Journal of Hazardous Materials 299 (2015) 241–248

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Effect of colloids on the occurrence, distribution and photolysis of

emerging organic contaminants in wastewaters
Caixia Yan a,b,c , Minghua Nie a,d , Yi Yang b,e,∗ , Junliang Zhou b , Min Liu e ,
Mohammed Baalousha c , Jamie R. Lead c
a
School of Geography and Environment, Key Laboratory of Poyang Lake Wetland and Watershed Research, Ministry of Education, Jiangxi Normal
University, No. 99, Ziyang Road, Nanchang 330022, China
b
State Key Laboratory of Estuarine and Coastal Research, East China Normal University, 3663 North Zhongshan Road, Shanghai 200062, China
c
(SmartState) Center for Environmental Nanoscience and Risk, Arnold School of Public Health, University of South Carolina, 921 Assembly Street, Columbia
29208,USA
d
Department of Environmental Science and Engineering, Fudan University, 220Handan Road, Shanghai 200433, China
e
Key Laboratory of Geographic Information Science of the Ministry of Education, Department of Geosciences, East China Normal University, 3663 North
Zhongshan Road, Shanghai 200062, China

h i g h l i g h t s

• Colloidal fractions in wastewaters were isolated using cross flow ultrafiltration.

• EOCs exhibited a pseudo - first - order degradation kinetics in all water samples.
• Photolysis of EOCs in permeate were accelerated, while inhibited in the retentates.
• EOCs with higher degradation rates were detected at low level in natural water.

a r t i c l e i n f o a b s t r a c t

Article history: The effect of colloids on the occurrence, phase distribution and photolysis of twenty-seven emerging
Received 26 February 2015 organic contaminants (EOCs) was studied in domestic and livestock wastewaters (DW and LW), respec-
Received in revised form 2 June 2015 tively. Filtered water (<1 ␮m) was separated into permeate (<1 kDa) and retentate (1 kDa-1 ␮m) by cross
Accepted 9 June 2015
flow ultrafiltration. Results indicated that total concentration of EOCs ranged from 1220 ng L−1 in per-
Available online 18 June 2015
meate of DW to 5065 ng L−1 in retentate of LW. The average EOC fraction associated with colloids was
13.5% and 14.4% in DW and LW. Most of the EOCs exhibited pseudo-first-order degradation kinetics in
Keywords:
all water samples. Control experiments using glass and quartz reactors showed that UV light was more
Emerging organic contaminants (EOCs)
Colloids
effective on the photolysis of most EOCs. The EOCs photolysis in the three fractions of DW and LW could be
Ultraviolet (UV) accelerated or inhibited compared to ultrapure water with the enhancement factor ranging from −0.94
Photolysis to 7.33. The impact of colloids on the photolysis of EOCs depended on the compound and the source of
water. The photolysis of most EOCs in permeates and filtrates was generally accelerated, while inhibited
in the retentates, which could be attributed to the relatively high dissolved organic carbon contents in
retentates.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction released to aquatic systems via wastewater treatment plants

(WWTPs), livestock wastewater, manufacturing, hospital effluents,
Emerging organic contaminants (EOCs) such as pharmaceuti- and urban runoff [1–4]. Recently, EOCs have become a matter of
cals, personal care products, hormones and sterols, surfactants, concern due to their frequent detection in multiple aquatic environ-
plasticizers, etc., are a group of biologically active, and widely ments, such as wastewater [5,6], surface water [7–9], ground water
spread compounds in the aquatic environment. EOCs are being [10,11], and drinking water [12,13], as well as due to their high toxi-
city, persistence, and inefficient removal by traditional wastewater
treatment processes [14–16]. Despite the well-known detrimen-
∗ Corresponding author. tal environmental and human health effects of EOCs [17,18]; they
E-mail address: [email protected] (Y. Yang). are still not regulated and their environmental fate and behav-

http://dx.doi.org/10.1016/j.jhazmat.2015.06.022
0304-3894/© 2015 Elsevier B.V. All rights reserved.
242 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248

