Chemical Engineering & Processing: Process Intensification 169 (2021) 108635

Contents lists available at ScienceDirect

Chemical Engineering and Processing - Process

Intensification
journal homepage: www.elsevier.com/locate/cep

Evaluation of dynamic behavior of hydrophilic and photocatalytic

self-cleaning PES nanofiltration membrane post treating biologically
treated palm oil mill effluent (POME)
Hadis Zangeneh 1, Ali Akbar Zinatizadeh 2, 3, 4, *, Zahra Rahimi 2, Sirus Zinadini 2, 3,
Morteza Sadeghi 1, Mohamed Abdulgader 5, 6
1
Department of Chemical Engineering, Isfahan University of Technology, Isfahan, 84156-83111, Iran
2
Department of Applied Chemistry, Faculty of Chemistry, Razi University, Kermanshah, Iran
3
Environment Research Center (ERC), Razi University, Kermanshah, Iran
4
Department of Environmental Sciences, College of Agriculture and Environmental Sciences, University of South Africa, P. O. Box 392, Florida, 1710, South Africa
5
School of Engineering and Built Environment, Griffith University, Nathan Campus, QLD,4111, Australia
6
Department of Environmental Science, Faculty of Engineering & Technology, Sebha University, Libya

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

Keywords: The photocatalytic nanocomposite NF membranes blended with K-B-N triple TiO2 were fabricated employing
Fouling conventional wet phase inversion method. The dynamic performance of the photocatalytic membranes was
photocatalytic NF membrane evaluated by filtering biologically treated palm oil mill effluent (POME) at dark and light conditions. The per­
self-cleaning
formance of the chosen membrane was checked using central composite design (CCD) of response surface
K-B-N-triple TiO2
methodology (RSM) under continuous visible light irradiation and various variables in a dynamic mode. Ac­
cording to experimental results, the 0.5 wt. % membrane indicated less flux reduction without scarifying COD
rejection capacity as a function of time while filtering POME under continuous visible light irradiation. The
largest values of FRR and COD removal efficiency were reported to be 97 % and 100 %, respectively, at the
highest feed flow rate of 150 L/min, the lowest operating pressure of 5 bar, and the lowest COD concentration of
1000 mg/L. While, the highest value of flux (33.9 kg/m2h) was obtained at operating pressure of 9 bar, feed flow
rate of 50 L/h and COD concentration of 1000 mg/L. By judging from results, the photocatalytic NF membrane
by benefiting from both photocatalytic and self-cleaning properties has enough potential to be applied in
practical applications.

1. Introduction technologies have been employed to purify the polluted water, including
bioprocesses, adsorption processes, membrane-based separation pro­
Water scarcity issue has become a ubiquitous challenge all over the cesses, coagulation and photocatalytic approaches [1–8].
world, especially, in the developing countries. There are many acceler­ Malaysia is the second-largest producer of palm oil in the word after
ating factors, which are responsible to extremely intensify over- Indonesia. The palm oil manufacturing company has a crucial effect on
exploitation of water bodies and therefore the demand for clean water the economy of Malaysia. However, this company produces large
including over-growth of world’s population, urbanization, industriali­ qualities of palm oil mill effluent (POME) which is a potential contam­
zation, agricultural sectors as well as global warming alarming. On the inant to environment due to having a wide variety of compounds, high
other hand, the clean water used at various applications is discharged as organic contents and nutrients with thick nature and dark brown color
polluted water (wastewater), and hence, health burden of weak quality [9, 10]. Therefore, should raw palm oil mill effluent (POME) be released
water accounts for a serious threat for public life and ecosystems. into the aquatic ecosystems; the incurable environmental consequences
Therefore, these disastrous challenges come into force the researchers to will arise [11]. To combat the addressed issues, the biological treatment
purify wastewater with the aim of reuses. In this regard, the enormous methods are explored to noticeably reduce biodegradable organic

* Corresponding author:
E-mail addresses: [email protected], [email protected] (A.A. Zinatizadeh).

https://doi.org/10.1016/j.cep.2021.108635
Received 10 April 2021; Received in revised form 5 September 2021; Accepted 7 September 2021
Available online 17 September 2021
0255-2701/© 2021 Elsevier B.V. All rights reserved.
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

