolol
beta blocker
Talinolol
50 ngL−1
50 ngL−1 50 ngL−1
cr
antidepressant
us
Venlafaxine
ip t
anticonvulsant
50 ngL−1
50 ngL−1
lipid-lowering drug
Sigma
Warfarin
50 ngL−1
Tramadol
opioid analgesic
Sigma
Tramadol 13C, D3
50 ngL−1
AstraZeneca
Amitriptyline D3
50 ngL−1
Venlafaxine D6
50 ngL−1
Trimipramine D3
50 ngL−1
Valsartan D9
100 ngL−1
Candesartan angiotensin receptor
Sigma
Irbesartan
antagonist
Sigma
M
Eprosartan
an
Bezafibrate
Valsartan
Sigma
81
nium acetate were purchased from Merck. Formic acid
82
(LC-MS grade) and Na2 EDTA (ACS reagents) were
83
obtained from Sigma and water (HPLC-grade) from
98
The specific surface area of the photocatalyst par-
84
VWR. The standards were provided by different sup-
99
ticles was determined by the Brunauer-Emmett-Teller
85
pliers as listed in Tab. 1. The treated wastewater was
100
(BET) method. This property was analyzed at 77 K by
86
kindly provided by the WWTP Kaditz located in Dres-
101
nitrogen adsorption-desorption in a Micromeritics TriS-
87
den, Germany, operated by Stadtentwsserung Dresden
102
tar analyzer (Micromeritics). Before performing ad-
88
GmbH. This treatment plant currently cleans the sewage
103
sorption experiments, samples (0.5 g) were outgassed
89
of 650,000 people and has a design capacity of 740,000
104
at 26.7 Pa and 350 °C for 6 h.
90
inhabitant equivalents. The yearly average sewage vol-
105
The morphology for both particle types were ana-
91
ume is about 55·106 m3 . The WWTP consist of primary
106
lyzed with a scanning electron microscope (SEM) op-
92
clarifier, activated sludge reactor and secondary clarifier
107
erated at 10 kV and 25 kV.
93
[28]. The sample was taken as a 24 h flow proportional
108
The UV-Vis diffuse reflectance spectra were obtained
94
composite effluent sample on June 24th 2014, stored at
109
using a Shimadzu UV-Vis spectrophotometer 2101PC
95
4 °C and analyzed on the next day. Further characteris-
110
in the range of 190 to 600 nm. It was equipped with a
96
tics of the sewage sample are summarized in Tab. 2.
111
diffuse reflectance attachment and we used BaSO4 as a
112
reference.
97
d
te
ce p
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
2.2. Nanoparticle characterization
3
Page 4 of 12
time intervals and centrifuged for one hour to remove
131
the catalyst nanoparticles. All experiments were carried
COD (Chemical Oxygen Demand)
37 mgL−1
132
out at room temperature. As controls, experiments were
4 mgL−1
133
carried out in the absence of catalyst nanoparticles or
134
without UVA irradiation. All other parameters in the control experiments were kept unchanged.
BOD (Biochemical Oxygen Demand) Nitrogen
ip t
130
Table 2: Treated wastewater parameters.
12.0 mgL−1
135
TKN (Total Kjeldahl Nitrogen)
<5.0 mgL−1
136
The effluent sample from the WWTP Kaditz was
Nammonium
0.31 mgL−1
137
filtered by a filter paper (VWR pore sizes 5-13 µm)
Nnitrite
0.03 mgL−1
138
to remove suspended particulate matter. −1
cr
Ntotal
Thereafter,
7.40 mgL−1
139
1 gL
Ninorganic
7.74 mgL−1
140
of the treated wastewater. The suspension was then ex-
141
posed under continuous stirring to UVA-radiation. Af-
142
terwards, the samples were centrifuged and analyzed by
Ptotal
0.86 mgL−1
Pphosphate
0.56 mgL−1
143
144
7.5
145
the SPE-HPLC-MS/MS method. In contrast to the previous experiment, we artificially
spiked millipore water with carbamazepine. Therefore,
M
pH
an
Phosphor
of catalyst was added to a volume of 100 mL
us
Nnitrate
2.3. Photocatalytic degradation experiments
146
50 mL of 12 mgL−1 carbamazepine solution with 1 gL−1
147
of catalyst was exposed under continuous stirring to
148
UVA-radiation. The samples were filtered (Rotilabo ny-
149
lon, pore size 0.2 µm) and centrifuged. Absorbance
d
113
The photocatalytic degradation was carried out in
115
borosilicate beakers (VWR) with 3.3 mm wall thick-
150
measurements were performed with a Varian CARY-
116
ness and 5 cm diameter.
151
100 UV-VIS Spectrophotometer.
152
limit for carbamazipine in this setup is 300 µgL−1 . 2.4. Analytic method
Under constant stirring 1
ce p
−1
te
114
of TiO2 P25 or ZnO was added to the samples.
The quantification
117
gL
118
Prior to illumination, the solutions containing the cat-
119
alyst were stirred in the dark for 30 min to achieve an
153
120
adsorption-desorption equilibrium of the pharmaceuti-
154
The analysis of the WWTP effluent sample and the
121
cals on the photocatalyst surface. Afterwards, the sam-
155
degradation experiments were conducted with a SPE-
122
ples were exposed to UVA-radiation. The illuminating
156
HPLC-MS/MS-method, which has been described in
123
device (UMEX) was equipped with six Philips 8 W mer-
157
the work of Gurke et al. [29]. Briefly summarised, 1
124
cury fluorescent tubes (Emax at 365 nm).
158
mL of the degradation experiment sample was taken,
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
125
The distance between the energy source and the pho-
159
adjusted to a pH of 3 by using formic acid and spiked
126
tocatalytic reactor was 15 cm. At this position the UVA
160
with 100 µL of the Internal Standard (IS) solution (10
127
intensity was determined by an UV34 Lux Meter (PCE)
161
µg/L). The samples were extracted using an Abimed
128
ranging from 1.5 - 1.6 mWcm−2 . Aliquots of 2 mL
162
ASPEC XL (Gilson) with Oasis HLB 10 mg Extrac-
129
of the reaction solution were withdrawn at determined
163
tion Cartridges (Waters). The eluates were evaporated
4
Page 5 of 12
to dryness under a gentle air stream at 50 °C and re-
195
where A(aq) is the investigated organic molecule in
165
dissolved in 250 µL mixture of solvent A and solvent
196
solution and A(ad) represents the adsorbed molecule.
166
B (80/20, v/v). The solvents A (97/3/0.05; v/v/v) and
197
M(aq) represents the mineralized components that are
167
B (5/95/0.05; v/v/v) were a composition of 2 mM am-
198
formed during the reaction. Temporal changes of the
168
monium acetate solution, acetonitrile, and formic acid.
199
molecule concentration in the solution CA(aq) (in 1 m−3 )
169
A LC-MS/MS system, consisting of a Dionex-HPLC
200
and the concentration on the photocatalyst particle sur-
170
composed of an UltiMate3000 Pump and Autosampler
201
face CA(ad) (in 1 m−2 ) are given by
171
(Thermo Fischer Scientific) with a Chromeleon 7 Chro-
172
matography Data System (Dionex Softron) and coupled
173
to an API 4000 tandem mass spectrometer (AB Sciex)
174
equipped with an electrospray ionization source (ESI),
175
was used for the analyses of the samples. The chro-
176
matographic separation was performed with a Synergi
202
177
2.5u HydroRP 100A, 100 x 2.0 mm and a C18 security
203
178
guard 4 mm x 2 mm (both Phenomenex) using a multi
204
179
step gradient out of solvent A and B with a total run-
180
time of 15 min. For the analyses, an injection volume
181
of 20 µL was chosen. The mass spectrometric analyses
jads = kads (1 − Θ) CA(aq)
(4)
182
were performed in multiple reaction monitoring (MRM)
jdes = kdes CA(ad)
(5)
183
mode with positive electrospray ionization. The Analyst
jreac = kreac CA(ad)
(6)
184
data system 1.6 (AB Sciex) was applied for MS control
185
and for the peak area evaluation, regression analysis of
186
calibration curves and calculation of concentrations.
cr
d dt C A(aq)
(2)
= as ( jdes − jads )
(3)
an
These equations include the specific surface area of the nanoparticles aS (surface area per solution volume) and different molecule fluxes
M
2.5. Photocatalytic degradation model
= jads − jdes − jreac
us
d dt C A(ad)
d
te
ce p
187
ip t
164
205
These molecule fluxes are mainly determined by the
206
corresponding concentrations and the three different
207
rate constants, describing the reaction (kreac ), adsorption
208
(kads ) and desorption (kdes ) processes. For the adsorption
209
process is also the surface coverage important (Θ).
