Zeolite 4A supported CdS/g-C3N4 type-II heterojunction: A novel visible-light-active ternary nanocomposite for potential photocatalytic degradation of cefoperazone

Article abstract of DOI:10.1016/j.molliq.2021.117479

The CdS/g-C₃N₄ heterojunction photocatalyst supported on 4A zeolite was successfully synthesized using a simple chemical precipitation method. The physicochemical characteristics of the as-prepared ternary composite were assessed using X-Ray diffraction (XRD), field emission- scanning electron microscopy (FE-SEM), energy dispersive X-Ray (EDX), transmission electron microscopy (TEM), N₂ adsorption–desorption, differential reflectance spectroscopy (UV–Vis-DRS), and photoluminescence (PL) techniques. The results confirmed the successful synthesis of the CdS/g-C₃N₄/4AZ nanocomposite and introduction of the CdS and g-C₃N₄ on the substrate of 4A zeolite. Cefoperazone (CFP) antibiotic was tested as the model pollutant to assess the photocatalytic performance of the synthesized nanocomposite under visible light irradiation. The response surface methodology (RSM) and artificial neural network (ANN) showed desirable reasonability for the prediction of the CFP degradation efficiency. More than 93% of CFP with a concentration of 17 mg L^-1 degraded in the presence of the 0.4 g L^-1 of the catalyst at pH of 9 after 80 min treatment time (RSM-based optimization results). The pH of the solution, irradiation time, catalyst dosage, and the initial concentration of the CFP affected degradation efficiency with a percentage impact of 37, 29, 19, and 15 %, respectively (ANN-based modeling results). The addition of 1 mM of isopropanol, benzoquinone, and sodium oxalate reduced the CFP degradation efficiency from 93.23% to 85.18, 41.16, and 32.47%, respectively, proving the decisive role of the °O₂^– and h⁺ in the photodegradation process. The kinetic studies indicated the following of the process from the Langmuir-Hinshelwood's pseudo-first-order model (k_app = 3.71 × 10^-2 min^?1). The structure of the identified by-products using GC-MS analysis confirmed that CFP mainly decomposed through the cleavage of C-S, C-N, and N-N bonds. Moreover, the formation of the aliphatic compounds and carboxylic acids as by-products confirmed nearly complete mineralization of the CFP to non-toxic products.

Full text of DOI:10.1016/j.molliq.2021.117479

Journal of Molecular Liquids 342 (2021) 117479
Contents lists available at ScienceDirect
Journal of Molecular Liquids
journal homepage: www.elsevier.com/locate/molliq
Zeolite 4A supported CdS/g-C3N4 type-II heterojunction: A novel
visible-light-active ternary nanocomposite for potential photocatalytic
degradation of cefoperazone
b
b
Naime AttariKhasraghi a, Karim Zare a, Ali Mehrizad b, , Nasser Modirshahla , Mohammad A. Behnajady
⇑
a Department of Chemistry, Science and Research Branch, Islamic Azad University, Tehran, Iran
b Department of Chemistry, Tabriz Branch, Islamic Azad University, Tabriz, Iran
a r t i c l e i n f o
a b s t r a c t
Article history:
The CdS/g-C3N4 heterojunction photocatalyst supported on 4A zeolite was successfully synthesized using
a simple chemical precipitation method. The physicochemical characteristics of the as-prepared ternary
composite were assessed using X-Ray diffraction (XRD), field emission- scanning electron microscopy
(FE-SEM), energy dispersive X-Ray (EDX), transmission electron microscopy (TEM), N2 adsorption–des-
orption, differential reflectance spectroscopy (UV–Vis-DRS), and photoluminescence (PL) techniques.
The results confirmed the successful synthesis of the CdS/g-C3N4/4AZ nanocomposite and introduction
of the CdS and g-C3N4 on the substrate of 4A zeolite. Cefoperazone (CFP) antibiotic was tested as the
model pollutant to assess the photocatalytic performance of the synthesized nanocomposite under visi-
ble light irradiation. The response surface methodology (RSM) and artificial neural network (ANN)
showed desirable reasonability for the prediction of the CFP degradation efficiency. More than 93% of
CFP with a concentration of 17 mg L-1 degraded in the presence of the 0.4 g L-1 of the catalyst at pH of
9 after 80 min treatment time (RSM-based optimization results). The pH of the solution, irradiation time,
catalyst dosage, and the initial concentration of the CFP affected degradation efficiency with a percentage
impact of 37, 29, 19, and 15 %, respectively (ANN-based modeling results). The addition of 1 mM of iso-
propanol, benzoquinone, and sodium oxalate reduced the CFP degradation efficiency from 93.23% to
85.18, 41.16, and 32.47%, respectively, proving the decisive role of the °O–2 and h+ in the photodegradation
process. The kinetic studies indicated the following of the process from the Langmuir-Hinshelwood’s
pseudo-first-order model (kapp = 3.71 ꢀ 10-2 minꢁ1). The structure of the identified by-products using
GC-MS analysis confirmed that CFP mainly decomposed through the cleavage of C-S, C-N, and N-N bonds.
Moreover, the formation of the aliphatic compounds and carboxylic acids as by-products confirmed
nearly complete mineralization of the CFP to non-toxic products.
Received 4 August 2021
Revised 31 August 2021
Accepted 3 September 2021
Available online 8 September 2021
Keywords:
CdS/g-C3N4/4A zeolite
Cefoperazone
Photocatalytic process
RSM
ANN
Visible-light
Ó 2021 Elsevier B.V. All rights reserved.
1. Introduction
classical degradation techniques such as adsorption and biodegra-
dation. Recent studies on the development of the effective purifica-
Recently, the unsafe direct discharge of the pharmaceutical-
containing wastewaters compounds attracted global concerns.
Wastewaters containing pharmaceutical effluents from domestic,
hospitals, and pharmaceuticals industries lead to serious damages
to physicochemical characteristics and also microorganisms of the
receptor water [1–3]. Cephalosporins are the frequently used
antibiotics to treat bacteria-caused infections. The extensive usage
of these b-lactam antibiotics may cause severe negative impacts on
flora and fauna such as increased bacteria resistance or disturbance
on the photosynthesis [4,5]. These organic effluents are resistant to
tion techniques come an end to great attracts to advanced
oxidation processes (AOPs) as the favorable techniques to solve
the environmental problems. The non-selective decomposition of
the wide diversity of the organic molecules using produced power-
ful radicals is the basic principle of the AOPs. Photocatalysis is one
of the most powerful AOPs with a high mineralization degree,
which utilized a semiconductor as the catalyst under ultraviolet
or visible light irradiation. In this regard, wide ranges of the semi-
conductors such as metal oxides, metal sulfides, and their compos-
ites have been applied as the catalysts [6–9]. The nanostructured
metal sulfides such as CdS [10–13], ZnS [14], CuS [15], FeS2 [16],
and their composites are desired photocatalysts owing to the high
absorption coefficient and suitable band gap energy. CdS acts as
⇑
Corresponding author.
E-mail address: mehrizad@iaut.ac.ir (A. Mehrizad).
https://doi.org/10.1016/j.molliq.2021.117479
0167-7322/Ó 2021 Elsevier B.V. All rights reserved.

