Occurrence and removal of pharmaceuticals, caffeine and
DEET in wastewater treatment plants of Beijing, China
Qian Sui, Jun Huang, Shubo Deng, Gang Yu*, Qing Fan
POPs Research Centre, Department of Environmental Science & Engineering, Tsinghua University, Beijing 10084, China
a r t i c l e i n f o
Received 14 January 2009
Received in revised form
5 July 2009
Accepted 8 July 2009
Available online 15 July 2009
a b s t r a c t
The occurrence and removal of 13 pharmaceuticals and 2 consumer products, including
antibiotic, antilipidemic, anti-inflammatory, anti-hypertensive, anticonvulsant, stimulant,
insect repellent and antipsychotic, were investigated in four wastewater treatment plants
(WWTPs) of Beijing, China. The compounds were extracted from wastewater samples by
solid-phase extraction (SPE) and analyzed by ultra-performance liquid chromatography
coupled with tandem mass spectrometry(UPLC–MS/MS). Most of the targetcompounds were
detected, with the concentrations of 4.4 ngL?1–6.6 mgL?1and 2.2–320 ngL?1in the influents
and secondary effluents, respectively. These concentrations were consistent with their
consumptions in China, and much lower than those reported in the USA and Europe. Most
from ?12% to 100% were achieved during the secondary treatment. In the tertiary treatment,
different processes showed discrepant performances. The target compounds could not be
eliminated by sand filtration, but the ozonation and microfiltration/reverse osmosis (MF/RO)
processes employed in two WWTPs were very effective to remove them, showing their main
contributions to the removal of such micro-pollutants in wastewater treatment.
ª 2009 Elsevier Ltd. All rights reserved.
With the progress of sensitive analytical techniques, the
frequent detection of various pharmaceuticals in the aquatic
environment has received global concerns of both the
academic community and the public (Daughton and Ternes,
1999; Jones et al., 2005). After intake by humans or animals,
the pharmaceuticals will be partially converted to metabo-
lites, however, partially excreted unchanged or as conjugates,
and finally delivered to the wastewater treatment plants
(WWTPs). As there is no unit specifically designed to remove
these compounds, the elimination by most WWTPs seems to
be inefficient (Ternes, 1998; Castiglioni et al., 2006; Lishman
et al., 2006; Nakada et al., 2006; Santos et al., 2007; Vieno et al.,
2007b; Xu et al., 2007; Gulkowska et al., 2008; Paxeus, 2004).
Together with treated wastewater, these compounds are
released to the aquatic environment, and consequently found
to contaminate the receiving water bodies (Lindqvist et al.,
2005; Kasprzyk-Hordern et al., 2009), or even raw water sour-
ces of drinking water treatment plant (Ternes et al., 2002;
Vieno et al., 2007a; Radjenovic et al., 2008). Meanwhile, results
of toxicology studies have revealed that some pharmaceuti-
cals are suspected to have direct toxicity to certain aquatic
organisms (Ferrari et al., 2003; Jjemba, 2006; Grung et al., 2008;
Quinn et al., 2008). Besides, their continual but undetectable
effects could accumulate slowly, and finally lead to irrevers-
ible change on wildlife and human beings (Daughton and
Ternes, 1999). Therefore, the occurrence and behavior of
pharmaceuticals in the WWTPs, which are both the sink and
source of the compounds, should be focused on. So far,
concentrations of pharmaceuticals from various therapeutic
classes in the WWTPs have been well documented in the
* Corresponding author. Tel.: þ86 10 62787137; fax: þ86 10 62794006.
E-mail address: firstname.lastname@example.org (G. Yu).
0043-1354/$ – see front matter ª 2009 Elsevier Ltd. All rights reserved.
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
water research 44 (2010) 417–426
North America (Thomas and Foster, 2005; Lishman et al.,
2006), Japan(Nakadaet al.,2006) and someEuropeancountries
(Ternes, 1998; Castiglioni et al., 2006; Santos et al., 2007; Vieno
et al., 2007b; Jones et al., 2007; Paxeus, 2004). Reported species
and concentrations ofpharmaceuticals varied fromcountry to
country, and plant to plant, owing to the different usage
patterns. Meanwhile, the removal efficiencies of pharmaceu-
ticals also varied much (Nakada et al., 2006; Gulkowska et al.,
2008), indicating that the removal could be affected by both
the compound-specific properties, and the factors concerning
specific WWTPs, such as types of treatment processes, solids
temperature, etc. In recent years, very few studies about the
situation in China have been reported. Only one specific
therapeutic class, antibiotics, has been investigated by limited
previous studies (Xu et al., 2007; Gulkowska et al., 2008; Chen
et al., 2008). Therefore, it is necessary and important to
investigate the occurrence and removal of pharmaceuticals
from different therapeutic classes in the WWTPs of China.
Due to the low efficiency of conventional wastewater
treatment processes, some advanced treatment technologies
have been evaluated. Ozonation was found to be effective to
remove pharmaceuticals in real municipal WWTPs of Japan
(Nakada et al., 2007; Okuda et al., 2008) and Germany (Ternes
et al., 2003). Nanofiltration (NF) and reverse osmosis (RO)
membrane filtration, the well-proven technologies to remove
applied at bench, pilot and full scale (Khan et al., 2004; Nghiem
et al., 2005; Drewes et al., 2005; Al-Rifai et al., 2007; Watkinson
et al., 2007; Comerton et al., 2008; Radjenovic et al., 2008).
Retention behavior of pharmaceuticals during the processes
associated with physicochemical properties of pharmaceuti-
cals, membranes as well as the solution chemistry, and
mechanisms of pharmaceutical rejection have been discussed
in Kimura et al. (2004), Nghiem et al. (2005), Nghiem and
Coleman (2008) and Comerton et al. (2008). Recently, consid-
ering the requirement of reclaimed water, several advanced
treatment facilities have been installed in the WWTPs of
Beijing. However, the removal efficiency of micro-pollutants,
such as pharmaceuticals, has not been evaluated yet.