188
To study the photocatalytic oxidation process we used
189
a suspension of photocatalytic nanoparticles due to the
210
190
large specific surface area in such a system. Addition-
211
191
ally, it is possible to model the degradation process in
212
192
this case efficiently as shown by Eckert et al. [30]. The
213
kads CA(aq) , kdes CA(ad) ), the surface concentration in the
193
simplified reaction formula to describe the degradation
214
system of equations (2) and (3) can be eliminated. The
194
process is
215
solution of Eq. (2) in the limit of small coverage Θ 1
216
reads CA(ad) = CA(aq) kads (kdes + kreac )−1 . Insertion of this
217
expression into Eq. (3) yields C˙ A(aq) = −kappCA(aq) with
218
the solution
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
A(aq) A(ad) → M(aq)
(1)
In the present case of a fast establishing adsorptiondesorption equilibrium with quasi-stationary C˙ concentration C˙ = 0 (i.e. A(ad)
surface
A(ad)
5
Page 6 of 12
Table 3: Adsorption and desorption rate constants derived from the
nanoparticles.
where the apparent degradation rate constant is given
219
catalysts
kads (ms−1 )
kdes (s−1 )
TiO2
5.3·10−9
4.5·10−4
ZnO
2.2·10−7
5.9·10−3
by
kapp = as kads kreac (kdes + kreac )−1 .
(8)
cr
220
experiments with carbamazepine using photocatalytic TiO2 and ZnO
(7)
ip t
CA(aq) (t) = CA(aq),0 exp(−kapp t)
Thus, for kdes kreac , we find kapp = as kads . This
236
present as monodisperse nanospheres. In the ZnO pow-
222
means that the degradation rate becomes adsorption-
237
der more complex polydisperse rectangular structures
223
limited and does not depend on the reaction rate con-
238
were found.
224
stant.
242
243
est level at 365 nm. In contrast, the absorption potential
244
of TiO2 reaches its maxima around 310 nm. This indi-
245
cates that with ZnO the light source can be utilized more
240
A specific surface area of 56 m g
was obtained
for the TiO2 P25 nanopowder with the BET method.
For the ZnO nanopowder a specific surface area of
230
5.23 m2 g−1 was found. In suspension it is expected an
231
additional clustering, especially for TiO2 , which will re-
232
duce the active surface area.
te
229
ce p ZnO
1.5 μm
TiO2
10kV, 5mm, x50000
observed that ZnO absorbance already reached its high-
efficiently.
100 80 60 40
max
20 min 0 436
25kV, 9mm, x50000
Lamp TiO2 ZnO
Absorbance
228
3.1. Characterization of TiO2 and ZnO nanoparticles 2 −1
of the radiation lamp is plotted in the figure. It can be
M
227
241
TiO2 are included in Fig. 2. Also the emission spectra
Light intensity (E365nm = 100)
226
3. Results and discussion
d
225
The UV-Vis diffuse reflectance spectra of ZnO and
an
239
us
221
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
405
365 313 Wavelength λ (nm)
Figure 1: SEM images of colloidal TiO2 and ZnO. Figure 2: Light output of the used mercury fluorescent tubes normalized at 365 nm compared to the diffuse reflectance spectra of TiO2 233
This difference can also be qualitatively observed in
234
the SEM images (Fig. 1). TiO2 particles show a finer
235
structure than the ones in the ZnO sample. TiO2 is
and ZnO.
246
6
Page 7 of 12
TiO2
simulation experiment
b)
c)
d)
cr
a)
0
20
40
60 0 Time t (min)
20
us
50
0
effluent C(0) = 1.4 µgL−1
0 100
ip t
50
40
an
Degradation ratio C(t)/C(0) (%)
100
isolated C(0) = 12.0 mgL−1
ZnO
60
Figure 3: Measured concentration during the degradation of carbamazepine under UVA irradiation with 1 gL−1 catalysts TiO2 (a+c) and ZnO
M
(b+d). In the upper row, carbamazepine was dissolved in millipore water (a+b) and measured by UV/VIS spectrophotometry. In the lower row the results of the degradation of carbamazepine in the treated effluent analysed by HPLC-MS/MS-method are shown. The full lines represent fits for carbamazepine dissolved in millipore water according to the presented model using the values of kads and kdes determined from experiments in the
247
3.2. Influence of WWTP effluent
te
d
dark (Table 3).
262
ceuticals, two different control experiments were con-
263
ducted without nanoparticles and without UVA irradia-
Carbamazepine’s photodegradation was evaluated in
249
a real WWTP effluent sample, alongside other occur-
264
tion for 45 min. In the case of the control exposed to
250
ring pharmaceuticals as listed in Tab. 1. The effluent of
265
UVA-radiation no significant change in the monitored
251
a WWTP represents a challenging matrix for photocat-
266
concentrations occurred. After irradiation the average
252
alytic degradation. Different parameters like ion con-
267
change over all monitored drugs was 0.8 % (SD 4.8 %).
253
centration, pH value or the manifold mixture of organic
268
However for the controls with nanoparticles kept in the
254
and inorganic molecules lead us to conjecture about the
269
dark we observed an initial drop of the concentrations
255
concentration development over time that may be al-
270
before they stabilised, so that the changes in concentra-
256
tered compared to a ’clean’ laboratory experiment. To
271
tion after the initial 30 min of adsorption is 0.3 % (SD
257
examine this assumption we selected the anticonvulsant
272
4.9 %) for TiO2 and 0.2 % (SD 5.7 %) for ZnO. This
258
carbamazepine dissolved in millipore water, as a com-
273
effect is due to the adsorption on the catalyst’s surface
259
parison ’clean’ degradation experiment.
274
until the adsorption-desorption equilibrium is reached.
ce p
248
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
260
To assure that only the combination of nanoparticles
275
This is consistent with the idea that the UVA-radiation
261
and UV-radiation causes the degradation of the pharma-
276
per se does not induce the degradation process nor the
7
Page 8 of 12
nanoparticles without UVA irradiation.
100 deviation TiO2 ZnO
278
Carbamazepine’s concentration was considerably re-
279
duced over time in the presence of the catalysts and
280
UVA-radiation. The same behaviour was observed for
281
carbamazepine present in the wastewater effluent sam-
282
ple. In Fig. 3 we compare the four experiments with
283
carbamazepine dissolved in millipore water and present
284
in the effluent. It can be noticed that the degradation
285
by ZnO is significantly faster than with TiO2 in both
286
setups which is in high accordance with the conducted
287
simulation of the degradation. These results are shown
288
in Tab. 3 and demonstrate that carbamazepine can ad-
maceuticals (Tab. 1) measured by HPLC-MS/MS-method. The grey
289
sorb more than 40 times faster on the ZnO than on the
290
TiO2 surface. Because of this, ZnO presents a higher
291
degradation rate compared to TiO2 regardless its larger
an
Degradation ratio C(t)/C(0) (%)
277
292
surface area.
80 60
20
10
20 30 40 Time t (min)
50
60
us
0
cr
ip t
40
0
Figure 4: Average degradation ratio over time of the 14 selected phar-
area corresponds to the standard deviation.
3.3. Degradation of pharmaceuticals in the WWTP ef-
M
301
302
fluent
To determine the degradation in an effluent sample of
When the parameters determined from the millipore
303
294
water experiment (Fig. 3 a, b) are compared to the out-
304
the WWTP Dresden Kaditz, 55 target pharmaceuticals
295
come from the effluent (Fig. 3 c, d) similar results can
305
based on their prescription numbers were selected based
296
be found. This shows that the effluent background does
306
on a previous study of Gurke et al. [29]. For monitoring
297
not reduce the efficiency of the selected photocatalytic
307
the degradation process, the initial concentration (Ci )
298
particles, and the method to model the degradation in
308
of the pharmaceuticals needs to be significantly higher
299
the millipore water experiment can also be applied for
309
than the corresponding detection limit. Therefore, we
300
pharmaceuticals investigated in the effluent.
310
set the lower limit to be Ci > 0.3 µgL−1 . After an initial
311
analysis of the effluent sample, 14 pharmaceuticals were
312
selected to be monitored in the degradation experiment
313
(Tab. 1).
ce p
te
d
293
Table 4:
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Apparent rate constants average and adsorption rate
constants average for the 14 studied pharmaceuticals (Tab. 1) with the photocatalysts TiO2 and ZnO.
