Professional Documents
Culture Documents
Hadis Zangeneh, Ali Akbar Zinatizadeh, Sirus Zinadini, Mostafa Feyzi, Detlef
W. Bahnemann
PII: S1383-5866(18)30896-7
DOI: https://doi.org/10.1016/j.seppur.2018.09.030
Reference: SEPPUR 14928
Please cite this article as: H. Zangeneh, A.A. Zinatizadeh, S. Zinadini, M. Feyzi, D.W. Bahnemann, Preparation and
characterization of a novel photocatalytic self-cleaning PES nanofiltration membrane by embedding a visible-driven
photocatalyst boron doped-TiO2-SiO2/CoFe2O4 nanoparticles, Separation and Purification Technology (2018), doi:
https://doi.org/10.1016/j.seppur.2018.09.030
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Preparation and characterization of a novel photocatalytic self-cleaning PES
TiO2-SiO2/CoFe2O4 nanoparticles
Hadis Zangeneh1, Ali Akbar Zinatizadeh1,*, Sirus Zinadini1, Mostafa Feyzi2, Detlef W.
Bahnemann3
1
Environmental Research Center (ERC), Department of Applied Chemistry, Razi University,
Kermanshah, Iran.
2
Department of Physical Chemistry, Razi University, Kermanshah, Iran.
2
Institut fuer Technische Chemie, Leibniz Universität Hannover, Callinstrasse 3, D-
30167 Hannover, Germany.
* Corresponding author:
Email: zinatizadeh@gmail.com
1
Nomenclature:
List of symbol
2
Abstract
Novel self-cleaning PES nanofiltration membranes were successfully synthesized via the
phase inversion method with different amounts of the boron doped-TiO2-SiO2 cobalt ferrite
by the rejection of direct red 16 and COD removal of the biologically treated palm oil mill
effluent (POME). The addition of the B-TiO2-SiO2/CoFe2O4 nanoparticles into the polymeric
casting solution improved porosity, morphology, structure, pure water flux (PWF),
into the membrane pores (irreversible fouling). The prevention of the interaction between
organic foulant and hydrophilic membrane improves antifouling properties and wettability of
nanoparticle had the highest pure water flux and flux recovery ratio (FRR) with the best
surface which led to slightly reduction of PWF, FRR and separation performance. The
obtained results were also confirmed by contact angle measurement, scanning electron
3
1. Introduction
Water pollution has recently become a very serious challenge due to large volumes of
are treated by conventional techniques such as physical, chemical and biochemical processes
[2]. However, the above methods are not efficient due to production of large amount of
sludge, harmful byproducts, bio-resistant species, and complex operational procedures [3].
Membrane filtration has been considered as a promising technology in water and industrial
wastewater treatment fields due to its low cost, high separation performance, high effluent
quality, ease of setup, insignificant chemical consumption [4-5]. Nevertheless, this method
pores, which eventually leads to decline in separation performance, increase in the operation
Polymeric membranes are widely applied in water purification and wastewater treatment [8].
Polyethersulfone (PES) is one of the most popular polymeric membranes due to its excellent
chemical resistance, mechanical strength, high rigidity, thermal and biological stability [9-
10]. However, the hydrophobic nature of polyethersulfone membrane has limited its
application due to strong interaction between hydrophobic contaminants and the membrane
surface result in internal (pore clogging) and external (surface deposition) membrane fouling
monomers [15], cold plasma treatment, blending with hydrophilic polymers, surface-coating
nanoparticles (e.g. SiO2, ZrO2, Al2O3, TiO2, Fe3O4, and etc.) due to reproducibility,
4
convenient operation, capable of industrialization, good separation performances, excellent
thermal and chemical resistance and adaptability to the harsh wastewater environments and
TiO2 nanoparticles, as one of the most effective inorganic materials, improve permeability
and surface hydrophilicity, enhance mineralization of pollutants under ultraviolet (UV) light,
aggregation of TiO2 NPs in PES matrix due to its high surface energy [20], low photocatalyst
activity due to high recombination rate [21-22] and decomposing effect of UV light on the
PES membrane structure in long time operation [23] are main drawbacks of the pure
TiO2/PES membrane. Song, et al. [24] reported degradation of the structure of TiO2-
PVDFPEG membrane after 12 h UV irradiation. Hence, many strategies have been developed
to overcome the TiO2 drawbacks including doping of metal, non-metal elements [22] loading
of noble metals [22] and combining with other semiconductors [22]. Doping TiO2 with metal
or nonmetal ions has been considered as one of the most promising ways to develop TiO2-
based photocatalysts to extend its light absorption into the visible region [22].