In the present study, we investigated the contamination
levels of 13 pharmaceuticals and 2 consumer products from 8
classes (i.e. antibiotic, antilipidemic, anti-inflammatory, anti-
hypertensive, anticonvulsant, stimulant, insect repellent and
antipsychotic) in four WWTPs of Beijing, China, which have
different advanced treatment units, and evaluated the elimi-
nation efficiencies of the target pharmaceuticals. To the best
of our knowledge, this is the first report on the occurrence and
removal of pharmaceuticals and consumer products from
multiple classes in the WWTPs of China, especially for the
situation during the advanced treatment processes.
2. Materials and methods
All the standards including chloramphenicol (CP), nalidixic
acid (NA), trimethoprim (TP), bezafibrate (BF), clofibric acid
(CA), gemfibrozil (GF), diclofenac (DF), indometacin (IM),
ketoprofen (KP), mefenamic acid (MA), metoprolol (MTP),
carbamazepine (CBZ), caffeine (CF), N,N-diethyl-meta-tolua-
mide (DEET) and sulpiride (SP) (Appendix) were of analytical
grade (>90%), and purchased from Sigma–Aldrich (Steinheim,
Germany). Isotopically labeled compounds, used as internal
Aldrich, and3D-mecoprop from Dr. Ehrenstorfer (Augsburg,
Germany). HPLC grade methanol, acetone, dichloromethane,
hexane, as well as formic acid were provided by Dikma (USA),
were prepared by diluting the stock solutions before each
analytical run. All the solutions were stored at 4?C in the dark.
2.2. Sample collection
Four full-scale municipal WWTPs, referred as A, B, C and D,
were selected in our study. These WWTPs employ similar
conventional treatment processes: primary treatment to
remove particles coupled with secondary biological treat-
ment. For the secondary biological treatment processes,
WWTPs A and D employ anaerobic/anoxic/oxic (A2/O) acti-
vated sludge process, anoxic/oxic (A/O) activated sludge
process is adopted in WWTP B, and WWTP C employs oxida-
tion ditch (OD). Other detailed information on each WWTP,
such as inhabitantsserved,dailyflow,HRTand SRTareshown
in Table 1. Part of the secondary effluents was further treated
in WWTPs A, B and D, by the processes of ultrafiltration
(UF)/ozone, sand filtration (SF) and microfiltration/reverse
osmosis (MF/RO), respectively. In WWTP A, a dead-end
ultrafiltration system (Zenon GE) is used. The whole system
has 6 trains of Zee-Weed 1000 membrane. Each train contains
9 cassettes of 57–60 modules per cassette. The membrane,
with the pore size of 0.02 mm, is made by PVDF. The module is
operated in an outside/in configuration at a constant flow of
23 L (m2h)?1
and the total treatment capacity reaches
80,000 m3d?1. The membrane is hydraulically backwashed at
a constant flow rate of 34 (m2h)?1, and 29 times per day. The
backwash phase lasts for 1 min. Maintenance cleaning is
conducted once per day. Membranes are soaked in the sodium
hypochlorite solution(50 mgL?1) for 25 min. Forthe ozonation
process, gaseous ozone is generated from an ozone generator
(MitsubishiElectric).The ozonedosageand contact timein the
reaction tank is 5 mgL?1and 15 min, respectively. The pH of
the wastewater before ozonation ranges 6.5–8.0 and shows no
significant change after ozonation. As the heart of the
advanced treatment in WWTP D, a spiral-wound crossflow
module is employed for the reverse osmosis (RO) membrane
filtration. The RO membrane (Filmtec, DOW) is made from
a thin-film composite polyamide material. Each module is
designed to operate at a water flux of 1.3 m3h?1, and a product
water recovery of 75–80%. The trans-membrane pressure is
between 0.04 and 0.06 MPa, and the salt rejection remained at
the level of 99%. Every 3–6 months, normally when the trans-
membrane pressure reaches above 0.06 MPa, the membrane is
cleaned with 0.1%(w) sodium hydroxide solution (for organic
foulants), 2%(w) citric acid (for inorganic foulants) and 0.5%(w)
formaldehyde (as biocide). Schematic diagram of treatment
processes in the four WWTPs is shown in Fig. 1.
water research 44 (2010) 417–426
The samples were collected once from the four WWTPs
during June and July 2008, with no compensation for HRT. All
of them were collected as grab samples in duplicate (500 mL
for influents and 1000 mL for the others) in prewashed amber
glass bottles, kept in the cooler and transported to the labo-
ratory. Immediately after delivery to the laboratory, they were
filtered through prebaked (400?C, >4 h) glass microfiber filters
(GF/F, Whatman) to remove particles and stored at 4?C before
2.3. Sample extraction and analysis
The method for the extraction and analysis of pharmaceuti-
cals and consumer products is presented elsewhere (Sui et al.,
in press) and briefly described here. After the solid-phase
extraction (SPE) cartridges (Oasis, HLB, 200 mg, 6 mL) were
conditioned, wastewater samples, added with internal stan-
dards and adjusted to pH¼7, were introduced to the cartridge
via a PTFE tube, at a flow rate of 5–10 mLmin?1. After washing
by 5 mL of 5.0% methanol solution, the cartridge was dried
under vacuum for 2 h and eluted with 5 mL of methanol. The
extract was then concentrated to 0.4 mL under a gentle
nitrogen stream and stored at 4?C for analysis. Concentra-
tions of the target compounds were analyzed using ultra-
performance liquid chromatography coupled with tandem
mass spectrometry (UPLC–MS/MS). Analytes were separated
using Waters Acquity UPLC system (Waters Corporation, USA)
equipped with Acquity UPLC BEH C18 column (50?2.1 mm,
particle size of 1.7 mm), and detected by Quattro Premier XE
tandem quadrupole mass spectrometry (Waters Corp., USA)
equipped with an electrospray ionization source. The analysis
was carried out in multiple reaction monitoring (MRM) mode,
b WWTP B
c WWTP C
Fig. 1 – Schematic diagram of the treatment processes in the four WWTPs and sampling site location (C).
