Journal of Environmental Chemical Engineering 4 (2016) 4453–4465

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

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

Degradation of oxytetracycline using microporous and mesoporous

photocatalyst composites: Uniform design to explore factors
Xiao-Dan Xiea , Kefu Zhoua,* , Bor-Yann Chenb , Chang-Tang Changc,*
a
Department of Environmental and Ecological Engineering, Xiamen University, China
b
Department of Chemical and Materials Engineering, National I-Lan University, Taiwan
c
Department of Environmental Engineering, National I-Lan University, Taiwan

A R T I C L E I N F O A B S T R A C T

Article history:
Received 23 July 2016 Microporous ZSM-5 and mesoporous MCM-41 loaded with TiO2 were synthesized to remove
Received in revised form 6 October 2016 oxytetracycline in aqueous solutions. The adsorption of oxytetracycline showed that MCM-41 had
Accepted 8 October 2016 higher adsorption capacity than ZSM-5 and that this capacity changed little after TiO2 loading. Four
Available online 11 October 2016 factors affecting photocatalytic degradation were investigated by uniform design; these factors included
the initial pH of oxytetracycline solution, reaction temperature, dosage of catalysts, and TiO2 content of
Keywords: catalysts. Single-factor experiment confirmed that experimental values agreed with predicted ones by
Oxytetracycline the regression formula of the uniform design. After recycling and reuse for five cycles, the composites
Composite
retained their degradation performance. Both adsorption and photocatalytic degradation experiments
Uniform design
indicated that MCM-41 was more suitable than ZSM-5 for loading TiO2 to remove oxytetracycline
Toxicity assessment
because of the former’s better absorption properties. The biomass at the stationary growth phase and the
maximum specific growth rate of Escherichia coli were used to assess toxicity during photocatalytic
degradation, which linearly decreased with increased oxytetracycline degradation efficiency.
ã 2016 Elsevier Ltd. All rights reserved.

1. Introduction Nanometer TiO2 loading on zeolite possesses strong adsorption

Oxytetracycline (OTC, Fig. 1) is a tetracycline antibiotic, which is
widely used in medicine and the livestock and fishery industries.
OTC undergoes minimal metabolism and is mainly excreted via
urine as prototype into sewage or directly into surface water [1–3].
The broad-spectrum antimicrobial properties of OTC limit
traditional biological treatment, so that it can be detected at
nanogram to microgram levels in the sewage treatment plant
effluent [4,5]. The contamination with antibiotics has introduced
antibiotic resistance genes in the environment [6,7].
Many advanced oxidation processes have been developed for
removing antibiotics from sewage. Photocatalytic degradation by
nanometer TiO2 is one of the research hotspots [8]. However,
nanometer TiO2 is difficult to recover and reuse. The potentially
toxic effects of nanoparticles bring danger to the ecological
environment [9,10]. Consequently, the composites with nanometer
TiO2 loaded on various materials were developed [11–14].
* Corresponding authors at: Department of Environmental Engineering, National
I-Lan University, Taiwan.
E-mail addresses: zhkefu@xmu.edu.cn (K. Zhou), ctchang73222@gmail.com
(C.-T. Chang).
ability as well as photocatalytic properties. Pore sizes of different
zeolites vary from 10
10 to 10
8 meter, which closely relates to TiO2
loading position and adsorption capacity of composites. Therefore,
to remove specific contaminants, it is necessary to take pore size of
photocatalytic composites into consideration.
The application of experimental design methods in studying
photocatalytic process, such as orthogonal experiment [15],
central composite design [16–18], and surface response method
[19–21], can reduce the amount of experiments and bring out
statistically significant results. However, with the methods
mentioned above, as the number of factors and levels increase,
the number of experiments always exponentially increased.
Uniform design were employed to select representative sample
sets, thereby reducing the experimental times and the computa-
tion. Experimental time is minimized to equal the number of
levels. Meanwhile, experimental points were evenly distributed
within the test range. More information can be extracted from a
relatively small number of experiments by uniform design, which
is suitable for multi-level and multi-factor problems [22]. In recent
years, uniform design was widely used in many research field, such
as engineering mechanical [23–25], analysis chemistry [26,27],
material science [28,29] and toxicology [30–32]. However, it is has
never been applied to study photocatalytic process, so we try to