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
catalyst in visible-light region with a narrow band gap (ꢂ2.1–
2.4 eV), low toxicity, fast charge transfer, high chemical and ther-
mal durability. Unfortunately, the widespread application of the
CdS is restricted because of the high recombination rate of the gen-
erated charge carriers [17]. Coupling with other semiconductors,
doping, and surface modification are some of the techniques used
to tackle this problem [18–22]. Graphitic carbon nitride (g-C3N4)
with high availability, thermal and chemical stability as well as
appropriate band gap (ꢂ2.7 eV) is one of the suitable candidates
to form heterojunctions with other narrow band gap semiconduc-
tors such as CdS nanostructures to lowering the charge carrier’s
recombination and intensifying the photocatalytic performance
[23–25]. Coupling of g-C3N4 nanosheets with CdS resulting in the
formation of the type-II or Z-scheme heterojunctions with acceler-
ated charge transfer rate and decreased charge recombination rate
[4,10,26,27]. The limited accessibility to catalyst active sites of the
nanostructured semiconductors due to the accumulation of the
nanocatalyst particles is another crucial issue that should be
addressed. In this regard, stabilizing of the nanostructured cata-
lysts on the suitable substrates with ordered and porous structure
possess valuable characteristics in catalytic activities [3,28]. The
fundamental of the synthesis of the photocatalyst semiconductors
within the pores of the zeolites has been investigated by Herron
et al. [29]. Zeolite 4A with the chemical formula of Na12 [AlO2.
SiO2]12ꢃ27H2O, high porosity, thermal, mechanical, and chemical
stability provides suitable substrates for the embedding of the cat-
alyst particles [30]. Based on our knowledge, the embedding of the
CdS/g-C3N4 photocatalyst on the substrate of the 4A zeolite and its
application as photocatalyst has not been reported yet. Different
approaches have been reported in order to synthesis of the CdS/
g-C3N4 composite, including solvothermal [31], ultrasound-
assisted technique [32], and hydrothermal [33]. Most of these tech-
niques are energy and time-consuming. Co-precipitation technique
provides a simple and low-cost procedure to preparation of the
wide range of the nanomaterials.
The large-scale application of the various treatment techniques
required detailed information about the interactions of the opera-
tional variables to reach the highest performance. Evaluating the
impact of each influential parameter with the classical one-
factor-at-a-time method is very time-consuming, and even the
interaction of the variables cannot be described in this method.
Applying the developed optimization and modeling methods such
as response surface methodology (RSM) and artificial neural net-
work (ANN) can effectively decrease the number of tests [3,34].
Based on the above mentioned the most important research
topics of this research study can be described as follows: (1)
increasing the photocatalytic performance of the CdS by coupling
with g-C3N4 nanosheets and stabilizing on the substrate of the zeo-
lite 4A; (2) potential application of the synthesized nanocomposite
for photodecomposition of the cefoperazone (CFP) antibiotic as the
representative of cephalosporins; (3) optimization and modeling of
the degradation process of the CFP using RSM and ANN methods
and comparison of the two used approaches; (4) investigation
the impact of the scavengers on the produced radicals and
electron-hole pairs; (5) kinetic studies of the degradation process;
and (6) identification of by-products from photocatalytic degrada-
tion of CFP using GC-MS analysis.
was purchased from Wellona Pharma Co. (India). All the experi-
ments were carried out using doubled-distilled water.
2.2. Synthesis of CdS/g-C3N4/4AZ ternary composite
The g-C3N4 nanosheets were prepared by thermal polymeriza-
tion of 5 g of melamine at 520 °C for 4 h. A two-step synthesis tech-
nique was applied to the preparation of the CdS/g-C3N4/4AZ
composite. In the first step, 1 g of CdCl2ꢃH2O was dissolved in
50 mL distilled water in a 250 mL flask under vigorous stirring.
Then 0.08 g of synthesized g-C3N4 powder homogeneously dis-
persed in 20 mL ethanol under ultrasonic irradiation (ultrasonic
bath (Ultra 8060, England) with a frequency of 36 kHz, volumetric
capacity of 3 L and output intensity of 150 W) and added to the
above solution. Then 50 mL of sodium sulfide solution with con-
centration of 0.1 M was added dropwise to the mixture and stirred
for 1 h. The as-obtained product was centrifuged for 10 min at
5000 rpm followed by washing with ethanol and distilled water.
In the next step, 1 g of 4A zeolite was dispersed in 100 mL distilled
water and sonicated for 30 min. The resultant powder from the
first step was added to the zeolite solution and stirred for another
2 h. Finally, the as-prepared CdS/g-C3N4/4AZ composite was dried
and calcinated at inert atmosphere for 1 h at 500 °C in a furnace.
2.3. Characterization of the CdS/g-C3N4/4AZ composite
The XRD (X-ray diffraction) analysis was applied to observe the
crystalline structure of the as-synthesized CdS/g-C3N4/4AZ com-
posite applying a PHILIPS (PW1730, Nederland) with Cu-ka radia-
tion (k = 1.54 A°, 40 kV, 30 mA). The FE-SEM (scanning electron
microscopy), and TEM (transmission electron microscopy) analysis
were used to study the morphology of the prepared nanocompos-
ite using TESCAN, MIRA III and LEO 906 E (100 kV), respectively.
The chemical composition and surface properties of the prepared
nanocomposite were evaluated using EDX (Energy Dispersive X-
Ray) and BET (Brunauer, Emmett and Teller) analysis using a TES-
CAN, MIRA III equipped with SAMX detector and BELSORP MINI II,
respectively. The photoluminescence (PL) spectra of the as-
synthesized nanocomposite were recorded using Shimadzu spec-
trofluorometer (CARY ECLIPSE/Vaian) with the excitation wave-
length of 325 nm.
2.4. Assessment the photocatalytic performance of the CdS/g-C3N4/4AZ
composite
The aqueous solution of cefoperazone (CFP) was chosen as the
model effluent to investigate the photocatalytic performance of
the synthesized CdS/g-C3N4/4AZ nanocomposite under visible light
irradiation. A 300 W halogen lamp (OSRAM, Germany) was
selected as a visible light source with a glass optical filter for cut
off the components with k > 420 nm. In each experiment, the
desired amount of the as-synthesized ternary composite was
added to the 100 mL of the CFP solution with a specified concentra-
tion and pH value. Before starting each run, to establish the adsorp-
tion/desorption equilibrium, the antibiotic-photocatalyst mixture
was magnetically stirred in the dark for 60 min. The pH of the solu-
tion was adjusted using NaOH and HCl solutions with a concentra-
tion of 0.1 M. At specific time intervals, 5 mL of the solution was
extracted from photo-reactor and the residual amount of CFP
antibiotic was measured spectrophotometrically using a single
beam spectrophotometer (Shimadzu UV- Mini-1240) at kmax = 229-
nm. Eq. (1) was used to calculate the degradation percentages of
the CFP.
2. Material and methods
2.1. Chemicals
All the used analytical grade chemicals were provided by Merck
Co. (Germany) and used without any purification. Cefoperazone
sodium (CFP) with the chemical structure illustrated in Fig. S1
½CFPꢄ0 ꢁ ½CFPꢄt
Degradationð%Þ ¼
ꢀ 100
ð1Þ
½CFPꢄ0
2