Table 1 – Information of the WWTPs investigated.
HRT (h)SRT (d)Secondary treatment Tertiary treatment
water research 44 (2010) 417–426
and in general, two precursorion/product ion transitions were
monitored for one compound with the purpose of quantifi-
cation and confirmation.
2.4. Quality control
For each sampling, 500 mL Milli-Q water in an amber glass
bottle as a field blank was brought to the WWTPs, exposed to
the environment where the samples were taken from, and
then delivered back to the laboratory with samples. For each
set of samples (normally 10 samples), at least one procedural
blank was prepared from ultra-pure water in the laboratory.
Both the field blanks and procedural blanks were run identi-
cally to the wastewater samples, and the concentrations of
target compounds were below the limit of quantification
(LOQ). The absolute recoveries, calculated by comparing the
concentrations of target compounds in spiked and unspiked
wastewaters, were proved to be 73–102% and 50–95% in the
effluent and influent for most compounds, respectively. While
for several compounds (i.e. sulpiride, gemfibrozil, mefenamic
acid), the absolute recoveries were not satisfactory. However,
usedforpositive and negative ion moderespectively wereable
to compensate for the loss of most analytes, and relative
recoveries were 67–130% for all the analytes in the effluent
and 79–140% in the influents except mefenamic acid (251%)
and nalidixic acid (178%). Therefore, the concentrations of
these two compounds in the wastewater influents were not
quantitatively determined and reported. The LOQs were
0.3–5.5 ngL?1and 0.7–20 ngL?1in the effluent and influent,
respectively. Detailed information about the calibration,
recoveries, LOQ, matrix effects, etc. were described in Sui et al.
(in press), and briefly listed in Table 2. As duplicate samples
were collected at each sampling site, mean concentrations
were adopted. In most cases, deviations of duplicate samples
were less than 20%. For some tertiary effluent samples, low
3D-mecoprop, the surrogate standards
concentrations of some target compounds (i.e. caffeine, DEET,
carbamazepine) resulted in slightly higher deviations.
3. Result and discussion
As shown in Fig. 2, 12 target compounds were detected in all
the influent samples from the four WWTPs, while ketoprofen
was below LOQ in all wastewater samples. The most
abundant compounds detected were the consumer products,
caffeine(3.4–6.6 mgL?1) and
(0.6–1.2 mgL?1), probably due to the large consumption of
drinks containing caffeine (i.e. coffee, tea, etc.) and wide
application of insect repellent during the summer time when
we sampled. Diclofenac, trimethoprim, sulpiride, carbama-
zepine, indometacin and metoprolol showed relatively high
concentrations (Fig. 2). A similar composition distribution
was observed among all the influents of the four WWTPs.
The concentrations of target pharmaceuticals except
diclofenac and trimethoprim, were much lower than those
reported in the European and North American countries
(Thomas and Foster, 2005; Lishman et al., 2006; Vanderford
and Snyder, 2006; Santos et al., 2007; Gomez et al., 2007; Vieno
et al., 2007b; Huerta-Fontela et al., 2008). For instance, the
concentrations of ketoprofen in the wastewater influents
were recorded to be 2.0 ?0.6 mgL?1in Finland (Lindqvist et al.,
2005), 200 ngL?1in Australia (Al-Rifai et al., 2007), and
300–1360 ngL?1in Spain (Santos et al., 2007), while in the
influents of four WWTPs in Beijing, it could not be detected.
Concerning gemfibrozil, which is used to lower cholesterol
and triglyceride levels in the blood, the contamination level
found in the present study was 24–140 ngL?1, even 1 or 2 order
of magnitude lower than those in the USA (4770 ngL?1,
Vanderford and Snyder, 2006) and Canada (418 ngL?1,
Table 2 – Instrumental quantification limit (IQL), limit of quantification (LOQ), absolute recovery (AR), relative recovery (RR)
and matrix suppression of target compounds.
LOQ (ngL?1) AR (n ¼6, %)
Compounds IQL (pg)
RR (n ¼6, %)
Matrix effect (%)
a Value in the brackets refers to the deviation of the recovery.
water research 44 (2010) 417–426
Lishman et al., 2006). The low levels of target pharmaceuticals
were probably due to the lower per capita consumption in
China than in the countries with higher socioeconomic
statuses, where medical care is more prevalent (Thomas and
Foster, 2005). The per capita consumption rate of gemfibrozil
in China is estimated to be 0.036 mgperson?1d?1(Table 3),
lower than those in Germany (0.2 mgperson?1d?1, Ternes,
1998) and Canada (0.2 mgperson?1d?1, Lishman et al., 2006).
Since the levels of target pharmaceuticals were somewhat
different from those of European and North American
countries, we theoretically calculate the concentration of
pharmaceuticals in the wastewater influent by the following
equation (Lindqvist et al., 2005; Nakada et al., 2006)
Cpred¼T ? e% ? I ? 1012
365 ? P ? Q
where Cpredis the predicted concentration of the pharma-
ceutical in wastewater influent (ngL?1); T is the total produc-
tion of a pharmaceutical both for human and animal use in
China per year (tonyear?1), P is the population of China, e% is
the amount of the pharmaceutical excreted unchanged, I is
the number of inhabitants served and Q is the influent flow
(m3d?1). The predicted concentrations of gemfibrozil, diclo-
fenac, indometacin, ketoprofen, carbamazepine, and sulpir-
ide were comparable to those measured in the influents
(Table 3). Much lower measured concentration than predicted
concentration of chloramphenicol was probably because it
had been forbidden for use in food and aquaculture in China
since 2005, and the available data about the production of
pharmaceuticals were based on the year of 2004. It should be
noticed that since there is no available data on the total
consumption of any pharmaceutical, we used figures for total
production of individual pharmaceutical instead. Therefore,
the differences between the amounts actually produced and
applied as well as the amount used in human and veterinary
medicine could not be distinguished, which might result in
overestimation of the theoretical concentration. Neverthe-
less, the comparability between the predicted concentrations
and measured concentrations illustrates the overall reason-
ability of the approach.