ior is poorly understood. Therefore, it is essential to quantify the aline (SQ); (3) macrolides (MLs) including erythromycin (ETM);
concentrations of EOCs, determine their distribution among water (4) estrogens including estrone (E1), 17␤-estradiol (E2), estriol
components (e.g. truly dissolved vs. colloid-bound), and to under- (E3), 17␣-ethinylestradiol (EE2), and bisphenol A (BPA); (5) lipid
stand the fate and behavior of EOCs in the aquatic environment. regulators (LRs) including gemfibrozil (GFB); (6) non-steroidal
Colloids are heterogeneous mixture of particles in the size range anti-inflammatory drugs (NSAIDs) including indomethacin (IDM),
from 1 kDa to 1 ␮m with different chemical compositions, including paracetamol (PRCT), diclofenac (DCF), and ibuprofen (BRF); (7) anti-
iron oxyhydroxides, aluminosilicates, humic substances, viruses, histaminics (AHTs) including omeprazole (OMR), and cimetidine
bacteria, etc. [19]. Recently, colloids have been shown to play an (CMT); (8) antihypertensive (AHPs) including nimodipine (NMD),
important role in determining the distribution, fate, and behavior and nifedipine (NFD); and (9) anticonvulsants (ACs) including car-
of EOCs in aquatic systems, which is generally attributed to the bamazepine (CMZ), and diazepam (DZP). In addition, these EOCs
unique properties (e.g. concentrated organic carbon content and showed an acceptable recovery through a cross-flow ultrafiltration
high surface area) and high affinity of colloid towards contami- system (recoveries: 64.24–117.82%; SD < 5.38%), suggesting that
nants [20–22]. For instance, we have demonstrated that colloids cross flow ultrafiltration was a reliable technique to study these
act as a sink for EOCs both in river and estuarial water [23,24]. In 27 EOCs [25]. The physicochemical properties and main applica-
addition, the fate of EOCs in estuarial water has been shown to be tion of these 27 compounds are presented in Table S1. CAP-d5 ,
controlled by the physicochemical properties of colloids (e.g. size, SMX-d4 , RTM-d7 , BPA-d16 , E2-d2 , DCF-d4 , CMZ-d10 , CMT-d3 , and
zeta-potential, and fluorescence intensity) and the estuary salinity NFD-d6 were used as internal standards. All compounds including
[25]. However, the effect of natural colloids on the degradation of internal standards were purchased from Dr. Ehrenstorfer (GmbH,
EOCs is still poorly understood. Germany). All solvents used in the study were of HPLC grade.
Among all degradation processes of EOCs, photodegradation is
recognized as an important and efficient elimination pathway and 2.2. Sample collection
could affect the environmental persistence of EOCs in surface water
and WWTPs (especially for pharmaceuticals) [26,27]. Once irradi- Water samples were collected in April, 2013 from (1) a domestic
ated, some EOCs undergo direct photolysis due to the absorbed wastewater (DW) from the final effluent of an urban WWTP, which
photons, or indirect photolysis promoted by reactive species gener- is discharged into Huangpu River (one of the tributaries of Yangtze
ated by natural photo sensitizers such as dissolved organic matter River), and (2) a livestock wastewater (LW) from the upstream of
(DOM), Cl− , NO3 − and Fe3+ in natural waters [27,28]. In addi- Huangpu River impacted by concentrated animal feeding opera-
tion, these aquatic constituents can inhibit the photolysis of EOCs. tions (CAFOs) in Shanghai, China (Fig. S1). Water samples were
For example, chloride was observed to accelerate the photolysis collected in 50 L stainless steel barrels (pre-cleaned sequentially
of phenicol antibiotics under UV–vis irradiation ( > 200 nm), yet by acetone, deionized water, and Milli-Q water). Once transported
have no effect on the acceleration of the photolysis of phenicol to the laboratory, the water samples were filtered immediately
antibiotics under simulated sunlight ( > 290 nm) [29]; whereas, through 1 ␮m glass fiber filters (PAUL, USA), which were held in
chloride inhibits the photolysis of 17␣-ethynylestradiol under a 293 mm stainless steel sanitary filter holder (Handwheel Wrench
UV–vis ( = 254 nm) [30]. Likewise, as the important components of 316Holder, Millipore).
colloids [31], humic acid (HA) enhances the photolysis of estriol as a Colloids, solid material within the size range of 1 kDa to 1 ␮m,
photosensitizer under weak irradiation (e.g. sunlight), but inhibits were isolated from filtrates (<1 ␮m) by cross-flow ultrafiltration
the photolysis of estriol under the high incident light intensity (e.g. (Pellicon System, Millipore), using a Millipore 1 kDa regener-
8-fold sunlight) due to its light screening and free radical quench- ated cellulose Pellicon 2 PLAC ultrafiltration membrane (0.5 m2 )
ing effects [32]. Nevertheless, little attention has been paid to the according to the method reported by Maskaoui and Zhou [36]. A
photolysis of multiple EOCs under sunlight in the presence of natu- concentration factor (cf) (ratio of the initial feedwater volume to
ral colloids. Furthermore, most previous research on the photolysis the retentate volume) of 10 was chosen due to the minimized com-
procedures was conducted with unrealistically high initial EOCs plex sorptive losses during the ultrafiltration with cf between 10
concentrations (at the mg L −1 level), compared to environmental and 15 [37]. The cross-flow ultrafiltration was performed in a sam-
concentrations (e.g. ␮g L−1 level) [33], which may have a significant pling mode. The retentate flow (1 kDa-1 ␮m) was directed back to
impact on the degradation of EOCs. The photodegradation rate of the feed container as the colloidal phase, while the permeate flow
some EOCs decreases with increased initial EOCs concentrations (<1 kDa) was collected separately as the truly dissolved phase. The
[34,35]. feed pressure was kept at 20 psi, and a permeate flow rate was set
The objectives of this study were to: (1) investigate the occur- to 30–40 mL min−1 . The ultrafiltration system was first conditioned
rence, and phase distribution of 27 EOCs in both domestic and with re-circulating 0.1 M NaOH solution for 1 h. Before loading the
livestock wastewater; (2) determine the effect of ultraviolet (UV) sample, the system was thoroughly rinsed with 10 L Milli-Q water
light on the direct photolysis behavior of 27 EOCs at environ- and 2 L feedwater (filtrates).
mentally relevant, low initial concentration (␮g L−1 level); and (3)
determine the impact of colloids on the photolysis of EOCs. 2.3. Sample treatment and analysis

Samples were processed immediately after filtration and ana-

2. Materials and methods lyzed in triplicate following the method described in [25]. Briefly,
aliquots of water samples (retentates: 0.2 L; filtrates/permeates:
2.1. Chemicals and standards 1 L) were extracted by solid phase extraction (SPE) using cartridges
(Oasis HLB, 200 mg, 6 mL, Waters) at a flow rate of 5–10 mL min−1 .
Twenty seven EOCs were selected for this study due to their All samples were spiked with 20 ng internal standards before SPE
relatively high consumption and occurrence in the environment. for quality control. Prior to extraction, the SPE cartridges were pre-
They were grouped into 9 classes as follows: (1) chlorampheni- conditioned by 6 mL of methanol, followed by 10 mL of ultrapure
cols (CPs) including chloramphenicol (CAP), thiamphenicol (TAP), water. Once extracted, EOCs were eluted using 10 mL methanol,
and florfenicol (FF); (2) sulfonamides (SAs) including sulfadiazine and then concentrated to 0.5 mL by evaporation under nitrogen.
(SD), sulfapyridine (SP), sulfamethoxazole (SMX), sulfathiazole About 0.5 mL ultrapure water was added to each vial to reach the
(ST), sulfamerazine (SM), sulfamethazine (SMT), and sulfaquinox- final volume of 1.0 mL before analysis of EOCs.
 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248 243