http://dx.doi.org/10.1016/j.jece.2016.10.012
2213-3437/ã 2016 Elsevier Ltd. All rights reserved.
4454 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465
Fig. 1. Stick and ball model (a) and Connolly molecular surface structure (b) of OTC.
prove the usability of the experimental design model in this
research field.
In this research, the adsorption and photocatalytic degradation
capability of microporous and mesoporous composites, TiO2
loading on ZSM-5 and MCM-41 zeolite were prepared and
compared. The performance of photocatalytic degradation was
explored by uniform design with OTC simulated wastewater.
Uniform design was adopted to study photocatalytic degradation
process for the first time. The predicted results of uniform design
and actual results of single factor experiments were compared. To
assess the toxic effect of OTC and its subsequent degradation
solution to Escherichia coli, two calculating methods, the maximum
specific growth rate and the biomass at stationary growth phase of
inhibition rate were compared.
2. Materials and methods
2.1. Materials
Oxytetracycline hydrochloride (OTC, 95%–100.5%) was pur-
chased from Panreac. The raw ZSM-5 zeolite and TiO2 (P25)
powder, as well as the chromatographically pure acetonitrile and
triethylamine were purchased from ACROS. Tetraethyl orthosili-
cate (TEOS, 98%) and hexadecyltrimethylammonium bromide
(CTMABr, 99%) were purchased from Aldrich. Ethanol (99.5%),
ethylenediaminetetraacetic acid disodium (EDTA-2Na), ammoni-
um acetate, HCl, H2SO4, formic acid, NaOH, and ammonia water
(28%) were analytical grade and purchased from SHIYAKU. Ultra-
pure water with a resistivity more than 18.00 MV cm was prepared
by OTUN ultrapure water system.
2.2. Preparation and characterization of the composites
MCM-41 was synthesized according to our previous method
[33]. Briefly, 4.6 g of CTMABr was dissolved in 95 mL of distilled
water. A homogeneous transparent solution was made by
continuously stirring with a laboratory stirrer before ammonia
water was added to the solution. The mixture was stirred for
15 min before 8.8 mL of TEOS was added. The mixture was stirred
for 12 h at room temperature. The white precipitate was filtered
and washed consecutively with distilled water and dried at 100  C
for 8 h. The material was finally calcined at 550  C for 5 h to burn
out the template to obtain MCM-41.
Composites were synthesized by the solid-dispersion method
[34]. TiO2 was uniformly dissolved in ethanol after ultrasonic
dispersion for 30 min. ZSM-5 or MCM-41 was added into the
suspension with rapid stirring at room temperature for 1 h to
enable TiO2 to be fully adsorbed onto the surface of ZSM-5 or MCM-
41. The suspension was dried at 100  C for 8 h to remove the solvent
and calcined at 400  C for 2 h. The products were finally ground
into ultrafine powder and referred to as xTZ or xTM, where x was
the weight percent of TiO2 in the composite.
The crystal phases of the composites were determined with an
X-ray diffractometer (XRD, Rigaku Ultima IV). Their shape and fine
structure were observed by scanning election microscopy (SEM;
Mac Science, MXP18) and transmission electron microscopy (TEM;
JEOL, JEM-2000 FX). The element distribution of composites was
analyzed by energy dispersive spectroscopy (EDS; HORIBA, EMX)
and inductively coupled plasma-atomic emission spectroscopy
(ICP-AES, Agilent, 4100).The specific surface area and pore
distribution was studied by an N2 adsorption/desorption analyzer
(Micromeritics, ASAP 2020N).
2.3. Adsorption experiment
The OTC solution of 10 mg L
1 was freshly prepared in ultra-
pure water. The initial pH was unadjusted and ranged from 4.63 to
4.91. The amount of material used was 0.1, 0.2, 0.3, 0.4, and 0.5 g,
whereas the solution volume was 100 mL. The materials were
mixed in a group of glass conical flasks on a constant bath shaker
shielded from light at 125 rpm at 25  C. Throughout the reaction of
80 min, samples were taken at predetermined time and filtrated
using filter membrane (0.2 mm) before analysis.
2.4. Photocatalytic degradation experiment
Photocatalytic experiments were performed in cylindrical
reactor with the diameter of 7 cm and height of 40 cm, which
were shielded from external light. Ultraviolet (UV) lamp (254 nm,
0.97 mW cm
2; APUV-12F, Sparxic) was used as the single light
source. The 1 L solution was continuously stirred by a magnetic
agitator at 100 rpm. The pH was adjusted with 1 mol L
1 HCl or
1 mol L
1 NaOH. The temperature was controlled by a constant-
temperature water circulator. The initial pH of OTC solution,
reaction temperature, dosage of catalysts, and the TiO2 content in
catalysts were changed according to the uniform design. The
device is shown schematically in Fig. S1. Throughout the reaction
for 60 min, samples were taken at predetermined times and
filtered with a filter membrane (0.2 mm) before analysis.
2.5. Recycling experiment
The photocatalyst was recycled and reused 5 times to study its
stability. In the first cycle, 0.01 g TiO2 and the composite with an
equal amount of TiO2 were added to the 10 mg L
1 OTC solution
without adjusting pH. Subsequently, the 1 L mixture was continu-
ously stirred by a magnetic agitator at 100 rpm. The temperature
was controlled at 25  C by a constant-temperature water circulator.
After the reaction for 150 min, samples were purified with a filter
membrane (0.2 mm) before analysis. At the end of the first cycle,
the photocatalyst was filtered by filter paper, dried at 100  C, and
calcified at 400  C for 2 h to remove adsorbed intermediates. The
same procedure was employed after every cycle of usage.
 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465 4455

Fig. 2. SEM images of (a) ZSM-5, (b)10TZ, (c) MCM-41, and (d)10TM, and TEM images of (e) ZSM-5, (f)10TZ, (g) MCM-41, and (h)10TM.
4456 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465
2.6. Analysis of OTC
The OTC was monitored by a HPLC system (Ultimate 3000,
Thermofisher) consisting of an Acclaim 120 C18 reverse phase
column (4.6 mm  250 mm, 5 mm). The isocratic mobile phase was
88% A with ammonium acetate (0.25 mol L
1), EDTA-2Na solution
(0.1 mol L
1), and triethylamine at volume ratio of 100/10/1,
whereas 12% B was acetonitrile. The flow rate was 1.0 mL min
1.
The detection wavelength was 280 nm.
A HPLC–MS system with positive ion electrospray ionization
source and ACQUITY CSH C18 reverse phase column (1.7 mm,
2.1 10 mm) was used to analysis degradation intermediates of
OTC. The flow rate was 0.2 mL min
1. Gradient conditions of mobile
phase A (0.1% formic acid aqueous solution) and B (0.1% formic acid
acetonitrile solution) were listed in Table S1. The value of capillary
voltage, ion source temperature, desolvation temperature, des-
olvation gas flow rate, and cone voltage was 2.5 kV, 150  C, 500  C,
1000 L h
1, and 16 V, respectively. Detection mode was multiple
reaction monitoring.
2.7. Toxicity tests
The toxicity of photocatalytic degradation solution was
evaluated with E. coli DH5a. The strain was activated in 50 mL
of Luria–Bertani (LB) broth (10 g L
1 peptone, 5 g L
1 yeast, 10 g L
1
NaCl) at 37  C for 12 h in a constant-temperature shaking incubator
at 125 rpm. A certain amount of activated E. coli and 5 mL of the
photocatalytic degradation solution was added into 50 mL of LB
broth such that the initial optical density (OD) is 0.10. The mixture
was cultured at 37  C for 12 h in a constant-temperature shaking
incubator at 125 rpm. Meanwhile, 5 mL of sterile distilled water
was used instead of the photocatalytic degradation solution as
control. Samples of 0.5 mL were taken at predetermined times
from each culture flask to measure their OD; the growth curve of E.
coli was recorded.
Fig. 3. EDS spectra of (a) ZSM-5, (b)10TZ, (c) MCM-41, and (d)10TM.
3. Results and discussion
3.1. Characterization of composites
3.1.1. SEM, TEM and element analysis
The morphologies of ZSM-5, MCM-41, 10TZ, and 10TM were
investigated by SEM (Fig. 2). The particle size of ZSM-5 ranged from
0.5 mm to 3 mm; almost all MCM-41 particles were smaller than
0.5 mm. No major changes were observed in the ZSM-5 structure;
TiO2 was uniformly dispersed without significant aggregation.
However, the TiO2 particle size of 21 nm was larger than the pore
size of ZSM-5 (5.1-5.6 Å, as measured by Blasioli et al. [35]) or
MCM-41 (approximately 3 nm, as measured in our pervious study
[33]). Therefore, the TiO2 particles are attached to the outer surface
rather than the inner surface of ZSM-5 or MCM-41. It can be
observed from TEM images (Fig. 2) that MCM-41 with lattice
spacing of 2.5 nm was successfully synthesized. Lattice spacing of
ZSM-5 is 0.39 nm. TiO2 particles in 10TZ and 10TM composites are
uniformly distributed. EDS spectra (Fig. 3) shows that the major
elements in ZSM-5 and MCM-41 are silicon and oxygen. Aluminum
(tag manually added in Fig. 3) is undetectable in ZSM-5. Both 10TZ
and 10TM contain titanium. However, EDS elemental analysis is
semi-quantitative with ratio of surface elements. Therefore, ICP-
AES was used to further analyze elements ratio. The results showed
that, titanium content of 10TZ and 10TM was 5.54% and 5.71%
respectively, which is close to the expected 10% TiO2, indicating
little TiO2 loss during preparation process.
3.1.2. N2 adsorption-desorption isotherms, pore and BET surface area
The surface area of ZSM-5, MCM-41, and the composites were
determined based on the N2 adsorption-desorption. According to
the classification of IUPAC, the shape of hysteresis loop in N2
adsorption/desorption curve can be divided into H1, H2, H3, and
H4 four types [36]. The isotherms of 10TZ composite in Fig. 4(a) is
type H4, indicated narrow pores in material which is caused by
bond of disc- and plate-shaped particles [37]. The isotherms of
 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465 4457