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
In this equation, [CFP]0 and [CFP]t represent the initial and
residual concentration of the CFP in the solution (mg L-1),
respectively.
To probe the degradation mechanism of the CFP in the photo-
catalytic process, the effect of the addition of the three different
scavengers, including isopropanol (IPA), benzoquinone (BQ), and
sodium oxalate (NaOx) was investigated on the degradation
performance.
An Agilent 6890 gas chromatograph with a 30 m to 0.25 mm
HP-5MS capillary column along with an Agilent 5973 mass spec-
trometer (Canada) was applied to determine the produced inter-
mediates from photodecomposition of CFP.
2.5. Optimization and modeling of the process
The response surface methodology (RSM) and artificial neural
network (ANN) were used for the optimization and modeling of
the involved photocatalytic process whose details of these
approaches are given in the Supplementary Text 1.
Fig. 1. The XRD patterns of the (a) CdS, (b) g-C3N4, and (c) CdS/g-C3N4/4AZ.
as 21.15 and 18.56 nm, confirming the decrease in the size of the
CdS in the presence of the 4A zeolite.
2.5.1. Central composite design
To decrease the number of experiments, investigate the impact
of the influential parameters, and optimization of the process, a
series of the experiments were designed applying Design-
Expert11 (DX11) software based on the central composite design
(CCD, as a subset of RSM). Table 1 listed the domain of the exper-
imental parameters.
Fig. 2(a) shows the morphology of the as-synthesized CdS/g-
C3N4/4AZ composite. The spherical-shaped CdS particles and also
g-C3N4 nanosheets can be observed in FE-SEM images. Further-
more, the successful introduction of the CdS nanoparticles and g-
C3N4 sheets on the surface of the 4A zeolite can be seen. The
SEM image of the commercial 4A zeolite was also shown in Fig. 2
(b), showing a typical cubic morphology with homogeneous size
distribution. Using the SEM images (Fig. 2a and S2) and Digimizer
software the mean particle size of the CdS in the pure CdS, CdS/g-
C3N4, and CdS/g-C3N4/4AZ composite was obtained as 59.81, 53.23,
and 48.30 nm. These findings confirmed that in the presence of the
g-C3N4 and 4A zeolite the number of the CdS nucleation site
decreased resulted in declined particle diameter. The chemical
composition of the prepared CdS/g-C3N4/4AZ ternary composite
was observed by EDX analysis, as shown in Fig. 2(c). As can be seen,
the elemental composition of the synthesized composite contains
all the Cd, S, C, N, O, Na, Si, and Al atoms, confirming the successful
formation of the CdS/g-C3N4/4AZ composite.
2.5.2. Artificial neural network
Ranking of the operational parameters influence on the degra-
dation efficiency of the CFP was performed employing MATLAB
R2020 b software based on ANN toolbox. Similar to the CCD
method, the catalyst dosage, initial CFP concentration, pH, and
time were the input variables of the ANN model and the CFP degra-
dation efficiency (%) was the output of the model.
3. Results and discussion
3.1. Characterization of the CdS/g-C3N4/4AZ nanocomposite
The TEM analysis was conducted to better investigate the mor-
phology of the as-prepared CdS/g-C3N4/4AZ composite, as exhib-
ited in Fig. 3. The results of the TEM analysis confirmed the
successful synthesis of the nanocomposite below 100 nm. Further-
more, the presence of the spherical-shaped CdS particles and gC3N4
sheets on the surface of the 4A zeolite cubic particles was also
confirmed.
The porosity and pore structure of the CdS/g-C3N4 and CdS/g-
C3N4/4AZ were investigated by N2 adsorption–desorption iso-
therms and the results was illustrated in Fig. 4. The as-
synthesized nanocomposites showed a type III isotherm (IUPAC),
confirming the low adsorption of the N2 (22.17 and 44.78 cm3
gꢁ1 for CdS/g-C3N4 and CdS/g-C3N4/4AZ, respectively) on the sur-
face of the catalyst [35]. The specific surface area of the CdS/g-
C3N4 and CdS/g-C3N4/4AZ nanocomposites (SBET) was obtained as
0.92 and 6.43 m2 gꢁ1, respectively, indicating the increase in avail-
The crystallographic structure of the as-prepared CdS/g-
C3N4/4AZ nanocomposite was studied by XRD analysis as depicted
in Fig. 1. The peaks related to pure CdS and g-C3N4 were also shown
in Fig. 1. As can be seen, the successful formation of the CdS/g-
C3N4/4AZ was indicated by the results. The peaks related to hexag-
onal CdS (JCPDS Card No. 41-1049) and g-C3N4 nanosheets (JCPDS
Card No. 87-1526) can be observed in the XRD pattern of the CdS/
g-C3N4/4AZ with lower intensity, confirming the presence of both
phase in the composite [4]. The diffraction peaks located at about
18.60° (420), 30.75° (642), and 32.85° (840) are related to the
presence of the 4A zeolite (JCPDS Card No. 43-0142) in the struc-
ture of the nanocomposite [30]. The observed decrease in the
intensity of the CdS and g-C3N4 peaks is related to their interac-
tions with 4A zeolite. The crystallite size of the CdS in pure CdS
and composite structure using the Scherrer equation was obtained
Table 1
The domain of experimental parameters.
Parameters
Real and coded values of levels
ꢁ2
0.1
5
ꢁ1
0.2
10
5
0
+1
+2
A-[Catalyst]0 (g L-1
)
0.3
15
7
0.4
20
9
0.5
25
11
B-[CFP]0 (mg L-1
C-pH
)
3
D-Time (min)
20
40
60
80
100
3

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Fig. 2. The FE-SEM images of (a) CdS/g-C3N4/4AZ and (b) 4A zeolite; (c) the EDX spectrum of the CdS/g-C3N4/4AZ.
Fig. 3. The TEM analysis of the CdS/g-C3N4/4AZ nanocomposite.
able active surface sites of the ternary composite with embedding
of the CdS/g-C3N4 on the substrate of the 4A zeolite.
The photocatalytic property of the CdS/g-C3N4/4AZ nanocom-
posite was investigated by the band-gap energy (Eg), which was
determined by UV–Vis spectrophotometer, as shown in Fig. 5.
According to Fig. 5, the Tauc-Mott’s equation was applied to calcu-
late the Eg of the catalyst using extrapolating of the linear parts of
Fig. 4. The N2 adsorption-desorption isotherms of the CdS/g-C3N4 and CdS/g-C3N4/
4AZ.
the (Ah
m
)2 versus h
m
according to (Ah
m
)2 = B (h
m-Eg), where A is the
tional constant, and Eg is the band gap energy (eV) [36]. The Eg
value for the CdS/g-C3N4/4AZ photocatalyst was achieved as
absorbance (a.u.), h
m
is the photon energy (eV), B is the propor-
4