Table 3 – Outputs, per capita consumption, predicted concentrations (PECs) and measured concentrations (MECs) of some
pharmaceuticals in the wastewater influents of WWTPs investigated.
a From CMEIN (2005).
b From Bolton and Null (1981), Ternes (1998), Khan and Ongerth (2004), Niwa et al. (2005), Nakada et al. (2006), Jjemba (2006).
c n.d. ¼Not detected.
Concentration (ng L-1)
Pharmaceuticals & Consumer Products
Concentration (ng L-1)
Pharmaceuticals & Consumer Products
Fig. 2 – Concentrations of target pharmaceuticals in
wastewater influents (a) and secondary effluents (b) of four
WWTPs in Beijing.
water research 44 (2010) 417–426
3.2. Secondary effluent
Similarto the influentsamples,ketoprofen wasbelowthe LOQ
in all the secondary effluent samples. Nalidixic acid and
chloramphenicol were detected only in one WWTP, with the
concentration of 8.1 and 19 ngL?1, respectively. The mean
concentrations of the other 12 compounds ranged from 5 to
200 ngL?1(Fig. 2). Diclofenac, N,N-diethyl-meta-toluamide,
concentrations in the secondary effluents. Carbamazepine
and metoprolol followed, with the concentrations ranging
from 69 to 120 ngL?1, and 60 to 108 ngL?1, respectively. Other
compounds, such as caffeine, gemfibrozil and mefenamic
acid, occurred at the lowest levels. Despite of a wide variation
of trimethoprim from different WWTPs, the composition
profiles of target pharmaceuticals in secondary effluents from
the four WWTPs were quite similar (Fig. 3).
The concentration levels of most pharmaceuticals and
consumer products detected in the secondary effluent were
also lower than those reported in the Europe. They were over
100 ngL?1, in some cases even up to 500 ngL?1in the waste-
water effluents of the European countries (Santos et al., 2007;
Gomez et al., 2007; Ternes, 1998; Vieno et al., 2007b). While in
the present study, 10 out of 15 compounds were less than
100 ngL?1, and none of them exceeded 400 ngL?1in any
effluent samples (Fig. 2). Our results were in agreement with
thosein Japan(Nakada et al., 2006), Korea(Kimet al., 2007) and
some other cities of China (Xu et al., 2007; Gulkowska et al.,
2008; Chen et al., 2008). For instance, the concentrations of
chloramphenicol in the effluents of 4 WWTPs in Guangzhou
were <LOQ-17 ngL?1(Xu et al., 2007), while their concentra-
tions in our study were <LOQ-19 ngL?1. Carbamazepine was
detected in the wastewater effluents of Korea and Taiwan,
with the concentration of 73–729 and 290–960 ngL?1, respec-
tively (Chen et al., 2008; Kim et al., 2007), which were
comparable with those in the effluents of Beijing.
The environmental risk caused by these pharmaceuticals
can be calculated by dividing the measured environmental
concentration (MEC) by the predicted no-effect concentration
(PNEC) of individual compound (European Environment
Agency, 1998). Thanks to the low levels of pharmaceuticals
studied, the risk quotients for most compounds were below 1
in the wastewater effluents (Table 4). However, for diclofenac,
the dominant contributor in the wastewater effluent, the risk
quotient was higher than 1, implicating a risk to the aquatic
3.3. Removal efficiency of conventional treatment
The removal efficiency during the primary treatment was low,
indicating no significant adsorption of target compounds to
the particles removed in this stage (Fig. 4). Most of the phar-
maceuticals and consumer products have log Kowvalues of
less than 3.0, so they are not expected to adsorb significantly
to the particles. Other pharmaceuticals with higher Kow
values, such as gemfibrozil, have much lower pKavalues than
the pH of wastewater. Therefore, they are dissociated and
expected to be >98% in the aqueous phase (Thomas and
Foster, 2005), and not bound to the particles.
During the secondary treatment, the average removal rate
for different compounds ranged from ?12% to 100% (Fig. 4).
Caffeine, bezafibrate, trimethoprim and DEET were effectively
removed, with the average efficiency of 100? 0%, 88?12%,
76? 24% and 69?21%, respectively. These results were
comparable with those found in the previous studies (Ternes,
1998; Okuda et al., 2008; Thomas and Foster, 2005; Castiglioni
et al., 2006). Caffeine was proved to be readily biodegradable
(Okuda et al., 2008; Thomas and Foster, 2005; Huerta-Fontela
Table 4 – Measured concentrations in the effluent
samples, predicted no-effect concentrations and risk
quotients (MEC/PNEC) of target compounds.
a From Santos et al. (2007), Lindqvist et al. (2005), Grung et al.
(2008), Ferrari et al. (2003), Huschek et al. (2004).
Removal efficiency (%)
Fig. 4 – Removal efficiencies of target pharmaceuticals
during the conventional treatment.
Pharmaceutical Compostion (%)
Fig. 3 – Composition profiles of target pharmaceuticals in
secondary effluent samples from four WWTPs in Beijing.
water research 44 (2010) 417–426
et al., 2008; Gomez et al., 2007). The removal rate of bezafibrate
was found to be 87% in six Italian WWTPs in summer time
(Castiglioni et al., 2006), very similar to that observed in our
study. The concentrations of DEET were decreased by more
than 80% during the biological treatment in the WWTP of
Japan (Okuda et al., 2008), slightly better than our results. A
second group of pharmaceuticals, including three anti-
inflammatory drugs, clofibric acid, gembrozil, metoprolol and
sulpiride, had lower removal rates with large variation in
different WWTPs studied. For instance, 28–53% of diclofenac,
a representative of the anti-inflammatorydrugs,was removed
by secondary treatment in the WWTPs, which was between
26% in Finland (Lindqvist et al., 2005) and 69% in Germany
(Ternes, 1998). The elimination of these compounds may be
highly dependent on the configurations and operation condi-
tions of individual WWTP as well as wastewater characteris-
tics, and thus no definitive conclusion could be reached.