Coupling magnetic nanoparticles with TiO2 nanoparticles has indicated the synergistic
influence that leads to better photocatalytic performance and lower agglomeration. Besides,
spinel ferrites, with the general formula of MFe2 O4, where M is a divalent cation like Co
single-component metal oxides [26] due to lesser band gap energy, [27]. As a result, a rapid
charge separation and slow charge recombination occur, and any transferred charge carriers
may participate in the subsequent redox reactions thereby resulting in higher photocatalytic
activity [26-27]. Li et al., [28] prepared PES–TiO2 hybrid membranes by phase inversion
method. They reported that thermal stability, porosity and permeation properties of the
5
amount of TiO2, significant aggregation of TiO2 nanoparticles could be observed [28]. An
investigation carried out by Wu et al. [29] indicated that blending TiO2 NPs with PES
and anti-fouling ability at 0.5 wt.% TiO2 [29]. Vatanpour et al [30] studied the effect of three
different types of TiO2 NPs including P25, PC105 and PC500 with various sizes on
nanofiltration. They established that biofouling of the membranes was reduced by embedding
TiO2 NPs and decreasing nanoparticles size. Furthermore, they reported that the P25-TiO2
NPs had the higher pure water flux and the water flux recovery ratio due to the lower
tendency to the aggregation [30]. Sotto et al. [31] also reported that the overall membrane
In this work, we synthesized boron-doped TiO2 in order to narrow the band gap of TiO2 and
added SiO2 to suppress the recombination of electron-hole and increase thermally stable
anatase phase with high crystallinity and large surface area [25]. Finally we coupled B doped-
fabricated by phase inversion method with introducing three loading amounts of the boron
reported up to this date. The surface hydrophilicity and structure were analyzed using
scanning electron microscopy (SEM), atomic force microscopy (AFM) and water contact
powder milk solution. Nanofiltration performance of the selected optimum membrane was
assessed by rejection of direct red 16 and treating the biologically treated palm oil mill
effluent (POME).
6
2. Experimental
2.1 Materials
(TBOT, 99 wt %), boric acid (H3BO3), Tetraethyl orthosilicate (TEOS, 99 wt %), ethanol
temperature Tg=225 °C) as polymer and dimethyl acetamide (DMAc) as a solvent were
purchased from BASF Co., Germany. Polyvinylpyrrolidone (PVP) with 25,000 g/mol
molecular weight as a pore forming in the coagulation process was obtained from Merck Co.,
Germany. Distilled water was used as the non-solvent agent. Chemical substances which
were used for chemical oxygen demand (COD) experiments were supplied by Merck
Company.
nanocomposite
The magnetite particles (CoFe2O4) were obtained according to previous work. Briefly, 0.1 g
of the as-prepared magnetic particles were added into 10 mL ethanol and was processed
ultrasonically for 1h (solution A). Subsequently, a desired amount of TBOT, TEOS and
H3BO3 were dispersed in 5 mL ethanol (solution B) and next solution A was drop-wise added
to the solution B (solution C). After that, distilled water and HCl were mixed with the
remaining ethanol. In the next stage, the ethanolic solution which was a mixture of distilled
water and HCl was added drop-wise to the solution C under continuous stirring. The resulting
was placed in a 150 ◦C oven for 12 h and then cooled to room temperature. The magnetic B
doped –TiO2-SiO2 nanocomposite was washed with ethanol and water six times and then
7
dried under vacuum at 60 ◦C for 6 h. Finally, the products were calcined in argon at 500 °C
for 2.5 h.
The crystalline phases of the prepared samples were identified by X-ray diffraction (XRD)
using a Philips X’ Pert (40 kV, 30 mA) X-ray diffractometer. X-ray diffractometer with Cu
Kα radiation (λ=1.542 Å). The Fourier Transform Infra-Red (FT-IR) spectra were recorded
on KBr disks containing the powder samples with an FT-IR spectrometer (MAGNA-560).
The scanning electron microscope ((SEM, VEGA TESCAN (Czech Republic)) was applied
to observe surface morphology and particle size of the prepared NPs. The UV–Vis diffuse
reflectance spectra (DRS) of the prepared samples were obtained using a Shimadzu 1800
spectrometer in the range of 200–800 nm. The magnetic properties of prepared samples were
membranes
The desired amount of the B doped TiO2-SiO2/CoF2O4 nanoparticles (0.03, 0.15 and 0.3g
corresponding to 0.1, 0.5 and 1 wt. %) was dispersed in DMAs as a solvent under
composition was represented at Table 1. Then, 6 g of PES (20 wt. %) and 0.3 g of PVP (1 wt.
%) was added to casting solution under stirring for 24 h. Following that, the obtained
homogeneous solution was sonicated for 30 min for removing air bubbles and was casted
with 150 µm thickness at room temperature. Afterward, the film was submerged into distilled
water at 25 °C. The membranes were kept in fresh distilled water for 24 h to remove residual
8
solvent and complete phase inversion process. Finally, they were dried between two filter
paper sheets.
The surface roughness of the prepared membranes, was investigated by using atomic force
software (version 1.8). A piece 1 cm2 of the prepared membrane was placed on specimen
holders and scanned in the tapping mode in the air (8×8 µm areas). Water contact angle
device (G10, KRUSS, Germany) was used to measure the static contact angle formed
between the water and the membrane surface by the sessile drop method at 25 °C. The cross-
(Philips-X130 and Cambridge scanning electron microscopes) after the samples were frozen
Membrane type PES (wt%) PVP (wt%) B doped TiO2-SiO2 /CoF2O4 (wt %)
M0 20 1.0 0
M1 20 1.0 0.1
M2 20 1.0 0.5
M3 20 1.0 1.0
membranes
Filtration tests were carried out in batch mode using a batch, stirred dead-end filtration cell
with an effective membrane surface area of 12.56 cm2 and a volume capacity of 125 mL (Fig.