The target EOCs were quantified by ultra-performance liq- In this study, the pseudo-first-order model was introduced to
uid chromatography -tandem mass spectrometry (UPLC-MS/MS) present the degradation kinetics of EOCs as following:
in multiple reaction monitoring (MRM) mode according to the
ln(C/C 0 ) = −kt (4)
method described in [25]. The details of the liquid chromatogra-
phy and mass spectrometry were described in the Supplementary where C0 and C are the EOC concentrations at time 0 and t, respec-
Materials and Table S2. The quality control procedures were fol- tively; k is pseudo-first-order rate constant [39].
lowed throughout the study, including the limit of quantification
(LOQ), recovery experiments, and the analysis of procedural blanks. 3. Results and discussions
The details of the quality control were described in Supplementary
Material. 3.1. EOC concentrations in wastewaters

2.4. Photolysis experiments The EOC concentrations in the filtrate, permeate and retentate of
DW and LW are presented in Fig. 1. Only three of the 27 EOCs (SM,
The photolysis of EOCs was performed in triplicate in 30 mL E2, and OMR) were not detected in any water sample. The EOC
quartz reactors under simulated solar irradiation. The photodegra- concentrations were generally higher in LW compared with DW.
dation kinetics were evaluated using the different water fractions For DW, the total concentrations (sum of all analyzed EOCs) were
(filtrate, permeate, and retentate) to assess the effect of natural 1518 ng L−1 in filtrate, 1220 ng L−1 in permeate and 2419 ng L−1 in
colloids on the photolysis of EOCs in DW and LW wastewaters. retentate. While for LW, the total concentrations were 2637 ng L−1
All water samples were equilibrated to room temperature before in filtrate, 2447 ng L−1 in permeate and 5065 ng L−1 in retentate.
use. The system was equipped with a 150 W Xenon lamp (>250 nm, NSAIDs and SAs were observed to be the dominant contaminants
German Osram) placed 20 cm above the reaction solution. An air- for all three water fractions of DW and LW, respectively (e.g. in the
cooling system was used to maintain the temperature at 20 ± 2 ◦ C. filtrate sample, NSAIDs in DW: 35.27%; SAs in LW: 73.08%), which
Before irradiation, 200 ␮g L−1 of the target EOC compounds was could be attributed to the high contribution from IDM and SP as
spiked into each water sample. This level was much higher than the highest concentration in DW and LW. Estrogens were found to
the background of the original samples (see details in the following be the second most dominant compounds in both samples (e.g. in
discussions), while much lower than the previous studies (mg L−1 retentate samples, estrogens in DW and LW: 23.42% and 14.84%). In
level) [32]. Aliquots (0.5 mL) were collected at periodic intervals the present study, the total concentrations of EOCs in wastewaters
of 0, 30, 60, 120, 200, 300, and 480 min during the irradiation, (filtrates) were 4–75 times higher than those in Yangtze Estu-
synchronously spiked with methanol (Vwater :Vmethanol = 1:1) and ary (35.4–413.9 ng L−1 ) [25]. A number of studies confirmed that
20 ␮g L−1 internal standards. Subsequently, all samples were ana- the waste streams and effluents from WWTPs and livestock are
lyzed by UPLC-MS/MS as described above. Control experiments the major sources of EOCs with elevated concentrations [40–43].
with quartz reactor in dark (covered with foil) were performed Therefore, considering the sampling location (Fig. S1), the wastew-
identically to assess the loss due to the hydrolysis or adsorption to aters investigated in this study (WWTPs and CAFOs) might be a
the system. Another control experiment using glass reactors was significant input of EOCs to the Yangtze River.
conducted in parallel to assess the effect of UV light on the photol- The EOC concentrations in colloids are shown in Fig. 2 and Table
ysis of EOCs in ultrapure water. The glass vessels used in the study S3. Cc for the detected EOCs ranged from 0.32 ng L−1 to 25.87 ng L−1
absorb light below 300 nm, thus allowing a measure of the impor- in DW and from 0.03 ng L−1 to 147.5 ng L−1 in LW, respectively. Due
tance of the more energetic sub-300 nm radiation in the photolysis to the relative high concentration of NSAIDs for DW and SAs for LW
test [38]. in filtrates, a similar trend was found in Cc . In addition, the con-
centrations of estrogen in colloids were also high both in DW and
LW. In order to evaluate the contribution of each EOC group in col-
2.5. Colloid characterization loids, the contribution of EOCs was calculated including colloidal
and dissolved phases. The results revealed that the average level of
The concentration of dissolved organic carbon (DOC) in filtrates, the detected EOCs associated with colloids was 13.5% and 14.4% in
retentates and permeates were analyzed using liquiTOC II (Elemen- DW and LW, respectively (Fig. 2 and Table S3). Moreover, a higher
tar, Germany). contribution was observed in estrogens compared to other groups
of EOCs (21.2% for DW; 41.3% for LW) with the highest value of
2.6. EOCs in colloids and colloids/dissolved phase partition 64.5% for BPA in LW, which was consistent with our previous study
coefficient of EOCs [25]. Due to the high octanol-water partition coefficient (logKow ) of
estrogens, it was assumed that hydrophobicity might be an impor-
EOC concentration in colloids (Cc , ng L−1 ) was calculated using tant factor controlling the interaction between colloids and EOCs
the following equation: in the aquatic systems. Therefore, correlation analysis between
logKow value and Cc , and logKow value and colloidal contribution
C c = (C r – C p )/cf (1) was conducted. However, the significant positive correlation was
only found for the Cc values of DW and logKow values (R = 0.672,
where Cr and Cp are the EOC concentrations in retentate and per-
P = 0.001) (Fig. S2). It indicated that hydrophobicity might be an
meate, respectively; cf is the concentration factor.
important factor controlling the behavior of EOCs in the aquatic
To provide further insight of EOC sorption to colloids, the organic
environment, but more statistical data was needed.
carbon normalized partition coefficient (Kcoc , mL g−1 ) of each EOC
was calculated following the equation:
3.2. The colloids/dissolved phases of partition coefficient
C r /C p = 1 + K coc [COC] (2)
The logKcoc values of the EOCs in DW and LW ranged from 4.65 to
where [COC] is the total colloidal organic carbon concentration. 6.49 and from 4.36 to 6.51, respectively (Table 1). These levels were
The EOC contribution of colloids was determined as follows: 1–2 orders of magnitude lower than those investigated in Yangtze
Estuary [23,25]. A similar trend was found in the previous study
Contribution% = (C r – C p )/(C f ∗cf) × 100% (3) that the logKcoc values of estrogen in CAFO wastewater discharges
244 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248

Fig. 1. The detected EOC concentrations in filtrate (<1 ␮m), permeate (<1 kDa) and retentate (1 kDa-1 ␮m) samples.