Higher load of carbamazepine was found in the secondary
effluent than in the primary effluent, indicating negative
removal efficiency during the secondary treatment. Some
carbamazepine was found to be excreted as the form of
conjugates (Vieno et al., 2007b), which was biodegraded to
carbamazepine by enzymatic processes during the secondary
treatment, resulting in additional amounts of carbamazepine
in the secondary effluent. However, as the calculations of all
not sampled with a hydraulic lag in the present study, some
error might be brought in due to diurnal variation of the
concentration. Therefore, the present study only provided
a snapshot of the removal of pharmaceuticals and consumer
products in the WWTPs of Beijing. To better illustrate that,
24-h composite samples that are lagged by HRT should be
collected and analyzed in further studies.
It has been reported that high HRT (>12 h) and SRT (>10 d)
may contribute to an increased removal rate of pharmaceu-
ticals (Jones et al., 2007; Vieno et al., 2007b). In the present
study, the WWTP C, in which the HRT was higher than the
others, was the best in removing these compounds, due to
increased contact time of target compounds and the micro-
organisms. On the other hand, the different SRTs did not have
significant effects on the removal efficiency, probably because
the SRTs in all the four WWTPs were relatively high (>10 d),
and without large differences. In addition, it is noteworthy
that the WWTP C employed oxidation ditches, which showed
better removal of natural estrogens and estrogenic activity
than A/O (Hashimoto et al., 2007). It also could be the reason
for the higher removal efficiencies in the WWTP C. Further
investigation for different types of WWTPs is necessary to
confirm the results mentioned above.
3.4. Removal efficiency in advanced treatment processes
The removal efficiencies of the pharmaceuticals during theSF,
UF/ozonation, as well as MF/RO treatment in three corre-
sponding WWTPs are listed in Table 5.
Generally, sand filtration was not effective for these
compounds. Only trimethoprim, DEET and gemfibrozil were
removed slightly during this treatment process. It should be
noticed that these compounds were efficiently removed in the
secondary treatment, indicating that the biodegradation on
the biofilm present on the sand particle, rather than the
removal with particles, may be the main reason for their
elimination (Gobel et al., 2007).
The results showed that ozonation is effective in removing
most of the target compounds, probably due to the operation
conditions employed in WWTP A (ozone dosage: 5 mgL?1,
contact time: 15 min). Carbamazepine, diclofenac, indo-
methacin, sulpiride and trimethoprim were significantly
eliminated, with the removal rates of above 95%. The double
bond in the azepine ring of carbamazepine and pyrrole ring of
indomethacin, and the non-protonated amine of diclofenac
and trimethoprim were susceptible to ozone attack (Vieno
et al., 2007a; Nakada et al., 2007; Westerhoff et al., 2005). The
removal efficiencies of DEET and metoprolol were modest.
The amide group, which is not reactive with ozone, could be
Low removal efficiencies were found for bezafibrate, clofibric
acid, as well as caffeine. Only 14% of bezafibrate disappeared
in the ozone process, consistent with its low rate constants
with ozone (590?50 M?1S?1, Huber et al., 2003). The reaction
site of bezafibrate is the R-oxysubstituent (–O–C(CH3)2COOH)
on one of the aromatic rings. However, as the pKaof bezafi-
brate is 3.6, the R-oxysubstituent cannot be deprotonated and
consequently theoverallrateconstant atpH >4 ismuchlower
(Huber et al., 2003). It should be noticed that during the
ozonation, most of the pharmaceuticals were not mineralized
but transformed to the oxidation products. For instance, three
oxidation products containing quinazoline-based functional
groups were identified during the ozonation of CBZ (Mcdowell
et al., 2005).
The good performance of ozonation in the present study
was consistent with Ternes et al. (2003), Huber et al. (2005)
and Okuda et al. (2008). When 5 mgL?1ozone was applied to
the effluent of a municipal WWTP in Germany (contact time:
18 min), target compounds, such as trimethoprim, carba-
mazepine, indomethacin, clofibric acid, were removed by
more than 50% (Ternes et al., 2003). Huber et al. (2005)
conducted a pilot study on the oxidation of pharmaceuticals
during ozonation of conventional activated sludge (CAS) and
membrane bio-reactor (MBR) effluents with various ozone
dosages, and found that macrolide and sulfonamide antibi-
otics, estrogens, and acidic pharmaceuticals diclofenac,
Table 5 – Removal efficiencies (%) of target
pharmaceuticals and consumer products by advanced
treatment processes in studied WWTPs.
CompoundWWTP AWWTP BWWTP D
UF OzoneSF MF/RO
water research 44 (2010) 417–426
naproxen and indomethacin were oxidized by more than
90–99% for ozone doses ?2 mgL?1in all effluents.
The elimination by ultrafiltration in the WWTP A was low
for all the investigated compounds. The molecular weight cut-
off (MWCO) of UF membranes was much higher than 1000 Da,
thus UF membranes showed poor retention of all the inves-
tigated pharmaceuticals, of which the molecular weight are
less than 400 Da. The removal of individual target compound
was less than 50%, and might be due to the adsorption onto
the membrane. It has been also demonstrated that UF
membrane typically had less than 40% retention of 27 PPCPs,
and the mass balances calculated based on the concentration
of each compound in feed, permeate and retentate showed
the observed retention was significantly governed by adsorp-
tion (Yoon et al., 2006).