1Sa). Membrane cell was attached to a nitrogen gas line to force the liquid through the
filtration experiments.
The pure water flux was measured at pressure of 5 bar after the membrane was compressed at
6 bar to achieve steady permeate for 30 min. Each wet membrane was irradiated by a 100 W
9
tungsten lamp (visible light) about 60 min before that were used. The following equation (1)
M
J w,1 (1)
A. t
Where, M, A and t are the weight of permeated water (kg), membrane area and (m2) and
The antifouling and self-cleaning property of the prepared membrane was conducted using
following procedures. After measuring pure water flux (Jw1), a powder milk solution (8000
ppm) was fed in filtration cell and permeation flux (Jp (kg/m2h)) was measured. Next, the
fouled membrane was placed in deionized water (about 1h as rinsing time) to eliminate
loosely bound protein and the second water flux (Jw2) of washed membranes were again
measured. This experiment was repeated again with the same condition with difference that
the washing step of the fouled membrane using deionized water was done under visible light
irradiation (about 1h). The flux recovery ratio (FRR) and fouling resistance parameters,
including the total fouling ratio (Rt), reversible fouling ratio (Rr) and irreversible fouling ratio
J w,2
FRR (%) ( ) 100 (2)
J w,1
J w,2 J p
Rr ( ) 100 (3)
J w,1
J w,1 J w,2
Rir ( ) 100 (4)
J w,1
The membrane performance was investigated by treating an aqueous solution direct red 16
with the concentration of 50 ppm and biologically treated palm oil mill effluent (POME) with
COD concentration of 1000 mg/l in separate experiments. The biologically treated POME
was collected from effluent of second aerobic lagoon (Jugra palm oil mill, Kualalumpur,
10
Malaysia). Characteristics of direct red 16 and the biologically treated POME are presented in
Table 2.
Direct red 16
Molecular Formula C26H17N5Na2O8S2
Molecular Weight 637.55
Molecular Structure
Direct red 16 concentration was determined using a UV–vis spectrophotometer (Pharo 300,
Merck) at λmax = 526.6 nm. The DR16 removal efficiency at regular intervals was calculated
Cp
R (%) (1 ) 100 (5)
Cf
The standard methods (5220D) procedures were followed in order to measure COD of the
biologically treated POME [32]. The COD removal (CODrem) efficiency was calculated using
Eq. 6.
where, CODf and CODp are the COD concentration of the feed and the permeate solution,
respectively.
11
A UV–Vis spectrophotometer (Pharo 300, Merck) was used to evaluate the color removal
efficiency from the biologically treated POME. The biologically treated POME and treated
samples were scanned at 400-800 nm. The following equation was used to calculate the color
where, Af and Ap are the average absorption of the biologically treated POME and permeate
The long term performance of the bare (M0) and optimized membranes (M2) was carried out
in a plexiglass cross flow system equipped with halogen lamp (500 W, λ ˃ 400 nm) (Fig.
1Sb). The halogen lamp was held in a distance of 25 cm from the membrane cross flow setup.
The biologically treated POME samples (with COD of about 1000 mg/l) were filtrated 12 h
Fig. 1a, SEM image, indicates the morphology boron doped TiO2-SiO2/CoFe2O4 magnetic
nanoparticles. From the SEM image, it was found that the prepared boron doped TiO2-
UV–Vis diffuse reflectance spectra (DRS) of the pure TiO2 and boron doped TiO2-
SiO2/CoFe2O4 magnetic nanocomposite is revealed in Fig. 1b. The band gap of the prepared
photocatalyst was also determined by the Tauc plots (Fig. 1c) [34]. From the Fig. 1b, it can
be clearly detected that the absorption edge position of the boron doped TiO 2-SiO2/CoFe2O4
12
magnetic nanoparticles has a red shift to visible region and extended up to longer
wavelengths comparing to the pure TiO2. The absorption edge of the pure TiO2 and B-TiO2–
SiO2/CoFe2O4 samples were at about 390, 655 nm corresponding to 3.17 and 1.9 eV,
respectively. The PL results (Fig. 1d) showed the B-SiO2-TiO2/CoFe2O4 had the lowest PL
intensity because the couple of TiO2, SiO2 and CoFe2O4 as a heterojunction structure
suppress the electron-hole recombination rate and supply charge carriers separation [35].
(a)
80
B-TiO2-SiO2/CoFe2O4 Pure TiO2
70 Pure TiO2 B-TiO2-SiO2/CoFe2O4
60
(α*λν)^0.5
Absorbance (a.u)
50
40
30
20
10
(b) (c)
13
1000
B-TiO2
900
B-TiO2-SiO2
800
B-TiO2-SiO2/CoFe2O4
700
Intensity
600
500
400
300
200
100
0
400 450 500 550 600 650 700 750 800
Wavelength (nm)
(d)
Fig. 1. (a) SEM images, (b) DRS spectra, (c) Tauc plot and (d) PL spectra of the prepared
photocatalysts.