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Fig. 5. The (a) absorption curve and (b) band-gap energy of the CdS-g/C3N4/4AZ nanocomposite.
2.10 eV, illustrating the capability of the synthesized nanocompos-
ite to act as a visible-light responsive photocatalyst. Furthermore,
the Eg value of the CdS, g-C3N4, and CdS/g-C3N4 were obtained as
2.42, 2.7, 2.45 eV, as reported in our previous research work [4].
The decrease in band gap value and resulting increase in photocat-
alytic performance of the CdS/g-C3N4/4AZ nanocomposite could be
related to the structural properties of the 4A zeolite which acts as a
rough surface facilitating light reflection and scattering [3,30].
is statistically significant, indicating the proposed model is well fit-
ted to the experimental data. The insignificance of the lack of fit
also confirms it. The R2 value was obtained as 0.9744, indicating
the possibility of the prediction the 97.44% of the CFP removal vari-
ations in the photocatalytic degradation process using indepen-
dent parameters.
Another important factor to examine the adequacy of the model
is the normal plot of residuals, which shows the differences
between the predicted and experimental values. Fig. S3 (a) shows
the residual plots for CFP degradation in the photocatalytic pro-
cess, indicating the formation of the straight line in the normal
probability plot. Furthermore, the plot of residuals versus run in
Fig. S3 (b) doesn’t show a specific trend and the points are ran-
domly distributed with a zero mean, indicating the adequacy and
reasonability of the CCD model.
Since all the model statistics and diagnostic plots were OK, the
study continued by evaluating the interaction between the influen-
tial parameters.
Considering the aforementioned descriptions, it can be con-
cluded that three independent variables including the dosage of
the photocatalyst (A), pH (C), and irradiation time (D) and their
interactions are significant in predicting the CFP degradation effi-
ciency. Hence, the interaction between these parameters was illus-
trated as response surfaces (Fig. 6).
Fig. 6(a) shows the impact of the CdS/g-C3N4/4AZ content and
irradiation time on the degradation efficiency of CFP in the photo-
catalytic process. As Fig. 6(a) shows, the increase in the dosage of
the catalyst along with the increase in irradiation time leads to
the gradual increase in CFP degradation efficiency. This could be
related to the following reasons. Firstly, the increase in the number
of the photocatalyst particles results in increased physical adsorp-
tion of the pollutant molecules on the active surface sites of the
catalyst as well as the number of photons absorbed. Secondly,
the presence of more catalyst particles in the solution provides
additional reactive radical species to degrade of the constant
amount of the pollutant molecules. In fact, the greater catalyst/pol-
lutant ratio provides additional available active sites for the pro-
duction of the reactive radicals in the solution, resulting in
enhanced degradation efficiency. On the other hand, the increase
in catalyst dosage more than 0.4 g L-1 adversely affected on CFP
degradation efficiency, probably due to the accumulation of the
nanocomposite in high dosage and resulting decline in the avail-
able active surface sites [37].
3.2. The photocatalytic performance of the as-prepared CdS/g-C3N4/
4AZ nanocomposite towards the degradation of CFP
3.2.1. Process optimization by CCD
Table 2 shows the designed 30 experiments using the CCD for
photocatalytic degradation of CFP antibiotic in the presence of
the CdS/g-C3N4/4AZ nanocomposite as the catalyst.
After performing the proposed runs and importing the experi-
mentally results in the software, the relationship between CFP
degradation efficiency (%) and the influential parameters was sug-
gested in terms of coded factors:
Degradation ð%Þ ¼ þ 73:52 þ 2:17 A - 0:53 B þ 3:53 C
þ 4:83 D - 0:50 AB þ 2:11AC - 1:36 AD - 0:65 BC þ 0:54 BD
þ 5:95 CD - 1:21 A2 - 0:32 B2 þ 3:37 C2 þ 0:98 D2
ð2Þ
where +73.52 is a constant and the numbers in front of terms are
coefficients for linear, quadratic, and interaction effects (see general
form of Eq. (2) in the Supplementary Text 1).
According to Eq. (2), the dosage of the photocatalyst (A), pH (C),
and irradiation time (D) present more significant effects on the CFP
degradation efficiency. Running of the backward step in the DX11
software led to the elimination of nonsignificant variables in the
quadratic polynomial model and it was reduced to Eq. (3) [29]:
Degradation ð%Þ ¼ þ 73:16 þ 2:17 A þ 3:53 C þ 4:83 D
þ 2:11AC - 1:36 AD þ 5:95 CD - 1:17 A2 þ 3:42 C2 þ 1:03 D2
ð3Þ
As it was inference from Eq. (3), the variation of antibiotic con-
centration (B) and its quadratic and interaction impacts (B2, AB, BC,
and BD) had no effect on the CFP degradation efficiency.
Table S1 shows the analysis of variance (ANOVA) results for the
evaluation of the model’s fitness. As can be observed from the
results, the Fisher distribution (F test) value of the model (87.70)
The impact of the solution pH and catalyst dosage on the CFP
degradation efficiency was shown in Fig. 6 (b). As clearly seen,
5

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Table 2
The experimental matrix designed by CCD and results of the CFP degradation.
Run
A: [Catalyst]0(g L-1
)
B: [CFP]0 (mg L-1
)
C: pH
D: Time(min)
Degradation (%)
Experimental
Predicted
RSM
ANN
1
2
3
4
5
6
7
8
0.2* (ꢁ1)**
0.3 (0)
0.3 (0)
10 (ꢁ1)
15 (0)
5 (ꢁ1)
7 (0)
7 (0)
40 (ꢁ1)
60 (0)
60 (0)
72.15
75.11
71.08
72.91
74.22
88.09
65.58
93.51
75.31
73.18
70.61
61.54
95.09
75.38
76.39
70.44
73.21
79.99
86.43
70.01
92.23
73.23
72.16
72.11
63.61
88.45
73.78
67.95
73.16
76.41
72.59
73.16
73.16
73.16
73.08
87.84
63.54
93.88
73.16
72.59
70.48
63.54
93.66
73.16
75.45
70.48
74.82
79.76
88.92
73.16
93.66
72.83
73.16
73.06
64.15
87.84
75.45
67.60
73.16
74.82
72.87
75.83
71.80
73.63
74.94
88.81
66.30
94.23
70.03
73.90
71.33
62.26
95.81
76.10
77.11
71.16
73.93
80.71
87.15
70.73
92.95
73.95
72.88
72.83
64.33
89.17
74.59
68.67
73.88
77.13
25 (+2)
5 (ꢁ2)
20 (+1)
10 (ꢁ1)
10 (ꢁ1)
15 (0)
0.3 (0)
7 (0)
60 (0)
0.2 (ꢁ1)
0.2 (ꢁ1)
0.2(ꢁ1)
0.3 (0)
5 (ꢁ1)
9 (+1)
9 (+1)
11 (+2)
7 (0)
5 (ꢁ1)
5 (ꢁ1)
9 (+1)
9 (+1)
7 (0)
5 (ꢁ1)
5 (ꢁ1)
9 (+1)
3 (ꢁ2)
7 (0)
80 (+1)
80 (+1)
40 (ꢁ1)
60 (0)
9
0.3 (0)
15 (0)
60 (0)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
0.2 (ꢁ1)
0.4 (+1)
0.2 (ꢁ1)
0.4 (+1)
0.3 (0)
0.4 (+1)
0.4 (+1)
0.4 (+1)
0.3 (0)
20 (+1)
20 (+1)
20 (+1)
10 (ꢁ1)
15 (0)
10 (ꢁ1)
10 (ꢁ1)
20 (+1)
15 (0)
15 (0)
15 (0)
20 (+1)
15 (0)
15 (0)
40 (ꢁ1)
80 (+1)
40 (ꢁ1)
80 (+1)
60 (0)
40 (ꢁ1)
80 (+1)
40 (ꢁ1)
60 (0)
100 (+2)
60 (0)
80 (+1)
60 (0)
60 (0)
80 (+1)
60 (0)
80 (+1)
40 (ꢁ1)
20 (ꢁ2)
60 (0)
0.3 (0)
0.3 (0)
7 (0)
9 (+1)
7 (0)
0.4 (+1)
0.5 (+2)
0.3 (0)
0.2 (ꢁ1)
0.1 (ꢁ2)
0.2 (ꢁ1)
0.4 (+1)
0.3 (0)
7 (0)
10 (ꢁ1)
5 (ꢁ1)
15 (0)
7 (0)
20 (+1)
20 (+1)
15 (0)
15 (0)
10 (ꢁ1)
9 (+1)
5 (ꢁ1)
7 (0)
7 (0)
9 (+1)
0.3 (0)
0.4 (+1)
40 (ꢁ1)
* Real value.
** Coded value.
the increase in solution pH and catalyst dosage favored the degra-
dation efficiency of the CFP. This could be linked to the electrostatic
interactions between the catalyst surface and pH of solution
[38,39]. For CdS/g-C3N4/4AZ, the pHpzc, measured by the method
explained in the Supplementary Text 2, was 10.5 (i.e., composite’s
surface has positively charged at pH < 10.5 and negatively charged
at pH > 10.5, Fig. S4). CFP has two pKa values (pKa1 = 4.5 and
pKa2 = 6.5). In other words, this chemical compound is mainly
cationic (CFP+), zwitterionic (CFP ), and anionic (CFP-) at
pH < 4.5, 4.5 < pH < 6.5, and pH > 6.5, respectively [40]. Hence,
the electrostatic interactions between the positive charge of cata-
lyst’s surface and CFP at the pH range of 4.5–6.5 led to the
approach of the CFP to the CdS/g-C3N4/4AZ surface, which was fol-
lowed by an intense electrostatic attraction between the positive
charge of catalyst surface and the anionic CFP (CFP-) at the pH
range of 6.5–10.5. Furthermore, the alkaline pH intensifies the
hydrolysis of the CFP due to the instability of the beta-lactam ring
[41].
The interaction of irradiation time and pH on CFP degradation
was shown in Fig. 6(c). It is clear that the increase in duration
and pH have a synergic effect on the degradation of antibiotic, as
previously interpreted.
At the end of this section the RSM-based optimal conditions
was proposed to reach the highest CFP degradation efficiency
and the obtained results were summarized in Table 3. The degrada-
tion experiment of CFP was conducted under the optimum condi-
tions and the results showed that degradation of 93.23% of CFP was
occurred with the insignificant difference compared with the pre-
dicted value. The CFP degradation using CdS/g-C3N4 was performed
in the determined optimum condition and 81.34 % degradation
efficiency was achieved. The higher degradation rate in the pres-
ence of the CdS/g-C3N4/4A composite confirmed the better photo-
catalytic performance.
The kinetics of the CFP degradation in the photocatalytic pro-
cess was investigated at the determined RSM-based optimum con-
ditions utilizing the Langmuir–Hinshelwood model (described in
the Supplementary Tex 3) and Fig. S5 illustrates the plot of the
lnC0/Ct versus irradiation time (t). The results demonstrated that
the kinetics of the involved photocatalytic process obeys from
the Langmuir-Hinshelwood’s pseudo-first-order model. From the
slope of Fig. S5, the apparent rate constant (kapp) of the degradation
reaction of the CFP in the photocatalytic process was obtained as
3.71 ꢀ 10-2 minꢁ1
.
3.2.2. ANN modeling
The catalyst dosage, initial CFP concentration, pH, and irradia-
tion time were input variables of the ANN model and the CFP
degradation efficiency (%) was the output of the model. The exper-
imental data reported in Table 2 were normalized and used to
training, validating, and testing of the ANN model.
The number of neurons in the hidden layer was selected by
examining various numbers from 2 to 10 and the topology with
the number of neurons 1:9:4 in the input, hidden, and output lay-
ers was selected as the optimized ANN structure for CFP degrada-
tion prediction by the photocatalytic process. Table 4 listed the
connection weights of the layers of the developed ANN model.
To validate the designed ANN model, the values of the regres-
sion (R2) for the plots between outputs and targets were calculated
and reported in Fig. 7. The results showed good agreement
between the estimated values by the ANN model and the results
of the experiments, as reported in Table 2, confirming the reason-
ability of the model.
6