In contrast, MF/RO employed in WWTP D was very effec-
tive. In the effluent of MF/RO, all the target compounds except
caffeine were not detected. Generally, one or combination of
three basic mechanisms could be involved during the rejec-
tion of solute by NF/RO membrane: steric effect, charge
exclusion and adsorption (Radjenovic et al., 2008). For most
pharmaceuticals, the rejections were considered to be domi-
nated by steric interaction in ‘‘tight’’ NF or RO membrane
filtration (Nghiem et al., 2005; Radjenovic et al., 2008). As most
investigated compounds have molecular weights about
200–400 Da, smaller than MWCO of RO membrane applied,
excellent rejection of most pharmaceuticals by RO membrane
was observed in this study as well as in previous studies
(Kimura et al., 2004; Al-Rifai et al., 2007; Radjenovic et al.,
2008). Besides, membrane fouling and the presence of organic
matterin the wastewater effluentslikelycontributed to higher
rejections of pharmaceuticals, especially for some hydro-
phobic ionogenic compound (Nghiem and Coleman, 2008;
Comerton et al., 2008).
Nevertheless, the rejections of two compounds, caffeine
and mefenamic acid were slightly lower (i.e. 50–80% and
0–50%, respectively). The concentration of mefenamic acid in
feed wastewaters of MF/RO membrane process was very low,
only a bit higher than its LOQ in the wastewater effluent,
which could be the reason for the low rejection rate. The low
retention of caffeine in the present study was inaccordance
with Drewes et al. (2005). They found that in two full-scale RO
facilities, target EDCs and PPCPs were efficiently rejected to
below detection limit except for caffeine, still detected in the
permeates. The physiochemical properties might explain the
low rejection rate of caffeine. As a representative of hydro-
philic and non-ionic compounds, the rejection driven by
charge exclusion and adsorption is negligible, and steric
exclusion is solely responsible for the retention of caffeine
(Nghiem et al., 2005). However, the molecular weight of
caffeine is 195 Da, smaller than other target compounds, and
might resultin the decreased removal efficiencyduring the RO
membrane filtration process.
Compared to the other two, the WWTPs employing ozone
and RO membrane filtration as advanced treatment were
more efficient in removing pharmaceuticals. For these
WWTPs, the advanced treatment made a significant contri-
bution to the total elimination of most pharmaceuticals
(Fig. 5). Therefore, the utility of efficient advanced treatment
could be considered as a tool to reduce pharmaceuticals in the
municipal wastewater treatment plants. However, the prob-
lems of membrane fouling and further treatment or disposal
of retentate challenge the application of RO membrane
filtration (Van der Bruggen et al., 2008). For ozonation, as most
of the pharmaceuticals could not be mineralized, and oxida-
tion products are formed from parent pharmaceutical
compounds (Mcdowell et al., 2005), more research is required
to identify the oxidation products and their potential toxicity
during the partial oxidation process (Nakada et al., 2007).
Besides, economic feasibility should be evaluated by esti-
matingthe energy consumption
operation costs for both advanced treatment processes (Joss
et al., 2008).
13 out of 15 pharmaceuticals and consumer products from
eight classes were detected at four WWTPs in Beijing, China.
The concentrations of most compounds in the influent and
secondary effluent were lower than those reported in the USA
and Europe, but consistent with the production profile of the
Removal Contribution (%)
Contribution to removal efficiency (%)
Fig. 5 – Contributions of primary, secondary (or
conventional treatment) and tertiary treatment to the total
elimination of selected pharmaceuticals in WWTP A (a)
and WWTP D (b).
water research 44 (2010) 417–426
pharmaceuticals in China. According to the result of risk
assessment for the secondary effluent, only diclofenac might
pose a risk to the aquatic environment. The removal effi-
ciencies by the conventional treatment varied for different
compounds, depending on their chemical structures, physi-
ochemical properties, as well as the specific treatment
processes utilized at each WWTP. Further removal could be
achieved by adopting some advanced treatment processes,
such as ozonation and MF/RO. However, others, such as
sand filtration, showed low efficiency in removing these
compounds from secondary effluent.
This study was supported by the National Science Fund for
Distinguished Young Scholars (No. 50625823).
Supplementary information related to this article can be
found at doi:10.1016/j.watres.2009.07.010.
r e f e r e n c e s
Al-Rifai, J.H., Gabelish, C.L., Schafer, A.I., 2007. Occurrence of
pharmaceutically active and non-steroidal estrogenic
compounds in three different wastewater recycling schemes
in Australia. Chemosphere 69 (5), 803–815.
Bolton, S., Null, G., 1981. Caffeine: psychological effects, use and
abuse. Orthomolecular Psychiatry 10 (3), 202–211.
Castiglioni, S., Bagnati, R., Fanelli, R., Pomati, F., Calamari, D.,
Zuccato, E., 2006. Removal of pharmaceuticals in sewage
treatment plants in Italy. Environmental Science and
Technology 40 (1), 357–363.
Chen, H.C., Wang, P.L., Ding, W.H., 2008. Using liquid
chromatography–ion trap mass spectrometry to determine
pharmaceutical residues in Taiwanese rivers and
wastewaters. Chemosphere 72 (6), 863–869.
China Medicine Economic Information Net (CMEIN), 2005.
Chinese Medical Statistical Yearbook. Products section. China
Medicine Economic Information Net (CMEIN), Beijing, China.
Comerton, A.M., Andrews, R.C., Bagley, D.M., Hao, C.Y., 2008. The
rejection of endocrine disrupting and pharmaceutically active
compounds by NF and RO membranes as a function of
compound and water matrix properties. Journal of Membrane
Science 313 (1–2), 323–335.
Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal
care products in the environment: agents of subtle change?
Environmental Health Perspectives 107 (S6), 907–938.
Drewes, J.E., Bellona, C., Oekekoven, M., Xu, P., Kim, T.U., Amy, G.,
2005. Rejection of wastewater-derived micropollutants in
high-pressure membrane applications leading to indirect
potable reuse. Environmental Progress 24 (4), 400–409.