314
Based on this criteria the anticonvulsants carba-
315
mazepine (Ci = 1.29 µgL−1 ), gabapentin (Ci =
catalyst
kapp (s−1 )
kads (ms−1 )
316
11.30 µgL−1 ), lamotrigine (Ci = 0.98 µgL−1 ), and
TiO2
1.4·10−2
4.0·10−9
317
oxcarbazepine (Ci = 0.63 µgL−1 ), the antidepres-
ZnO
8.6·10−2
2.7·10−7
318
sant venlafaxine (Ci = 0.58 µgL−1 ), the beta block-
319
ers bisoprolol (Ci = 0.58 µgL−1 ), celiprolol (Ci =
320
0.35 µgL−1 ), and talinolol (Ci = 0.43 µgL−1 ), the lipid-
8
Page 9 of 12
lowering drug bezafibrate (Ci = 0.48 µgL−1 ), the opi-
356
lower selectivity of ZnO compared to TiO2 . In the ZnO
322
oid analgesic tramadol (Ci = 0.624 µgL ), as well as
357
experiments, all target pharmaceuticals are degraded in
323
the angiotensin receptor antagonists candesartan (Ci =
358
a similar way with just slight deviations. This character-
324
1.30 µgL ), eprosartan (Ci = 0.56 µgL ), irbesartan
359
istic is very important in real applications due to the dif-
325
(Ci = 1.50 µgL−1 ), and valsartan (Ci = 3.59 µgL−1 )
360
ferent mixtures of pharmaceuticals present in wastewa-
326
were analysed in the degradation experiment. Please
361
ter depending on regional consumption patterns or sea-
327
keep in mind that this data is a mere snapshot because
362
son of the year. It can be assumed that the treatment of
328
the concentration of micropollutants can significantly
363
pharmaceuticals present in WWTP effluents using ZnO
329
vary in sewage samples based on a diverse range of pa-
364
and UVA-radiation will result in a near complete degra-
330
rameters, like weather conditions.
365
dation of the pollutants in a relative short period of time.
366
4. Conclusions
cr
−1
331
The degradation experiments were traced for all tar-
332
get pharmaceuticals and are presented in the supple-
333
mentary material. For the SPE extraction just a small
334
sample volume of 1 mL was necessary, which allowed
367
335
a high temporal resolution monitoring (ten samples in
368
336
one hour) without disturbing the experiment by taking
369
an
−1
us
−1
ip t
321
337
out samples with a large volume. An overview of the
370
ents would be useful to prevent the contamination of
338
two photocatalytic materials is shown in Fig. 4. Af-
371
surface water. Semiconductor photocatalysis is recom-
339
ter 40 min, an average degradation of more than 95 %
372
mended whenever water pollutants present low degrad-
340
for the samples treated with ZnO was already observed.
373
ability and/or high chemical stability.
341
During the same period of time the pharmaceuticals
374
The pollutants are at least in parts mineralized due to
342
treated with TiO2 degraded by 40 %. The simulation
375
the generation of highly oxidative species and electron-
343
data shows that treatment with TiO2 would take over
376
hole pairs. Photocatalysis may be seen as a complemen-
344
four hours to achieve the same result as ZnO. These
377
tary method to the already existing technologies to im-
345
differences in degradation rates are even more signif-
378
prove the removal rates of pollutants, such as pharma-
346
icant taking into account the smaller surface area of
379
ceuticals. In this study, the applicability of a previously
347
ZnO. The resulting apparent rate constants and adsorp-
380
developed LC-MS/MS method by evaluating the degra-
348
tion rate constants are listed in Tab. 4 and are in accor-
381
dation of pharmaceuticals present in a real wastewater
349
dance with the results from carbamazepine dissolved in
382
sample by photocatalysis has been demonstrated. We
350
millipore water. In literature different new approaches
383
were able to conduct successfully degradation experi-
351
besides commercial nanoparticles are presented. The
384
ments of various pharmaceuticals present in the effluent
352
apparent rate constants found for example from TiO2
385
of a WWTP via photocatalysis by ZnO and TiO2 un-
353
nanowires is in average the same order of magnitude as
386
der UVA-radiation. Although ZnO shows higher degra-
354
the commercial TiO2 particles, but slower compared to
387
dation rates it poses a strong engineering challenge, to
355
the ZnO particles [31]. Furthermore, it was observed a
388
prevent ZnO contamination of the aquatic environment,
In general, pharmaceuticals are not fully removed by
urban WWTPs and can be detected in the effluents. Removing these type of pollutants from wastewater efflu-
M
d
te
ce p
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
9
Page 10 of 12
5. Acknowledgements
391
This work is funded by the European Union (ERDF)
392
and the Free State of Saxony via the ESF project
393
100098212 InnoMedTec. The authors are thankful to
394
Jackie Le, Pedro Martins, Sam Diamond, Thomas K¨ase-
395
berg, Quirina Roode-Gutzmer, Ignacio Gonzalez and
423
[6] J. a. Pereira, V. Vilar, M. Borges, O. Gonz´alez, S. Esplugas,
424
R. Boaventura, Photocatalytic degradation of oxytetracycline
425
using TiO2 under natural and simulated solar radiation, Solar
426
Energy 85 (2011) 2732–2740.
427
[7] O. A. H. Jones, N. Voulvoulis, J. N. Lester, Aquatic environ-
428
mental assessment of the top 25 English prescription pharma-
429
ceuticals., Water research 36 (2002) 5013–22.
ip t
390
due to its potential toxicity.
430
[8] W. C. Li, Occurrence, sources, and fate of pharmaceuticals in
431
aquatic environment and soil., Environmental pollution (Bark-
432
ing, Essex : 1987) 187 (2014) 193–201.
cr
389
433
[9] W. Mrozik, J. Stefa´nska, Adsorption and biodegradation of an-
Hoai Nga Le for their support and many valuable dis-
434
tidiabetic pharmaceuticals in soils., Chemosphere 95 (2014)
397
cussions. Additional we want to give acknowledgment
435
281–8.
398
to Rita Knoche and Norbert Lucke for providing the wastewater samples.
436
[10] T. Ternes, M. Meisenheimer, D. McDowell, F. Sacher, H.-J.
437
Brauch, B. Haist-Gulde, G. Preuss, U. Wilme, N. Zulei-Seibert,
438
Removal of pharmaceuticals during drinking water treatment.,
439
400
440
Supplementary information
441
Supplementary information related to this article can
443
402
be found online at doi:
444
d
445
403
References
Environmental science & technology 36 (2002) 3855–63.
[11] M. S. Fram, K. Belitz, Occurrence and concentrations of pharmaceutical compounds in groundwater used for public drinkingwater supply in California., The Science of the total environ-
M
401
442
an
399
us
396
446
ment 409 (2011) 3409–17.
[12] T. E. Doll, F. H. Frimmel,
Fate of pharmaceuticals -
photodegradation by simulated solar UV-light., Chemosphere 52 (2003) 1757–69. [13] N. Xekoukoulotakis, C. Drosou, C. Brebou, E. Chatzisymeon,
[1] C. D. Metcalfe, B. G. Koenig, D. T. Bennie, M. Servos,
448
E. Hapeshi, D. Fatta-Kassinos, D. Mantzavinos, Kinetics of UV-
405
T. Ternes, R. Hirsch, Occurrence of neutral and acidic drugs
449
A/TiO2 photocatalytic degradation and mineralization of the an-
406
in the effluents of Canadian sewage treatment plants., Environ-
450
tibiotic sulfamethoxazole in aqueous matrices, Catalysis Today
407
mental Toxicology and Chemistry 22 (2003) 2872–2880.
451
161 (2011) 163–168.
ce p
te
447
404
408
[2] M. Isidori, M. Bellotta, M. Cangiano, A. Parrella, Estrogenic
452
409
activity of pharmaceuticals in the aquatic environment., Envi-
453
ronment international 35 (2009) 826–9.
454
410
[14] R. Andreozzi, V. Caprio, A. Insola, R. Marotta,
Advanced
oxidation processes (AOP) for water purification and recovery, Catalysis Today 53 (1999) 51–59.
411
[3] M. Klavarioti, D. Mantzavinos, D. Kassinos, Removal of resid-
412
ual pharmaceuticals from aqueous systems by advanced oxida-
456
reactor for water purification, Catalysis Today 40 (1998) 73–83.
413
tion processes., Environment international 35 (2009) 402–17.
457
[16] L. Yang, L. Yu, M. Ray, Degradation of paracetamol in aqueous solutions by TiO2 photocatalysis., Water research 42 (2008)
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
455
[15] A. Ray, A. Beenackers, Development of a new photocatalytic
414
[4] N. Xekoukoulotakis, N. Xinidis, M. Chroni, D. Mantzavinos,
458
415
D. Venieri, E. Hapeshi, D. Fatta-Kassinos, UV-A/TiO2 photo-
459
416
catalytic decomposition of erythromycin in water: Factors af-
460
[17] E. Hapeshi, A. Achilleos, M. Vasquez, C. Michael, N. Xek-
417
fecting mineralization and antibiotic activity, Catalysis Today
461
oukoulotakis, D. Mantzavinos, D. Kassinos, Drugs degrading
418
151 (2010) 29–33.