contents. However, the most serious problem related to biologically the chosen addressed membrane was studied in a dynamic mode using
treated palm oil mill effluent (POME) is remained color, which highly cross flow membrane filtration set up under continuous visible light
restricts recovery of wastewater. Reason of behind this matter is back to irradiation to remove remained dye and COD from biologically treated
the existence of lignin and phenolic compounds as the main origin of POME. In contrast to our previous work, antifouling ability and rejection
colored compounds in treated wastewater that can not be degraded capacity of the photocatalytic NF membranes were checked using BSA
using biological processes [12–14]. Hence, it is indispensable to solution and activated sludge (AS) suspension as foulant model.
combine the biological treatment systems with other technologies like Furthermore, the performance of the photocatalytic membranes was
advanced oxidation possesses (AOPs) and membrane techniques for an examined by filtration of biologically treated POME in a dynamic mode
impactful decolonization. The membrane-based separation processes as at dark and light conditions to determine contribution percentage of
a robust and leading technology have drawn tremendous attention in the adsorption processes and photocatalytic reactions on the improvement
area of post-treatment to meet high quality water. Meanwhile, the AOPs of flux and COD rejection. Finally, the effect of various variables i.e.
have gained too much interest among researchers due to their capacity operating pressure (5, 7 and 9 bar), feed flow rate (50, 100, 150 L/min)
in degrading and mineralizing the pollutant compounds from aqueous and COD concentration (1000, 3000 and 5000 mg/l) on the performance
media [15]. More importantly, the AOPs neither produce dangerous of chosen photocatalytic NF membrane as three responses (flux, FRR and
residues nor convert organic contaminants from one phase to another COD removal efficiency) was evaluated using central composite design
one, therefore, these technologies are more environmentally friendly (CCD) of response surface methodology (RSM).
compared to others [16]. Amongst various types of the AOPs i.e. ho­
mogeneous AOPs and heterogeneous AOPs, the latter one commonly 2. Experimental
utilizing solid catalysts in combination with other approaches of AOPs,
namely H2O2, O3, persulfates, light, etc., has a distinguish advantage of 2.1. Materials
the convenience in the catalyst separation over the former [17, 18].
Based on these benefits, the most recent publications are focused on the To synthesize photocatalytic nanofiller (K-B-N- triple doped TiO2),
use of heterogeneous catalyst, for instance, g-C3N4-based catalysts in potassium nitrate (KNO3, Merck, Germany), boric acid (H3BO3, Merck,
AOPs [19]. Germany), ammonia (NH3, H2O, 37 wt. %, Merck, Germany), tetraiso­
Nanocomposite membranes benefiting from concurrent synergistic propyltitanate (C12H28O4Ti, 99 wt. %, Merck, Germany) and ethanol
effects of both organic and inorganic nanomaterials have been employed (C2H5OH, 99 wt. %, Merck, Germany) were utilized as purchased. To
in an extensive variety of applications [20–24]. More often, inclusion of prepare polymer dope solution, polyethersulfone (PES, MW of 58,000 g/
hydrophilic nanocomposites into the membrane body renders the mol, Tg of 225◦ C, Ultrason E 6020P), dimethylacetamide (DMAc, BASF,
membrane hydrophilic, thereby, enhancing permeability and anti­ Germany), and polyvinylpyrrolidone (PVP, MW of 25,000 g/mol,
fouling properties. More importantly, photocatalytic nanocomposite Merck, Germany) were used without any purification. Dichromate po­
membranes endow other additional noteworthy function via the pho­ tassium (K2Cr2O7), sulphuric acid (H2SO4, 98 %, Merck, Germany),
tocatalytic reactions to break down recalcitrant organic compounds, for mercuric sulphate (HgSO4, Merck, Germany) and silver sulfate (Ag2SO4,
instance, lignin and phenolic compounds, persistent and Merck, Germany) were purchased to measure chemical oxygen demand
non-biodegradable towards biological treatment methods [25]. TiO2 is (COD). Synthesis of photocatalytic self-cleaning nanofiller of K-B-N-tri­
remarkably more applicable relative to other photocatalysts due to its ple doped TiO2, preparation of nanofiltration membrane and assessment
high oxidizing capacity, therefore, its ability to degrade a widespread of performance of fabricated membranes are described in text S1 (sup­
range of contaminants, easy accessibility, cost-effectiveness, non-­ plementary file).
toxicity and harmlessness [26]. Nonetheless, the broad band gap of TiO2
as well as fast recombination of electron-hole pairs generated over 2.2. Assessment of performance of chosen photocatalytic nanocomposite
photocatalytic reactions highly have restricted its activity only under NF membranes using RSM
ultra-violet (UV) light irradiation [27,28]. The doping approach using
inclusion of metal and nonmetal atoms into the TiO2 network is specu­ In this part, the performance of 0.5 wt. % photocatalytic nano­
lated to effectively diminish above limiting factors, and to shift UV light composite NF membrane identified as chosen membrane based on pre­
irradiation into the visible region [26]. In line with this, several efforts vious conducted experiments was evaluated via filtration of biologically
have been made using various doping metal and non-metal agents treated palm oil mill effluent (POME) as dynamic mode in a cross flow
[29–33]. However, the usage of metal and non-mental doping agents filtration set up. In overall, the cross flow filtration set up is comprised of
alone creates some serious drawbacks [34–36]. Accordingly, simulta­ three stainless steel filtration cells and an extra plexiglass filtration cell.
neous incorporation of two or three types of non-metal and metal atoms In this study, only one of the stainless steel filtration cells were in use
is recommended [25]. Such multiple dopants give rise to reduction in under dark conditions (without irradiation of visible light) in order to
both charge carriers recombination and band gap energy, and therefore survey adsorptive behavior of synthesized photocatalytic membranes.
increase in photocatalytic activity under visible light irradiation [37]. Furthermore, the plexiglass filtration cell was employed under visible
According to given information, during recent two years, our research light irradiation conditions in order to check photocatalytic activity of
team was encouraged to successfully modify versatile photocatalyst of the chosen 0.5 wt. % photocatalytic membrane. To supply visible light, a
TiO2 by doping triple metal and non-metal in order to get visible light halogen lamp (500 W, λ ˃ 400 nm) placed in a 25 cm-distance on the top
driven photocatalytic nanocomposite membranes with of the plexiglass filtration cell was used. A picture of the cross flow
self-cleaning/-antifouling attributes and super-hydrophilicity to reme­ filtration set up rig drawn schematically is present in Figure S1.
diate biologically treated colored effluent [25, 38–40]. The performance of the chosen 0.5 wt. % photocatalytic membrane
In one of our previous published studies, we assessed the effect of was studied under three different variables viz. feed cross flow rate (Q in
triple K-B-N doped TiO2 nanocomposite on performance of the NF L/h), operating pressure (P in bar) and chemical oxygen demand con­
membrane at 0.5 wt. % optimum loading in view point of dye removal centration (COD in mg/L) each in three levels. Performance experiments
from biologically treated POME as a field sample in a static mode using a were done in such a way designed using central composite design (CCD)
dead end membrane filtration set up under continuous visible light of response surface methodology (RSM) in Design Expert Software (Stat-
irradiation [40]. The chosen photocatalytic membrane capacity was Ease Inc., version 10.0). CCD provides an estimation of simultaneous
analyzed as permeability, dye rejection and COD removal efficiency at interactive effects the three studied independent experimental variables,
various COD concentrations (1000, 3000, and 5000 mg/l) and operating and compromises of a core factorial to form a cubic comprising of 2k
pressure (3, 4.5 and 6 bar). In the following this work, the behavior of axial points, 2k cube axial points, and a single axial point.