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Fig. 6. Response surface plots for degradation efficiency of CFP as a function of (a) catalyst dosage-irradiation time, (b) catalyst dosage-pH, and (c) pH- irradiation time. (The
values of the other two parameters in each graph have been considered equal to the average of their ranges).
Table 3
The optimized values for highest CFP degradation (%) by photocatalytic process.
[Catalyst]0(g L-1
)
[CFP]0 (mg L-1
)
pH
Time (min)
80
Degradation (%)
Predicted
95.66
Experimental
93.23
0.4
17
9
Table 4
The matrix of the connection weights of the layers of the developed ANN model.
**
LW
IWj,i
Time
0.209
1.532
ꢁ0.214
ꢁ0.192
ꢁ2.039
ꢁ1.035
ꢁ2.138
ꢁ0.799
1.929
*
Neurons
j,i
Degradation (%)
pH
[CFP]0
[Catalyst]0
0.151
0.822
0.082
1.912
2.892
0.611
0.005
1.252
1.333
0.197
2.228
1.626
1.170
ꢁ0.208
1.933
1.683
0.489
0.474
ꢁ0.725
0.260
1.402
ꢁ0.133
ꢁ1.234
1.594
1
2
3
4
5
6
7
8
9
ꢁ0.223
0.130
0.884
ꢁ0.665
ꢁ0.308
0.028
ꢁ0.315
ꢁ0.634
ꢁ2.290
0.671
0.476
0.799
W
j,i*: weights between the hidden and the input layers.
W*j,*i: weights between the hidden and the output layers.
For better expression the impact of the variables on the degra-
dation efficiency of the CFP antibiotic in the photocatalytic process,
the percentage analysis of each factor was performed applying
Garson’s method (described in the Supplementary Tex 1), and the
result was shown as a radar chart in Fig. 8. As can be observed in
Fig. 8, the pH of the solution, irradiation time, nanocomposite
dosage, and the initial concentration of the CFP affected antibiotic
degradation efficiency with a percentage impact of 37, 29, 19, and
15 %, respectively.
7

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Fig. 7. Regression plots for validation of ANN model.
3.2.3. Comparison of the CCD and ANN
In this section, a comparative study was performed to compare
the accuracy of the two CCD and ANN approaches for the predic-
tion of the degradation efficiency of the CFP in the photocatalytic
process. As can be seen from Fig. S6, both methods are worthwhile,
but a closer look reveals that the ANN model is more susceptible to
predict the variation in CFP degradation by the photocatalytic pro-
cess compared.
3.3. Mechanism of the photocatalytic process
Three steps were taken to suggest a plausible photocatalytic
mechanism. (i) The PL analysis was carried out to study the sepa-
ration and recombination processes of the produced e--h+ pairs.
As Fig. 9 (a) shows, the ternary nanocomposite showed a weaker
emission peak compared with CdS/g-C3N4, confirming the decrease
in recombination rate of the photoexcited charge carriers. This can
be attributed to the presence of the zeolite in the composite tex-
ture, which leads to decreasing the recombination of the charge
carriers and consequently intensifying the photocatalytic activity
of the as-synthesized nanocomposite. Furthermore, the position
of the PL peak for CdS/g-C3N4/4AZ nanocomposites shifted to the
red region, confirming the increased visible-light activity of the
ternary composite. (ii) The role of the active species during the
Fig. 8. The relative importance of the input variables on the degradation of CFP
analyzed by ANN model.
8