The XRD, FT-IR and VSM result of the boron doped TiO2-SiO2/CoFe2O4 magnetic
nanoparticle was displayed in Fig. 1Sa-c, respectively. From the Fig. 1Sa, all the reflection
peaks are in accordance with the anatase TiO2 (PDF Card #89-4921) and spinel CoFe2O4
(PDF Card #22-1086 (reflection peaks at 29.9◦ (2 2 0), 35.2◦ (3 1 1), 42.8◦ (4 0 0), 53.3◦ (4 2
2), 56.7◦ (5 1 1) and 62.3◦ (4 4 0)). The average crystalline size of anatase of boron doped
Fig. 1Sb represents the FTIR spectra of boron doped TiO2-SiO2/CoFe2O4 magnetic
nanoparticles from 5000 to 400 cm-1. The observed peaks at 550 cm-1 was due to Ti–O
stretching vibration and the peaks at 3400 and 1650 cm-1 are corresponded with the stretching
and bending vibration –OH, respectively [38]. The peaks at 405 cm-1 and 583 cm-1 are related
to metal–oxygen (Fe–O and Co–O) bonds [36] and the peaks at 375–800 cm-1 are assigned to
The magnetization curves of boron doped TiO2-SiO2/CoFe2O4 was shown in Fig. 1Sc. The
VSM result indicates that the coercive force (Hc) is 1000. The saturation magnetization (Ms)
14
value of boron doped TiO2-SiO2/CoFe2O4 is about 21 emu/g which is lower compared to pure
Fig. 2 displayed the cross section SEM micrograph of the fabricated membranes and exhibits
the induced changes in the membrane morphology with adding nanoparticles. All the
fabricated membranes composed of a dense top layer, porous sub layer with a finger-like
macroporous structure resulting in the same asymmetric membrane structure [39]. From the
micrographs, it is clearly was seen that the finger-like structure and membrane porosity was
hydrophilicity of casting solution, which promotes fast exchange between solvent (DMAc)
and non-solvent (water) in coagulation bath thereby expanding the population of pores [40].
However, the addition of the modified TiO2 NPs with the amounts of more than 0.5 wt. %
leads to form a denser skin-layers with increased thickness and sub-layers with lower
porosities. Therefore, the pore radius and membrane porosity reduce due to the increase in the
viscosity of casting solution which hinders the exchange of solvent and non-solvent and
inhibits the formation of large pore radius. Also, the agglomeration of NPs occur on the
Two and three-dimensional AFM photographs exhibiting the surface roughness of the
prepared membrane as shown in Fig. 3. In these images, the brightest area demonstrates the
highest point of the membrane surface, and the dark regions illustrate the valleys or
membrane pores. The surface roughness of pristine PES membrane was higher than the
15
blended NPs membranes accordingly the AFM images. Incorporating the boron doped TiO2-
SiO2/CoFe2O4 magnetic nanoparticles in the casting solution from 0.1 to 0.5 %, led to
smoother membrane surface which implies on higher fouling resistance ability due to
noted that the agglomeration of NPs was occurred at higher concentration of the NPs (> 0.5
%) which increased the surface roughness at 1 % NPs blended membrane comparing to other
modified membranes [11, 39]. The same result was observed by Vatanpour et al., [41] for
PES membrane blended with TiO2 coated with multiwalled carbon nanotubes.
Table 3 presents roughness parameters fabricated membranes including roughness (Sa), root
mean square of the Z data (Sq) and the height difference between the highest peak and the
lowest valley (Sy) at a scan size of 2 μm × 2 μm. The unfilled PES membrane has the highest
surface roughness, and the roughness parameters were reduced by addition of NPs from 0.1
to 0.5 wt %. The Sa value reduced from 47.2 nm (puv, re PES) to 13.92 nm (0.1 % NPs-
Table 3. Surface roughness parameters of the prepared PES membranes resulted from
Roughness parameters
Membrane Sa (nm) Sq (nm) Sy (nm)
M0 47.11 59.60 304.46
M1 13.92 17.69 121.41
M2 12.60 15.97 117.45
M3 14.62 22.67 255.15
16
a b
c d
e f
g h
Fig. 2. The cross-section morphology SEM images of the fabricated PES membranes a and
b) bare PES (M0), c and d) 0.1 wt.% (M1), e and f) 0.5 wt.% (M2), g and d) 1 wt.% (M3) .
17
a) M0
b) M1
c) M2
18
d) M3
Fig. 3. Two and three dimensional AFM photographs of the prepared PES membranes with
different weight fraction NPs (a) bare PES (M0), (b) 0.1 wt.% (M1), (c) 0.5 wt.% (M2), (d) 1
wt.% (M3).