increasing TiO2 content, probably because TiO2 particles covered

10TZ Adsorption (a) the pores of MCM-41.
400 10TZ Desorption
10TM Adsorption
10TM Desorption 3.1.3. XRD
Quantily Adsorbed (cm /g STP)

The XRD patterns of the composites were compared to ZSM-5,

300 MCM-41, and TiO2 in Fig. 5. The anatase phase (101), (004), and
3

(200) of TiO2 can be clearly observed at the characteristic peak 2u

of 25.2 , 37.7, and 47.9 , where the intensity increased with the
200 rising TiO2 loading ratio. The characteristic peak at 2u = 27.3
represents the rutile phase (110), which is not obvious in
composites because of its low content. Characteristic low-angle
100 peaks of MCM-41 indicated that TM composites possessed a
hexagonal mesophase structure with a clear peak at 2u = 2.4 was
attributed to (100) and weaker reflections at 2u of 4.2 and 4.7,
which are assignable to (110) and (200). Similar to TM composites,
0
0.0 .2 .4 .6 .8 1.0
no significant shift of ZSM-5 characteristic peaks was observed in
the patterns of TZ composites compared to ZSM-5. Therefore, the
Relative Pressure (P/P0 )
structure of zeolite was not altered after TiO2 loading by the solid
dispersion method.

1.4
(b)
10TZ
1.2 10TM

(a)
1.0
Pore Volume (cm /g nm)

TiO2
.8
3

20TZ
.6
Realative Intensity

10TZ
.4

5TZ
.2

1TZ
0.0

ZSM-5
0 2 4 6 8 10

Pore Diameter (nm)

Fig. 4. (a) N2 adsorption–desorption isotherm, (b) pore volume and pore size
0 10 20 30 40 50 60 70 80 90
distribution of 10TZ and 10TM composites.
2θ ( ° )
10TM composite is type H1, indicated that pores with similar pore
shape and uniform pore size are orderly arranged [37]. Therefore,
MCM-41 with neat pores was successfully synthesized and its
structure was not altered in composite, which agreed with the small ang le (b)
results of TEM. The pore volumes of the 10TZ and 10TM materials
were 0.081 and 0.936 cm3 g
1, respectively. The main pore sizes of TiO2

the materials were 0.66 and 2.63 nm, respectively, as shown in

20TM
Fig. 4(b). The surface area of ZSM-5, MCM-41, and the composites
Realative Intensity

are listed in Table 1. The BET surface area of the ZSM-5 was 10TM
311 m2 g
1. With the increasing TiO2 weight ratio, the surface area
of composites was slightly increased from 311 m2 g
1 to 347 m2 g
1 5TM
then decreased to 295 m2 g
1. For the TZ composite, the perfect
dispersity was proved by increasing the surface area until heavier 2TM

loading caused particle agglomeration. The BET surface area of TM

composites decreased from 1100 m2 g
1 to 812 m2 g
1 with MCM-41

Table 1
BET surface area of ZSM-5, MCM-41, and composites.
0 10 20 30 40 50 60 70 80 90
Materials BET surface area (m2 g
1) Materials BET surface area (m2 g
1)
ZSM-5 311 MCM-41 1100
2θ ( ° )
2TZ 333 2TM 993
Fig. 5. XRD patterns of (a) ZSM-5, TiO2 and TZ composites, (b) MCM-41, TiO2, and
5TZ 347 5TM 934
TM composites. Characteristic peaks: anatase (black diamonds), rutile (empty
10TZ 327 10TM 817
diamonds), ZSM-5 (black circles), and MCM-41 (red triangles). (For interpretation of
20TZ 295 20TM 812
the references to colour in this figure legend, the reader is referred to the web
version of this article.)
4458 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465

3.2. Adsorption kinetics of OTC on materials Table 2

Adsorption kinetics parameters of OTC on materials.

