See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/307949032

Preparation and characterization of a novel

highly hydrophilic and antifouling
polysulfone/nanoporous TiO 2 nanocomposite
membrane

Article in Nanotechnology October 2016

DOI: 10.1088/0957-4484/27/41/415706

CITATIONS READS

0 87

7 authors, including:

Hossein Cheraghi Bidsorkhi Daryoush Emadzadeh

Sapienza University of Rome University of Ottawa
14 PUBLICATIONS 30 CITATIONS 29 PUBLICATIONS 384 CITATIONS

SEE PROFILE SEE PROFILE

W.J. Lau Ahmad Fauzi Ismail

Universiti Teknologi Malaysia Universiti Teknologi Malaysia
123 PUBLICATIONS 1,206 CITATIONS 681 PUBLICATIONS 10,047 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Pilot plant design and fabrication for Arsenic removal project View project

low cost ceramic membrane distillation for arsenic removal View project

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Hossein Cheraghi Bidsorkhi
letting you access and read them immediately. Retrieved on: 03 November 2016
 Nanotechnology

Nanotechnology 27 (2016) 415706 (11pp) doi:10.1088/0957-4484/27/41/415706

Preparation and characterization of a novel

highly hydrophilic and antifouling
polysulfone/nanoporous TiO2
nanocomposite membrane
H Cheraghi Bidsorkhi1,2, H Riazi3, D Emadzadeh4,5, M Ghanbari1,
T Matsuura5, W J Lau1 and A F Ismail1
1
Advanced Membrane Technology Research Centre (AMTEC), Universiti Teknologi Malaysia, 81310
Skudai, Johor, Malaysia
2
Research Center for Nanotechnology Applied to Engineering of Sapienza (CNIS), Sapienza University of
Rome, Rome, Italy
3
Department of Polymer Engineering and Color Technology, Amirkabir University of Technology,
PO Box 15875-4413, Tehran, Iran
4
Department of Chemical Engineering, Gachsaran Branch, Islamic Azad University, Gachsaran, Iran
5
Industrial Membrane Research Laboratory, Department of Chemical and Biological Engineering,
University of Ottawa, 161 Louis Pasteur St, Ottawa, ON K1N 6N5, Canada

E-mail: d.emadzadeh@iaug.ac.ir and afauzi@utm.my

Received 19 November 2015, revised 20 July 2016

Accepted for publication 11 August 2016
Published 8 September 2016

Abstract
In this research, novel ultraltration nanocomposite membranes were prepared by incorporating
self-synthesized nanoporous titanium dioxide (NTiO2) nanoparticles into polysulfone. The
surface of the nanoparticle was treated with a silane-based modier to improve its distribution in
the host polymer. Atomic-force microscopy, scanning electron microscopy, Fourier transform
infrared spectroscopy, BrunauerEmmettTeller, transmission electron microscopy, energy-
dispersive x-ray spectroscopy, porosity and contact angle tests were conducted to characterize
the properties of the particles as well as the fabricated nanocomposite membranes. The effects of
the nanoparticle incorporation were evaluated by conducting ultraltration experiments. It was
reported that the membrane pure water ux was increased with increasing NTiO2 loading owing
to the high porosity of the nanoparticles embedded and/or formation of enlarged pores upon
addition of them. The antifouling capacity of the membranes was also tested by ultraltration of
bovine serum albumin fouling solution. It was found that both water ux and antifouling
capacity tended to reach desired level if the NTiO2 added was at optimized loading.

Keywords: nanocomposite membrane, nanoporous titanium dioxide, antifouling, ultraltration

membrane
(Some gures may appear in colour only in the online journal)

1. Introduction particles are prevented. In other words, UF works based on

Ultraltration (UF) is a technique used for separation of
macromolecules and suspended solids from liquids. Small
molecules pass through the membrane together with the sol-
vent liquid while the passage of macromolecules and/or solid
size exclusion mechanism [1, 2]. UF has been used widely in
many areas including dairy, oil, paper and pulp industries. In
milk processing plants, for example, UF membranes are used
for whey concentration and milk dehydration. However, in
the food industry, they are used for recovering proteins and