The water permeability of the fabricated membranes was shown in Fig. 4a. From the Fig.,
pure water flux values were about 16.7, 27.2, 38.1 and 21 kg/m2.h for bare PES, 0.1 % NPs-
PES, 0.5 % NPs-PES and 1% NPs-PES, respectively. The results were evidence for the
significant impact of the NPs on hydrophilicity enhancement of the membrane surface [28,
29]. The highest pure water flux was achieved at 0.5 % of NPs while according to the results
of contact angle (Fig. 4b), the surface hydrophilicity of 1% of NPs was higher. At high
concentration of the NPs (1%), agglomeration of the nanoparticles on the membrane surface
leads to increase pore blocking/clogging and reduce membrane porosity [31] which was
confirmed by SEM photographs. From the Fig. 4b, all of the blended membranes had a
considerable reduction of contact angle values compared to unfilled PES membrane. The
reduction of interface energy with the addition of the hydrophilic NPs to the casting solution
increased migration of the hydrophilic NPs towards the blended membrane surfaces during
the phase inversion process which improved the membrane hydrophilicity [41-42].
19
45
40 38.1
35
kg/m2.h) 30
25.2
25
21
20
16.7
15
10
0
M0 M1 M2 M3
Membrane type
(a)
70
62.5
60
55.1
51.3
49.3
50
Contact angle (o)
40
30
20
10
0
M0 M1 M2 M3
Membrane type
(b)
Fig. 4. a) Pure Water flux after 60 min at 5 bar, and b) static contact angel analysis of the
prepared membranes.
20
3.3 Fouling resistance of the prepared membranes
The fouling resistance of membranes is one of the most important parameters showing
characteristics of the membrane in long term performance. Fig. 5 shows the trend of
permeation flux versus time during milk powder solution filtration for all of the prepared
membranes. From the Fig. 5, the decreasing trend was observed during filtration for all of the
membranes. But M3 showed higher flux after 90 min filtration resulting higher fouling
resistance. The permeation flux was 5.8 kg/m2 h for the bare PES, whereas this value was 9.5,
Fig. 5 shows a great reduction in the permeate flux when pure water is replaced by milk
powder solution and then, the flux becomes stable. This trend can be attributed to the
concentration polarization resulted from accumulation of the proteins near the membrane
M0 M1 M2 M3
17
15
Permation flux (kg/m2.h)
13
11
5
15 30 45 60 75 90
Time (min)
Fig. 5. Permeation flux of the prepared membranes versus time during milk powder solution
filtration.
21
In this study, the antifouling property of the prepared membrane was investigated by filtration
of 8000 ppm powder milk solution. Fig. 6a shows the FRR ratios of fabricated membranes
with different cleaning conditions after milk powder filtration. The FRR ratio of M2 was
maximum compared to the unfilled PES membrane and other modified membranes. The
value of FRR ratio for M2 was about 84.6, 98 and 99% cleaning with deionized water,
cleaning with deionized water under visible light or UV illumination, respectively. The FRR
value of the unfilled PES membrane was almost similar under different conditions. The
blended membrane due to higher hydrophilicity, smoother surface, and photocatalytic activity
Under UV or visible light illumination, the strong oxidant reagents like, hydroxyl radicals
(OH•), superoxide radical anions (O2•−) and hydroperoxyl radicals (HOO•) were generated on
the membrane surface and pores and degraded organic or bioorganic foulant to CO 2 and H2O
or decomposed to simple intermediates, thereby reducing membrane fouling [22, 44]. Also,
through producing a water layer on the membrane surface which increase hydrophilicity and
reduce adsorption of foulants [4, 44]. This should be noted that the polymeric membrane may
be destroyed under long time UV light irradiation [4]. Therefore, the visible driven
photocatalyst was the better modifier compare to pure TiO 2 for the polymeric membranes.
In more detail, the irreversible fouling ratio (Rir) was calculated with visible light cleaning
(Fig.6b). The irreversible fouling ratio showed a decreasing trend from the unfilled PES
(47.42%) to the M2 (3%) due to the photocatalytic degradation of organic foulant on the
membrane surface. The value of Rir/Rt ratio for M2 (0.06) was considerably low compared to
the other fabricated membranes. This value for the unfilled PES membrane was about 0.83.
22
The Rr/Rt ratio is related to loose adsorption of organic foulant on membrane surface and the
100
90
80
70
60 Cleaning with DI
FRR (%)
30
20
10
0
M0 M1 M2 M3
Membrane type
a))
100
Rt Rr Rir
90
80
Resistance parameters (%)
70
60 57
52
47.42 49
50 44 43
40 33
30
30
20 13
9.58 11
10 3
0
M0 M1 M2 M3
Membrane type
b))
23
3.4 Separation performance of the prepared membranes
Fig. 7a demonstrates the performance of the prepared membranes in terms of the rejection
value for direct red 16. The results showed that all the modified membranes indicated better
separation performance in comparison with the unfilled PES membrane. The highest removal
efficiency of direct red 16 was 98 % with the 0.5 wt. % of NPs (M2). From the literature,
surface was negatively charged at neutral pH. Furthermore, the repulsion interaction between
a negative membrane surface and dye molecules with negative charge was promoted the
the hydrophilic layer on membrane surface that inhibited the interaction between
hydrophobic materials like dye molecules and membrane surface. However, dye rejection
decreases when the NPs wt. % increases to 1 % due to increase of membrane pore radius
[48]. The digital photograph of direct red solution and permeates is shown in Fig. 7b.