European Environment Agency, 1998. Environmental Risk
Assessment – Approaches, Experiences and Information
Source. European Environment Agency, London, pp. 68–86.
Ferrari, B., Paxeus, N., Giudice, R.L., Pollio, A., Grric, J., 2003.
Ecotoxicological impact of pharmaceuticals found in treated
wastewaters: study of carbamazepine, clofibric acid, and
diclofenac. Ecotoxicology and Environmental Safety 55 (3),
Gobel, A., Mcardell, C.S., Joss, A., Siegrist, H., Giger, W., 2007. Fate
of sulfonamides, macrolides, and trimethoprim in different
wastewater treatment technologies. Science of the Total
Environment 372 (2–3), 361–371.
Gomez, M.J., Bueno, M.J.M., Lacorte, S., Fernandez-Alba, A.R.,
Aguera, A., 2007. Pilot survey monitoring pharmaceuticals and
related compounds in a sewage treatment plant located on the
Mediterranean coast. Chemosphere 66 (6), 993–1002.
Grung, M., Kallqvist, T., Sakshaug, S., Skurtveit, S., Thomas, K.V.,
2008. Environmental assessment of Norwegian priority
pharmaceuticals based on the EMEA guideline. Ecotoxicology
and Environmental Safety 71 (2), 328–340.
Gulkowska, A., Leung, H.W., So, M.K., Taniyasu, S., Yamashita, N.,
Yeung, L.W.Y., Richardson, B.J., Lei, A.P., Giesy, J.P., Lam, P.K.S.,
2008. Removal of antibiotics from wastewater by sewage
treatment facilities in Hongkong and Shenzhen, China. Water
Research 42 (1–2), 395–403.
Hashimoto, T., Onda, K., Nakamura, Y., Tada, K., Miya, A.,
Murakami, T., 2007. Comparison of natural estrogen removal
efficiency in the conventional activated sludge process and
the oxidation ditch process. Water Research 41 (10),
Huber, M.M., Canonica, S., Park, G.Y., Gunten, U.V., 2003.
Oxidation of pharmaceuticals during ozonation and advanced
oxidation processes. Environmental Science and Technology
37 (5), 1016–1024.
Huber, M.M., Gobel, A., Joss, A., Hermann, N., Loffler, D.,
Mcardell, C.S., Ried, A., Siegrist, H., Ternes, T.A., Gunten, U.V.,
2005. Oxidation of pharmaceuticals during ozonation of
municipal wastewater effluents: a pilot study. Environmental
Science and Technology 39 (11), 4290–4299.
Huerta-Fontela, M., Galceran, M.T., Martin-Alonso, J., Ventura, F.,
2008. Occurrence of psychoactive stimulatory drugs in
wastewaters in north-eastern Spain. Science of the Total
Environment 397 (1–3), 31–40.
Huschek, G., Hansen, P.D., Maurer, H.H., Krengel, D., Kayser, A.,
2004. Environmental risk assessment of medicinal products
for human use according to European Commission
recommendations. Environmental Toxicology 19 (3), 226–240.
Jjemba, P.K., 2006. Excretion and ecotoxicity of pharmaceutical
and personal care products in the environment. Ecotoxicology
and Environmental Safety 63 (1), 113–130.
Jones, O.A.H., Voulvoulis, N., Lester, J.N., 2005. Human
pharmaceuticals in wastewater treatment processes. Critical
Reviews in Environmental Science and Technology 35 (4),
Jones, O.A.H., Voulvoulis, N., Lester, J.N., 2007. The occurrence
and removal of selected pharmaceutical compounds in
a sewage treatment works utilising activated sludge
treatment. Environmental Pollution 145 (3), 738–744.
Joss, A., Siegrist, H., Ternes, T.A., 2008. Are we about to upgrade
wastewater treatment for removing organic micropollutants.
Water Science and Technology 57 (2), 251–255.
Kasprzyk-Hordern, B., Dinsdale, R.M., Guwy, A.J., 2009. The
removal of pharmaceuticals, personal care products,
endocrine disruptors and illicit drugs during wastewater
treatment and its impact on the quality of receiving waters.
Water Research 43 (2), 363–380.
Khan, S.J., Wintgens, T., Sherman, P., Zaricky, J., Schafer, A.I.,
2004. Removal of hormones and pharmaceuticals in the
advanced water recycling demonstration plant in Queensland,
Australia. Water Science and Technology 50 (5), 15–22.
Khan, S.J., Ongerth, J.E., 2004. Modelling of pharmaceutical
residues in Australian sewage by quantities of use and
fugacity calculations. Chemosphere 54 (3), 355–367.
water research 44 (2010) 417–426
Kim, S.D., Cho, J., Kim, I.S., Vanderford, B.J., Snyder, S.A., 2007.
Occurrence and removal of pharmaceuticals and endocrine
disruptors in South Korean surface, drinking, and waste
waters. Water Research 41 (5), 1013–1021.
Kimura, K., Toshima, S., Amy, G., Watanabe, Y., 2004. Rejection of
neutral endocrine disrupting compounds (EDCs) and
pharmaceutical active compounds (PhACs) by RO membranes.
Journal of Membrane Science 245 (1–2), 71–78.
Lishman, L., Smyth, S.A., Sarafin, K., Kleywegt, S., Toito, J.,
Peart, T., Lee, B., Servos, M., Beland, M., Seto, P., 2006.
Occurrence and reductions of pharmaceuticals and personal
care products and estrogens by municipal wastewater
treatment plants in Ontario, Canada. Science of the Total
Environment 367 (2–3), 544–558.
Lindqvist, N., Tuhkanen, T., Kronberg, L., 2005. Occurrence of
acidic pharmaceuticals in raw and treated sewages and in
receiving waters. Water Research 39 (11), 2219–2228.