Fig. 6 shows the time courses of OTC concentration when ZSM- Materials qe (mg g
1) kpos (g mg
1 min
1) t1/2 (min) R2
5, MCM-41, 10TZ, and 10TM were used as adsorbents. To evaluate ZSM-5 1.89 0.249 2.13 0.966
the adsorption rate, a pseudo-second-order model was applied to 10TZ 1.91 0.154 3.40 0.977
the experimental data as Eqs. (1) and (2): MCM-41 8.59 0.019 6.23 0.990
10TM 6.39 0.055 2.85 0.969
dqt
¼ kpso ðqe
qt Þ2 ð1Þ
dt
and the TiO2 content in catalysts (X4). An OTC degradation
efficiency of 60 min was considered as a response (YTZ and YTM).
t t 1 According to the uniform design table U11(116) and its usage table
¼ þ ð2Þ
qt qe kpso q2e (Tables S2 and S3), a 4-factor and 11-level uniform design leads to a
where kpso is the pseudo-second-order adsorption rate constant; qt total number of 11 experiments with a deviation of 0.3528 [38].
and qe are the amounts of OTC adsorbed per gram of zeolite at t or The values of different factors and responses are listed in Table 3.
after reaching the adsorption equilibrium, respectively. Notably, In the photocatalytic degradation process, the degradation
kpso is an apparent rate constant and a complex function of the efficiency of tetracyclines usually increases with the initial pH,
initial concentration of OTC (C0), which is different from the temperature, and catalyst dosage until the optimum is reached
adsorption equilibrium constant. Furthermore, t1/2 is often used to before the efficiency declines [39–45]. The quadratic equations X1,
evaluate the adsorption rate, which is the time required to adsorb X2, and X3 were anticipated to simulate their effects. The TiO2
half the amount of qe. t = t1/2; the equation: qt = qe/2 were inserted content in catalysts will not contribute negatively to the normal
into Eq. (2) to obtain t1/2 in Eq. (3): degradation efficiency. However, with the increased TiO2 content,
the growth of the degradation efficiency slows down, which is
1 expected to fit with the quadratic equation of X4. By considering
t1=2 ¼ ð3Þ
kpso qe the possible interactions between each factor, six interactive items
The values of qe, kpso, and t1/2 were calculated for each of the were introduced to the equation. Finally, a multivariate non-linear
experimental conditions as shown in Table 2. model was built to represent the effects of each factor on the
The equilibrium adsorption of ZSM-5, 10TZ, MCM-41, and 10TM response:
was 1.89, 1.91, 8.59, and 6.39 mg g
1, respectively. The equilibrium Y TZ orY TM ¼ b0 þ b1 X 1 þ b2 X 2 þ b3 X 3 þ b4 X 4 þ b5 X 1 2 þ b6 X 2 2
adsorption quantity of ZSM-5 was low because micropores in the
þ b7 X 3 2 þ b8 X 4 2 þ b9 X 1 X 2 þ b10 X 1 X 3 þ b11 X 1 X 4
ZSM-5 zeolite (pore diameter 6.6 Å in this study) were not þ b12 X 2 X 3 þ b13 X 2 X 4 þ b14 X 3 X 4 ð4Þ
appropriate for large molecular structures [35], such as OTC
molecules with volume of 325.5 Å3 (calculated from Connolly where b0, b1, b2, . . . , b14 are coefficients. This equation contains 15
solvent-excluded volume modle, Fig. 1). Loading TiO2 slightly unknown coefficients, which are more than the 11 total
improved the adsorption performance. For mesoporous MCM-41, experimental times. Thus, the variables need to be screened. After
loading TiO2 weakened the adsorption capacity, thereby indicating excluding the items with non-significant contributions by the
that the low adsorption properties of TiO2 limited the adsorption enter method (IBM SSPS Statistics 19), the regression equations
capacity of TM composite. The t1/2 values of all materials were less are:
than 7 min, thereby confirming the rapid adsorption and removal
of OTC from the aqueous solution. Y TZ ¼
23:63 þ 3:28X 1 þ 19:68X 2 þ 2:40X 3
0:21X 1 2

1:49X 2 2
0:73X 3 2
0:58X 4 2
1:55X 3 X 4 ð5Þ
3.3. Uniform design on photocatalytic degradation of OTC

The factors investigated in our study were the initial pH of OTC Y TM ¼
39:45 þ 2:41X 1 þ 3:54X 2 þ 6:86X 3
0:31X 1 2
solution (X1), reaction temperature (X2), dosage of catalysts (X3),
0:21X 2 2
1:15X 3 2
0:65X 4 2
1:54X 3 X 4 ð6Þ

The standardized coefficients of Eqs. (5) and (6) are compared

in Fig. 7. For TZ composites, the influence of temperature is the
largest, followed by the interaction of dosage and TiO2 content; for
TM composites, the influence of dosage is the largest, but the
interaction of dosage and TiO2 content also have a significant
impact. The dosage and TiO2 content have a strong interaction,
thereby indicating that the actual dosage of TiO2 significantly
affects the degradation efficiency.
The statistical parameters of the regression equations are listed
in Table 4. The Sig. values of two regression equations are both less
than 0.05, thereby indicating that both equations are significant.
The correlation coefficient and the residual standard deviation
proved that the predicted data of the TZ composite fits better than
that of the TM composite.

3.4. Prediction and validation of single factor experiment

The values of three factors were fixed while the remaining

factor was changed in a certain range as shown in Table 5. The
Fig. 6. Adsorption of OTC on ZSM-5, 10TZ, MCM-41, and 10TM. values of degradation efficiency were predicted by Eqs. (5) and (6)
 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465 4459

Table 3
Experiment parameters of uniform design.

Runs Experiment Design Experiment Conditions Results


1
X1 X2 X3 X4 pH T ( C) Dosage (g L ) TiO2 (%) YTZ (%) YTM (%)

1 1 3 5 7 2 15 0.08 12 44.49 79.27

2 2 6 10 3 3 30 0.18 4 38.56 51.14
3 3 9 4 10 4 45 0.06 18 43.19 61.96
4 4 1 9 6 5 5 0.16 10 27.34 77.34
5 5 4 3 2 6 20 0.04 2 53.07 66.47
6 6 7 8 9 7 35 0.14 16 89.51 98.64
7 7 10 2 5 8 50 0.02 8 38.05 66.17
8 8 2 7 1 9 10 0.12 0 15.45 46.04
9 9 5 1 8 10 25 0 14 25.52 25.66
10 10 8 6 4 11 40 0.1 6 63.99 74.32
11 11 11 11 11 12 55 0.2 20 79.55 84.01

conditions, thereby limiting the photodegradation of OTC [46]. By

4 contrast, alkaline degradation is originally beneficial to photo-
X1 X2 X3 X3 X4
degradation because the quantum performance is increased at
3 higher pH values; negatively charged OTC molecules with a high
electrical density on the ring system tend to attract reactive species
2 such as OH, which facilitate OTC photolysis [47]. However, given
Standardized coefficients