100 96 98 96
90
90
80
Dye removal (%)
70
60
50
40
30
20
10
0
M0 M1 M2 M3
Membrane type
(a)
24
(b)
Fig. 7. a) Dye rejection performance of the prepared membranes (5 bar, original pH, 50 ppm
direct red 16, after 60 min filtration) and b) The photograph of original solution of direct red
The performance of the prepared membranes was evaluated in the terms of the dye removal
and COD removal from the biologically treated POME (Fig.8a) in a dead end system.
Accordingly Fig.8a, the dye removal was about 88, 94, 99 and 92 % for M0, M1, M2 and M3
while their COD removal efficiencies were 75, 84, 98 and 80 %, respectively. From the
results, all of the embedded membranes especially M2 (0.5 wt% of NPs) indicated excellent
performances and higher separation properties. The presence of the wide range of organic
compound with high molecular weight in the biologically treated POME wastewater could
not allow passing through the small pores of prepared membranes [49].
25
Dye removal (%) COD removal (%)
99 98
100 94 92
88
84
80
80 75
Removal efficiency (%)
60
40
20
0
M0 M1 M2 M3
Membrane type
(a)
(b)
Fig. 8. a) Performance of the prepared membranes for post treatment of the biologically treated
POME (initial COD=1000 mg/l, pressure=5 bar and original pH= 5 after 120 min) in a dead end
system, and b) The photograph of the biologically treated POME and permeation of the M2
membrane.
26
In order to investigate the effect of photocatalyst activity on the membrane performance, the
performance of M2 as the optimum membrane was investigated in the cross flow system. The
permeation flux of M2 at presence or absence of visible light irradiation during 120 min
filtration of the biologically treated POME is shown in Fig. 9. The permeation flux was
decreased during filtration process in the condition without light irradiation. It might be due
to the presence of the large amounts of foulant components in the biologically treated POME.
It can be attributed to deposit or adsorb more organic compounds on the membrane surface
which leads to significant concentration polarization and serious membrane fouling [50-51].
However, the permeation flux decline for the M2 under visible light irradiation was
negligible. The COD removal for M2 was about 100 and 98% at presence and absence of
25
20
15
10
0
30 50 70 90 110
Time (min)
Fig. 9. Permeation flux of M2 versus filtration time during post treatment of the biologically
treated POME in a cross flow system (initial COD=1000 mg/l, pressure=5 bar, Q=150 L/h at
original pH).
27
3.5 Long term filtration performance
The permeation flux of the M2 (0.5 wt %) and M0 membranes during 12 h filtration of the
biologically treated POME in a cross flow system are shown in Fig. 10. As can be seen in
Fig. 10, the permeation flux of the M2 decreases about 5.5 and 2 kg/m2.h after 12 h at
presence or absence visible light irradiation, respectively, while the value was 9.8 kg/m2.h for
M0. The flux decline of the M0 was significant during filtration process. The long term results
showed that the hydrophilic nature and photocatalytic role of the added NPs significantly
improved the fouling resistance or self-cleaning properties of the modified membrane. Also,
the COD removal efficiency of M0 and M2 was 78 and 100% which was not changed after 12
h. From the results, it can be claimed that the polymeric structure of the composite membrane
22
20
18
16
14
12
10
8
6
1 3 5 7 9 11
Time (h)
Fig. 10. Permeation flux of the M0 and M2 membranes versus filtration time during post
treatment of the biologically treated POME in a cross flow system (initial COD=1000 mg/l,
28
4. Conclusion
properties and fouling resistance of the blended membranes were improved compared to the
bare PES membrane. The modified membrane (M2) showed a high pure water flux of 38.1
kg/m2.h and high rejection for direct red 16 and post treatment of the biologically treated
POME as well as good self-cleaning performance and antifouling resistance. After a visible
light cleaning, the modified membrane enabled to reach a high recovery of membrane
performance after fouling, confirming their self-cleaning ability. In this work, the dye
removal efficiency of direct red 16 and the biologically treated POME were also improved
remarkably and reached 98% and 98 % after 60 min and 120 min, respectively. This
improvement was induced by coupling high photocatalyst properties and hydrophilic nature
of the used NPs. Long term performance for the M2 showed high permeation flux after 12 h
5. Acknowledgement
The authors would like to acknowledge Iran National Science Foundation (INSF) for the full
financial support provided for this research work. The authors would also like to thank Razi
29
References:
and evaluation its efficiency in removal of organic pollutants from water, Sep. Purif. Technol.
50 (2006) 147–155.
[2] Y. Yang, H. Wang, J. Li, B. He, T. Wang, S. Liao, Novel Functionalized Nano-TiO2
Loading Electrocatalytic Membrane for Oily Wastewater Treatment, Environ. Sci. Technol.
46 (2012) 6815−6821.
[3] R. W. Baker, Research needs in the membrane separation industry: Looking back, looking
[4] S.J. You, G.U. Semblante, S.C. Lu, R.A. Damodar, T.C. Wei, Evaluation of the
poly(acrylic acid) membrane with self-assembled TiO2, J. Hazard. Mater. 237– 238 (2012)
10– 19.
(2006) 6104-6109.