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Fig. 9. The PL spectra of the CdS/g-C3N4 and CdS/g-C3N4/4AZ (a); The impact of the scavengers on the CFP degradation efficiency by the photocatalytic process under RSM-
based optimized conditions (b).
photocatalytic CFP degradation over CdS/g-C3N4/4AZ was moni-
tored using the addition 1 mM of the IPA, BQ and NaOx as °OH,
°O–2, and h+ scavengers, respectively. The results presented in
Fig. 9(b) show that IPA as the hydroxyl radical scavenger applied
the less scavenging effect with about 8 % decline, whereas the pres-
ence of the BQ and NaOx decreased the CFP degradation efficiency
from 93.23 to 41.16 and 32.47%, respectively, confirming that the
superoxide anion radicals and photogenerated holes were the main
responsible species in the decomposition of the CFP by the photo-
catalytic process [3]. The claim has been authenticated from the
photoluminescence studies in an alkaline solution of terephthalic
acid as described in the Supplementary Text 4 [42]. Fig. S7 showed
that there was no significant change in the intensity of the PL spec-
trum during the time, indicating hydroxyl radicals are inefficacious
in the involved photocatalytic process and the produced h+ can
directly oxidized CFP molecules [21,43]. (iii) The relative position
of conduction and valence band edges of CdS and g-C3N4 was
established based on our previous study (ECB/EVB of CdS and g-
C3N4 was ꢁ0.36/1.98 eV and ꢁ1.11/1.54 eV, respectively) [4].
Undoubtedly, the position of energy band alignments favors the
formation of staggered gap (type-II) heterojunction [44]. Unless
Fig. 10. The proposed schematic illustration for the mechanism of the studied
there is a compelling reason to violate it.
photocatalytic process.
According to the above description, Fig. 10 illustrates a sche-
matic mechanism diagram for photodegradation of CFP antibiotic
using CdS/g-C3N4/4AZ nanocomposite as the catalyst. As shown,
the irradiation of the visible light to CdS/g-C3N4/4AZ semiconduc-
tor resulted in the excitation of CdS and g-C3N4. The excited elec-
trons (e-) moved from conduction band (CB) of the g-C3N4 to CB
of CdS. At the same time, the holes (h+) transferred from the
valance band (VB) of CdS to VB of g-C3N4. Since CB position of
CdS (-0.36 eV) is more negative than the standard redox potential
of O2/°O–2 (-0.33 eV) [45], superoxide anion radicals can be easily
produced from O2 molecules reduction. On the other hand, VB
position of g-C3N4 (+1.54 eV) is more negative than the standard
redox potential of H2O/°OH (+2.27 eV) and OH–/°OH (+2.38 eV)
[46], as a result, photogenerated holes cannot oxidize H2O or OH–
to hydroxyl radicals and they play a direct role in CFP molecules
eradication.
other configurations such as Z-scheme. The main reason can be
incompatibility the relative position of conduction and valence
band edges of semiconductors with standard redox potential of
O2/°O–2, H2O/°OH, OH–/°OH, H+/H2, etc. In addition, the presence
of a mediator between the CdS and g-C3N4 can lead to the Z-
scheme structure [6,7,27,45–47].
Since none of these exceptions has been observed in the present
study, it can be firmly accepted that a type-II heterojunction is pre-
ferred, which agrees with previously reported articles [48,49].
3.4. Intermediates determination during the CFP photocatalysis
The produced intermediates during degradation of the CFP in
the photocatalytic process in the presence of the CdS/g-C3N4/4AZ
nanocomposite were identified using GC-MS analysis. The list of
the major produced by-products during degradation of CFP with
m/z ratio and related retention time (Rt) were summarized in
Table 5. The structure of the produced intermediate compounds
confirmed that the CFP molecules decomposed mainly through
The above interpretations confirm the formation of the type-II
heterojunction, which lowering the recombination of the produced
e--h+ pairs and intensifying the photocatalytic activity of the as-
synthesized nanocomposite. Not to mention that depending on
the prevailing conditions, the CdS/g-C3N4 composites can have
9

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
Table 5
The recognized intermediates during degradation of the CFP by the photocatalytic process.
m/z
Rt
2.82
Chemical structure
Name and chemical formula*
No.
1
59
Ethyl Formate
(C3H6O2)
101
59
3.14
3.53
5-Hydroxy-4-oxopentanoic acid
(C5H8O4)
2
3
2,3-Dimethyl-2-pentanol
(C7H16O)
31
60
3.67
7.99
1-Methoxy-2-methyl-2-propanol
(C5H12O2)
4
5
3-Pyrrolidinecarboxylic acid
(C5H9NO2)
203
201
20.30
23.11
2-Methylthiazole-4-carbonitrile
(C5H4N2S)
6
7
1,3-dimethyl-1H-pyrazol-5(4H)-one
(C5H8N2O)
* Based on national institute of standards and technology (NIST) library.
the cleavage of the C-S, C-N and, N-N bonds. Production of the dif-
ferent intermediates with various structures confirmed the com-
plexity of the CFP degradation procedure. The formation of the
aliphatic compounds and carboxylic acids indicated the mineral-
ization of the CFP to non-toxic products.
4. Conclusion
The CdS/g-C3N4 type II heterojunction supported on 4A zeolite
was successfully synthesized as a fascinating visible light respon-
sive composite semiconductor. The photodecomposition of the
CFP antibiotic using the as-prepared catalyst under visible light
irradiation was studied and the impact of the operating variables
optimized using RSM and ANN approaches. The highest degrada-
tion efficiency of CFP was obtained as 93.23 % under optimized
condition ([Catalyst]0 = 0.4 g L-1, [CFP]0 = 17 mg L-1, pH = 9,
time = 80 min). The results of the ANN model indicated that pH
with a percentage impact of 37 % was the most influential variable
in the degradation of the CFP. Based on the results of the scavenger
experiments, it was concluded that the produced superoxide radi-
cals and holes were the main responsible species in degradation
process. The pseudo-first-order model of the photodegradation
process was confirmed by the result of the kinetic studies. Finally,
the structure of the formed intermediated during the degradation
of the CFP showed that the decomposition of the CFP mainly
occurred through the cleavage of the of C-S, C-N, and N-N bonds.
3.5. Stability of the CdS/g-C3N4/4AZ nanocomposite
Finally, the stability and reusability of the as-prepared CdS/g-
C3N4/4AZ nanocomposite as the important factor in long-term
application of the catalysts was evaluated and the results are pre-
sented in Fig. S8. The degradation experiments were performed at
the determined optimal conditions ([Catalyst]0
= 0.4 g ,
L-1
[CFP]0 = 17 mg L-1, pH = 9, time = 80 min). The nanocomposite
was used at 3 repetitive run after washing with distilled water
and drying. As can be seen, the insignificant decrease in the perfor-
mance of the nanocomposite was observed after 3 run, confirming
good stability of the nanocomposite and capability to reuse in
practical applications. Furthermore, the XRD patterns of the fresh
and used catalyst were recorded and reported in Fig. S 8, confirm-
ing no meaningful changes in the crystalline structure of the
catalyst.
CRediT authorship contribution statement
Naime AttariKhasraghi: Investigation, Writing – original draft.
Karim Zare: Resources, Supervision. Ali Mehrizad: Supervision,
Software, Formal analysis, Writing – review & editing. Nasser
10