Mcdowell, D.C., Huber, M.M., Wagner, M., Gunten, U.V.,
Ternes, T.A., 2005. Ozonation of carbamazepine in drinking
water: identification and kinetic study of major oxidation
products. Environmental Science and Technology 39 (20),
Nakada, N., Tanishima, T., Shinohara, H., Kiri, K., Takada, H.,
2006. Pharmaceutical chemicals and endocrine disrupters in
municipal wastewater in Tokyo and their removal during
activated sludge treatment. Water Research 40 (17), 3297–3303.
Nakada, N., Shinohara, H., Murata, A., Kiri, K., Managaki, S.,
Sato, N., Takada, H., 2007. Removal of selected
pharmaceuticals and personal care products (PPCPs) and
endocrine-disrupting chemicals (EDCs) during sand filtration
and ozonation at a municipal sewage treatment plant. Water
Research 41 (19), 4372–4382.
Nghiem, L.D., Schafer, A.I., Elimelech, M., 2005. Pharmaceutical
retention mechanisms by nanofiltration membrane.
Environmental Science and Technology 39 (19), 7698–7705.
Nghiem, L.D., Coleman, P.J., 2008. NF/RO filtration of the
hydrophobic ionogenic compound triclosan: transport
mechanisms and the influence of membrane fouling.
Separation and Purification Technology 62 (3), 709–716.
Niwa, T., Inoue, S., Shiraga, T., Takagi, A., 2005. No inhibition of
cytochrome P450 activities in human liver microsomes by
sulpiride, an antipsychotic drug. Biological and
Pharmaceutical Bulletin 28 (1), 188–191.
Okuda, T., Kobayashi, Y., Nagao, R., Yamashita, N., Tanaka, H.,
Tanaka, S., Fujii, S., Konishi, C., Houwa, I., 2008. Removal
efficiency of 66 pharmaceuticals during wastewater treatment
process in Japan. Water Science and Technology 57 (1), 65–71.
Paxeus, N., 2004. Removal of selected non-steroidal anti-
inflammatory drugs (NSAIDs), gemfibrozil, carbamazepine,
beta-blockers, trimethoprim and triclosan in conventional
wastewater treatment plants in five EU countries and their
discharge to the aquatic environment. Water Science and
Technology 50 (5), 253–260.
and chronic toxicity of eleven pharmaceuticals (and their
solvents) found in wastewater effluent on the cnidarian, Hydra
attenuata. Science of the Total Environment 389 (2–3), 306–314.
Radjenovic, J., Petrovic, M., Ventura, F., Barcelo, D., 2008. Rejection
of pharmaceuticals in nanofiltration and reverse osmosis
membrane drinking water treatment. Water Research 42 (14),
Santos, J.L., Aparicio, I., Alonso, E., 2007. Occurrence and risk
assessment of pharmaceutically active compounds in
wastewater treatment plants. A case study: Seville city
(Spain). Environment International 33 (4), 596–601.
Sui, Q., Huang, J., Deng, S. B., Yu, G. Rapid determination of
pharmaceuticals from multiple therapeutic classes in
wastewater by solid-phase extraction and ultra-performance
liquid chromatography tandem mass spectrometry. Chinese
Science Bulletin, in press.
Ternes, T.A., 1998. Occurrence of drugs in German sewage
treatment plants and rivers. Water Research 32 (11),
Ternes, T.A., Meisenheimer, M., Mcdowell, D., Sacher, F.,
Brauch, H.J., Haist-gulde, B., Preuss, G., Wilme, U.,
Zulei-seibert, N., 2002. Removal of pharmaceuticals during
drinking water treatment. Environmental Science and
Technology 36 (17), 3855–3863.
Ternes, T.A., Stuber, J., Herrmann, N., Mcdowell, D., Ried, A.,
Kampmann, M., Teiser, B., 2003. Ozonation: a tool for removal
of pharmaceuticals, contrast media and musk fragrances from
wastewater? Water Research 37 (8), 1976–1982.
Thomas, P.M., Foster, G.D., 2005. Tracking acidic
pharmaceuticals, caffeine, and triclosan through the
wastewater treatment process. Environmental Toxicology and
Chemistry 24 (1), 25–30.
Vanderford, B.J., Snyder, S.A., 2006. Analysis of pharmaceuticals
in water by isotope dilution liquid chromatography/tandem
mass spectrometry. Environmental Science and Technology
40 (23), 7312–7320.
Van der Bruggen, B., Manttari, M., Nystrom, M., 2008.
Drawbacks of applying nanofiltration and how to avoid them:
a review. Separation and Purification Technology 63 (2),
Vieno, N.M., Harkki, H., Tuhkanen, T., Kronberg, L., 2007a.
Occurrence of pharmaceuticals in river water and their
elimination in a pilot-scale drinking water treatment plant.
Environmental Science and Technology 41 (14), 5077–5084.
Vieno, N., Tuhkanen, T., Kronberg, L., 2007b. Elimination of
pharmaceuticals in sewage treatment plants in Finland. Water
Research 41 (5), 1001–1012.
Watkinson, A.J., Murby, E.J., Costanzo, S.D., 2007. Removal of
antibiotics in conventional and advanced wastewater
treatment: implications for environmental discharge and
wastewater recycling. Water Research 41 (18), 4164–4176.
Westerhoff, P., Yoon, Y., Snyder, S., Wert, E., 2005. Fate of
endocrine-disruptor, pharmaceutical, and personal care
product chemicals during simulated drinking water treatment
processes. Environmental Science and Technology 39 (17),
Xu, W.H., Zhang, G., Li, X.D., Zou, S.C., Li, P., Hu, Z.H., Li, J., 2007.
Occurrence and elimination of antibiotics at four sewage
treatment plants in the Pearl River Delta, South China. Water
Research 41 (19), 4526–4534.
Yoon, Y., Westerhoff, P., Snyder, S.A., Wert, E.C., 2006.
Nanofiltration and ultrafiltration of endocrine disrupting
compounds, pharmaceuticals and personal care products.
Journal of Membrane Science 270 (1–2), 88–100.
water research 44 (2010) 417–426