2
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

Range and level of the studied variables as original and symbol units Morphology the pure TiO2 and K-B-N-TiO2 nanostructures was also
are presented in Table S2. In overall, 20 experimental runs with 5 rep­ displayed in the Figure 1a-b, respectively. The particle size distribution
lications were designed based on formula of 2k+2k +cp, where, k and cp of the prepared nanocomposites was also estimated by ImageJ 1.44p
demonstrate the number of variables and replicates for each the central software as shown in Table 1.
point, respectively.
The considered responses were initial pure water flux (kg/m2.h), flux 3.2. Investigation of physicochemical characteristics of the fabricated NF
recovery ratio (FRR, %), COD removal efficiency (%). membranes
Before starting NF filtration experiments, the 0.5 wt. % chosen
photocatalytic membrane was firstly undergone hydraulic compaction The variations in cross-sectional morphological characteristics of the
at 5 bare for 1 h by filtration of pure water to reach stabilized perme­ reference and photocatalytic NF membranes fabricated using a preva­
ability. At steady-state conditions, the flux was calculated directly by lent technique of wet phase inversion process were analyzed by SEM
weight measurement of the permeate stream passed through an effective images. The obtained results in our previous published work are pre­
membrane surface area (40 m2) over a pre-determined course time sented here [40] (Figure 2). As seen in the Figure, the nanocomposite
under continuous visible light irradiation. membranes containing photocatalytic nanofiller up to 0.5 wt. % had
The flux was calculated using Equation 2. After measuring flux, distinctive substructure comprising of enlarged finger-like pores with
assigned to J1, for 1 h, the POME was replaced in feed tank to investigate high density of pores and large void volumes, resigning under a thin skin
antifouling properties as FRR and COD removal efficiency. region. Changes in morphological structure of the respective membranes
The FRR is known as a significant indicator of antifouling feature, so are related to variations in phase behaviors of the casting solutions and
that the higher FRR, the more antifouling properties. The water flux over dynamic features i.e. in-diffusion and out-diffusion mechanisms be­
filtration of fouling agent i.e. POME assigned to Jp was measured for 120 tween solvent and non-solvent during phase separation process [45].
min. In this stage, COD removal efficiency was determined based on The occurrence of the interior macro-pores in the porous sub-layer
Equation 7. After measuring Jp, the fouled membrane was exposed to corroborates thermodynamic instability, and therefore, facilitated
the visible light irradiation for 120 min to remove deposited foulant on instantaneous demixing of the benign polymer (PES) with non-solvent
the membrane surface using self-cleaning properties. Finally, the second (distilled water) [46,47]. Such favorable changes in morphological
pure water flux (J2) for the cleaned chosen photocatalytic membrane features provide appropriate confinements for photocatalytic nanofillers
was measured the same as the initial pure water flux (Equation 2). to be distributed in such a uniform pattern [48]. Whereas, the M-1.0
membrane bears a thick skin layer accompanied by narrow finger-like
3. Results and discussion pores. The reason behind this observation was supported by the in­
crease in viscosity of the casting solutions at the highest extent of pho­
3.1. Characterization of the prepared photocatalyst tocatalytic nanofiller (1.0 wt. %) that results in delayed demixing of the
PES polymer and non-solvent [46]. Further, the increase in viscosity of
According to the our previous published paper [40], the results ob­ the casting polymer solutions induce the nanofillers to aggregate as
tained from characterization analysis of the pure TiO2 and the chosen cluster, and thereby, reduce the extent of their exposure in the photo­
photocatalyst with optimum formulation of K-B-N-TiO2 are presented in catalytic nanocomposite NF membrane [49].
Figure S2. The topographical characteristics of the photocatalytic membrane
From XRD pattern (Figure S2a), the photocatalytic nanofiller of K-B- skin surface were qualitatively and quantitatively analyzed using 3-
N-TiO2 showed similar anatase phase to pure TiO2 with the 2ɵ values of dimentional AFM photos and the roughness parameters, respectively.
25.3◦ , 36.9◦ , 37.8◦ , 48.0◦ , 55.1◦ and 68.8◦ ascribed to crystal planes of The acquired outcomes are presented in Figure 3 and Table 2. According
(101), (103), (0 0 4), (200), (211), and (1 1 6), respectively [41,42]. to the Figure, the number of the darkest and brightest regions are
Compared to the pure TiO2, the intensity of peaks was higher, especially decreased after incorporation of photocatalytic nanofiller inside the
in the case of (1 0 1), that might be a sign of reduction in particle size. In benign polymer matrix. The darkest regions are related to valleys or
addition to anatase phase, the pure TiO2 revealed rutile phase, while pores of the membranes, while, the brightest regions are related to peaks
formation of such a phase was prevented in photocatalytic nanofiller of the membranes surface. The reported observations are in agreement
due to the presence of K, N and B. Typically, the anatase phase results in with the surface roughness parameters (Table 2). Although, the surface
much more photocatalytic properties relative to the rutile phase [26]. roughness parameters of all photocatalytic membranes were improved
The weak peaks appeared at about 27.9◦ and 30.6◦ were in the relation relative to the reference membrane, the M-1.0 membrane revealed
with formation of B2O3 on the photocatalytic nanofiller [43]. distinctive trend ascribed to non-uniformity of the nanofillers distribu­
From FTIR spectra (Figure S2b), peak observed at 469 cm− 1 was tion and their subsequent agglomeration [48]. Therefore, the reduction
attributed to the Ti-O-Ti stretching vibration. In the FTIR spectra of K-N- in the surface roughness of the membrane reflects the improvement in
B TiO2, peaks of 1100 and 1200 cm− 1 were corresponded with stretching the antifouling properties.
vibration of B-O in structures of B2O3 and Ti–O–B, respectively. In The measurement of overall porosities of the photocatalytic nano­
addition, occurrence of 1440 cm− 1 peak was in the relation with ni­ composite NF membranes provides quantitative information concerning
trogen atoms contributed in TiO2 network [44]. changes created in the porous character of the membrane after
The optical properties of pure TiO2 and modified TiO2 photocatalytic amendment. According to the obtained results (Figure 4), the amended
nanostructures were also evaluated as shown in the Figure S2c-e. The membranes revealed progressive increase in the porosity. However,
DRS spectra (Figure S2c) implies that the introduction of optimum further amount of photocatalytic nanofillers and their subsequent ag­
concentration of dopant agents (K, B and N elements) into TiO2 lattice gregation within polymer matrix; which took place in the case of M-1.0;
make the modified TiO2 nanostructure responsive in visible light with reduced the porous nature, and therefore, the void volumes of the
band gap energy of 2.4 eV (Tauc plot, Figure 2Sd). The PL spectra also amended membrane.
verify the effect of dopants on the reduction of recombination rate as the
highest PL intensity which implies the fastest recombination of e− /h+ 3.3. Investigation of performance of the prepared membranes
pairs was related to pure TiO2 whereas the PL intensity of modified TiO2
catalysts was far less compared to bare TiO2 and the introduction of 3.3.1. Pure water flux, rejection capacity and antifouling tests under a
optimum concentration of dopant agents (K, B and N elements) into TiO2 static mode in a dead end membrane set up
lattice noticeably reduced the PL intensity and the recombination rate Results of water contact angle (WCA) measure derived for the
(Figure 2Se). reference and amended photocatalytic nanocomposite membranes are

3
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

Figure 1. SEM images of (a) Pure TiO2 and (b) K-B-N-TiO2.

increasing in FRR percentages [50]. In contrast, the FRR diminished at

Table 1
the highest concentration of photocatalytic nanofiller (1.0 wt. %) that
The result of particle size distribution.
was justifiable considering the results of roughness.
Particle size (nm) Number of particle As can be seen from the Figure, the rejection efficiencies showed a
Pure TiO2 K-B-N-TiO2
progressive increase which were in a good agreement with the reduction
0-10 17 10 in roughness parameters. Surprisingly, the COD rejection efficacy were
10-20 33 29 as high as 99.9 %, independent of the extent of photocatalytic nanofiller
20-30 69 97
30-40 84 131
while filtering activated sludge as fouling agent compared with other
40-50 114 77 fouling agent i.e. BSA rejection. This may be due to different nature of
50-60 23 29 the used fouling agents.
60-70 11 8 To further examine resistivity of the synthesized membranes against
70-80 9 7
adherence of fouling agents on the surface of the respective membranes,
80-90 8 3
90-100 7 4 the fouling resistance parameters namely irreversible fouling resistance
(Rir), reversible fouling resistance (Rr) and total fouling resistance (Rt)
were calculated.
given in Table 3. From the Table, it evidently gets observed that hy­ Generally, the more Rr, the more antifouling property, showing
drophilic character was directly proportional with the extent of photo­ effectiveness of relevant nanofillers. The increase in Rr indicates anti­
catalytic nanofiller. Analysis of the data obtained from pure water flux fouling agents become deposited on the membrane surface loosely, and
indicates that it remarkably increased from 13.2 kg/m2.h for the refer­ can be eliminated using simple physical methods. While, the increase in
ence membrane to 22.4 kg/m2.h for M-0.1 and then to 39.2 kg/m2.h for Rir is in the relation with tight attachment of antifouling agents on the
M-0.5. However, at higher loading of photocatalytic nanofiller for M- membrane surface and not removable simply.
1.0, diminish in the pure water flux was observed ascribed to formation The changes trend in values of fouling resistance ratios obtained by
of nanofillers’ aggregates. The changes trend in the permeate of the the synthesized membranes using two different fouling agents i.e. BSA
photocatalytic nanocomposite NF membranes is consistent with solution and AS suspension are presented in Figure 6 a and b. Among the
improvement in porosity and void volume. In our previous published synthesized membranes, the M-0.5 membrane revealed the highest Rr
papers, the photocatalytic membranes blended with 0.5 wt. % of C, N, S- and the lowest Rir filtering both fouling agents substantiating the hy­
triple doped TiO2-ZnO, and of K-B-N-triple TiO2 showed the highest pure drophilic and smooth surface enables the concerned membrane to
water flux of 41.8 kg/m2.h and 37.3 kg/m2.h, respectively [25,40]. exclude irreversible deposition of the organic fouling agents.
The results of rejection percentage and FRR using 500 mg/l BSA Although progressive increase in loading of photocatalytic nanofiller
solution and activated sludge (AS) suspension are presented in Figure 5 a led to the betterment in hydrophilic properties, the antifouling charac­
and b. From the Figure, all photocatalytic nanocomposite NF mem­ teristics were weakened at the extent of 1.0 wt. % due to the increase in
branes revealed much higher rejection towards both fouling agents as the surface roughness. The results of this study was comparable with the
compared to those obtained for the reference membrane. These previous work conducted using powder milk solution and biologically
observed findings were in an inverse proportional with WCAs’ results. treated POME [40].
The FRR for the reference PES membrane is smaller than the pho­
tocatalytic membranes. The introduction of photocatalytic nanofiller 3.3.2. Evaluation of performance of the prepared membranes under a
into PES matrix led to betterment in hydrophilic character of the dynamic mode in a cross flow membrane set up
membrane surface (see the results of WCA, Table 3) and mitigation of In previous sections, the performance of the synthesized membranes
membrane roughness (see the results of AFM photos, Figure 3 and were surveyed under a static stream mode in a dead end filtration set up
Table 2). The reduction in ups and downs of membrane surface (i.e. filtering BSA and activated sludge (AS) solutions. In this section, the
reduction of roughness) gives rise to decreasing deposition of fouling performance of the prepared NF membranes would be checked in a
agents in/on the pores and surface of membrane which in turns dynamic stream regime while filtering biologically treated palm oil mill