Fig. 2. EOCs concentration in colloids and the corresponding colloidal contribution in wastewaters.

Table 1 were lower than those in tributaries and mainstream of Huangpu

EOC partition coefficients between colloids and permeates (logKcoc ).
River [24]. The phenomenon of the lower sorption capacity of col-
Compounds DW LW Compounds DW LW loids in wastewater was also corroborated by the calculated mass
CAP 4.81 – EE2 – 5.47 balance that relative low fraction of EOCs was bound with colloids
TAP 5.26 5.49 BPA 6.14 6.51 in the wastewaters compared to Yangtze Estuary [23,25]. It indi-
FF 5.15 5.16 GFB – 5.08 cates that the EOCs sorption to colloids is primarily dependent on
SD 5.93 5.27 IDM 5.42 5.45 the sources of water, and that colloids in natural water have a rela-
SP 5.18 5.31 PRCT 5.5 5.36
tively high sorption capacity compared to those in wastewater. In
SMX 5.44 5.21 DCF 5.6 5.72
ST 5.75 – BRF – 5.51 our non-published data, it showed that the EOCs sorption capacity
SMT 5.65 5.33 CMT 4.65 5.34 of the low molecular weight (LMW) colloids (i.e. 1–5 kDa colloids)
SQ 6.49 4.69 NMD 5.72 – was much higher than that of the high molecular weight (HMW)
ETM 5.63 5.66 NFD 5.44 –
colloids(i.e. 300 kDa-1 ␮m colloids) (Table S4). In addition, most of
E1 5.4 – CMZ 5.34 4.36
E3 – 6.11 DZP 5.8 5.23 the HMW colloids could be removed in WWTPs through aggrega-
tion and degradation [30,44,45], which could result in the dominant
 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248 245