462
photocatalytically: Kinetics and mechanisms of ofloxacin and
3480–8.
419
[5] V. Cunningham, S. Binks, M. Olson, Human health risk assess-
463
atenolol removal on titania suspensions., Water research 44
420
ment from the presence of human pharmaceuticals in the aquatic
464
(2010) 1737–46.
421
environment., Regulatory toxicology and pharmacology : RTP
465
422
53 (2009) 39–45.
[18] A. Fujishima, T. Rao, D. Tryk, Titanium dioxide photocatalysis,
10
Page 11 of 12
466
Journal of Photochemistry and Photobiology C: Photochemistry
509
[30] H. Eckert, M. Bobeth, S. Teixeira, K. K¨uhn, G. Cuniberti, Mod-
467
Reviews 1 (2000) 1–21.
510
eling of photocatalytic degradation of organic components in
468
[19] F. Thevenet, O. Gua¨ıtella, J. Herrmann, A. Rousseau, C. Guil-
511
water by nanoparticle suspension, Chemical Engineering Jour-
469
lard, Photocatalytic degradation of acetylene over various tita-
512
nal (2014).
470
nium dioxide-based photocatalysts, Applied Catalysis B: Envi-
513
[31] A. Hu, X. Zhang, K. D. Oakes, P. Peng, Y. N. Zhou, M. R.
ronmental 61 (2005) 58–68.
Servos, Hydrothermal growth of free standing TiO2 nanowire
[20] M. Faramarzpour, M. Vossoughi, M. Borghei, Photocatalytic
515
membranes for photocatalytic degradation of pharmaceuticals,
473
degradation of furfural by titania nanoparticles in a floating-bed
516
Journal of Hazardous Materials 189 (2011) 278–285.
474
photoreactor, Chemical Engineering Journal 146 (2009) 79–85.
477
idences from spectroscopic studies, Journal of Luminescence
478
131 (2011) 1975–1981.
479
[22] M. Lazar, S. Varghese, S. Nair, Photocatalytic Water Treatment
480
by Titanium Dioxide: Recent Updates, Catalysts 2 (2012) 572–
481
601.
482
[23] A. Fujishima, X. Zhang, Donald A. Trykc, TiO2 photocataly-
483
sis and related surface phenomena, Surface Science Reports 63
485
(2008) 515–582. [24] L. Pan, X. Liu, Z. Sun, C. Q. Sun, Nanophotocatalysts via
M
484
486
microwave-assisted solution-phase synthesis for efficient pho-
487
tocatalysis, Journal of Materials Chemistry A 1 (2013) 8299. [25] R. Chalasani, S. Vasudevan,
Cyclodextrin-Functionalized
d
488
Fe3O4@TiO2: Reusable, Magnetic Nanoparticles for Photocat-
490
alytic Degradation of Endocrine-Disrupting Chemicals in Water
491
Supplies, ACS Nano 7 (2013) 4093–4104.
te
489
492
[26] R. Wang, X. Wang, X. Xi, R. Hu, G. Jiang, Preparation and Photocatalytic Activity of Magnetic Fe3O4/SiO2/TiO2 Composites,
494
Advances in Materials Science and Engineering 2012 (2012) 1–
ce p
493
495
cr
interaction of colloidal TiO2 nanoparticles with lysozyme: Ev-
us
[21] A. Kathiravan, M. Asha Jhonsi, R. Renganathan, Photoinduced
an
475 476
ip t
514
472
471
8.
496
[27] L. K. Adams, D. Y. Lyon, P. J. Alvarez, Comparative eco-
497
toxicity of nanoscale TiO2, SiO2, and ZnO water suspensions,
498
Water Research 40 (2006) 3527–3532.
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
499
[28] C. Marx, N. G¨unther, S. Schubert, R. Oertel, M. Ahnert,
500
P. Krebs, V. Kuehn, Mass flow of antibiotics in a wastewater
501
treatment plant focusing on removal variations due to opera-
502
tional parameters, Science of The Total Environment 538 (2015)
503
779–788.
504
[29] R. Gurke, J. Rossmann, S. Schubert, R. Oertel, T. Sandmann,
505
M. R¨oßler, J. Fauler, Predictive approach of pharmaceuticals
506
in an urban sewage system by analytical quantification and pre-
507
scription data of the main regional health insurance, Journal of
508
Chromatography B (accepted) (2015).
11
Page 12 of 12
S2213-3437(15)30045-2 http://dx.doi.org/doi:10.1016/j.jece.2015.10.045 JECE 838
To appear in: Received date: Revised date: Accepted date:
2-7-2015 26-10-2015 31-10-2015
Please cite this article as: Sara Teixeira, Robert Gurke, Hagen Eckert, Klaus K¨uhn, Joachim Fauler, Gianaurelio Cuniberti, Photocatalytic degradation of pharmaceuticals present in conventional treated wastewater by nanoparticle suspensions, Journal of Environmental Chemical Engineering (2015), http://dx.doi.org/10.1016/j.jece.2015.10.045 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M an
100
ed
80
TiO2
60
1.5 μm
ZnO
ce pt
Degradation ratio C(t)/C(0) (%)
us
cr
ip t
Graphical Abstract
40
Ac
20 0
0
10
20 30 40 Time t (min)
50
60
Page 1 of 12
*Manuscript
Photocatalytic degradation of pharmaceuticals present in conventional treated wastewater by nanoparticle suspensions Sara Teixeiraa , Robert Gurkeb,c , Hagen Eckerta,d,∗, Klaus K¨uhna , Joachim Faulerb , Gianaurelio Cunibertia,d,e a Institute
for Materials Science and Max Bergmann Center of Biomaterials, TU Dresden, 01062 Dresden, Germany of Clinical Pharmacology, Faculty of Medicine Carl Gustav Carus, TU Dresden, 01307 Dresden, Germany c Research Association Public Health Saxony and Saxony Anhalt, Faculty of Medicine Carl Gustav Carus, TU Dresden, 01307 Dresden, Germany d Dresden Center for Computational Materials Science (DCCMS), TU Dresden, 01062 Dresden, Germany e Center for Advancing Electronics Dresden, TU Dresden, 01062 Dresden, Germany
cr
ip t
b Institute
us
Abstract
Pharmaceuticals have become an important public health issue as environmental pollutants over the last years. After
an
ingestion, pharmaceuticals are partly excreted unchanged. They can reach the wastewater treatment plant (WWTP) via the sewer network. Because the conventional treatments are ineffective in their removal, new methods should be approached, for example semiconductor photocatalysis. Several of the hitherto published studies analyzed the
M
degradation of model pollutants but for the degradation of pharmaceuticals in unspiked real wastewater further investigations are required. Therefore, we want to focus on the removal of pharmaceuticals in an actual effluent from a WWTP and investigate the effluent background effect. This study shows the heterogeneous photocatalytic degrada-
d
tion of 14 pharmaceuticals with initial concentrations Ci > 0.3 µgL−1 present in a WWTP effluent. We found that
te
UVA (1.5 mWcm−2 , intensity peak at 365 nm) irradiation of TiO2 P25 (A s = 56 m2 g−1 ) or ZnO (A s = 5.23 m2 g−1 ) nanoparticles leads to considerable degradation of the analyzed pharmaceuticals. With ZnO nanoparticles, 40 min
ce p
UVA irradiation was sufficient to degrade over 95 % of these pharmaceuticals (kapp = 8.6 · 10−2 s−1 ). Using TiO2 P25 on the other hand, it would take more than six times longer to reach the same level (kapp = 1.4 · 10−2 s−1 ). Carbamazepine dissolved in millipore water served as a comparison model. Also in this system ZnO presents faster degradation.
Keywords: Photocatalysis, pharmaceuticals, ultraviolet radiation, wastewater
1
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
1. Introduction
5
ceuticals in drinking water the World Health Organiza-
6
tion already reviewed scientific evidence to address this
2
Pharmaceuticals emerging in the aquatic ecosystems
7
issue. They are mostly introduced in the sewage sys-
3
have become an important public health issue over the
8
tem through excretion of unmetabolized compounds af-
4
past few years. To evaluate the impact of those pharma-
9
ter medical use or inappropriate disposal [1, 2, 3, 4] and
10
then transported into the wastewater treatment plants
11
(WWTPs). However, conventional WWTPs are not
12
designed to treat water polluted with pharmaceuticals
author. Tel.: +49 351 463-31461 Email address: [email protected] (Hagen
∗ Corresponding
Eckert) Preprint submitted to Journal of Environmental Chemical Engineering
October 26, 2015
Page 2 of 12
carbamazepine.
present at trace levels and therefore, the applied treat-
48
14
ments are ineffective in their removal [5, 6].