4
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

Figure 2. The cross-sectional SEM photos obtained at magnifications of 20 µm for the fabricated NF membranes (a) the nascent PES, (b) 0.1 wt.%, (c) 0.5 wt.%, and
(d) 1 wt.%.

effluent (POME) at constant working pressure of 5 bar, feed flow rate of and reduced consequently that to radical-ion species (O2− ) in order to
◦

150 L/h and COD concentration of 1000 mg/l, not examined in our produce oxidative species of OH◦ [26]. The photo-produced holes show
previous work [40]. their roles in photocatalytic reactions employing two paths, they either
Compared to the previous work carried out using with the same react with water molecules to generate OH◦ or degrade directly organic
photocatalytic nanofiller, the novelty of the present study is reflected in compounds into short chain intermediate compounds. Furthermore, the
this section and following section. The membranes performance were oxidative species i.e. O2− and OH◦ produced during photocatalytic re­
◦

exanimated as pure water flux and COD removal efficiency without/ actions give rise to decomposition of organic fouling agent on the
with visible light irradiation to determine contribution percentage of membrane surface, thereby, reduce the fouling phenomena.
adsorption and photocatalytic properties of the synthesized membranes. The hydrophilic and smooth nature of the photocatalytic membranes
The archived outcomes are shown in Figure S3a and b. From the surface can be considered as other causes to alleviate fouling and
Figure S3a, all the fabricated membranes revealed a declining trend in enhance flux. Amongst various membranes, the M-0.5 possessed the
flux values upon filtration of biologically treated POME. This reduction highest flux value that might be in the relation with uniform distribution
was sever at absence of visible light irradiation. of photocatalytic nanofiller, therefore the more exposure to the visible
Observation of such a declining behavior was justified by residing light, and being undergone more photocatalytic reactions.
many foulants, which exist in the field sample of biologically treated The performance of the fabricated membranes in terms of COD
POME, on the membranes surface through adsorption mechanism [40]. removal efficiency was revealed in Figure S3b. As can be seen from the
Therefore, the membranes exhibited high resistance towards the Figure, the COD removal efficiency remained more or less unaffected
permeation of water as a result of the fouled membrane surface. How­ without/with visible light irradiation in the case of the reference
ever, this reduction in flux was significantly mitigated after visible light membrane (M-0.0). In contrary, the photocatalytic membranes exhibi­
irradiation compared to counterpart conditions (dark conditions). ted the marked increase in the COD removal efficiency after irradiating
The presence of visible light was ineffective on the performance of visible light. More importantly, the COD removal showed a rising trend
the reference membrane since that was lack of photocatalytic nanofiller. by increasing exposure time of visible light. Such observations were
The photocatalytic membranes were activated at the presence of visible supported by photocatalytic reactions of photocatalytic nanofillers
light via excitation of photocatalytic nanofiller continued by production embedded inside the polymer matrix. The COD removal efficiency
of electron and hole pairs. The electrons exited to the higher electron reached to maximum value (100 %) after a 150-min filtration for all the
layers reacted with O2 molecules available at the surrounding media, photocatalytic membranes, in contract to 71 % for the reference

5
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

Figure 3. AFM images of the fabricated NF membranes (a) the nascent PES, (b) 0.1 wt.%, (c) 0.5 wt.%, and (d) 1 wt.%.

significant contrast between the values reported as flux and COD

Table 2
removal efficiency at absolute dark and light conditions. Most impor­
The extent of roughness parameters of the membrane skin surface obtained for
tantly, continuously visible light irradiating had the highest roles in
the synthesized membranes.
improvement of flux and COD removal efficiency for the M-0.5 while
Membrane code Roughness parameters, nm filtering biologically treated POME in comparison with when the
Ra Rq Rz
membrane had experienced only dark conditions for early 30 min. As
M-0.0 (Reference) 47.1 59.6 304.5 mentioned above, the visible light irradiation was not impactful in the
M-0.1 27.1 29.8 187.7
case of the reference membrane.
M-0.5 19.2 21.7 99.8
M-1.0 34.9 48.3 234.9 All these observations were supported by photo-enhanced reactions,
and therefore, activated free radicals generation due to the presence of
the photocatalytic nanofillers leading to more degradation of organic
membrane. Based on the obtained results, the M-0.5 displayed the best pollutants and insignificant reduction of flux. In the following, the effect
performance amongst the others, therefore, was considered as a chosen of three variables i.e. feed flow rate (Q), operating pressure (P) and feed
membrane and employed to do rest of the experiments. concentration (COD) on the performance of the chosen membrane (M-
Changes in flux and COD removal efficiency were studied for the M- 0.5) would be discussed in details as dynamic mode in the cross flow
0.5 having the best performance amongst the others, and compared with filtration set up.
the reference membrane to determine how much different reactions i.e.
adsorption processes at dark conditions, photocatalytic reactions at light 3.3.3. Statistical evaluation of performance tests of the chosen membrane
irradiation conditions and combination of both reactions would be (M-0.5) under a dynamic stream using RSM
effective on the mentioned functional parameters. The results obtained As mentioned already, the performance experiments of the chosen
from this investigation for both M-0.5 and the reference membranes are membrane were designed using CCD of response surface methodology
presented in Figure S4 a and b. Based on the Figure, there was a (RSM) by means of Design Expert Software. In this study, the

6
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

Figure 4. Overall porosities obtained for the reference and photocatalytic nanocomposite membranes at various mass loadings of K-N-B-triple TiO2 photocatalyst.

performance of the chosen membrane was assessed at three different

Table 3
levels of operating pressure (5, 7, 9 bar), feed flow rate (50, 100, 150 L/
WCAs and PWFs obtained for the reference and photocatalytic nanocomposite
h) and COD concentration (abbreviated as COD Conc.; 500, 7000, 1000
membranes at various mass loadings of K-N-B-triple TiO2 photocatalyst.
mg/L) as variables in dynamic mode using the cross flow filtration under
Parameters Membrane code
continuous visible light irradiation. In order to illustrate curvature ef­
M-0.0 M-0.1 M-0.5 M-1.0
fects of variables and their interactions on the responses, the experi­
WCA, ◦ 63.2 57.5 52.6 50.6 mental data were fitted to higher degree polynomial equations i.e. cubic.
PWF, kg/m2.h 13.7 22.4 39.2 30.1
However, the original models were modified by elimination of

Figure 5. The results of rejection efficiency and FRR obtained for photocatalytic nanocomposite NF membranes filtering (a) BSA and (b) activated sludge
(AS) suspension.