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
Journal of Molecular Liquids 342 (2021) 117479
[17] M. Nagamine, M. Osial, K. Jackowska, P. Krysinski, J. Widera-Kalinowska,
Tetracycline photocatalytic degradation under CdS treatment, J. Mar. Sci. Eng.
8 (2020) 483.
Modirshahla: Methodology, Conceptualization. Mohammad Ali
Behnajady: Visualization, Writing – review & editing.
[18] A.G. Khosroshahi, A. Mehrizad, Optimization, kinetics and thermodynamics of
photocatalytic degradation of acid red 1 by Sm-doped CdS under visible light, J.
Mol. Liq. 275 (2019) 629–637.
Declaration of Competing Interest
_
[19] A.I. Vaizog˘ullar, Ternary CdS/MoS2/ZnO photocatalyst: synthesis,
characterization and degradation of ofloxacin under visible light irradiation,
J. Inorg. Organomet. Polym Mater. 30 (2020) 4129–4141.
[20] T. Senasu, T. Chankhanittha, K. Hemavibool, S. Nanan, Visible-light-responsive
photocatalyst based on ZnO/CdS nanocomposite for photodegradation of
reactive red azo dye and ofloxacin antibiotic, Mater. Sci. Semicond. Process.
123 (2021) 105558.
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
Acknowledgements
[21] B. Shao, X. Liu, Z. Liu, G. Zeng, W. Zhang, Q. Liang, Y. Liu, Q. He, X. Yuan, D.
Wang, Synthesis and characterization of 2D/0D g-C3N4/CdS-nitrogen doped
hollow carbon spheres (NHCs) composites with enhanced visible light
photodegradation activity for antibiotic, Chem. Eng. J. 374 (2019) 479–493.
The authors would like to thank Tabriz Branch and Tehran Science
and Research Branch, Islamic.
_
[22] S. Yılmaz, A. Ünverdi, M. Tomakin, I. Polat, E. Bacaksız, Surface modification of
Azad University for providing facilities and technical support.
CBD-grown CdS thin films for hybrid solar cell applications, Optik 185 (2019)
256–263.
[23] Y. Yan, H. Yang, Z. Yi, R. Li, T. Xian, Design of ternary CaTiO3/g-C3N4/AgBr Z-
scheme heterostructured photocatalysts and their application for dye
photodegradation, Solid State Sci. 100 (2020) 106102.
Appendix A. Supplementary material
[24] Chunyan Du, Zhuo Zhang, Shiyang Tan, Guanlong Yu, Hong Chen, Lu Zhou, Lie
Yu, Yihai Su, Yin Zhang, Fangfang Deng, Shitao Wang, Construction of Z-
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.molliq.2021.117479.
scheme
g-C3N4/MnO2/GO
ternary
photocatalyst
with
enhanced
photodegradation ability of tetracycline hydrochloride under visible light
radiation, Environ. Res. 200 (2021) 111427, https://doi.org/10.1016/j.
envres.2021.111427.
References
[25] G. Tang, F. Zhang, P. Huo, S. Zulfiqarc, J. Xu, Y. Yan, H. Tang, Constructing novel
visible-light-driven ternary photocatalyst of AgBr nanoparticles decorated 2D/
2D heterojunction of g-C3N4/BiOBr nanosheets with remarkably enhanced
photocatalytic activity for water-treatment, Ceram. Int. 45 (2019) 19197–
19205.
[26] S. Cao, Z. Jiao, H. Chen, F. Jiang, X. Wang, Carboxylic acid-functionalized
cadmium sulfide/graphitic carbon nitride composite photocatalyst with well-
combined interface for sulfamethazine degradation, J. Photochem. Photobiol.,
A 364 (2018) 22–31.
[27] I.A. Mkhalid, R.M. Mohamed, A.A. Ismail, M. Alhaddad, Z-scheme g-C3N4
nanosheet photocatalyst decorated with mesoporous CdS for the
photoreduction of carbon dioxide, Ceram. Int. 47 (2021) 17210–17219.
[28] A.M. Taddesse, T. Bekele, I. Diaz, A. Adgo, Polyaniline supported CdS/CeO2/
Ag3PO4 nanocomposite: An ‘‘AB” type tandem nn heterojunctions with
[1] S. Majumder, S. Chatterjee, P. Basnet, J. Mukherjee, ZnO based nanomaterials
for photocatalytic degradation of aqueous pharmaceutical waste solutions–a
contemporary review, Environ. Nanotechnol. Monit. Manage. 14 (2020)
100386, https://doi.org/10.1016/j.enmm.2020.100386.
[2] X. Shi, A. Karachi, M. Hosseini, M.S. Yazd, H. Kamyab, M. Ebrahimi, Z. Parsaee,
Ultrasound wave assisted removal of Ceftriaxone sodium in aqueous media
with novel nano composite g-C3N4/MWCNT/Bi2WO6 based on CCD-RSM
model, Ultrason. Sonochem. 68 (2020) 104460.
[3] A. Ataei, A. Mehrizad, K. Zare, Photocatalytic degradation of cefazoline
antibiotic using zeolite-supported CdS/CaFe2O4 Z-scheme photocatalyst:
Optimization and modeling of process by RSM and ANN, J. Mol. Liq. 328
(2021) 115476.
[4] N. AttariKhasraghi, K. Zare, A. Mehrizad, N. Modirshahla, M.A. Behnajady,
Achieving the enhanced photocatalytic degradation of ceftriaxone sodium
enhanced photocatalytic activity, J. Photochem. Photobiol.,
113005.
A 406 (2021)
using CdS-gC
3 N 4 nanocomposite under visible light irradiation: RSM
[29] N. Herron, Y. Wang, M.M. Eddy, G.D. Stucky, D.E. Cox, K. Moller, T. Bein,
Structure and optical properties of cadmium sulfide superclusters in zeolite
hosts, J. Am. Chem. Soc. 111 (1989) 530–540.
[30] F. Ganji, K. Mohammadi, B. Roozbehani, Efficient synthesis of 4A zeolite based-
NiCo2O4 (NiCo2O4@ 4A) nanocomposite by using hydrothermal method, J.
Solid State Chem. 282 (2020) 121111.
[31] D. Chen, Z. Wang, D. Yue, G. Yang, T. Ren, H. Ding, Synthesis and visible
photodegradation enhancement of CdS/g-C3N4 photocatalyst, J. Nanosci.
Nanotechnol. 16 (2016) 471–479.
modeling and optimization, J. Inorg. Organomet. Polym Mater. (2021) 1–11.
[5] J. Xu, J. Zhu, M. Chen, Simultaneous removal of ceftriaxone sodium and Cr (VI)
by
a novel multi-junction (pn junction combined with homojunction)
composite photocatalyst: BiOI nanosheets modified cake-like anatase-rutile
TiO2, J. Mol. Liq. 320 (2020) 114479.
[6] D. Kong, X. Ruan, J. Geng, Y. Zhao, D. Zhang, X. Pu, S. Yao, C. Su, 0D/3D ZnIn2S4/
Ag6Si2O7 nanocomposite with direct Z-scheme heterojunction for efficient
photocatalytic H2 evolution under visible light, Int. J. Hydrogen Energy 46
(2021) 28043–28052.
[32] Dasari Ayodhya, Guttena Veerabhadram, Ultrasonic synthesis of g-C3N4/CdS
composites and their photodegradation, catalytic reduction, antioxidant and
antimicrobial studies, Mater. Res. Innovations 24 (4) (2020) 210–228.
[33] Nhu-Nang Vu, Serge Kaliaguine, Trong-On Do, Synthesis of the g-C3N4/CdS
nanocomposite with a chemically bonded interface for enhanced sunlight-
driven CO2 photoreduction, ACS Appl. Energy Mater. 3 (7) (2020) 6422–6433.
[34] Ali Mehrizad, Parvin Gharbani, Application of central composite design and
artificial neural network in modeling of reactive blue 21 dye removal by
photo-ozonation process, Water Sci. Technol. 74 (1) (2016) 184–193.
[35] J. Rouquerol, F. Rouquerol, P. Llewellyn, G. Maurin, K.S. Sing, Adsorption by
powders and porous solids: principles, methodology and applications,
Academic press; 2013.
[36] M.S. Seyed Dorraji, A.R. Amani-Ghadim, M.H. Rasoulifard, S. Taherkhani, H.
Daneshvar, The role of carbon nanotube in zinc stannate photocatalytic
performance improvement: experimental and kinetic evidences, Appl. Catal. B
205 (2017) 559–568.
[37] Gongduan Fan, Rongsheng Ning, Jing Luo, Jin Zhang, Pei Hua, Yang Guo,
Zhongsheng Li, Visible-light-driven photocatalytic degradation of naproxen by
Bi-modified titanate nanobulks: synthesis, degradation pathway and
mechanism, J. Photochem. Photobiol., A 386 (2020) 112108, https://doi.org/
10.1016/j.jphotochem.2019.112108.
[38] M. Shokri, G. Isapour, S. Shamsvand, B. Kavousi, Photocatalytic degradation of
ceftriaxone in aqueous solutions by immobilized TiO2 and ZnO nanoparticles:
investigating operational parameters, J. Mater. Environ. Sci. 7 (2016) 2843–
2851.
[39] C.-H. Chiou, C.-Y. Wu, R.-S. Juang, Influence of operating parameters on
photocatalytic degradation of phenol in UV/TiO2 process, Chem. Eng. J. 139
(2008) 322–329.
[40] Y. Chen, L. Ruiping, G. Yan, T. Hailin, H. Yingping, C. Junsong, F. Yanfen, Y.
Changying, Green and efficient degradation of cefoperazone sodium by
Bi4O5Br 2 leading to the production of non-toxic products: performance and
degradation pathway, J. Environ. Sci. 100 (2021) 203–215.
[7] Z. Shao, X. Meng, H. Lai, D. Zhang, X. Pu, C. Su, H. Li, X. Ren, Y. Geng, Coralline-
like Ni2P decorated novel tetrapod-bundle Cd0.9Zn0.1S ZB/WZ homojunctions
for highly efficient visible-light photocatalytic hydrogen evolution, Chin. J.
Catal. 42 (2021) 439–449.
[8] C. Su, D. Zhang, X. Pu, Z. He, X. Hu, L. Li, G. Hu, Magnetically separable NiFe2O4/
Ag3VO4/Ag2VO2PO4 direct Z-scheme heterostructure with enhanced visible-
light photoactivity, J. Chem. Technol. Biotechnol. (2021) 1–10.
[9] D. Kong, H. Fan, D. Yin, D. Zhang, X. Pu, S. Yao, C. Su, AgFeO2 nanoparticle/
ZnIn2S4 microsphere p–n heterojunctions with hierarchical nanostructures for
efficient visible-light-driven H2 evolution, ACS Sustain. Chem. Eng. 9 (2021)
2673–2683.
[10] N. Zhang, X. Li, Y. Wang, B. Zhu, J. Yang, Fabrication of magnetically
recoverable Fe3O4/CdS/g-C3N4 photocatalysts for effective degradation of
ciprofloxacin under visible light, Ceram. Int. 46 (2020) 20974–20984.
[11] X. Ning, G. Lu, Photocorrosion inhibition of CdS-based catalysts for
photocatalytic overall water splitting, Nanoscale 12 (2020) 1213–1223.
[12] J.Y. Li, Y.H. Li, M.Y. Qi, Q. Lin, Z.R. Tang, Y.J. Xu, Selective organic
transformations over cadmium sulfide-based photocatalysts, ACS Catal. 10
(2020) 6262–6280.
[13] S.G. Kumar, R. Kavitha, P. Nithya, Tailoring the CdS surface structure for
photocatalytic applications, J. Environ. Chem. Eng. 8 (2020) 104313.
[14] J. Bai, Y. Li, P. Jin, J. Wang, L. Liu, Facile preparation 3D ZnS nanospheres-
reduced graphene oxide composites for enhanced photodegradation of
norfloxacin, J. Alloy. Compd. 729 (2017) 809–815.
[15] M. Arshad, Z. Wang, J.A. Nasir, E. Amador, M. Jin, H. Li, Z. Chen, Z. ur Rehman,
W. Chen, Single source precursor synthesized CuS nanoparticles for NIR
phototherapy of cancer and photodegradation of organic carcinogen, J.
Photochem. Photobiol., B 214 (2021) 112084.
[16] Q. Guo, G. Tang, W. Zhu, Y. Luo, X. Gao, In situ construction of Z-scheme FeS2/
Fe2O3 photocatalyst via structural transformation of pyrite for photocatalytic
degradation of carbamazepine and the synergistic reduction of Cr (VI), J.
Environ. Sci. 101 (2021) 351–360.
11