Con-
49
As photocatalysts we chose TiO2 and ZnO and com-
15
sequently, they reach the aquatic system and can be
50
pared the degradation efficiencies of both photocata-
16
found in surface and ground water [7, 8], soil and sed-
51
lysts. Despite several semiconductors have been studied
17
iments [8, 9] and even in drinking [10, 11] and tap wa-
52
for applications in wastewater decontamination, ZnO
18
ter [8, 12]. Although, normally pharmaceuticals do not
53
and TiO2 are frequently the most studied photocata-
19
present acute toxic effects on aquatic organisms due to
54
lysts because of their interesting optical properties, low
20
their low concentrations, in the range of ng to µg per
55
cost, and availability [22]. Although ZnO is usually
21
liter, concerns have been raised for chronic exposure,
56
described as the most active semiconductor [23], TiO2
22
due to their continuous input into the environment, act-
57
is used more frequently because it is more stable than
23
ing as slightly persistent pollutants [2, 4, 13].
58
ZnO in aqueous solution [24]. We used the photocat-
us
cr
ip t
13
For these reasons, diverse efforts have been made
59
alysts as nanoparticles in a slurry mixture to maximise
25
to remove pharmaceuticals from wastewater, such as
60
the surface area of the system. The upscaling of such
26
membrane filtration, activated carbon adsorption and
61
27
advanced oxidation processes (AOPs). AOPs are re-
62
28
commended when water pollutants have a high chem-
63
29
ical stability, allowing to achieve almost the total miner-
64
think that this can be solved in the near future, for ex-
30
alization of contaminants to carbon dioxide, water and
65
ample through the use of magnetic core nanoparticles
31
inorganic compounds or, at least, allow their partial oxi-
66
[25, 26]. Whenever photocatalytic systems are applied
32
dation to become more biodegradable and/or less harm-
67
in an actual wastewater treatment plant, a risk assess-
33
ful [3, 14].
68
ment regarding the material output into the environment
69
is necessary, due to their photo activity, size distribution
an
24
a setup provides a challenge regarding the separation of the nanoparticles from water after the treatment. In the
te
d
M
light of an active research regarding this problem, we
Different techniques involve the generation of hy-
35
droxyl radicals, which are nonselective and have twice
70
and potential toxicity for aquatic organisms in the case
36
the oxidizing power of chlorine [4, 6, 15, 16, 17]. Het-
71
of ZnO [27].
37
erogeneous semiconductor photocatalysis has become
38
an attractive method to remediate environmental con-
72
2. Experimental part
39
tamination due to its high photocatalytic activity, nontoxicity and photostability [3, 15, 18, 19, 20, 21]. How-
73
40
2.1. Chemicals and Materials
41
ever, most of the studies do not use unspiked wastew-
74
For the degradation experiments, TiO2 P25 (kindly
42
ater from sewage treatment plants but aqueous solu-
75
provided by Evonik), ZnO (IOLITEC Ionic Liq-
43
tions of model compounds or surface waters. Therefore,
76
uids Technologies GmbH), and carbamazepine (Sigma
44
we want to degrade pharmaceuticals in effluent samples
77
Aldrich) were used in this work.
45
from a WWTP and investigate the effluent background
78
tained from a Millipore Milli-Q System (Water, Milli-
46
effect. To do so, we additionally investigated the degra-
79
pore). For the SPE-HPLC-MS/MS analysis, acetoni-
47
dation process in millipore water artificially spiked with
80
trile, methanol (HiperSolv, HPLC-grade), and ammo-
ce p
34
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Water was ob-
2
Page 3 of 12
Table 1: Drugs (LLoQ - Lower Limit of Quantification)
Analyte
Drug class
Provider
Internal standard
LLoQ
Sigma
Carbamazepine D10
50 ngL−1
Pfizer
Gabapentin D10
200 ngL−1
Lamotrigine
Sigma
Lamotrigine 13C, 15N4
50 ngL−1
Oxcarbazepine
Cerilliant
Carbamazepine D10
Wyeth
Venlafaxine D6
Merck
Oxprenolol
LGC Standards
Oxprenolol
LGC Standards
Venlafaxine D6
Carbamazepine Gabapentin
Bisoprolol Celiprolol
beta blocker
Talinolol
50 ngL−1
50 ngL−1 50 ngL−1
cr
antidepressant
us
Venlafaxine
ip t
anticonvulsant
50 ngL−1
50 ngL−1
lipid-lowering drug
Sigma
Warfarin
50 ngL−1
Tramadol
opioid analgesic
Sigma
Tramadol 13C, D3
50 ngL−1
AstraZeneca
Amitriptyline D3
50 ngL−1
Venlafaxine D6
50 ngL−1
Trimipramine D3
50 ngL−1
Valsartan D9
100 ngL−1
Candesartan angiotensin receptor
Sigma
Irbesartan
antagonist
Sigma
M
Eprosartan
an
Bezafibrate
Valsartan
Sigma
81
nium acetate were purchased from Merck. Formic acid
82
(LC-MS grade) and Na2 EDTA (ACS reagents) were
83
obtained from Sigma and water (HPLC-grade) from
98
The specific surface area of the photocatalyst par-
84
VWR. The standards were provided by different sup-
99
ticles was determined by the Brunauer-Emmett-Teller
85
pliers as listed in Tab. 1. The treated wastewater was
100
(BET) method. This property was analyzed at 77 K by
86
kindly provided by the WWTP Kaditz located in Dres-
101
nitrogen adsorption-desorption in a Micromeritics TriS-
87
den, Germany, operated by Stadtentwsserung Dresden
102
tar analyzer (Micromeritics). Before performing ad-
88
GmbH. This treatment plant currently cleans the sewage
103
sorption experiments, samples (0.5 g) were outgassed
89
of 650,000 people and has a design capacity of 740,000
104
at 26.7 Pa and 350 °C for 6 h.
90
inhabitant equivalents. The yearly average sewage vol-
105
The morphology for both particle types were ana-
91
ume is about 55·106 m3 . The WWTP consist of primary
106
lyzed with a scanning electron microscope (SEM) op-
92
clarifier, activated sludge reactor and secondary clarifier
107
erated at 10 kV and 25 kV.
93
[28]. The sample was taken as a 24 h flow proportional
108
The UV-Vis diffuse reflectance spectra were obtained
94
composite effluent sample on June 24th 2014, stored at
109
using a Shimadzu UV-Vis spectrophotometer 2101PC
95
4 °C and analyzed on the next day. Further characteris-
110
in the range of 190 to 600 nm. It was equipped with a
96
tics of the sewage sample are summarized in Tab. 2.
111
diffuse reflectance attachment and we used BaSO4 as a
112
reference.
97
d
te
ce p
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
2.2. Nanoparticle characterization
3
Page 4 of 12
time intervals and centrifuged for one hour to remove
131
the catalyst nanoparticles. All experiments were carried
COD (Chemical Oxygen Demand)
37 mgL−1
132
out at room temperature. As controls, experiments were
4 mgL−1
133
carried out in the absence of catalyst nanoparticles or
134
without UVA irradiation. All other parameters in the control experiments were kept unchanged.
BOD (Biochemical Oxygen Demand) Nitrogen
ip t
130
Table 2: Treated wastewater parameters.
12.0 mgL−1
135
TKN (Total Kjeldahl Nitrogen)
<5.0 mgL−1
136
The effluent sample from the WWTP Kaditz was
Nammonium
0.31 mgL−1
137
filtered by a filter paper (VWR pore sizes 5-13 µm)
Nnitrite
0.03 mgL−1
138
to remove suspended particulate matter. −1
cr
Ntotal
Thereafter,
7.40 mgL−1
139
1 gL
Ninorganic
7.74 mgL−1
140
of the treated wastewater. The suspension was then ex-
141
posed under continuous stirring to UVA-radiation. Af-
142
terwards, the samples were centrifuged and analyzed by
Ptotal
0.86 mgL−1
Pphosphate
0.56 mgL−1
143
144
7.5
145
the SPE-HPLC-MS/MS method. In contrast to the previous experiment, we artificially
spiked millipore water with carbamazepine. Therefore,
M
pH
an
Phosphor
of catalyst was added to a volume of 100 mL
us
Nnitrate
2.3. Photocatalytic degradation experiments
146
50 mL of 12 mgL−1 carbamazepine solution with 1 gL−1
147
of catalyst was exposed under continuous stirring to
148
UVA-radiation. The samples were filtered (Rotilabo ny-
149
lon, pore size 0.2 µm) and centrifuged. Absorbance
d
113
The photocatalytic degradation was carried out in
115
borosilicate beakers (VWR) with 3.3 mm wall thick-
150
measurements were performed with a Varian CARY-
116
ness and 5 cm diameter.