the strong electric repulsion, the extremely low adsorption of

1 negatively charged OTC molecules on catalysts hampered further
0
-1
-2
X1
2
X2
2
-3
pH Temperature
Factors
Fig. 7. Comparison of standardized coefficients.
and plotted in Fig. 8. The actual single factor experiment points are
basically aligned with the predicted trends. Furthermore, the
degradation performance of the TM composite was better than
that of the TZ composite under most conditions.
The photocatalytic degradation of OTC under different pH
conditions is shown in Fig. 8(a). Overall, the effect of pH on the
degradation efficiency is not evident. For the TZ composite, the
degradation efficiency first increased before slightly decreasing.
The highest degradation efficiency of 76.38% was achieved at pH 7.
The experimental values were basically equal to the predicted
values. For the TM composite, the degradation efficiency achieved
a maximum of 92.14% at pH 5 and decreased more obviously than
that of the TZ composite in alkaline conditions. OTC molecules
with low quantum performance are relatively stable in acidic
photocatalytic degradation [40].
The photocatalytic degradation of OTC under different temper-
ature conditions is shown in Fig. 8(b). The actual values of TZ and
TM composites were both first increased but later decreased. The
degradation efficiency was greatly improved when the tempera-
TZ ture rose from 5  C to 25  C, whereas the high temperature of 45  C
caused the low degradation efficiency. The recombination of
TM
X3
2
X4 2 electrons and holes was inhibited under higher temperatures,
which promoted photocatalytic degradation [48]. The dissolved
Dosage Loading Interaction of
oxygen content of the solution decreased with increasing
dosage and loading
temperature and the negative effect simultaneously occurred
[49]. In addition, the effect of temperature on the TZ composite is
obviously larger than that on the TM composite. The adsorption of
tetracyclines on zeolite is an endothermic process; slightly more
sorption was observed as the temperature increased [50,51].
Meanwhile, the photocatalytic degradation of TiO2 and zeolite
composites was a synergistic reaction of adsorption and photo-
catalysis [52]. Therefore, the excellent adsorption performance of
the TM composite relieved the impact of temperature on photo-
catalytic degradation.
The effect of photocatalyst dosage on the OTC photocatalytic
degradation is shown in Fig. 8(c). Excessive catalyst particles
increased the turbidity of the reaction system and the scattering
effect of light, that is, the shielding effect. Under this condition, less
activated catalyst particles produced less reactive species and
decreased the degradation efficiency [43,49]. The effect of dosage
on the TM composite is greater than that on the TZ composite. A
probable reason is that the increasing dosage enhanced the

Table 4
Statistical parameters of the regression equation.

Analysis of variances Model summary

Sum d.f.a Mean of square Fa Sig.a R R2 R2adja RSEa

of
square
TZ Regression 5179 8 647 46.5 0.021 0.997 0.995 0.973 3.731
Residual 28 2 14 – – – – – –
Total 5207 10 – – – – – – –
TM Regression 3976 8 497 20.3 0.048 0.994 0.988 0.939 4.947
Residual 49 2 24 – – – – – –
Total 4024 10 – – – – – – –
a
d.f.: degree of freedom, F: Fisher coefficien, Sig.: significance, R2adj: adjusted R2, RSE: residual standard deviation.
4460 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465

Table 5
Conditions of single factor experiments.

pH T ( C) Dosage (g L
1) TiO2 (%)

3, 5, 7, 9, 11a 25 0.10 10 Fig. 5(a)
2–12b
5 5, 15, 25, 35, 45a 0.10 10 Fig.5(b)
5–55b
a
5 25 0, 0.02, 0.05, 0.10, 0.20 10 Fig. 5(c)
0–0.20b
5 25 0.10 0, 2, 5,10, 20a Fig. 5(d)
0–20b
a
Points of single factor experiments.
b
Forecast range.

Fig. 8. Predicted and actual values of single factor experiment, factors: (a) the initial pH of OTC solution (X1), (b) reaction temperature (X2), (c) dosage of catalysts (X3), and (d)
TiO2 content in catalysts (X4).
 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465 4461

adsorption capacity, but the excessive smaller TM particles 105

produced a larger shielding effect. TiO2
The effect of TiO2 content in composites on OTC photocatalytic 10TZ

The removal efficiency of OTC (%)

degradation is shown in Fig. 8(d). ZSM-5 without TiO2 contributes 100 10TM
the same OTC removal as the absence of a photocatalyst (UV only),
thereby indicating that the ZSM-5 zeolite possesses poor photo-
catalytic function and adsorption capacity. MCM-41 has excellent 95
adsorption performance and promoted the photodegradation by
20%, which can be explained by the adsorption of OTC and its
intermediate products. For TZ and TM composites, the degradation 90
efficiency of both increased with the TiO2 content when the TiO2
content is below 10%. When the TiO2 content reached 20%, the
degradation efficiency had no distinct changes. As discussed above, 85
the pore size of ZSM-5 and MCM-41 is smaller than particle size of
TiO2. TiO2 particles located on the outer surface occupy adsorption
sites, and the agglomeration of TiO2 particles reduces utilization of 80
UV light [40,42]. Therefore, according to the single factor 1 2 3 4 5

experiment, 10% can be chosen as the optimal TiO2 content, which Recycling times
is close to the predicted optimal content of 15% and 12.5% for TZ
Fig. 10. Degradation efficiency of OTC in different recycling times.
and TM composites, respectively.
The comparison of predicted and experimental values is shown
in Fig. 9. The comparison points of uniform design are basically on
predicted line, whereas the single factor experimental points show
a certain bias. The result of correlation analysis was a significance
value below 0.01; the Pearson correlation coefficient of the TZ and
TM composites is 0.987 and 0.971, respectively. Therefore, the
regression equations obtained by uniform design can forecast 100

100
relatively accurate results for both TZ and TM composites. 80

Iμ (%) 60

40
3.5. Recycling of materials 80
20

0
As seen in Fig. 10, the degradation performance of the TZ 60 0 30 60 90 120 150
Iμ (%)

composite dropped by only 5% after recycling for five times, Time (min)

whereas that of the TM composite remained over 99% without any 40

apparent decline, thereby indicating that the reused composites
were stable. The decreased OTC degradation efficiency was 20
probably caused by catalyst surface contamination, increased UV
particle sizes after several heat treatment processes, and the 0
TZ
decreased specific surface area and adsorption activity [39,53]. TM
(a)
Compared with the composites, the photocatalytic effect of TiO2
was relatively obviously declined, thereby indicating that the 0 20 40 60 80 100
loading on zeolite improved the photocatalyst reusability. C/C 0 (%)

100

Uniform design - TZ
Single factor test - TZ 100

80 Uniform design - TM 100 80

Single factor test - TM
IX (%)

60
Predicted degradation (%)

40
80
20
60 0
60 0 30 60 90 120 150
IX (%)

Time (min)

40
40

20

UV
20 0 TZ
TM
(b)

0 20 40 60 80 100
0
0 20 40 60 80 100 C/C 0 (%)

Actual degradation (%) Fig. 11. Toxicity assessment by (a) the maximum specific growth rate and (b) the
biomass at stationary growth phase (insets: changes of inhibition rate with the
Fig. 9. Comparison of predicted values and actual experiment data. time).
4462 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465

Table 6
The intermediate products analyzed by HLPC-MS.

m/z Relative abundance (%) Possible molecular formula Possible molecular structure