[6] F. Shi, Y. Ma, J. Ma, P. Wang, W. Sun, Preparation and characterization of PVDF/TiO 2
hybrid membranes with different dosage of nano-TiO2, J. Membr. Sci. 389 (2012) 522–531.
[7] A. Razmjou, J. Mansouri, V. Chen, The effects of mechanical and chemical modification
of TiO2 nanoparticles on the surface chemistry, structure and fouling performance of PES
30
[9] M. Safarpour, V. Vatanpour, A. Khataee, Preparation and characterization of graphene
oxide/TiO2 blended PES nanofiltration membrane with improved antifouling and separation
[10] Q. Shi, Y. Su, S. Zhu, C. Li, Y. Zhao, Z. Jian, A facile method for synthesis of pegylated
Influence of nanofiller shape on characterization and fouling resistance, Appl. Surf. Sci.349
(2015) 66-77.
surface properties of PES ultrafiltration membranes to reduce the fouling resistance using
property of polyethersulfone ultrafiltration membranes using NaX zeolite and titanium oxide
[14] T. Noeiaghaei, J.-O. Kim, S.-R. Chae, Recent advances in nano-hybrid membranes for
[15] L.Y. Ng, A.W. Mohammad, C.P. Leo, N. Hilal, Polymeric membranes incorporated with
[16] D. Rana, T. Matsuura, Surface modifications for antifouling membranes, Chem. Rev.110
(2010) 2448–2471.
[17] M.S. Muhamad, M.R. Salim, W.J. Lau, Preparation and characterization of PES/SiO 2
composite ultrafiltration membrane for advanced water treatment, Korean J. Chem. Eng.,32
(2015) 2319-2329.
31
[18] J. Shen, H. Ruan, L. Wu, C. Gao, Preparation and characterization of PES–SiO2
[19] Z. Xu, T. Wu, J. Shi, K. Teng, W. Wang, M. Mab, J. Li, X. Qian, C. Li, J. Fan,
oxide and TiO2 for water treatment, J. Membr. Sci. 520 (2016) 281–293.
mixed matrix membrane via insitu colloidal precipitation method, Desalination. 295 (2012)
61–69.
Superhydrophilic graphene-loaded TiO2 thin film for self-cleaning applications, ACS. Appl.
oxidation of organic dyes and pollutants in wastewater using different modified titanium
[24] H. Song, J. Shao, Y. He, B. Liu, X. Zhong, Natural organic matter removal and flux
degradation of Acid Black 1(AB 1) under UV light, Appl. Catal. A: Gen. 475 (2014) 235–
241.
32
[26] C. Leng, J. Wei, Z. Liu, R. Xiong, C. Pan, J. Shi, Facile synthesis of PANI-modified
[27] Y. Yao J. Qin, H. Chen F. Wei, X. Liu, J. Wang, S. Wang, One-pot Approach for
[28] J.-F. Li, Z.-L. Xu, H. Yang, L.-Y. Yu, M. Liu, Effect of TiO2 nanoparticles on the
surface morphology and performance of microporous PES membrane, Appl. Surf. Sci. 255
(2009) 4725–4732.
[29] G. Wu, S. Gan, L. Cui, Y. Xu, Preparation and characterization of PES/TiO 2 composite
[30] V. Vatanpour S.S. Madaeni A. Khataee, E. Salehi S. Zinadini H. Ahmadi Monfared, TiO2
embedded mixed matrix PES nanocomposite membranes: Influence of different sizes and
[31] A. Sotto, A. Boromand, R. Zhang, P. Luis, J. M. Arsuaga, J. Kim, B.V. Bruggen, Effect
and fouling resistance of PES–TiO2 membranes, J. Colloid Interface. Sci. 363 (2011) 540–
550.
[32] APHA-AWWA-WEF, Standard Methods for the Examination of Water and Wastewater.
nonmetal doped TiO2 (K-B-N-TiO2) for post treatment of biologically treated palm oil mill
33
[34] A. Slave, Optical characterization of TiO2-Ge nanocomposite films obtained by reactive
Synthesis and characterization of a mixture of CoFe2O4 and MgFe2O4 from layered double
hydroxides: Band gap energy and magnetic responses, J. Magn. Magn. Mater. 369 (2014)
249-259.
210.
Linear Alkyl Benzene (LAB) industrial wastewater under visible light, Mater. Sci. Semicond.
Novel high flux antifouling nanofiltration membranes for color removal containing
34
[42] S. Rajesh, A. Fauzi Ismail, D. R. Mohan, Structure-property interplay of poly(amide-
[45] E. Bilgin Simsek, Solvothermal synthesized boron doped TiO2 catalysts: Photocatalytic
novel antifouling mixed matrix PES membrane by embedding graphene oxide nanoplates, J.
coagulation bath during phase inversion for preparation of ZnFe 2O4/SiO2/PES nanofiltration
membrane: A novel method for flux enhancement and fouling resistance, J. Ind. Eng Chem.
46 (2017) 9-18.
35
[50] M. Al-Aseeri, Q. Bu-Ali, S. Haji, N. Al-Bastaki, Removal of Acid Red and sodium
chloride mixtures from aqueous solutions using nanofiltration, Desalination. 206 (2007) 407–
413.