151
100 UV-VIS Spectrophotometer.
152
limit for carbamazipine in this setup is 300 µgL−1 . 2.4. Analytic method
Under constant stirring 1
ce p
−1
te
114
of TiO2 P25 or ZnO was added to the samples.
The quantification
117
gL
118
Prior to illumination, the solutions containing the cat-
119
alyst were stirred in the dark for 30 min to achieve an
153
120
adsorption-desorption equilibrium of the pharmaceuti-
154
The analysis of the WWTP effluent sample and the
121
cals on the photocatalyst surface. Afterwards, the sam-
155
degradation experiments were conducted with a SPE-
122
ples were exposed to UVA-radiation. The illuminating
156
HPLC-MS/MS-method, which has been described in
123
device (UMEX) was equipped with six Philips 8 W mer-
157
the work of Gurke et al. [29]. Briefly summarised, 1
124
cury fluorescent tubes (Emax at 365 nm).
158
mL of the degradation experiment sample was taken,
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
125
The distance between the energy source and the pho-
159
adjusted to a pH of 3 by using formic acid and spiked
126
tocatalytic reactor was 15 cm. At this position the UVA
160
with 100 µL of the Internal Standard (IS) solution (10
127
intensity was determined by an UV34 Lux Meter (PCE)
161
µg/L). The samples were extracted using an Abimed
128
ranging from 1.5 - 1.6 mWcm−2 . Aliquots of 2 mL
162
ASPEC XL (Gilson) with Oasis HLB 10 mg Extrac-
129
of the reaction solution were withdrawn at determined
163
tion Cartridges (Waters). The eluates were evaporated
4
Page 5 of 12
to dryness under a gentle air stream at 50 °C and re-
195
where A(aq) is the investigated organic molecule in
165
dissolved in 250 µL mixture of solvent A and solvent
196
solution and A(ad) represents the adsorbed molecule.
166
B (80/20, v/v). The solvents A (97/3/0.05; v/v/v) and
197
M(aq) represents the mineralized components that are
167
B (5/95/0.05; v/v/v) were a composition of 2 mM am-
198
formed during the reaction. Temporal changes of the
168
monium acetate solution, acetonitrile, and formic acid.
199
molecule concentration in the solution CA(aq) (in 1 m−3 )
169
A LC-MS/MS system, consisting of a Dionex-HPLC
200
and the concentration on the photocatalyst particle sur-
170
composed of an UltiMate3000 Pump and Autosampler
201
face CA(ad) (in 1 m−2 ) are given by
171
(Thermo Fischer Scientific) with a Chromeleon 7 Chro-
172
matography Data System (Dionex Softron) and coupled
173
to an API 4000 tandem mass spectrometer (AB Sciex)
174
equipped with an electrospray ionization source (ESI),
175
was used for the analyses of the samples. The chro-
176
matographic separation was performed with a Synergi
202
177
2.5u HydroRP 100A, 100 x 2.0 mm and a C18 security
203
178
guard 4 mm x 2 mm (both Phenomenex) using a multi
204
179
step gradient out of solvent A and B with a total run-
180
time of 15 min. For the analyses, an injection volume
181
of 20 µL was chosen. The mass spectrometric analyses
jads = kads (1 − Θ) CA(aq)
(4)
182
were performed in multiple reaction monitoring (MRM)
jdes = kdes CA(ad)
(5)
183
mode with positive electrospray ionization. The Analyst
jreac = kreac CA(ad)
(6)
184
data system 1.6 (AB Sciex) was applied for MS control
185
and for the peak area evaluation, regression analysis of
186
calibration curves and calculation of concentrations.
cr
d dt C A(aq)
(2)
= as ( jdes − jads )
(3)
an
These equations include the specific surface area of the nanoparticles aS (surface area per solution volume) and different molecule fluxes
M
2.5. Photocatalytic degradation model
= jads − jdes − jreac
us
d dt C A(ad)
d
te
ce p
187
ip t
164
205
These molecule fluxes are mainly determined by the
206
corresponding concentrations and the three different
207
rate constants, describing the reaction (kreac ), adsorption
208
(kads ) and desorption (kdes ) processes. For the adsorption
209
process is also the surface coverage important (Θ).
188
To study the photocatalytic oxidation process we used
189
a suspension of photocatalytic nanoparticles due to the
210
190
large specific surface area in such a system. Addition-
211
191
ally, it is possible to model the degradation process in
212
192
this case efficiently as shown by Eckert et al. [30]. The
213
kads CA(aq) , kdes CA(ad) ), the surface concentration in the
193
simplified reaction formula to describe the degradation
214
system of equations (2) and (3) can be eliminated. The
194
process is
215
solution of Eq. (2) in the limit of small coverage Θ 1
216
reads CA(ad) = CA(aq) kads (kdes + kreac )−1 . Insertion of this
217
expression into Eq. (3) yields C˙ A(aq) = −kappCA(aq) with
218
the solution
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
A(aq) A(ad) → M(aq)
(1)
In the present case of a fast establishing adsorptiondesorption equilibrium with quasi-stationary C˙ concentration C˙ = 0 (i.e. A(ad)
surface
A(ad)
5
Page 6 of 12
Table 3: Adsorption and desorption rate constants derived from the
nanoparticles.
where the apparent degradation rate constant is given
219
catalysts
kads (ms−1 )
kdes (s−1 )
TiO2
5.3·10−9
4.5·10−4
ZnO
2.2·10−7
5.9·10−3
by
kapp = as kads kreac (kdes + kreac )−1 .
(8)
cr
220
experiments with carbamazepine using photocatalytic TiO2 and ZnO
(7)
ip t
CA(aq) (t) = CA(aq),0 exp(−kapp t)
Thus, for kdes kreac , we find kapp = as kads . This
236
present as monodisperse nanospheres. In the ZnO pow-
222
means that the degradation rate becomes adsorption-
237
der more complex polydisperse rectangular structures
223
limited and does not depend on the reaction rate con-
238
were found.
224
stant.
242
243
est level at 365 nm. In contrast, the absorption potential
244
of TiO2 reaches its maxima around 310 nm. This indi-
245
cates that with ZnO the light source can be utilized more
240
A specific surface area of 56 m g
was obtained
for the TiO2 P25 nanopowder with the BET method.
For the ZnO nanopowder a specific surface area of
230
5.23 m2 g−1 was found. In suspension it is expected an
231
additional clustering, especially for TiO2 , which will re-
232
duce the active surface area.
te
229
ce p ZnO
1.5 μm
TiO2
10kV, 5mm, x50000
observed that ZnO absorbance already reached its high-
efficiently.
100 80 60 40
max
20 min 0 436
25kV, 9mm, x50000
Lamp TiO2 ZnO
Absorbance
228
3.1. Characterization of TiO2 and ZnO nanoparticles 2 −1
of the radiation lamp is plotted in the figure. It can be
M
227
241
TiO2 are included in Fig. 2. Also the emission spectra
Light intensity (E365nm = 100)
226
3. Results and discussion
d
225
The UV-Vis diffuse reflectance spectra of ZnO and
an
239
us
221
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
405
365 313 Wavelength λ (nm)
Figure 1: SEM images of colloidal TiO2 and ZnO. Figure 2: Light output of the used mercury fluorescent tubes normalized at 365 nm compared to the diffuse reflectance spectra of TiO2 233
This difference can also be qualitatively observed in
234
the SEM images (Fig. 1). TiO2 particles show a finer
235
structure than the ones in the ZnO sample. TiO2 is
and ZnO.
246
6
Page 7 of 12
TiO2
simulation experiment
b)
c)
d)
cr
a)
0
20
40
60 0 Time t (min)
20
us
50
0
effluent C(0) = 1.4 µgL−1
0 100
ip t
50
40
an
Degradation ratio C(t)/C(0) (%)
100
isolated C(0) = 12.0 mgL−1
ZnO
60
Figure 3: Measured concentration during the degradation of carbamazepine under UVA irradiation with 1 gL−1 catalysts TiO2 (a+c) and ZnO
M
(b+d). In the upper row, carbamazepine was dissolved in millipore water (a+b) and measured by UV/VIS spectrophotometry. In the lower row the results of the degradation of carbamazepine in the treated effluent analysed by HPLC-MS/MS-method are shown. The full lines represent fits for carbamazepine dissolved in millipore water according to the presented model using the values of kads and kdes determined from experiments in the
247
3.2. Influence of WWTP effluent
te
d
dark (Table 3).