7
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

Figure 6. Results of different fouling resistance parameters obtained for the photocatalytic NF membranes filtering (a) BSA and (b) activated sludge (AS) suspension.

insignificant terms according to obtained statistical results. The results (5000 mg/l).
obtained from analysis of variance (ANOVA) are given in Table 4. Ac­ The maximum amount of flux (33.9 kg/m2h) was found at operating
cording to statistical studies, models that have probability (p-values) of pressure of 9 bar and feed flow rate of 50 L/min at the lowest amount of
less than 0.05 are significant, otherwise, they are insignificant. The high COD concentration i.e.1000 mg/l. The effective impact of feed flow rate
F-value (the ratio of model’s mean square to the mean square error) is (B model term) on the flux was attributed to the increase in cross flow
indicating high significance of the studied model. In addition, the high velocity from 0.35 to 1.0 m/s corresponded with feed flow rates of 50 to
determination coefficient value of R2 substantiates the high degree of 150 L/h, respectively [25]. The values of feed flow rates of 50 and 150
accuracy and fitness of the used model. High Adjusted R2 confirms the L/h correspond with Reynolds numbers (Re) of 1270 to 4300 which are
remained terms in the model equation add value to the model. Adequate related to laminar and turbulent flows, respectively [25]. It is apparent
precision expresses the ratio of signal to noise, and ratios greater than 4 from the Figure S5a that increasing effect of operating pressure (A) and
indicate adequate model discrimination. The coefficient of variation (C. feed flow rate (B) on the flux was more evident at the lowest amount of
V.) or relative standard deviation is the standard deviation expressed as COD concentration (1000 mg/l). Figure S6 reveals the 3-dimentional
a percentage of the mean. The low values of C.V. guarantee the accuracy response surface plots achieved for flux as function of operating pres­
and reliability of experimental results. Positive coefficients of model sure and COD concentration at various feed flow rates. As shown in the
terms are showing the increasing effect on the studied response, Figure, the increase in COD concentration and reduction in feed flow
whereas, the negative ones have a deteriorative effect on the response. rate reveals a mitigating trend in flux values while filtering biologically
Based on the given explanations and ANOVA Table, all regression treated POME under continuous visible light irradiation. The mecha­
models presented here are highly significant. nism behind this behavior is occupation of reaction sites of the mem­
The effect of different variables on the studied responses is visualized brane surface by excess COD concentration, followed by reduction in
using 3-dimensional response surface plots. Figure S5 reveals the 3- photocatalytic processes and therefore reduction in generation of
dimentional response surface plots achieved for flux as function of oxidative agents i.e. hydroxyl and superoxide radicals [25]. Conse­
operating pressure and feed flow rate at various COD concentrations. As quently, the most of contaminant molecules remain unreacted, and
clear from the Figure, changes trend in flux of the chosen membrane hence some small organic compounds can pass through the membrane
under continuous visible light irradiation are almost similar at different pores, thereby, COD removal efficiency and FRR would be deteriorated.
COD concentrations, ranging from 13.9 to 33.9 kg/m2.h. In contrast, the increase in feed flow rate (B model term) gives rise to
From the Table 4, model terms of A, B, BC and ABC with positive mitigation in membrane fouling via diminish in concentration polari­
coefficients were significant, while, model terms of C, AB, A2B, and AB2 zation [51].
with negative coefficients showed a declining effect on the response. Should feed flow rate increases, the cross flow velocity increases, and
Furthermore, model term A (operating pressure) with having the highest therefore, adsorption and deposition of fouling agents and formation of
positive coefficient value was the most significant model term, influ­ gel layer on the membrane surface are inhibited [51,52]. These results
encing flux as response. As can be seen from the Figure, by increasing are ascribed to enhanced mass transfer of rejected molecules back to
operating pressure from 5 to 9, flux was increased significantly. Ac­ bulk solution as well as increased wall shear stresses as a result of
cording to the obtained results, COD concentration (model term C) reduction of the local hydrodynamics [51]. The relationship between
showed decreasing effect on the flux especially at its highest amount cross flow velocity and flux is determined by the following equation

Table 4
The results of ANOVA for models obtained for the evaluated responses.
Flux=22.99+6.90A+4.90B-3.37C-1.81AB+2.24BC+2.31ABC-3.64A2B-4.34AB2
F-value Probability Lack of fit probability R2 Adj. R2 Adeq. precision C.V. % PRESS

30.40 < 0.0001 0.2816 0.9298 0.8788 17.306 7.48 287.32

FRR=+63.00-10.80A+15.00B-12.70C-5.75AB
2
F-value Probability Lack of fit probability R Adj. R2 Adeq. precision C.V. % PRESS
14.73 < 0.0001 0.0553 0.8403 0.7832 15.195 11.16 1669.51
COD removal=+59.35-8.10A+13.00B-12.70C-5.12AB+6.88BC 4.86A2+6.14B2+6.14C2+5.12ABC-5.38A2B
F-value Probability Lack of fit probability R2 Adj. R2 Adeq. precision C.V. % PRESS
56.16 < 0.0001 0.1505 0.9842 0.9667 32.965 4.35 790.94