36
Table’s Captions:
Table 3. Surface roughness parameters of the prepared PES membranes resulted from
37
Figure’s Captions
Fig. 1. a) SEM images, b) DRS spectra, c) Tauc plot and d) PL spectra of the prepared
photocatalysts.
Fig. 2. The cross-section morphology SEM images of the fabricated PES membranes a and
b) bare PES (M0), c and d) 0.1 wt.% (M1), e and f) 0.5 wt.% (M2), g and d) 1 wt.% (M3) .
Fig. 3. Two and three dimensional AFM photographs of the prepared PES membranes with
different weight fraction NPs (a) bare PES (M0), (b) 0.1 wt.% (M1), (c) 0.5 wt.% (M2), (d) 1
wt.% (M3).
Fig. 4. a) Pure Water flux after 60 min at 5 bar, and b) static contact angel analysis of the
prepared membranes.
Fig. 5. Permeation flux of the prepared membranes versus time during milk powder solution
filtration.
Fig. 7. a) Dye rejection performance of the prepared membranes (5 bar, original pH, 50 ppm
direct red 16, after 60 min filtration) and b) The photograph of original solution of direct red
Fig.8. a) Performance of the prepared membranes for post treatment of the biologically treated
POME (initial COD=1000 mg/l, pressure=5 bar and original pH= 5 after 120 min) in a dead end
system, and b) The photograph of the biologically treated POME and permeation of the M2
membrane.
Fig. 9. Permeation flux of M2 versus filtration time during post treatment of biologically
treated POME in a cross flow system (initial COD=1000 mg/l, pressure=5 bar, Q=150 L/h at
original pH).
38
Fig.10. Permeation flux of the M0 and M2 membranes versus filtration time during post
treatment of biologically treated POME in a cross flow system (initial COD=1000 mg/l,
39
Table 1.
Membrane type PES (wt%) PVP (wt%) B doped TiO2-SiO2 /CoF2O4 (wt %)
M0 20 1.0 0
M1 20 1.0 0.1
M2 20 1.0 0.5
M3 20 1.0 1.0
Table 2.
Direct red 16
Molecular Formula C26H17N5Na2O8S2
Molecular Weight 637.55
Molecular Structure
Table 3.
Roughness parameters
Membrane Sa (nm) Sq (nm) Sy (nm)
M0 47.11 59.60 304.46
M1 13.92 17.69 121.41
M2 12.60 15.97 117.45
M3 14.62 22.67 255.15
40
a)
80
B-TiO2-SiO2/CoFe2O4 Pure TiO2
70 Pure TiO2 B-TiO2-SiO2/CoFe2O4
60
Absorbance (a.u)
(α*λν)^0.5
50
40
30
20
10
0
250 350 450 550 650 750 1.5 2 2.5 3 3.5 4 4.5 5 5.5
Wavelength (nm) Eg (eV)
b) c)
1000
B-TiO2
900
B-TiO2-SiO2
800
B-TiO2-SiO2/CoFe2O4
700
Intensity
600
500
400
300
200
100
0
400 450 500 550 600 650 700 750 800
Wavelength (nm)
d)
Fig. 1.
41
a b
c d
e f
g h
Fig. 2.
42
a) M0
b) M1
c) M2
43
d) M3
Fig. 3.
44
45
40 38.1
35
30
kg/m2.h)
25.2
25
21
20
16.7
15
10
0
M0 M1 M2 M3
Membrane type
a)
70
62.5
60 55.1
51.3
49.3
50
Contact angle (o)
40
30
20
10
0
M0 M1 M2 M3
Membrane type
b)
Fig. 4.
45
M0 M1 M2 M3
17
15
Permation flux (kg/m2.h)
13
11
5
15 30 45 60 75 90
Time (min)
Fig. 5.
46
100
90
80
70
60 Cleaning with DI
FRR (%)
30
20
10
0
M0 M1 M2 M3
Membrane type
a)
100
Rt Rr Rir
90
80
Resistance parameters (%)
70
60 57
52
47.42 49
50 44 43
40 33
30
30
20 13
9.58 11
10 3
0
M0 M1 M2 M3
Membrane type
b)
Fig. 6.
47
100 96 98 96
90
90
80
Dye removal (%)
70
60
50
40
30
20
10
0
M0 M1 M2 M3
Membrane type
a)
b)
Fig. 7.
48
Dye removal (%) COD removal (%)
99 98
100 94 92
88
84
80
80 75
Removal efficiency (%)
60
40
20
0
M0 M1 M2 M3
Membrane type
a)
b)
Fig. 8.
49
Wihout light With visible light irradiation
30
20
15
10
0
30 50 70 90 110
Time (h)
Fig. 9.
22
20
18
16
14
12
10
8
6
1 3 5 7 9 11
Time (h)
Fig. 10.
50
Graphical abstract:
51
Highlights:
B doped TiO2-SiO2/CoFe2O4 visible driven nanophotocatalyst was a desirable
The used NPs can be degraded adsorbed organic foulants on the membrane surface or
The interaction between organic foulant and hydrophilic membrane surface was
modified membranes
52