262
ceuticals, two different control experiments were con-
263
ducted without nanoparticles and without UVA irradia-
Carbamazepine’s photodegradation was evaluated in
249
a real WWTP effluent sample, alongside other occur-
264
tion for 45 min. In the case of the control exposed to
250
ring pharmaceuticals as listed in Tab. 1. The effluent of
265
UVA-radiation no significant change in the monitored
251
a WWTP represents a challenging matrix for photocat-
266
concentrations occurred. After irradiation the average
252
alytic degradation. Different parameters like ion con-
267
change over all monitored drugs was 0.8 % (SD 4.8 %).
253
centration, pH value or the manifold mixture of organic
268
However for the controls with nanoparticles kept in the
254
and inorganic molecules lead us to conjecture about the
269
dark we observed an initial drop of the concentrations
255
concentration development over time that may be al-
270
before they stabilised, so that the changes in concentra-
256
tered compared to a ’clean’ laboratory experiment. To
271
tion after the initial 30 min of adsorption is 0.3 % (SD
257
examine this assumption we selected the anticonvulsant
272
4.9 %) for TiO2 and 0.2 % (SD 5.7 %) for ZnO. This
258
carbamazepine dissolved in millipore water, as a com-
273
effect is due to the adsorption on the catalyst’s surface
259
parison ’clean’ degradation experiment.
274
until the adsorption-desorption equilibrium is reached.
ce p
248
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
260
To assure that only the combination of nanoparticles
275
This is consistent with the idea that the UVA-radiation
261
and UV-radiation causes the degradation of the pharma-
276
per se does not induce the degradation process nor the
7
Page 8 of 12
nanoparticles without UVA irradiation.
100 deviation TiO2 ZnO
278
Carbamazepine’s concentration was considerably re-
279
duced over time in the presence of the catalysts and
280
UVA-radiation. The same behaviour was observed for
281
carbamazepine present in the wastewater effluent sam-
282
ple. In Fig. 3 we compare the four experiments with
283
carbamazepine dissolved in millipore water and present
284
in the effluent. It can be noticed that the degradation
285
by ZnO is significantly faster than with TiO2 in both
286
setups which is in high accordance with the conducted
287
simulation of the degradation. These results are shown
288
in Tab. 3 and demonstrate that carbamazepine can ad-
maceuticals (Tab. 1) measured by HPLC-MS/MS-method. The grey
289
sorb more than 40 times faster on the ZnO than on the
290
TiO2 surface. Because of this, ZnO presents a higher
291
degradation rate compared to TiO2 regardless its larger
an
Degradation ratio C(t)/C(0) (%)
277
292
surface area.
80 60
20
10
20 30 40 Time t (min)
50
60
us
0
cr
ip t
40
0
Figure 4: Average degradation ratio over time of the 14 selected phar-
area corresponds to the standard deviation.
3.3. Degradation of pharmaceuticals in the WWTP ef-
M
301
302
fluent
To determine the degradation in an effluent sample of
When the parameters determined from the millipore
303
294
water experiment (Fig. 3 a, b) are compared to the out-
304
the WWTP Dresden Kaditz, 55 target pharmaceuticals
295
come from the effluent (Fig. 3 c, d) similar results can
305
based on their prescription numbers were selected based
296
be found. This shows that the effluent background does
306
on a previous study of Gurke et al. [29]. For monitoring
297
not reduce the efficiency of the selected photocatalytic
307
the degradation process, the initial concentration (Ci )
298
particles, and the method to model the degradation in
308
of the pharmaceuticals needs to be significantly higher
299
the millipore water experiment can also be applied for
309
than the corresponding detection limit. Therefore, we
300
pharmaceuticals investigated in the effluent.
310
set the lower limit to be Ci > 0.3 µgL−1 . After an initial
311
analysis of the effluent sample, 14 pharmaceuticals were
312
selected to be monitored in the degradation experiment
313
(Tab. 1).
ce p
te
d
293
Table 4:
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Apparent rate constants average and adsorption rate
constants average for the 14 studied pharmaceuticals (Tab. 1) with the photocatalysts TiO2 and ZnO.
314
Based on this criteria the anticonvulsants carba-
315
mazepine (Ci = 1.29 µgL−1 ), gabapentin (Ci =
catalyst
kapp (s−1 )
kads (ms−1 )
316
11.30 µgL−1 ), lamotrigine (Ci = 0.98 µgL−1 ), and
TiO2
1.4·10−2
4.0·10−9
317
oxcarbazepine (Ci = 0.63 µgL−1 ), the antidepres-
ZnO
8.6·10−2
2.7·10−7
318
sant venlafaxine (Ci = 0.58 µgL−1 ), the beta block-
319
ers bisoprolol (Ci = 0.58 µgL−1 ), celiprolol (Ci =
320
0.35 µgL−1 ), and talinolol (Ci = 0.43 µgL−1 ), the lipid-
8
Page 9 of 12
lowering drug bezafibrate (Ci = 0.48 µgL−1 ), the opi-
356
lower selectivity of ZnO compared to TiO2 . In the ZnO
322
oid analgesic tramadol (Ci = 0.624 µgL ), as well as
357
experiments, all target pharmaceuticals are degraded in
323
the angiotensin receptor antagonists candesartan (Ci =
358
a similar way with just slight deviations. This character-
324
1.30 µgL ), eprosartan (Ci = 0.56 µgL ), irbesartan
359
istic is very important in real applications due to the dif-
325
(Ci = 1.50 µgL−1 ), and valsartan (Ci = 3.59 µgL−1 )
360
ferent mixtures of pharmaceuticals present in wastewa-
326
were analysed in the degradation experiment. Please
361
ter depending on regional consumption patterns or sea-
327
keep in mind that this data is a mere snapshot because
362
son of the year. It can be assumed that the treatment of
328
the concentration of micropollutants can significantly
363
pharmaceuticals present in WWTP effluents using ZnO
329
vary in sewage samples based on a diverse range of pa-
364
and UVA-radiation will result in a near complete degra-
330
rameters, like weather conditions.
365
dation of the pollutants in a relative short period of time.
366
4. Conclusions
cr
−1
331
The degradation experiments were traced for all tar-
332
get pharmaceuticals and are presented in the supple-
333
mentary material. For the SPE extraction just a small
334
sample volume of 1 mL was necessary, which allowed
367
335
a high temporal resolution monitoring (ten samples in
368
336
one hour) without disturbing the experiment by taking
369
an
−1
us
−1
ip t
321
337
out samples with a large volume. An overview of the
370
ents would be useful to prevent the contamination of
338
two photocatalytic materials is shown in Fig. 4. Af-
371
surface water. Semiconductor photocatalysis is recom-
339
ter 40 min, an average degradation of more than 95 %
372
mended whenever water pollutants present low degrad-
340
for the samples treated with ZnO was already observed.
373
ability and/or high chemical stability.
341
During the same period of time the pharmaceuticals
374
The pollutants are at least in parts mineralized due to
342
treated with TiO2 degraded by 40 %. The simulation
375
the generation of highly oxidative species and electron-
343
data shows that treatment with TiO2 would take over
376
hole pairs. Photocatalysis may be seen as a complemen-
344
four hours to achieve the same result as ZnO. These
377
tary method to the already existing technologies to im-
345
differences in degradation rates are even more signif-
378
prove the removal rates of pollutants, such as pharma-
346
icant taking into account the smaller surface area of
379
ceuticals. In this study, the applicability of a previously
347
ZnO. The resulting apparent rate constants and adsorp-
380
developed LC-MS/MS method by evaluating the degra-
348
tion rate constants are listed in Tab. 4 and are in accor-
381
dation of pharmaceuticals present in a real wastewater
349
dance with the results from carbamazepine dissolved in
382
sample by photocatalysis has been demonstrated. We
350
millipore water. In literature different new approaches
383
were able to conduct successfully degradation experi-
351
besides commercial nanoparticles are presented. The
384
ments of various pharmaceuticals present in the effluent
352
apparent rate constants found for example from TiO2
385
of a WWTP via photocatalysis by ZnO and TiO2 un-
353
nanowires is in average the same order of magnitude as
386
der UVA-radiation. Although ZnO shows higher degra-
354
the commercial TiO2 particles, but slower compared to
387
dation rates it poses a strong engineering challenge, to
355
the ZnO particles [31]. Furthermore, it was observed a
388
prevent ZnO contamination of the aquatic environment,
In general, pharmaceuticals are not fully removed by
urban WWTPs and can be detected in the effluents. Removing these type of pollutants from wastewater efflu-
M
d
te
ce p
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
9
Page 10 of 12
5. Acknowledgements
391
This work is funded by the European Union (ERDF)
392
and the Free State of Saxony via the ESF project
393
100098212 InnoMedTec. The authors are thankful to
394
Jackie Le, Pedro Martins, Sam Diamond, Thomas K¨ase-
395
berg, Quirina Roode-Gutzmer, Ignacio Gonzalez and
423
[6] J. a. Pereira, V. Vilar, M. Borges, O. Gonz´alez, S. Esplugas,
424
R. Boaventura, Photocatalytic degradation of oxytetracycline
425
using TiO2 under natural and simulated solar radiation, Solar
426
Energy 85 (2011) 2732–2740.