8
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

[52]. 4. Conclusions
dC
N = JC − D + p(ζ) + q(τ) The photocatalytic nanocomposite NF membranes blended with
dy
various loadings of K-B-N triple TiO2 nanofiller were fabricated
where, D stands for Brownian diffusion coefficient, p(ζ) stands for employing conventional wet phase inversion method to investigate
migration of the solutes/particles, and q(τ) stands for effect of the local membranes performance under both static and dynamic modes using
hydrodynamics on the mass flux. filtration of three foulng agents i.e. BSA solution, activated sludge (AS)
Above-mentioned mechanisms are also dominant for reduction in suspension and biologically treated POME. The highlighted conclusions
other responses i.e. COD removal efficiency and FRR. are as follows. The best antifouling attribute obtained in static mode was
The 3-dimensional response surface plot of the changes trend in FRR related to M-0.5 with having the highest FRR of 93.7 % and 98 % after
towards operating pressure (A) and feed flow rate (B) at various COD filtration of BSA solution and AS suspension, respectively. Also, the
concentrations is displayed in Figure S7 Based on the portrayed Figure, referenced membrane showed lower flux reduction and higher COD
changes trend in FRR was similar at different COD concentrations. From rejection as a function of time while filtering biologically treated POME
ANOVA results, model terms of A (operating pressure), C (COD Conc.,) under continuous visible light irradiation in dynamic mode. Based on
and AB have negative effects on the response of FRR, whereas, only achieved results from CCD of RSM, the largest values of FRR and COD
model term of B (feed flow rate) shows a positive effect. As clear from removal efficiency was reported to be 97 % and 100 % at the lowest COD
the Figure, FRR was increased with the reduction in model term A concentration of 1000 mg/L, the highest feed flow rate of 150 L/min and
representing operating pressure and the increase in model term B rep­ the lowest operating pressure of 5 bar, respectively. While, the highest
resenting feed flow rate which was in the relation with above-mentioned value of flux (33.9 kg/m2h) was obtained at operating pressure of 9 bar,
explanations. At the highest feed flow rate (150 L/h), FRR values were feed flow rate of 50 L/h and COD concentration of 1000 mg/L. As a final
decreased significantly with the increase in the operating pressure conclusion, based on results obtained from our previous work and the
showing membrane fouling. present study, the photocatalytic NF membrane can be considered as a
These results can be ascribed to pore blocking accelerated by in­ promising contender to be applied in practical applications.
crease in operating pressure. While, at the lowest pressure (5 bar), the
increment in feed flow rate showed a positive effect on the FRR. 3- CRediT author statement
dimentinal surface response plots of FRR as a function of COD concen­
tration and operating pressure at various feed flow rates are displayed in Hadis Zangeneh: Conceptualization, Methodology.
Figure S8. Ali Akbar Zinatizadeh: Supervision, Designing, Reviewing and
In contrast to the Figure S7, changes trend of FRR at various feed Editing.
flow rates are different from each other that may be due to interference Zahra Rahimi: Conceptualization, Writing.
of other environmental causes such as temperature and moisture, not Sirus Zinadini: Co-supervision, Reviewing and Editing.
considered in this study. Morteza Sadeghi: Reviewing and Editing.
The FRR values were increased significantly with reduction in COD Mohamed Abdulgader: Reviewing and Editing.
concentration along with the increase in feed flow rate. The maximum
FRR value of 97 % was achieved at the lowest COD concentration (1000 Declaration of Competing Interest
mg/l) and operating pressure (5 bar), and the highest feed flow rate of
150 L/min. The authors declare that they have no known competing financial
The 3-dimentional response surface plots of COD removal efficiency interests or personal relationships that could have appeared to influence
as a function of operating pressure and COD concentration at various the work reported in this paper.
feed flow rates is indicated in Figure S9. As obvious from the Figure,
variations tendency in COD removal are similar at various feed flow Acknowledgments
rate. However, curvatures observed may be due to interference of other
environmental causes such as temperature and moisture, not considered The authors would like to gratefully thank financial support pro­
in this study. From ANOVA outcomes, model terms B, BC, B2, C2, ABC vided by Razi University (Kermanshah, Iran), Water and Wastewater
are known as positive terms in modified model, while, A, C, AB, A2, A2B Company (Kermanshah, Iran), and Vice President for Science and
are known as negative terms. Technology Water; Drought, Erosion and Environment Working Group
The COD removal efficiency was incremented by simultaneous (Tehran, Iran) to conduct this research studY.
reduction in COD concentration and operating pressure. The observed
results are justifiable based on stated reasons in before paragraphs. The Supplementary materials
3-dimentional surface response plots of COD removal efficiency as a
function of feed flow rate and operating pressure at various COD con­ Supplementary material associated with this article can be found, in
centrations is denoted in Figure S10. the online version, at doi:10.1016/j.cep.2021.108635.
The COD removal efficiency increased with the increase in feed flow
rate, while, reduction in operating pressure, ascribed to above- References
mentioned reasons. In overall, the largest value of COD removal effi­
ciency was reported to be 100 % at the lowest COD concentration of [1] N. Niknezhad, N. Azimi, S. Ahmadi, Ultrasound-assisted adsorption of Cr (VI) ions
from aqueous solution: Experimental study and CFD simulation, J. Appl. Res.
1000 mg/L, the highest feed flow rate of 150 L/min and the lowest Water Wastewater. 8 (2021) 41–49. https://arww.razi.ac.ir/article_1531.html
operating pressure of 5 bar. (accessed August 26, 2021).
The obtained results in this study were in a good accordance with our [2] Z. Rahimi, A.A. Zinatizadeh, S. Zinadini, Preparation and characterization of a high
antibiofouling ultrafiltration PES membrane using OCMCS-Fe3O4 for application in
previous published works done using chosen photocatalytic nano­
MBR treating wastewater, J. Appl. Res. Water Wastewater. 1 (2014) 13–17.
composite membranes modified by C, N-doped TiO2-CdS and C, N, S [3] M. Pirsaheb, M. Mohamadi, A.M. Mansouri, A.A. Zinatizadeh, S. Sumathi,
triple doped TiO2-ZnO under continuous visible light irradiation in cross K. Sharafi, Process modeling and optimization of biological removal of carbon,
nitrogen and phosphorus from hospital wastewater in a continuous feeding &
flow filtration system filtering waste stabilization pound’s effluent and
intermittent discharge (CFID) bioreactor, Korean J. Chem. Eng. 32 (2015)
biologically treated POME, respectively [25,38]. 1340–1353, https://doi.org/10.1007/s11814-014-0365-z.
[4] L. Lei, D. Huang, C. Zhang, R. Deng, S. Chen, Z. Li, F dopants triggered active sites
in bifunctional cobalt sulfide@nickel foam toward electrocatalytic overall water