0957-4484/16/415706+11$33.00 1 2016 IOP Publishing Ltd Printed in the UK

Nanotechnology 27 (2016) 415706
carbohydrates from palm oil mill efuent [3, 4]. In drinking
water plants, UF membranes are replacing conventional
technologies such as sand ltration, since it is able to remove
bacteria and viruses to deliver clean drinkable water even in
the municipal scale [5]. In many petrochemical and steel-
making plants, large quantities of oily wastewater are pro-
duced and their recovery is vital. Cleansing by UF mem-
branes gives the chance of removing tiny oil droplets of
submicron size, whereas such level of purity is not achievable
by other conventional techniques like gravity separation and
air oatation [6].
Immersion precipitation is a technique by which UF
membranes are prepared. A polymer solution is cast to form a
thin layer which is then transferred to a coagulation bath
immediately or after a delayed time [7]. Polyacrylonitrile,
polyethersulfone, polysulfone (PSF), polyvinylidine uoride
and cellulose acetate are widely used in UF membrane
manufacturing [8, 9]. It should be noted that most of the
membranes made of the aforementioned materials face foul-
ing problems. Fouling is categorized into reversible fouling
which can be obviated by membrane cleaning and irreversible
fouling which cannot be prevented. Fouling is caused by
different reasons including adsorption of foulants, blocking of
pores and cake formation [10]. Currently, the most simple yet
effective method to reduce membrane fouling is to enhance its
surface hydrophilicity by blending it with hydrophilic poly-
mers such as polyethylene glycol and polyvinylpyrrolidone
(PVP) or grafting its surface with hydrophilic bran-
ches [11, 12].
Addition of inorganic nanoparticles is another approach
to overcome membrane shortages including low ux, bio-
fouling, low hydrophilicity and inferior mechanical properties
[13, 14]. Many attempts have been made by introducing
various types of nanoparticles (based on halloysite nanotubes,
titania, silver, iron, zirconium, silica and alumina) into poly-
meric matrices to enhance membrane properties such as
selectivity, permeability, antifouling properties, etc [1518].
Zodrow et al [19] incorporated nanosilver particles into PSF
to improve resistance of the membrane against virus pene-
tration and biofouling. UF membranes made of such modied
PSF were used in drinking water purication instruments.
Silica (SiO2) nanoparticles were also added to the PSF
membrane, which resulted in facile operation and high ther-
mal/chemical resistance [19]. It is proved that these mem-
branes possessing a hybrid structure are efcient for
separating oil droplets from oily wastewater. Iron nano-
particles on the other hand were considered as potential
nanollers to develop membranes sensitive to magnetic elds
[20]. Beneting from different behavior of a solute/polymer
pair in a magnetic eld, the membrane performance can be
improved.
Titanium dioxide (TiO2) is a commonly used nano-
particle in UF membrane manufacturing [21, 22]. It is a
hydrophilic inorganic nanoparticle with a high level of che-
mical stability, which endows membranes with photocatalysis
features [23]. Furthermore, TiO2 is commercially available
and is not very expensive. Usually TiO2 forms strong inter-
actions with host polymers leading to improvement in
2
H C Bidsorkhi et al
mechanical properties and thermal stability. Its presence in
the membranes changes pore size and pore size distribution
[24]. It could also improve membrane hydrophilicity, per-
meability and solute rejection [25]. Addition of TiO2 is also
reported to play a role in decreasing fouling tendency of the
membranes and guarantees stable long-term performance in
industrial plants [26].
In this research, a spindle-shaped anatase nanoporous
TiO2 (NTiO2) is added to PSF to fabricate novel type of UF
nanocomposite membranes. In comparison with membranes
containing spherical non-porous TiO2 particles, higher por-
osity of the nanoporous particles, which is conrmed by the
UF experiments with the bovine serum albumin (BSA)
solution, enhances the ux of the membranes more sig-
nicantly. In the permeation experiments and antifouling
tests, effects of the presence of NTiO2 on ux value, rejection
and recovery ratio were evaluated.
2. Experimental
2.1. Materials
PSF (MW: 65 000 g mol1) was purchased from Shuguang
Chemical Factory, China. Polyvinylpyrrolidone (PVP, MW:
55 000 g mol1) and n-methyl2-pyrrolidone (NMP, purity:
95%) were supplied by Sigma-Aldrich. Nanoporous TiO2 was
synthesized according to the method of Ye et al [27]. [1-(2-
amino-ethyl)-3-aminopropyl] trimethoxysilane (AAPTS) was
purchased from Merck Co. Bovine serum albumin (BSA,