role of LMW colloids in the natural water, and further gave rise to was calculated by comparing the photodegradation rates in fil-
the higher sorption capacity of colloids in the natural water. trates/permeates/retentates and the rates in the control group
(quartz reactor with ultrapure water) as follows:
3.3. Photolysis analysis
Enhancementfactor(EF) = (ki − kq )/kq (5)
3.3.1. Kinetics of direct photolysis
where kq is the pseudo-first-order rate constant of EOCs in con-
The photolysis of EOCs was investigated in water samples spiked
trol group of quartz reactor; ki = kf , kp or kr (which were the
with an environmental relevant concentration of EOCs (200 ␮g L−1 )
pseudo-first-order rate constant of EOCs in filtrates, permeates and
as the initial concentration. No significant hydrolysis or sorption
retentates, respectively). When EF > 0, it indicates an enhancement
occurred during the photolysis of EOCs in the control experi-
effect; EF < 0, indicates an inhibition effect.
ment in dark indicated by the high recovery for all selected EOCs
The EF values of EOCs are shown in Fig. 3. Compared to the
(73.8–115.2%). The photodegradation kinetics of EOCs in both glass
ultrapure water, the photodegradation rates of some EOCs were
(filter UV radiation) and quartz (allows UV radiation) systems are
accelerated with the EF values ranging from 0.02 (CAP in perme-
shown in Fig. S3. In the glass reactors, only 6 compounds under-
ate of DW) to 7.33 (ETM in permeate of DW), while inhibited for
went significant (degradation ratio over 50%) decomposition over
some EOCs with the EF values ranging from −0.03 (CAP in perme-
480 min, including CAP, GFB, DCF, OMR, NFD and NMD, and no
ate of LW) to −0.94 (SMX in retentate of LW). In DW sample, 48.1%,
decomposition was observed for nearly half of EOCs. While in
55.6% and 40.7% of the 27 analytes showed enhancement effect
the quartz system, a significant degradation (degradation ratio
in filtrate, permeate and retentate; while in LW sample 37%, 37%
over 50%) of 20 compounds was observed including CPs, SAs, LRs,
and 18.5% showed enhancement, respectively. More compounds
NSAIDs, AHTs, AHPs, and E1 (excluding SQ and PRCT). Estrogens
showed enhancement effects in DW compared to LW. Moreover,
(excluding E1) and ACs were relatively stable. These results suggest
the higher average photodegradation rates of EOCs were found in all
that: (i) most EOCs undergo degradation by direct photolysis under
the water fractions of DW than those of LW. This might be attributed
UV irradiation in the aquatic environment (e.g. TAP, FF, SMX, ST, E1,
to the different compositions in wastewaters with different sources
CMT etc.); (ii) some EOCs could decompose with non-UV irradia-
and DOC content in DW is lower than in LW. Nevertheless, EF values
tion (e.g. CAP, GFB, DCF, OMR, NFD and NMD); while (iii) some EOCs
of some EOCs were consistently higher than zero (SP, ST, E2, EE2,
are relatively persistent and do not undergo any photolysis in the
and CMT) or less than zero (TAP, FF, SMX, GFB, IDM, and BRF) in all
aquatic system (e.g. E2, E3, EE2, BPA, CMZ and DZP).
three fractions of both DW and LW, indicating the compound spe-
In addition, the degradation kinetics of most EOCs followed a
cific photolysis behavior of EOCs. In general, our results revealed
pseudo-first-order model. The pseudo-first-order rate constant of
that the impact of colloids on the photolysis depended on both the
EOCs in control group of the glass (kg ) and quartz (kq ) reactors are
compound and the water chemistry.
shown in Table S5, and the corresponding half-lives of EOCs cal-
Among all fractions of water sample, the difference of the degra-
culated from the equation are summarized in Table S6. No visible
dation rate of EOCs between filtrates and permeates were relatively
photolysis was observed for k values lower than 0.0001 min−1 (such
small, while the degradation rate of most EOCs was much less in
as TAP, SP, E1, ETM, CMZ, etc. in glass group). The photodegrada-
retentates (Fig. 3, e.g. CAP, SMT, ETM, DCF and OMR, etc.). In the
tion rate of EOCs in quartz reactor (kq ) are 1–125 folds higher than
previous study, similar results were also found showing that the
those in glass reactor (kg ), revealing a major role played by UV -
degradation rates of CAP in a wastewater were lower than that
light on the direct photolysis of EOCs. Moreover, except for NFD
in deionized water, and they were much higher in filtrate and
and OMR, the half-life in the glass reactor (tg ) of other compounds
permeate sample than those in retentate of wastewater samples
were more than the experimental time (480 min), while the half-
[46]. Therefore, in order to further elucidate the effect of colloids
life in the quartz reactor (tq ) were much shorter compared to tg
with different colloidal concentrations on the photolysis of EOCs,