10TZ 10TM
87 100 100 C4H6O2

C5H10O

111 50 40 C6H6O2

C7H10O

C8H14

127 74 29 C6H6O3

C7H10O2

C8H14O

165 22 23 C9H8O3

C10H12O2

C11H16O

179 10 16 C12H18O

C11H14O2
 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465 4463

Table 6 (Continued)
m/z Relative abundance (%) Possible molecular formula Possible molecular structure

10TZ 10TM
C10H10O3

225 15 18 C11H12O5

C12H16O4

461 0 0 C22 H24N2O9

3.6. Toxicity assessment
The biomass of the stationary growth phase (X) and the
maximum specific growth rate (m, h
1) can be used to evaluate the
growth of microorganisms and to assess the toxicity of additives
[54]. The OD value of E. coli DH5a at 12 h was considered as X, and
m was obtained according to Eq. (7):
d ln½OD 1 d½OD
m¼ ¼ ð7Þ
dt ½OD dt
where t is the culture time. On the curve of ln[OD] versus t, the
maximum slope is m. m and X of the initial OTC solution are defined
as m0 and X0, whereas those of the control group are mm and Xm,
correspondingly. The relative toxicity of the photodegradation
solution is expressed by the inhibition rates, Im and IX, as shown in
Eqs. (8) and (9).
mm
m
Im ð%Þ ¼  100 ð8Þ
mm
m0
Xm
X
IX ð%Þ ¼  100 ð9Þ
Xm
X0
During photolysis and two photocatalytic degradation process-
es, the toxicity appeared to gradually decline as shown in the insets
of Fig. 11. For both photocatalytic degradation processes, Im and IX
have a good linear relationship with the OTC degradation efficiency
as shown in Figs.11(a) and 8(b), thereby proving the agreement of
Im and IX with the toxicity assessment of photocatalytic degrada-
tion. For photolysis, the linear relationship of IX and the OTC
degradation efficiency is better than that of Im, which has a certain
positive bias, mainly in the first half of photolysis. Previous studies
suggested that the main photolysis products of tetracyclines still
contain the characteristic tetra-phenyl structure [44,55,56].
Furthermore, the loss and change of their substituent groups
causes lower steric resistance and easier penetration into the cells
of microorganisms [56,57], which subsequently leads to higher
inhibition of growth rate, as shown by positive bias in Fig. 11(a).
However, after 12 h of culture and proliferation, the ability to retain
antibacterial activity and the potential long-term effects on
antibiotic resistance limit the inhibition rate of photoproducts
[58], as shown by the good liner relationship between IX and
degradation efficiency in Fig. 11(b). By comparing the two methods
for toxicity assessment, IX can be calculated from less data and the
error is relatively smaller, whereas Im provides more information
on the acute toxicity. Intermediate products were detected using a
HPLC–MS system and listed in Table 6. Only intermediate products
of low molecular weight were detected (Fig. S8), in degradation
solution using both 10TZ and 10TM composites. It proved that four-
membered ring structure was destroyed during degradation using
both 10TZ and 10TM composites. It is consistent with the results of
toxicity assessment, which indicated no toxic intermediates.
4. Conclusion
TZ and TM composites were synthesized to remove OTC in
aqueous solutions. Characterization showed that TiO2 was loaded
on the outer surface of ZSM-5 and MCM-41, which decreased the
specific surface area, without the change in crystal form of any raw
materials. The adsorption of OTC showed that mesoporous MCM-
41 had a higher adsorption capacity than microporous ZSM-5, and
the adsorption property minimally changed after TiO2 loading.
Among the four factors investigated by uniform design, the
reaction temperature greatly influenced the TZ composites,
whereas the dosage of catalyst decided the performance of TM
composites. The interaction between dosage and TiO2 content,
namely, the TiO2 dosage, was a significant factor that cannot be
neglected. The single factor experiment confirmed that the
4464 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465
experimental values agreed with the predictions by the regression
formula of uniform design. After recycling and reuse for five times,
the remaining degradation efficiency was in the order of TM >
TZ > TiO2. The adsorption and photocatalytic degradation experi-
ments both indicated that MCM-41 was more suitable than ZSM-5
for loading TiO2 to remove OTC because it had better absorption
properties. For both photocatalytic degradation processes of the TZ
and TM composites, the toxicity decreased with the OTC
degradation efficiency in a good linear relationship.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jece.2016.10.012.
References
[1] M.O. Elema, K.A. Hoff, H.G. Kristensen, Bioavailability of oxytetracycline from
medicated feed administered to Atlantic salmon (Salmo salar L.) in seawater,
Aquaculture 143 (1996) 7–14.
[2] P. Nielsen, N. Gyrd-Hansen, Bioavailability of oxytetracycline, tetracycline and
chlortetracycline after oral administration to fed and fasted pigs, J. Vet.
Pharmacol. Ther. 19 (1996) 305–311.
[3] Z. Lin, M. Li, R. Gehring, J.E. Riviere, Development and application of a
multiroute physiologically based pharmacokinetic model for oxytetracycline
in dogs and humans, J. Pharm. Sci. 104 (2015) 233–243.
[4] K. Kümmerer, Antibiotics in the aquatic environment – a review – part II,
Chemosphere 75 (2009) 435–441.
[5] I. Michael, L. Rizzo, C.S. McArdell, C.M. Manaia, C. Merlin, T. Schwartz, C. Dagot,
D. Fatta-Kassinos, Urban wastewater treatment plants as hotspots for the