MW: 66 000 g mol1) from Sigma-Aldrich was used as the
foulant. Acetic acid (HAc) and tetrabutyl titanate (TBT) were
purchased from Beijing Chemical Reagent Co. and Beijing
Great Wall Chemical Reagent Co., respectively.
2.2. Membrane fabrication
Nanoporous anatase TiO2 particles were obtained through
solvothermal reaction of the TBT and HAc solution. Under
continuous agitation, 0.2 ml of TBT was added dropwise to
10 ml of HAc. The resultant white suspension was transferred
to a 20 ml Teon-sealed stainless-steel autoclave and heated
to 200 C and kept for 24 h. The autoclave was allowed to
cool to room temperature and the nal product was obtained
by centrifugation. The resultant material was washed with
ethanol several times and subsequently dried at 60 C over-
night. To remove remaining organic compounds, the product
was calcined at 400 C for 30 min [27].
To modify the surface of the synthesized nanoparticles,
silane-coupling agent (AAPTS) bearing NH2 functional
groups was used. To synthesize modied NTiO2, 1.0 g of
nanoparticles was added into 50 ml toluene containing 1.0 g
of AAPTS. The mixture was then stirred at 95 C overnight
followed by ltration and subsequent washing using toluene,
ethanol, ethanol/water (1:1 v/v) and water. The nal product
was then dried at 100 C in a vacuum oven overnight [28, 29].
The silane-modied nanoparticles are named NTiO2 there-
after. Dope solution containing 17.5 wt% PSF, 0.5 wt% PVP
Nanotechnology 27 (2016) 415706
and 82 wt% NMP was rst prepared before a pre-determined
amount of NTiO2 (0.5%, 1% and 2% of PSF) was added to
form a homogenous mixture. The mixture was degassed for
24 h at 25 C to eliminate trapped air bubbles. The obtained
polymer/nanoparticle suspension was cast on a glass plate
and rapidly transferred to a water coagulation bath at room
temperature. The polymeric lm was peeled off from the glass
plate spontaneously upon completion of phase inversion. The
membranes were stored in a fresh water bath for 24 h to
assure complete removal of the residual solvent. Finally, the
membranes were kept in deionized water prior to their use.
2.3. Characterizations
Transmission electron microscopy (TEM model JEM-100 CX
II) was applied to observe the nanoporous structure of the
inorganic particles. Scanning electron microscopy (Shimadzu
table top SEM) was used to observe the surface morphology
of the membranes. The membranes were fractured in liquid
nitrogen to examine the cross-sectional morphology of them.
Both surface and cross-section of the membranes were gold
sputtered before being mounted to the SEM equipment.
Fourier transform infrared spectroscopy (FTIR) spectra were
recorded in the range of 4000400 cm1 using UATR, Per-
kin-Elmer. The tests were done on the NTiO2 nanoparticles,
PSF and NTiO2PSF nanocomposite membranes. The test
was carried out to evaluate the surface area of unmodied and
modied TiO2 particles. ASAP 2010, Micromeritics, was
used to perform the test. The surface roughness of the
membranes was assessed by an qtomic force microscope
(NanoSurf Easy Scan).
The crystalline structures of the NTiO2 nanoparticles,
PSF and NTiO2PSF nanocomposite membranes were eval-
uated by x-ray diffractometer (D/max-rB 12KW Rigaku).
The spectra were recorded in a 2 range of 1080 with a
step size 0.1 and a rate 1 s step1. A goniometer (OCA 15
Plus, Data Physics) was used to measure the contact angle of
the membranes based on the sessile drop technique. The
droplets were placed on ve different points of the membrane
and the average value was reported as the contact angle. After
immersing the membranes in deionized water for a sufcient
amount of time, the water at the membrane surface was
blotted with tissue paper. The wet and dry weights of mem-
brane were measured and the porosity was calculated as fol-
lows
Ww - Wd
Porosity (%) = , (1 )
rwater *A*L
where Ww, Wd, water, A and L are the weight of the membrane
in the wet and dry states, density of water, the membrane
surface area and the membrane thickness, respectively.
2.4. Performance study
The performance of the membranes was evaluated by UF
experiments using a laboratory-made cross-ow UF cell. The
pressure and the feed ow rate were kept constant at 1 bar and
400 ml min1, respectively. A disc-like membrane with an
3
H C Bidsorkhi et al
effective area of 12.57 cm2 was used. At rst, the pure water
was allowed to pass through the membrane for 1 h. Subse-
quently, the pure water feed was replaced by the BSA solu-
tion (100 ppm) which was ltered for 1 h. Finally, pure water
was used again as the feed during the third hour of the
experiment. The permeate ux ( J) of the membranes were
calculated by:
J = V AT, (2 )