N. AttariKhasraghi, K. Zare, A. Mehrizad et al.
[41] B.F. Abramovic´, M.M. Uzelac, S.J. Armakovic´, U. Gašic´, D.D. Cetojevic´-Simin, S.
Journal of Molecular Liquids 342 (2021) 117479
ˇ
[46] Y. Guo, J. Li, Z. Gao, X. Zhu, Y. Liu, Z. Wei, W. Zhao, C. Sun, A simple and effective
method for fabricating novel p–n heterojunction photocatalyst g-C3N4/
Bi4Ti3O12 and its photocatalytic performances, Appl. Catal. B 192 (2016) 57–
71.
´
Armakovic, Experimental and computational study of hydrolysis and
photolysis of antibiotic ceftriaxone: degradation kinetics, pathways, and
toxicity, Sci. Total Environ. 768 (2021) 144991.
[42] Wan-Kuen Jo, Thillai Sivakumar Natarajan, Influence of TiO2 morphology on
the photocatalytic efficiency of direct Z-scheme g-C3N4/TiO2 photocatalysts
for isoniazid degradation, Chem. Eng. J. 281 (2015) 549–565.
[43] Y. Zhao, Y. Wang, H. Shi, E. Liu, J. Fan, X. Hu, Enhanced photocatalytic activity of
ZnSe QDs/g-C3N4 composite for ceftriaxone sodium degradation under visible
light, Mater. Lett. 231 (2018) 150–153.
[44] Jingxiang Low, Jiaguo Yu, Mietek Jaroniec, Swelm Wageh, Ahmed A. Al-
Ghamdi, Heterojunction photocatalysts, Adv. Mater. 29 (20) (2017) 1601694,
https://doi.org/10.1002/adma.v29.2010.1002/adma.201601694.
[45] S. Chen, Y. Hu, S. Meng, X. Fu, Study on the separation mechanisms of
photogenerated electrons and holes for composite photocatalysts g-C3N4-
WO3, Appl. Catal. B 150 (2014) 564–573.
[47] L. Qian, Y. Hou, Z. Yu, M. Li, F. Li, L. Sun, W. Luo, G. Pan, Metal-induced Z-
scheme CdS/Ag/g-C3N4 photocatalyst for enhanced hydrogen evolution under
visible light: the synergy of MIP effect and electron mediator of Ag, Molecular
Catalysis 458 (2018) 43–51.
[48] J. Fu, B. Chang, Y. Tian, F. Xi, X. Dong, Novel C3N4–CdS composite
photocatalysts with organic–inorganic heterojunctions: in situ synthesis,
exceptional activity, high stability and photocatalytic mechanism, J. Mater.
Chem. A 1 (2013) 3083–3090.
[49] S.W. Cao, Y.P. Yuan, J. Fang, M.M. Shahjamali, F.Y. Boey, J. Barber, S.C.J. Loo, C.
Xue, In-situ growth of CdS quantum dots on g-C3N4 nanosheets for highly
efficient photocatalytic hydrogen generation under visible light irradiation,
Int. J. Hydrogen Energy 38 (2013) 1258–1266.
12