release of antibiotics in the environment: a review, Water Res. 47 (2013)
957–995.
[6] P. Gao, M. Munir, I. Xagoraraki, Correlation of tetracycline and sulfonamide
antibiotics with corresponding resistance genes and resistant bacteria in a
conventional municipal wastewater treatment plant, Sci. Total Environ.
421–422 (2012) 173–183.
[7] J. Xu, Y. Xu, H. Wang, C. Guo, H. Qiu, Y. He, Y. Zhang, X. Li, W. Meng, Occurrence
of antibiotics and antibiotic resistance genes in a sewage treatment plant and
its effluent-receiving river, Chemosphere 119 (2015) 1379–1385.
[8] A.Y. Tong, R. Braund, D.S. Warren, B.M. Peake, TiO2-assisted photodegradation
of pharmaceuticals-a review, Cent. Eur. J. Chem. 10 (2012) 989–1027.
[9] Z. Clemente, V. Castro, C. Jonsson, L. Fraceto, Ecotoxicology of nano-TiO2-an
evaluation of its toxicity to organisms of aquatic ecosystems, Int. J. Environ.
Res. 6 (2011) 33–50.
[10] F. Gottschalk, T. Sonderer, R.W. Scholz, B. Nowack, Modeled environmental
concentrations of engineered nanomaterials (TiO2, ZnO Ag, CNT, fullerenes)
for different regions, Environ. Sci. Technol. 43 (2009) 9216–9222.
[11] D. Kanakaraju, J. Kockler, C.A. Motti, B.D. Glass, M. Oelgemöller, Titanium
dioxide/zeolite integrated photocatalytic adsorbents for the degradation of
amoxicillin, Appl. Catal. B Environ. 166–167 (2015) 45–55.
[12] M. Ito, S. Fukahori, T. Fujiwara, Adsorptive removal and photocatalytic
decomposition of sulfamethazine in secondary effluent using TiO2-zeolite
composites, Environ. Sci. Pollut. Res. 21 (2014) 834–842.
[13] G. Williams, B. Seger, P.V. Kamat, TiO2-graphene nanocomposites. UV-assisted
photocatalytic reduction of graphene oxide, ACS nano 2 (2008) 1487–1491.
[14] S. Salaeh, D. Juretic Perisic, M. Biosic, H. Kusic, S. Babic, U. Lavrencic Stangar, D.
D. Dionysiou, A. Loncaric Bozic, Diclofenac removal by simulated solar assisted
photocatalysis using TiO2-based zeolite catalyst; mechanisms, pathways and
environmental aspects, Chem. Eng. J. 304 (2016) 289–302.
[15] J.H. Sun, Y.K. Wang, R.X. Sun, S.Y. Dong, Photodegradation of azo dye Congo Red
from aqueous solution by the WO3-TiO2/activated carbon (AC) photocatalyst
under the UV irradiation, Mater. Chem. Phys. 115 (2009) 303–308.
[16] J. Dasgupta, A. Singh, S. Kumar, J. Sikder, S. Chakraborty, S. Curcio, H.A. Arafat,
Poly (sodium-4-styrenesulfonate) assisted ultrafiltration for methylene blue
dye removal from simulated wastewater: optimization using response surface
methodology, J. Environ. Chem. Eng. 4 (2016) 2008–2022.
[17] R. Daghrir, P. Drogui, A. Dimboukou-Mpira, M.A. El Khakani,
Photoelectrocatalytic degradation of carbamazepine using Ti/TiO2
nanostructured electrodes deposited by means of a pulsed laser deposition
process, Chemosphere 93 (2013) 2756–2766.
[18] R. Daghrir, P. Drogui, I. Ka, M.A. El Khakani, Photoelectrocatalytic degradation
of chlortetracycline using Ti/TiO2 nanostructured electrodes deposited by
means of a Pulsed Laser Deposition process, J. Hazard. Mater. 199–200 (2012)
15–24.
[19] Y. Abdollahi, A. Zakaria, K.A. Matori, K. Shameli, H. Jahangirian, M. Rezayi, T.
Abdollahi, Interactions between photodegradation components, Chem. Cent. J.
6 (2012) 100.
[20] D. Lutic, I. Cretescu, Optimization study of rhodamine 6G removal from
aqueous solutions by photocatalytic oxidation, Rev. Chim. 67 (2016) 134–138.
[21] M. Kovacic, S. Salaeh, H. Kusic, A. Suligoj, M. Kete, M. Fanetti, U.L. Stangar, D.D.
Dionysiou, A.L. Bozic, Solar-driven photocatalytic treatment of diclofenac
using immobilized TiO2-based zeolite composites, Environ. Sci. Pollut. Res. Int.
23 (2016) 17982–17994.
[22] K.T. Fang, D.K. Lin, P. Winker, Y. Zhang, Uniform design: theory and application,
Technometrics 42 (2000) 237–248.
[23] W. Yan, H.L. Wei, H.Q. Fei, G.X. Li, Heat transfer and friction characteristics of
rotor-assembled strand heat exchanger studied by uniform design
experiment, Adv. Mech. Eng. 7 (2015).
[24] S. Ho, Y.L. Lee, H.T. Kang, C.J. Wang, Optimization of a crankshaft rolling process
for durability, Int. J. Fatigue 31 (2009) 799–808.
[25] X.T. Liao, Q. Li, X.J. Yang, W. Li, W.G. Zhang, A two-stage multi-objective
optimisation of vehicle crashworthiness under frontal impact, Int. J.
Crashworthiness 13 (2008) 279–288.
[26] Y.B. Ji, Q.S. Xu, Y.Z. Hu, Y.V. Heyden, Development, optimization and validation
of a fingerprint of Ginkgo biloba extracts by high-performance liquid
chromatography, J. Chromatogr. A 1066 (2005) 97–104.
[27] Y. Huang, H. Lin, R. Song, Y. Tian, Z.J. Zhang, Optimization of dispersive liquid–
liquid microextraction for analysis of levonorgestrel in water samples using
uniform design, Anal. Methods 3 (2011) 857–864.
[28] B. Tang, S. Zhang, X. Zhou, D. Wang, Y. Yuan, Regression analysis for complex
doping of X8R ceramics based on uniform design, J. Electron. Mater. 36 (2007)
1383–1388.
[29] S.L. Zhang, Z.X. Zhang, Z.X. Xin, K. Pal, J.K. Kim, Prediction of mechanical
properties of polypropylene/waste ground rubber tire powder treated by
bitumen composites via uniform design and artificial neural networks, Mater.
Des. 31 (2010) 1900–1905.
[30] S.S. Liu, X.Q. Song, H.L. Liu, Y.H. Zhang, J. Zhang, Combined photobacterium
toxicity of herbicide mixtures containing one insecticide, Chemosphere 75
(2009) 381–388.
[31] J. Zhang, S.S. Liu, R.N. Dou, H.L. Liu, J. Zhang, Evaluation on the toxicity of ionic
liquid mixture with antagonism and synergism to Vibrio qinghaiensis sp.-Q67,
Chemosphere 82 (2011) 1024–1029.