(most tq values lower than 400 min) (Table S6). In this study, some
three photolysis types were summarized in this study, according to
EOCs with high octanol–water partition coefficients (logKow ) such
the k values in ultrapure water, permeates, filtrates and retentates
as estrogens, LRs and SQ showed relatively low photodegradation
(Fig. 4).
rate and lower R2 values for the pseudo-first-order equation, but no
significant correlation was found between the logKow values and kq
values. Notably, no related information (no calculated k and t val- • Type I (full promotion): kq < kr < kf < kp (IA), kq < kr < kp < kf (IB) or
ues) about NFD, CMZ and DZP is shown in Table S5 and Table S6, kq < kp < kf < kr (IC);
because NFD was completely degraded within 30 min in all exper- • Type II (transition): kr < kf < kq < kp (IIA), kr < kq < kf < kp (IIB) or
iments, while CMZ and DZP were persistent in all experiments. kr < kq < kp < kf (IIC);
• Type III (complete inhibition): k < kf < kp < kq (IIIA) or
3.3.2. Effect of colloids on EOCs photolysis kr < kp < kf < kq (IIIB)
The photodegradation kinetics of EOCs in filtrate, permeate and
retentate of DW and LW are shown in Fig. S4 and S5 and the pho- where kf , kp and kr are the pseudo-first-order rate constants of
todegradation rates are presented in Fig. S6. Results revealed that EOCs in filtrates, permeates and retentates. Type I means that the
the photodegradation rates of CAP, ST, E1, DCF, OMR, NMD, and photolysis is fully promoted in all three water fractions compared
NFD were relatively high in these fractions of water. While exclud- to ultrapure water; Type II means that the photodegradation rate
ing DCF, the concentrations of CAP, ST, E1, OMR, NMD and NFD in ultrapure water is higher than in the retentate, while lower than
in LW (Fig. 1) and Yangtze Estuary [25] were very low, indicating in the permeate or filtrate; Type III means that the photolysis is
that the photolysis process could be one of the efficient removal completely inhibited in all three water fractions. The subclasses of
pathways for these EOCs in the aquatic environment. DW was an each type (“A”–“C”) are the possible distribution of the photodegra-
exception probably due to the lack of irradiation in WWTP systems. dation rates of EOCs in different water fractions. Fig. 4 shows the
In all three fractions of DW and LW, most of the EOCs decom- potential and relative degradation rate of EOCs for the different
position also exhibited a pseudo-first-order degradation pattern water fractions to help understand the photolysis types easier. As
(Table S5) and the corresponding half-lives were calculated as Fig. 4 shown, from ultrapure water to retentate sample (the col-
shown in Table S6. In order to further understand the impact of loidal concentration increased), an “ascending stage” followed by a
colloids on the photolysis of EOCs, an enhancement factor (EF) “descending stage” existed in all three photolysis types. The corre-
246 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248

Fig. 3. Enhancement or inhibition effect of filtrates, permeates and retentate compared to the control group (quartz reactor with ultrapure water).

sponding photolysis type of each EOC in DW and LW were classified of water sources on the photolysis of these compounds. Except for
and shown in Table S5. type “IC”, kr values are lower than kp and kf values in all the other
For DW, about 50.0%, 18.2% and 31.8% of the selected EOCs types, which means that colloids generally acted as an inhibitor
showed their photolysis type as Type I, II and III, respectively; while except for these EOCs (e.g. E2, E3, EE2, PRCT and CMT in DW; E2
for LW, the corresponding percentages were 22.7%, 22.7% and 54.5% in LW). Accordingly, the related half-lives of EOCs in each water
(Fig. S7). As Type I is a promotion type while Type III is an inhibi- sample were different from those in the ultrapure water, and gener-
tion one, the degradation rates were accelerated for most EOCs in ally decreased as follows: DW–Per. < DW–Fil. < blank < DW - Ret. < L
DW, and inhibited for nearly half of the EOCs in LW. Furthermore, –Fil. < LW–Per. < LW–Ret.
the photolysis type of some EOCs was classified as Type I and Type According to [19], only low molecular organic acids (e.g. amino
II in DW, and the counterparts were Type II (SMT, ETM and NMD) acid and fulvic acid) and inorganic ions (e.g. Cl− , NO3 − , CO3 2− , Na+ ,
or III (CAP, SD, SM, E1, E3 and PRCT) in LW. In contrast, the pho- Ca2+ , Fe3+ ) were associated with permeates. Many studies have
tolysis types for some EOCs including TAP, FF, SP, E2 and BRF were reported that NO3 − , Fe3+ , Cl− and humic substances could pro-
consistent both in DW and LW, indicating non-significant influence mote the photochemical loss of some EOCs, such as TAP, ETM, E3