427
[7] O. A. H. Jones, N. Voulvoulis, J. N. Lester, Aquatic environ-
428
mental assessment of the top 25 English prescription pharma-
429
ceuticals., Water research 36 (2002) 5013–22.
ip t
390
due to its potential toxicity.
430
[8] W. C. Li, Occurrence, sources, and fate of pharmaceuticals in
431
aquatic environment and soil., Environmental pollution (Bark-
432
ing, Essex : 1987) 187 (2014) 193–201.
cr
389
433
[9] W. Mrozik, J. Stefa´nska, Adsorption and biodegradation of an-
Hoai Nga Le for their support and many valuable dis-
434
tidiabetic pharmaceuticals in soils., Chemosphere 95 (2014)
397
cussions. Additional we want to give acknowledgment
435
281–8.
398
to Rita Knoche and Norbert Lucke for providing the wastewater samples.
436
[10] T. Ternes, M. Meisenheimer, D. McDowell, F. Sacher, H.-J.
437
Brauch, B. Haist-Gulde, G. Preuss, U. Wilme, N. Zulei-Seibert,
438
Removal of pharmaceuticals during drinking water treatment.,
439
400
440
Supplementary information
441
Supplementary information related to this article can
443
402
be found online at doi:
444
d
445
403
References
Environmental science & technology 36 (2002) 3855–63.
[11] M. S. Fram, K. Belitz, Occurrence and concentrations of pharmaceutical compounds in groundwater used for public drinkingwater supply in California., The Science of the total environ-
M
401
442
an
399
us
396
446
ment 409 (2011) 3409–17.
[12] T. E. Doll, F. H. Frimmel,
Fate of pharmaceuticals -
photodegradation by simulated solar UV-light., Chemosphere 52 (2003) 1757–69. [13] N. Xekoukoulotakis, C. Drosou, C. Brebou, E. Chatzisymeon,
[1] C. D. Metcalfe, B. G. Koenig, D. T. Bennie, M. Servos,
448
E. Hapeshi, D. Fatta-Kassinos, D. Mantzavinos, Kinetics of UV-
405
T. Ternes, R. Hirsch, Occurrence of neutral and acidic drugs
449
A/TiO2 photocatalytic degradation and mineralization of the an-
406
in the effluents of Canadian sewage treatment plants., Environ-
450
tibiotic sulfamethoxazole in aqueous matrices, Catalysis Today
407
mental Toxicology and Chemistry 22 (2003) 2872–2880.
451
161 (2011) 163–168.
ce p
te
447
404
408
[2] M. Isidori, M. Bellotta, M. Cangiano, A. Parrella, Estrogenic
452
409
activity of pharmaceuticals in the aquatic environment., Envi-
453
ronment international 35 (2009) 826–9.
454
410
[14] R. Andreozzi, V. Caprio, A. Insola, R. Marotta,
Advanced
oxidation processes (AOP) for water purification and recovery, Catalysis Today 53 (1999) 51–59.
411
[3] M. Klavarioti, D. Mantzavinos, D. Kassinos, Removal of resid-
412
ual pharmaceuticals from aqueous systems by advanced oxida-
456
reactor for water purification, Catalysis Today 40 (1998) 73–83.
413
tion processes., Environment international 35 (2009) 402–17.
457
[16] L. Yang, L. Yu, M. Ray, Degradation of paracetamol in aqueous solutions by TiO2 photocatalysis., Water research 42 (2008)
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
455
[15] A. Ray, A. Beenackers, Development of a new photocatalytic
414
[4] N. Xekoukoulotakis, N. Xinidis, M. Chroni, D. Mantzavinos,
458
415
D. Venieri, E. Hapeshi, D. Fatta-Kassinos, UV-A/TiO2 photo-
459
416
catalytic decomposition of erythromycin in water: Factors af-
460
[17] E. Hapeshi, A. Achilleos, M. Vasquez, C. Michael, N. Xek-
417
fecting mineralization and antibiotic activity, Catalysis Today
461
oukoulotakis, D. Mantzavinos, D. Kassinos, Drugs degrading
418
151 (2010) 29–33.
462
photocatalytically: Kinetics and mechanisms of ofloxacin and
3480–8.
419
[5] V. Cunningham, S. Binks, M. Olson, Human health risk assess-
463
atenolol removal on titania suspensions., Water research 44
420
ment from the presence of human pharmaceuticals in the aquatic
464
(2010) 1737–46.
421
environment., Regulatory toxicology and pharmacology : RTP
465
422
53 (2009) 39–45.
[18] A. Fujishima, T. Rao, D. Tryk, Titanium dioxide photocatalysis,
10
Page 11 of 12
466
Journal of Photochemistry and Photobiology C: Photochemistry
509
[30] H. Eckert, M. Bobeth, S. Teixeira, K. K¨uhn, G. Cuniberti, Mod-
467
Reviews 1 (2000) 1–21.
510
eling of photocatalytic degradation of organic components in
468
[19] F. Thevenet, O. Gua¨ıtella, J. Herrmann, A. Rousseau, C. Guil-
511
water by nanoparticle suspension, Chemical Engineering Jour-
469
lard, Photocatalytic degradation of acetylene over various tita-
512
nal (2014).
470
nium dioxide-based photocatalysts, Applied Catalysis B: Envi-
513
[31] A. Hu, X. Zhang, K. D. Oakes, P. Peng, Y. N. Zhou, M. R.
ronmental 61 (2005) 58–68.
Servos, Hydrothermal growth of free standing TiO2 nanowire
[20] M. Faramarzpour, M. Vossoughi, M. Borghei, Photocatalytic
515
membranes for photocatalytic degradation of pharmaceuticals,
473
degradation of furfural by titania nanoparticles in a floating-bed
516
Journal of Hazardous Materials 189 (2011) 278–285.
474
photoreactor, Chemical Engineering Journal 146 (2009) 79–85.
477
idences from spectroscopic studies, Journal of Luminescence
478
131 (2011) 1975–1981.
479
[22] M. Lazar, S. Varghese, S. Nair, Photocatalytic Water Treatment
480
by Titanium Dioxide: Recent Updates, Catalysts 2 (2012) 572–
481
601.
482
[23] A. Fujishima, X. Zhang, Donald A. Trykc, TiO2 photocataly-
483
sis and related surface phenomena, Surface Science Reports 63
485
(2008) 515–582. [24] L. Pan, X. Liu, Z. Sun, C. Q. Sun, Nanophotocatalysts via
M
484
486
microwave-assisted solution-phase synthesis for efficient pho-
487
tocatalysis, Journal of Materials Chemistry A 1 (2013) 8299. [25] R. Chalasani, S. Vasudevan,
Cyclodextrin-Functionalized
d
488
Fe3O4@TiO2: Reusable, Magnetic Nanoparticles for Photocat-
490
alytic Degradation of Endocrine-Disrupting Chemicals in Water
491
Supplies, ACS Nano 7 (2013) 4093–4104.
te
489
492
[26] R. Wang, X. Wang, X. Xi, R. Hu, G. Jiang, Preparation and Photocatalytic Activity of Magnetic Fe3O4/SiO2/TiO2 Composites,
494
Advances in Materials Science and Engineering 2012 (2012) 1–
ce p
493
495
cr
interaction of colloidal TiO2 nanoparticles with lysozyme: Ev-
us
[21] A. Kathiravan, M. Asha Jhonsi, R. Renganathan, Photoinduced
an
475 476
ip t
514
472
471
8.
496
[27] L. K. Adams, D. Y. Lyon, P. J. Alvarez, Comparative eco-
497
toxicity of nanoscale TiO2, SiO2, and ZnO water suspensions,
498
Water Research 40 (2006) 3527–3532.
Ac
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
499
[28] C. Marx, N. G¨unther, S. Schubert, R. Oertel, M. Ahnert,
500
P. Krebs, V. Kuehn, Mass flow of antibiotics in a wastewater
501
treatment plant focusing on removal variations due to opera-
502
tional parameters, Science of The Total Environment 538 (2015)
503
779–788.
504
[29] R. Gurke, J. Rossmann, S. Schubert, R. Oertel, T. Sandmann,
505
M. R¨oßler, J. Fauler, Predictive approach of pharmaceuticals
506
in an urban sewage system by analytical quantification and pre-
507
scription data of the main regional health insurance, Journal of
508
Chromatography B (accepted) (2015).
11
Page 12 of 12