[32] H.L. Ge, S.S. Liu, B.X. Su, L.T. Qin, Predicting synergistic toxicity of heavy metals
and ionic liquids on photobacterium Q67, J. Hazard. Mater. 268 (2014) 77–83.
[33] G.B. Hong, R.T. Ruan, C.T. Chang, MCM-41 from spent glasses for volatile
organic compounds treatment, Chem. Eng. J. 215–216 (2013) 472–478.
[34] V. Durgakumari, M. Subrahmanyam, K.V. Subba Rao, A. Ratnamala, M.
Noorjahan, K. Tanaka, An easy and efficient use of TiO2 supported HZSM-5 and
TiO2 + HZSM-5 zeolite combinate in the photodegradation of aqueous phenol
and p-chlorophenol, Appl. Catal. A Gen. 234 (2002) 155–165.
[35] S. Blasioli, A. Martucci, G. Paul, L. Gigli, M. Cossi, C.T. Johnston, L. Marchese, I.
Braschi, Removal of sulfamethoxazole sulfonamide antibiotic from water by
high silica zeolites: a study of the involved host-guest interactions by a
combined structural, spectroscopic, and computational approach, J. Colloid
Interface Sci. 419 (2014) 148–159.
[36] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T.
Siemieniewska, Reporting physisorption data for gas/solid systems, Handbook
of Heterogeneous Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA, 2008.
[37] M.D. Donohue, G.L. Aranovich, Adsorption hysteresis in porous solids, J. Colloid
Interface Sci. 205 (1998) 121–130.
[38] K.T. Fang, R. Li, A. Sudjianto, Design and Modeling for Computer Experiments,
1st ed., CRC Press, 2005.
[39] C. Zhao, Y. Zhou, D.J. d. Ridder, J. Zhai, Y. Wei, H. Deng, Advantages of TiO2/5A
composite catalyst for photocatalytic degradation of antibiotic oxytetracycline
in aqueous solution: comparison between TiO2 and TiO2/5A composite
system, Chem. Eng. J. 248 (2014) 280–289.
[40] C. Zhao, H. Deng, Y. Li, Z. Liu, Photodegradation of oxytetracycline in aqueous
by 5A and 13X loaded with TiO2 under UV irradiation, J. Hazard. Mater. 176
(2010) 884–892.
[41] J.H.O.S. Pereira, A.C. Reis, D. Queirós, O.C. Nunes, M.T. Borges, V.J.P. Vilar, R.A.R.
Boaventura, Insights into solar TiO2-assisted photocatalytic oxidation of two
antibiotics employed in aquatic animal production oxolinic acid and
oxytetracycline, Sci. Total Environ. 463–464 (2013) 274–283.
[42] R. Li, Y. Jia, J. Wu, Q. Zhen, Photocatalytic degradation and pathway of
oxytetracycline in aqueous solution by Fe2O3-TiO2 nanopowder, RSC Adv. 5
(2015) 40764–40771.
[43] A. Nezamzadeh-Ejhieh, A. Shirzadi, Enhancement of the photocatalytic
activity of Ferrous Oxide by doping onto the nano-clinoptilolite particles
towards photodegradation of tetracycline, Chemosphere 107 (2014) 136–144.
[44] V.M. Mboula, V. Hequet, Y. Gru, R. Colin, Y. Andres, Assessment of the efficiency
of photocatalysis on tetracycline biodegradation, J. Hazard. Mater. 209 (2012)
355–364.
[45] R. Daghrir, P. Drogui, M. El Khakani, Photoelectrocatalytic oxidation of
chlortetracycline using Ti/TiO2 photo-anode with simultaneous H2O2
production, Electrochim. Acta 87 (2013) 18–31.
[46] C.V. Gómez-Pacheco, M. Sánchez-Polo, J. Rivera-Utrilla, J.J. López-Peñalver,
Tetracycline degradation in aqueous phase by ultraviolet radiation, Chem. Eng.
J. 187 (2012) 89–95.
[47] K. Li, A. Yediler, M. Yang, S. Schulte-Hostede, M.H. Wong, Ozonation of
oxytetracycline and toxicological assessment of its oxidation by-products,
Chemosphere 72 (2008) 473–478.
[48] C.C. Hsueh, B.Y. Chen, Comparative study on reaction selectivity of azo dye
decolorization by Pseudomonas luteola, J. Hazard. Mater. 141 (2007) 842–849.
[49] A. Nezamzadeh-Ejhieh, M. Khorsandi, Heterogeneous photodecolorization of
eriochrome black t using Ni/P zeolite catalyst, Desalination 262 (2010) 79–85.
[50] J. Kang, H.J. Liu, Y.M. Zheng, J.H. Qu, J.P. Chen, Application of nuclear magnetic
resonance spectroscopy, Fourier transform infrared spectroscopy UV-Visible
 X.-D. Xie et al. / Journal of Environmental Chemical Engineering 4 (2016) 4453–4465
spectroscopy and kinetic modeling for elucidation of adsorption chemistry in
uptake of tetracycline by zeolite beta, J. Colloid Interface Sci. 354 (2011)
261–267.
[51] M. Liu, L.A. Hou, S. Yu, B. Xi, Y. Zhao, X. Xia, MCM-41 impregnated with A zeolite
precursor: synthesis, characterization and tetracycline antibiotics removal
from aqueous solution, Chem. Eng. J. 223 (2013) 678–687.
[52] S. Fukahori, T. Fujiwara, Modeling of sulfonamide antibiotic removal by TiO2/
high-silica zeolite HSZ-385 composite, J. Hazard. Mater. 272 (2014) 1–9.
[53] M.V.P. Sharma, K. Lalitha, V. Durgakumari, M. Subrahmanyam, Solar
photocatalytic mineralization of isoproturon over TiO2/HY composite
systems, Sol. Energy Mater. Sol. Cells 92 (2008) 332–342.
[54] B.Y. Chen, H.L. Liu, Y.W. Chen, Y.C. Cheng, Dose-response assessment of metal
toxicity upon indigenous Thiobacillus thiooxidans BC1, Process Biochem. 39
(2004) 737–748.
4465
[55] S. Jiao, S. Zheng, D. Yin, L. Wang, L. Chen, Aqueous oxytetracycline degradation
and the toxicity change of degradation compounds in photoirradiation
process, J. Environ. Sci. 20 (2008) 806–813.
[56] F. Yuan, C. Hu, X. Hu, D. Wei, Y. Chen, J. Qu, Photodegradation and toxicity
changes of antibiotics in UV and UV/H2O2 process, J. Hazard. Mater. 185 (2011)
1256–1263.
[57] G. Lu, Y. Zhao, S. Yang, X. Cheng, Quantitative structure–biodegradability
relationships of substituted benzenes and their biodegradability in river water,
Bull. Environ. Contam. Toxicol. 69 (2002) 111–116.
[58] K.H. Wammer, M.T. Slattery, A.M. Stemig, J.L. Ditty, Tetracycline photolysis in
natural waters: loss of antibacterial activity, Chemosphere 85 (2011)
1505–1510.