Fig. 4. Summarized photolysis types of EOCs in different waters (the k values here only represent the potential and relative degradation rate of EOCs in the different water
fractions).
 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248 247

Table 2 Acknowledgments
Correlations between photodegradation rates and DOC in filtrates, permeates and
retentates (n = 6).
This study was funded by the National Natural Science Foun-
Compounds R P Compounds R P dation of China (41271473; 41171376 and 41130525). Additional
CAP −0.93 0.007** SMT −0.63 0.181 support was provided by the Fundamental Research Funds for the
TAP −0.92 0.009** ETM −0.67 0.147 Central Universities, the Open Foundation of East China Normal
FF −0.92 0.009** E1 −0.71 0.118 University and Open Foundation of State Key Laboratory of Poyang
SD −0.64 0.172 IDM −0.82 0.047*
Lake Wetland and Watershed Research, Ministry of Education
SP −0.78 0.068 DCF −0.90 0.015*
SMX −0.92 0.011* OMR −0.91 0.011* (Jiangxi Normal University). We also acknowledge the Smartstate
ST −0.70 0.119 NMD −0.76 0.079 Center for Environmental Nanoscience and Risk at the University
*
Significant correlation (P < 0.05). of South Carolina. We appreciate Dr. Jeffery Parks for the English
**
Very significant correlation (P < 0.01). revision.

Appendix A. Supplementary data

and p-aminobenzoic acid (PABA) in aqueous solutions under given
irradiation conditions [29,31,47,48].While in filtrates, the higher Supplementary data associated with this article can be found, in
molecular weight materials including proteins, polysaccharides, the online version, at http://dx.doi.org/10.1016/j.jhazmat.2015.06.
humic aggregates, and other metal oxide increased and were fur- 022
ther concentrated in retentates. It has been reported that high
molecular weight HAs were found to be a less effective sensitizer for References
the photolysis of atrazine compared with low molecular HAs, due
[1] D.J. Lapworth, N. Baran, M.E. Stuart, R.S. Ward, Emerging organic
to the relative low aromaticity, oxygen groups, and fluorophores in
contaminants in groundwater: a review of sources, fate and occurrence,
high molecular weight HAs [49]. Previous studies have suggested Environ. Pollut. 163 (2012) 287–303.
that DOM could either accelerate the photolysis of organic con- [2] X.L. Li, W. Zheng, W.R. Kelly, Occurrence and removal of pharmaceutical and
hormone contaminants in rural wastewater treatment lagoons, Sci. Total
taminants due to the produced reactive species such as 3 NOM* ,
1 DO and • OH from DOM [50], or inhibit the photolysis process Environ. 445–446 (2013) 22–28.
2 [3] P. Verlicchi, M. al Aukidy, A. Galletti, M. Petrovic, D. Barceló, Hospital effluent:
because of competitive absorption photons and the hydrophobic investigation of the concentrations and distribution of pharmaceuticals and
binding interaction between pollutants and DOM [51], especially at environmental risk assessment, Sci. Total Environ. 430 (2012) 109–118.
[4] J.L. Liu, M.H. Wong, Pharmaceuticals and personal care products (PPCPs): a
elevated concentration of DOM [52]. Consequently, in the present review on environmental contamination in China, Environ. Int. 59 (2013)
study, the photolysis of most EOCs would be enhanced by low 208–224.
molecular organic acids and inorganic ions in permeate, and this [5] F. Jelic, S.D. Fatone, M. Fabio, F. Petrovic, D. Cecchi Barcelo, Tracing
pharmaceuticals in a municipal plant for integrated wastewater and organic
effect in general decreased for some EOCs in filtrates (i.e. type IA, solid waste treatment, Sci. Total Environ. 433 (2012) 352–361.
IIA, IIB, IIIA), and finally inhibited in retentates for most of EOCs (i.e. [6] H.W. Leung, T.B. Minh, M.B. Murphy, J.C.W. Lam, M.K. So, M. Martin, P.K.S. Lam,
type II and type III) probably due to the existence of concentrated B.J. Richardson, Distribution, fate and risk assessment of antibiotics in sewage
treatment plants in Hong Kong, South China, Environ. Int. 42 (2012) 1–9.
DOM in this fraction. Furthermore, a negative correlation was found [7] C.X. Yan, Y. Yang, J.L. Zhou, M. Liu, M.H. Nie, H. Shi, L.J. Gu, Antibiotics in the
between the photodegradation rates of 14 EOCs and DOC contents surface water of the Yangtze Estuary: Occurrence, distribution and risk
in filtrate, permeate and retentate of DW and LW (Table 2), indi- assessment, Environ. Pollut. 175 (2013) 22–29.
[8] J.J. Jiang, C.L. Lee, M.D. Fang, Emerging organic contaminants in coastal
cating the significant role played by DOC on the photolysis of EOCs.
waters: anthropogenic impact, environmental release and ecological risk,
However, the photolysis of EOCs in the aquatic environment could Mar. Pollut. Bull. 85 (2014) 391–399.
be affected by inorganic ions, DOM and metal oxides in colloidal [9] M.H. Nie, C.X. Yan, W.B. Dong, M. Liu, J.L. Zhou, Y. Yang, Occurrence,
distribution and risk assessment of estrogens in surface water, suspended
fractions. Therefore, the impact of colloid properties and water
particulate matter and sediments of the Yangtze Estuary, Chemosphere 127
chemistry on the photolysis of EOCs during the irradiation process (2015) 109–116.
must be considered and requires further investigation. [10] S. Bartelt-Hunt, D.D. Snow, T. Damon-Powell, D. Miesbach, Occurrence of
steroid hormones and antibiotics in shallow groundwater impacted by
livestock waste control facilities, J. Contam. Hydrol. 123 (2011) 94–103.
[11] R. López-Serna, A. Jurado, E. Vázquez-Suñé, J. Carrera, M. Petrović, D. Barceló,
4. Conclusion Occurrence of 95 pharmaceuticals and transformation products in urban
ground waters underlying the metropolis of Barcelona, Spain. Environ. Pollut.
By determining the selected EOCs in the different fractions of 174 (2013) 305–315.
[12] M. Huerta-Fontela, M.T. Galceran, F. Ventura, Occurrence and removal of
typical wastewaters, we have been able to show that colloids play
pharmaceuticals and hormones through drinking water treatment, Water
an important role in determining the distribution, fate, and espe- Res. 45 (2011) 1432–1442.
cially the photodegradation of EOCs in aquatic systems. Under [13] M.F. Rahman, S. Peldszus, W.B. Anderson, Behaviour and fate of perfluoroalkyl
and polyfluoroalkyl substances (PFASs) in drinking water treatment: A
simulated solar radiation, UV light was proven to be more effective
review, Water Res. 50 (2014) 240–318.
than other wavelengths on the photolysis of most EOCs. Addition- [14] K.V. Thomas, L. Bijlsma, S. Castiglioni, A. Covaci, E. Emke, R. Grabic, F.
ally, the results of the photolysis of EOCs in different fractions of Hernández, S. Karolak, B. Kasprzyk-Hordern, R.H. Lindberg, M.L. de Alda, A.
wastewater showed that the impact of colloids on the photolysis Meierjohann, C. Ort, Y. Pico, J.B. Quintana, M. Reid, J. Rieckermann, S. Terzic,
A.L.N. van Nuijs, P. de Voogt, Comparing illicit drug use in 19 European cities
of EOCs was dependent on the compound, colloidal concentration, through sewage analysis, Sci. Total Environ. 432 (2012) 432–439.
and the source of water. Selected EOCs with higher degrada- [15] Z.H. Yan, X.F. Yang, G.H. Lu, J.C. Liu, Z.X. Xie, D.H. Wu, Potential environmental
tion rates were detected at low levels in natural samples, which implications of emerging organic contaminants in Taihu Lake, China:
Comparison of two ecotoxicological assessment approaches, Sci. Total
indicated that photodegradation could be an important removal Environ. 470–471 (2014) 171–179.
pathway of EOCs in aquatic systems. Information provided from [16] J. Martín, D. Camacho-Muñoz, J.L. Santos, I. Aparicio, E. Alonso, Occurrence of
this study will provide important data to better understand the pharmaceutical compounds in wastewater and sludge from wastewater
treatment plants: removal and ecotoxicological impact of wastewater
behavior and elimination of these compounds in aquatic envi- discharges and sludge disposal, J. Hazard. Mater. 239–240 (2012) 40–47.
ronments. However, more information on the colloidal properties [17] J. Sanchez-Avila, R. Tauler, S. Lacorte, Organic micropollutants in coastal
including composition, size, as well as DOM composition during the waters from NW Mediterranean Sea: sources distribution and potential risk,
Environ. Int. 46 (2012) 50–62.
photolysis procedure is needed to better understand the impact of
[18] C. Ávila, V. Matamoros, C. Reyes-Contreras, B. Piña, M. Casado, L. Mita, C.
colloids on the photolysis of EOCs in an aquatic system. Rivetti, C. Barata, J. García, J.M. Bayona, Attenuation of emerging organic
248 C. Yan et al. / Journal of Hazardous Materials 299 (2015) 241–248