9
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

splitting in neutral and alkaline media: Experiments and theoretical calculations, membrane, Compos. Part B Eng. 176 (2019), 107158, https://doi.org/10.1016/j.
J. Catal. 385 (2020) 129–139, https://doi.org/10.1016/j.jcat.2020.03.012. compositesb.2019.107158.
[5] K. Sharafi, A.M. Mansouri, A.A. Zinatizadeh, M. Pirsaheb, Adsorptive removal of [26] H. Zangeneh, A.A.L. Zinatizadeh, M. Habibi, M. Akia, M.H. Isa, Photocatalytic
methylene blue from aqueous solutions by pumice powder: Process modelling and oxidation of organic dyes and pollutants in wastewater using different modified
kinetic evaluation, Environ. Eng. Manag. J. 14 (2015) 1067–1078, https://doi.org/ titanium dioxides: A comparative review, J. Ind. Eng. Chem. 26 (2015) 1–36.
10.30638/eemj.2015.118. [27] Y. Gao, M. Hu, B. Mi, Membrane surface modification with TiO2-graphene oxide
[6] T. Naveed, N. Ahmed, S. Bhutto, N. Tunyo, D.Rais Hashmi, A sustainable solution for enhanced photocatalytic performance, J. Memb. Sci. 455 (2014) 349–356,
to treat textile effluent by employing combined coagulation, oxidation and https://doi.org/10.1016/j.memsci.2014.01.011.
ultrafiltration techniques, J. Appl. Res. Water Wastewater. 8 (2021) 66–70. htt [28] J. Wang, W. Mi, J. Tian, J. Dai, X. Wang, B X.L-C.P, Effect of calcinations of TiO2/
ps://arww.razi.ac.ir/article_1814.html (accessed August 26, 2021). ZnO composite powder at high temperature on photodegradation of methyl
[7] S. Abbasi, M. Mirghorayshi, S. Zinadini, A.A. Zinatizadeh, A novel single orange, Elsevier, 2013 (n.d.), https://www.sciencedirect.com/science/article/pii/
continuous electrocoagulation process for treatment of licorice processing S1359836812006038 (accessed September 9, 2019).
wastewater: Optimization of operating factors using RSM, Process Saf. Environ. [29] T. Hakoda, K. Matsumoto, A. Mizuno, K. Hirota, Role of metals loaded on a TiO2
Prot. 134 (2020) 323–332, https://doi.org/10.1016/j.psep.2019.12.005. surface in the oxidation of xylene in air using an electron beam irradiation/
[8] Y. Yang, G. Zeng, D. Huang, C. Zhang, D. He, C. Zhou, W. Wang, W. Xiong, B. Song, catalytic process, Appl. Catal. A Gen. 357 (2009) 244–249, https://doi.org/
H. Yi, S. Ye, X. Ren, In Situ Grown Single-Atom Cobalt on Polymeric Carbon Nitride 10.1016/j.apcata.2009.01.030.
with Bidentate Ligand for Efficient Photocatalytic Degradation of Refractory [30] S. Sun, J. Ding, J. Bao, C. Gao, Z. Qi, X. Yang, B. He, C. Li, Photocatalytic
Antibiotics, Small 16 (2020), 2001634, https://doi.org/10.1002/smll.202001634. degradation of gaseous toluene on Fe-TiO2 under visible light irradiation: A study
[9] P.E. Poh, W.-J. Yong, M.F. Chong, Palm Oil Mill Effluent (POME) Characteristic in on the structure, activity and deactivation mechanism, Appl. Surf. Sci. 258 (2012)
High Crop Season and the Applicability of High-Rate Anaerobic Bioreactors for the 5031–5037, https://doi.org/10.1016/j.apsusc.2012.01.075.
Treatment of POME, Ind. & Eng. Chem. Res. 49 (n.d.) 11732–11740. https://www. [31] N. Riaz, F.K. Chong, Z.B. Man, M.S. Khan, B.K. Dutta, Photodegradation of orange
academia.edu/19169976/Palm_Oil_Mill_Effluent_POME_Characteristic_in_High_ II under visible light using Cu-Ni/TiO2: Influence of Cu: Ni mass composition,
Crop_Season_and_the_Applicability_of_High_Rate_Anaerobic_Bioreactors_for_the_ preparation, and calcination temperature, Ind. Eng. Chem. Res. 52 (2013)
Treatment_of_POME (accessed October 20, 2020). 4491–4503, https://doi.org/10.1021/ie303255v.
[10] (PDF) Review of Current Palm Oil Mill Effluent (POME) Treatment Methods: [32] T. Sano, N. Mera, Y. Kanai, C. Nishimoto, S. Tsutsui, T. Hirakawa, N. Negishi,
Vermicomposting as a Sustainable Practice, (n.d.). https://www.researchgate.net/ Origin of visible-light activity of N-doped TiO2 photocatalyst: Behaviors of N and S
publication/213965918_Review_of_Current_Palm_Oil_Mill_Effluent_POME_Treat atoms in a wet N-doping process, Appl. Catal. B Environ. 128 (2012) 77–83,
ment_Methods_Vermicomposting_as_a_Sustainable_Practice (accessed October 20, https://doi.org/10.1016/j.apcatb.2012.06.034.
2020). [33] Y.H. Peng, G.F. Huang, W.Q. Huang, Visible-light absorption and photocatalytic
[11] N.S. Azmi, K.F.M. Yunos, Wastewater treatment of palm oil mill effluent (POME) activity of Cr-doped TiO2 nanocrystal films, Adv. Powder Technol. 23 (2012) 8–12,
by ultrafiltration membrane separation technique coupled with Adsorption https://doi.org/10.1016/j.apt.2010.11.006.
Treatment as Pre-treatment, Agric. Agric. Sci. Procedia. 2 (2014) 257–264, https:// [34] L. Zhang, J. Zhou, J. Li, G. Liu, X. Lin, B. Mao, R. Liu, S. Zhang, J.Q. Wang, Surface
doi.org/10.1016/j.aaspro.2014.11.037. structural reconstruction for optical response in iodine-modified TiO2
[12] A.L. Ahmad, S. Ismail, S. Bhatia, Water recycling from palm oil mill effluent photocatalyst system, J. Phys. Chem. C. 118 (2014) 13726–13732, https://doi.org/
(POME) using membrane technology, Desalination 157 (2003) 87–95, https://doi. 10.1021/jp503966r.
org/10.1016/S0011-9164(03)00387-4. [35] H. Li, S. Yin, Y. Wang, T. Sato, Efficient persistent photocatalytic decomposition of
[13] A.A.M. Azoddein, H. Bin Haris, F.A.M. Azli, Treatment of palm oil mill effluent nitrogen monoxide over a fluorescence-assisted CaAl2O4:(Eu, Nd)/(Ta, N)-codoped
(POME) using membrane bioreactor, Malaysian J. Anal. Sci. 19 (2015) 463–471. TiO2/Fe2O3, Appl. Catal. B Environ. 132–133 (2013) 487–492, https://doi.org/
https://www.researchgate.net/publication/282683583_Treatment_of_palm_oil_mi 10.1016/j.apcatb.2012.12.026.
ll_effluent_POME_using_membrane_bioreactor (accessed October 20, 2020). [36] K. Kondo, N. Murakami, C. Ye, T. Tsubota, T. Ohno, Development of highly
[14] W.L. Liew, M.A. Kassim, K. Muda, S.K. Loh, A.C. Affam, Conventional methods and efficient sulfur-doped TiO2 photocatalysts hybridized with graphitic carbon nitride,
emerging wastewater polishing technologies for palm oil mill effluent treatment: A Appl. Catal. B Environ. 142–143 (2013) 362–367, https://doi.org/10.1016/j.
review, J. Environ. Manage. 149 (2015) 222–235, https://doi.org/10.1016/j. apcatb.2013.05.042.
jenvman.2014.10.016. [37] F. tang Li, X. jing Wang, Y. Zhao, J. xing Liu, Y. juan Hao, R. hong Liu, D. shun
[15] Z.Z. Yang, C. Zhang, G.M. Zeng, X.F. Tan, H. Wang, D.L. Huang, K.H. Yang, J. Zhao, Ionic-liquid-assisted synthesis of high-visible-light-activated N-B-F-tri-doped
J. Wei, C. Ma, K. Nie, Design and engineering of layered double hydroxide based mesoporous TiO2 via a microwave route, Appl. Catal. B Environ. 144 (2014)
catalysts for water depollution by advanced oxidation processes: A review, 442–453, https://doi.org/10.1016/j.apcatb.2013.07.050.
J. Mater. Chem. A. 8 (2020) 4141–4173, https://doi.org/10.1039/c9ta13522g. [38] H. Zangeneh, Z. Rahimi, A.A. Zinatizadeh, S.H. Razavizadeh, S. Zinadini, L-
[16] H. Yi, M. Yan, D. Huang, G. Zeng, C. Lai, M. Li, X. Huo, L. Qin, S. Liu, X. Liu, B. Li, Histidine doped-TiO2-CdS nanocomposite blended UF membranes with
H. Wang, M. Shen, Y. Fu, X. Guo, Synergistic effect of artificial enzyme and 2D photocatalytic and self-cleaning properties for remediation of effluent from a local
nano-structured Bi2WO6 for eco-friendly and efficient biomimetic photocatalysis, waste stabilization pond (WSP) under visible light, Process Saf. Environ. Prot. 136
Appl. Catal. B Environ. 250 (2019) 52–62, https://doi.org/10.1016/j. (2020) 92–104, https://doi.org/10.1016/j.psep.2020.01.022.
apcatb.2019.03.008. [39] H. Zangeneh, A.A. Zinatizadeh, S. Zinadini, M. Feyzi, E. Rafiee, D.W. Bahnemann,
[17] S. Yuan, M. Wang, J. Liu, B. Guo, Recent advances of SBA-15-based composites as A novel L-Histidine (C, N) codoped-TiO2-CdS nanocomposite for efficient visible
the heterogeneous catalysts in water decontamination: A mini-review, J. Environ. photo-degradation of recalcitrant compounds from wastewater, J. Hazard. Mater.
Manage. 254 (2020), https://doi.org/10.1016/j.jenvman.2019.109787. 369 (2019) 384–397, https://doi.org/10.1016/j.jhazmat.2019.02.049.
[18] Y. Lin, H. Liu, C. Yang, X. Wu, C. Du, L. Jiang, Y. Zhong, Gama-graphyne as [40] H. Zangeneh, A.A. Zinatizadeh, S. Zinadini, M. Feyzi, D.W. Bahnemann, A novel
photogenerated electrons transfer layer enhances photocatalytic performance of photocatalytic self-cleaning PES nanofiltration membrane incorporating triple
silver phosphate, Appl. Catal. B Environ. 264 (2020), 118479, https://doi.org/ metal-nonmetal doped TiO2 (K-B-N-TiO2) for post treatment of biologically treated
10.1016/j.apcatb.2019.118479. palm oil mill effluent, React. Funct. Polym. 127 (2018) 139–152, https://doi.org/
[19] Y. Yang, X. Li, C. Zhou, W. Xiong, G. Zeng, D. Huang, C. Zhang, W. Wang, B. Song, 10.1016/j.reactfunctpolym.2018.04.008.
X. Tang, X. Li, H. Guo, Recent advances in application of graphitic carbon nitride- [41] Q. Ling, J. Sun, Q. Zhou, Preparation and characterization of visible-light-driven
based catalysts for degrading organic contaminants in water through advanced titania photocatalyst co-doped with boron and nitrogen, Appl. Surf. Sci. 254 (2008)
oxidation processes beyond photocatalysis: A critical review, Water Res 184 3236–3241, https://doi.org/10.1016/j.apsusc.2007.11.001.
(2020), https://doi.org/10.1016/j.watres.2020.116200. [42] H. Zangeneh, A.A. Zinatizadeh, M. Feyzi, S. Zinadini, D.W. Bahnemann,
[20] Z. Rahimi, A.A. Zinatizadeh, S. Zinadini, M.C.M. van Loosdrecht, β-cyclodextrin Application of a novel triple metal-nonmetal doped TiO2 (K-B-N-TiO2) for
functionalized MWCNTs as a promising antifouling agent in fabrication of photocatalytic degradation of Linear Alkyl Benzene (LAB) industrial wastewater
composite nanofiltration membranes, Sep. Purif. Technol. 247 (2020), 116979, under visible light, Mater. Sci. Semicond. Process. 75 (2018) 193–205, https://doi.
https://doi.org/10.1016/j.seppur.2020.116979. org/10.1016/j.mssp.2017.11.040.
[21] Z. Rahimi, A.A. Zinatizadeh, S. Zinadini, Milk processing wastewater treatment in a [43] D. Chen, D. Yang, Q. Wang, Z. Jiang, Effects of boron doping on photocatalytic
bioreactor followed by an antifouling O-carboxymethyl chitosan modified Fe3O4/ activity and microstructure of titanium dioxide nanoparticles, Ind. Eng. Chem. Res.
PVDF ultrafiltration membrane, J. Ind. Eng. Chem. 38 (2016) 103–112. 45 (2006) 4110–4116, https://doi.org/10.1021/ie0600902.
[22] Z. Rahimi, A.A. Zinatizadeh, S. Zinadini, Membrane bioreactors troubleshooting [44] Y. Cong, J. Zhang, F. Chen, M. Anpo, Synthesis and characterization of nitrogen-
through preparation of a high antifouling PVDF ultrafiltration mixed matrix doped TiO2 nanophotocatalyst with high visible light activity, J. Phys. Chem. C.
membrane blended with O-carboxymethyl chitosan-Fe3O4 nanoparticles, Environ. 111 (2007) 6976–6982, https://doi.org/10.1021/jp0685030.
Technol. (2018) 1–30. [45] P. Van De Witte, P.J. Dijkstra, J.W.A. Van Den Berg, J. Feijen, Phase separation
[23] Z. Rahimi, A.A. Zinatizadeh, Ultrasound-induced settleability and membrane processes in polymer solutions in relation to membrane formation, J. Memb. Sci.
filterability of activated sludge treating milk processing wastewater, Appl. Water 117 (1996) 1–31, https://doi.org/10.1016/0376-7388(96)00088-9.
Sci. 8 (2018) 176. [46] H. Strathmann, K. Kock, The formation mechanism of phase inversion membranes,
[24] S. Zinadini, V. Vatanpour, A.A. Zinatizadeh, M. Rahimi, Z. Rahimi, M. Kian, Desalination 21 (1977) 241–255, https://doi.org/10.1016/S0011-9164(00)88244-
Preparation and characterization of antifouling graphene oxide/polyethersulfone 2.
ultrafiltration membrane: application in MBR for dairy wastewater treatment, [47] H. Strathmann, K. Kock, P. Amar, R.W. Baker, The formation mechanism of
J. Water Process Eng. 7 (2015) 280–294. asymmetric membranes, Desalination 16 (1975) 179–203, https://doi.org/
[25] H. Zangeneh, A.A. Zinatizadeh, S. Zinadini, M. Feyzi, D.W. Bahnemann, 10.1016/S0011-9164(00)82092-5.
Preparation ultrafine L-Methionine (C,N,S triple doped)-TiO2-ZnO nanoparticles [48] S. Kar, A. Pal, M. Subramanian, J. Nuwad, R.C. Bindal, S. Chattopadhyay, P.
and their photocatalytic performance for fouling alleviation in PES nanocomposite K. Tewari, High-throughput mixed-matrix membrane with superior anti-bacterial