contaminants in a hybrid constructed wetland system under different [36] K. Maskaoui, J.L. Zhou, Colloids as a sink for certain pharmaceuticals in the
hydraulic loading rates and their associated toxicological effects in aquatic environment, Environ. Sci. Pollut. Res. 17 (2010) 898–907.
wastewater, Sci. Total Environ. 470–471 (2014) 1272–1280. [37] R. Wilding, J.L. Liu Zhou, Validation of cross-flow ultrafiltration for sampling
[19] J.R. Lead, K.J. Wilkinson, Aquatic Colloids and Nanoparticles: current of colloidal particles from aquatic systems, J. Colloid Interf. Sci. 280 (2004)
knowledge and future trends, Environ. Chem. 3 (2006) 159–171. 102–112.
[20] Y. Kalmykova, K. Björklund, A.M. Strömvall, L. Blom, Partitioning of polycyclic [38] S.D.W. Comber, Abiotic persistence of atrazine and simazine in water, Pestic.
aromatic hydrocarbons, alkylphenols, bisphenol a and phthalates in landfill Sci. 55 (1999) 696–702.
leachates and stormwater, Water Res. 47 (2013) 1317–1328. [39] M.H. Nie, Y. Yang, Z.J. Zhang, C.X. Yan, X.N. Wang, H.J. Li, W.B. Dong,
[21] Y.H. Zou, W. Zheng, Modeling manure colloid-facilitated transport of the Degradation of chloramphenicol by thermally activated persulfate in aqueous
weakly hydrophobic antibiotic florfenicol in saturated soil columns, Environ. solution, Chem. Eng. J. 246 (2014) 273–382.
Sci. Technol. 47 (2013) 5185–5192. [40] W.H. Xu, G. Zhang, X.D. Li, S.C. Zou, P. Li, Z.H. Hu, et al., Occurrence and
[22] S. Majumder, B. Nath, S. Sarkar, D. Chatterjee, G. Roman-Ross, M. Hidalgo, elimination of antibiotics at four sewage treatment plants in the Pearl River
Size-fractionation of groundwater arsenic in alluvial aquifers of West Bengal, Delta (PRD), South China. Water Res. 41 (2007) 4526–4534.
India: the role of organic and inorganic colloids, Sci. Total Environ. 468–469 [41] H.W. Leung, T.B. Minh, M.B. Murphy, J.C.W. Lam, M.K. So, M. Martin, P.K.S. Lam,
(2014) 804–812. B.J. Richardson, Distribution, fate and risk assessment of antibiotics in sewage
[23] Y. Yang, J. Fu, H. Peng, L. Hou, M. Liu, J.L. Zhou, Occurrence and phase treatment plants in Hong Kong, South China, Environ. Int. 42 (2012) 1–9.
distribution of selected pharmaceuticals in the Yangtze Estuary and its coastal [42] S. Managaki, A. Murata, H. Takada, B.C. Tuyen, N.H. Chiem, Distribution of
zone, J. Hazard. Mater. 190 (2011) 588–596. macrolides, sulfonamides, and trimethoprim in tropical waters: ubiquitous
[24] M.H. Nie, Y. Yang, M. Liu, C.X. Yan, H. Shi, W.B. Dong, J.L. Zhou, Environmental occurrence of veterinary antibiotics in the Mekong Delta, Environ. Sci.
estrogens in a drinking water reservoir area in Shanghai: occurrence, colloidal Technol. 41 (2007) 8004–8010.
contribution and risk assessment, Sci. Total Environ. 487 (2014) 785–791. [43] S. Cavenati, P.N. Carvalho, C.M. Almeida, M.C. Basto, M.T. Vasconcelos,
[25] C.X. Yan, Y. Yang, J.L. Zhou, M.H. Nie, M. Liu, M.F. Hochella Jr., Selected Simultaneous determination of several veterinary pharmaceuticals in
emerging organic contaminants in the Yangtze Estuary, China: a effluents from urban, livestock and slaughterhouse wastewater treatment
comprehensive treatment of their association with aquatic colloids, J. Hazard. plants using a simple chromatographic method, Water Sci. Technol. 66 (2012)
Mater. 283 (2015) 14–23. 603–611.
[26] S.K. Khetan, T.J. Collins, Human pharmaceuticals in the aquatic environment: [44] S. Kaniou, K. Pitarakis, I. Barlagianni, I. Poulios, Photocatalytic oxidation of
a challenge to green chemistry, Chem. Rev. 107 (2007) 2319–2364. sulfamethazine, Chemosphere 60 (2005) 372–380.
[27] L. Carlos, D.O. Mártire, M.C. Gonzalez, J. Gomis, A. Bernabeu, A.M. Amat, A. [45] K. Yapsakli, F. Çeçen, Effect of type of granular activated carbon on DOC
Arques, Photochemical fate of a mixture of emerging pollutants in the biodegradation in biological activated carbon filters, Process Biochem. 45
presence of humic substances, Water Res. 46 (2012) 4732–4740. (2010) 355–362.
[28] B. Razavi, S. Ben Abdelmelek, W. Song, K.E. O’shea, W.J. Cooper, [46] L. Deng, Y.S. Shao, N.Y. Gao, Y. Deng, S.Q. Zhou, X.H. Hu, Thermally activated
Photochemical fate of atorvastatin (lipitor) in simulated natural waters, persulfate (TAP) oxidation of antiepileptic drug carbamazepine in water,
Water Res. 45 (2011) 625–631. Chem. Eng. J. 228 (2013) 765–771.
[29] L.K. Ge, J.W. Chen, X.L. Qiao, J. Lin, X.Y. Cai, Light-source-dependent effects of [47] D. Vione, J. Feitosa-Felizzola, C. Minero, S. Chiron, Phototransformation of
main water constituents on photodegradation of phenicol antibiotics: selected human-used macrolides in surface water: kinetics, model
mechanism and kinetics, Environ. Sci. Technol. 43 (2009) 3101–3107. predictions and degradation pathways, Water Res 43 (2009) 1959–1967.
[30] D.W. Wang, Y. Li, G.P. Li, C. Wang, W.L. Zhang, Q. Wang, Modeling of [48] L. Mao, C. Meng, C. Zeng, Y. Ji, X. Yang, S. Gao, The effect of nitrate, bicarbonate
quantitative effects of water components on the photocatalytic degradation and natural organic matter on the degradation of sunscreen agent
of 17␣-ethynylestradiol in a modified flat plate serpentine reactor, J. Hazard. p-aminobenzoic acid by simulated solar irradiation, Sci. Total Environ. 409
Mater. 254–255 (2013) 64–71. (2011) 5376–5381.
[31] Y.D. Xu, D.B. Yue, Y. Zhu, Y.F. Nie, Fractionation of dissolved organic matter in [49] X.X. Ou, S. Chen, X. Quan, H.M. Zhao, Photoinductive activity of humic acid
mature landfill leachate and its recycling by ultrafiltration and evaporation fractions with the presence of Fe(III): the role of aromaticity and oxygen
combined processes, Chemosphere 64 (2006) 903–911. groups involved in fractions, Chemosphere 72 (2008) 925–931.
[32] Y. Chen, K. Zhang, Y.G. Zuo, Direct and indirect photodegradation of estriol in [50] L. Wang, H.M. Xu, W.J. Cooper, W.H. Song, Photochemical fate of beta-blockers
the presence of humic acid, nitrate and iron complexes in water solutions, Sci. in NOM enriched waters, Sci. Total Environ. 426 (2012) 289–295.
Total Environ. 463–464 (2013) 802–809. [51] J. Bachman, H.H. Patterson, Photodecomposition of the carbamate pesticide
[33] L. Prieto-Rodrigueza, S. Miralles-Cuevasa, I. Ollera, A. Agüerab, G. Li Pumad, S. carbofuran: kinetics and the influence of dissolved organic matter, Environ.
Malato, Treatment of emerging contaminants in wastewater treatment plants Sci. Technol. 33 (1999) 874–881.
(WWTP) effluents by solar photocatalysis using low TiO2 concentrations, J. [52] L. Chen, X.J. Tang, C.F. Shen, C. Chen, Y.X. Chen, Photosensitized degradation of
Hazard. Mater. 211–212 (2012) 131–137. 2,4 ,5-trichlorobiphenyl (PCB 31) by dissolved organic matter, J. Hazard.
[34] T.E. Doll, F.H. Frimmel, Fate of pharmaceuticals-photodegradation by Mater. 201–202 (2012) 1–6.
simulated solar UV-light, Chemosphere 52 (2003) l757–1769.
[35] B. Liu, X.L. Liu, Direct photolysis of estrogens in aqueous solutions, Sci. Total
Environ. 320 (2004) 269–274.