10
H. Zangeneh et al. Chemical Engineering and Processing - Process Intensification 169 (2021) 108635

properties: A facile approach towards development of point-of-use water [51] I. Koyuncu, D. Topacik, Effect of Cross Flow Velocity, Feed Concentration, and
purification device, Chem. Eng. J. 297 (2016) 193–206, https://doi.org/10.1016/j. Pressure on the Salt Rejection of Nanofiltration Membranes in Reactive Dye Having
cej.2016.03.147. Two Sodium Salts and NaCl Mixtures: Model Application, J. Environ. Sci. Heal. -
[49] Y. Yang, S. Qiao, R. Jin, J. Zhou, X. Quan, A novel aerobic electrochemical Part A Toxic/Hazardous Subst. Environ. Eng. 39 (2004) 1055–1068, https://doi.
membrane bioreactor with CNTs hollow fiber membrane by electrochemical org/10.1081/ESE-120028413.
oxidation to improve water quality and mitigate membrane fouling, Water Res 151 [52] Z. Cui, W. Peng, Y. Fan, W. Xing, N. Xu, Effect of cross-flow velocity on the critical
(2019) 54–63, https://doi.org/10.1016/j.watres.2018.12.012. flux of ceramic membrane filtration as a pre-treatment for seawater desalination,
[50] A. Rahimpour, S.S. Madaeni, A.H. Taheri, Y. Mansourpanah, Coupling TiO2 Chinese J. Chem. Eng. 21 (2013) 341–347, https://doi.org/10.1016/S1004-9541
nanoparticles with UV irradiation for modification of polyethersulfone (13)60470-X.
ultrafiltration membranes, J. Memb. Sci. 313 (2008) 158–169, https://doi.org/
10.1016/j.memsci.2007.12.075.

11