where V is the permeate volume, A is the effective area of the
membrane and T is the time required to collect V. The solute
rejection was obtained by
R = (1 - Cp Cf ) 100, (3 )
where Cp and Cf are the concentrations of BSA in the
permeate and feed streams, respectively, measured by the
UVvis spectroscope (Perkin-Elmer Lambda 25), at the
absorption peak of 280 nm.
2.5. Antifouling experiments
Flux recovery rate, FRR, was obtained for the membranes
according to the following equation
J
FRR = w2 100, (4 )
Jw1
where Jw1 and Jw2 are pure water ux at the 60 min (end of
the rst cycle of the pure water permeation experiment) and
130 min (beginning of the second cycle of the pure water
permeation experiment). The total, Rt, reversible, Rr, and
irreversible, Rir, fouling were determined according to the
following equations
Rt = (1 - Jp Jw1) 100, (5 )
R ir = Rt - Rr , (6 )
Jw2 - Jp
Rr = 100, (7 )
Jw1
where Jp is the permeation ux at 120 min (end of the UF
experiment).
3. Results and discussion
Figure 1 shows TEM micrographs of NTiO2. It is obvious that
the particles do not have a perfect spherical shape and their
surfaces resemble a porous structure. Figure 2 shows the
BrunauerEmmettTeller (BET) results of NTiO2 nano-
particles. The presence of rather uniform nanoporous with an
average diameter of 3.9 nm was conrmed by the corresp-
onding nitrogen adsorption/desorption measurement. The
BET data also revealed the high surface area of the NTiO2
compared with the unmodied nanoparticle. The surface areas
for the former and the latter are 112.25 m2 g1 and
50.50 m2 g1, respectively. High surface area of NTiO2 also
indicates its high porosity which will be discussed in the
following sections. Nanoporous nanoparticles have signicant
advantages over their non-porous or porous counterparts.
High specic surface area is their most important feature as it
Nanotechnology 27 (2016) 415706
Figure 1. TEM micrographs of NTiO2 at different scale bars, (a) 500 nm and (b) 20 nm.
Figure 2. The adsorption and desorption prole of NTiO2
nanoparticles for BET measurement.
Figure 3. FTIR spectra of NTiO2, PSF and NTiO2PSF membrane
(Inset: FTIR spectra of the bare and silane-modied TiO2).
4
H C Bidsorkhi et al
means more interaction sites and better antifouling properties.
They can have high crystalinity and large pore size as well
[30]. Hence, it can be expected that NTiO2 particles could
show a signicant antifouling improvement over non-porous
material.
In gure 3, FTIR spectra of NTiO2, PSF and NTiO2PSF
nanocomposite membrane are displayed. A silane-based
coupling agent was used to reduce surface energy of the
nanoparticles, minimizing their agglomeration in the mem-
brane. It has already been reported that establishment of
physical interactions upon addition of inorganic particles into
polymers could lead to change in the position and intensity of
peaks in FTIR spectrum [31]. In the NTiO2 spectrum, the
peak which appears at wavenumbers around 8301110 cm1
can be attributed to SiO covalent bond while the peak near
2950 cm1 is ascribed to alkyl groups of the modier [32].
Considering PSF and NTiO2PSF spectra, a stronger peak at
around 2980 cm1 is observed in the latter compared with the
former. Increase in the intensity of this peak can be related to
the presence of NTiO2 in the membrane [33]. These results
show that a strong interaction exists between the host polymer
and NTiO2.
To strongly support the presence of the modier at the
surface of nanoparticles, FTIR spectra of bare TiO2 and
NTiO2 are displayed in inset of gure 3. A new peak arising
at the wavenumbers around 1650 cm1 for NTiO2 is ascribed
to NH covalent bond formed at the surface of the particles
after the silane modication. An extra shoulder in FTIR
spectrum of the NTiO2 at around 920 cm1 signies occur-
rence of condensation reaction between silanol groups and
surface hydroxyl ones [34]. In addition, the appearance of two
peaks at 1154 and 1213 cm1 that is related to SiO and CN
bonds, respectively further supports the successful modica-
tion of nanoparticles by the AAPTS [32].
Figure 4 presents the AFM micrographs of the membrane
while table 1 shows the root average arithmetic roughness
values (Ra) of membrane top surface. As can be seen, increase
in the NTiO2 loading leads to roughness increment. The
observation is in good agreement with the ndings reported
elsewhere [35]. Although it is widely reported that the
Nanotechnology 27 (2016) 415706
Figure 4. 3D AFM images of the membrane top surface, (A) pure PSF, (B) 0.5% NTiO2PSF, (C) 1% NTiO2PSF and (D) 2% NTiO2PSF.
Table 1. Surface roughness and contact angle of the pure membrane
and nanocomposite membranes.
Membrane Roughness, Ra (nm) Contact angle ()
PSF 9.4 67.0
0.5% NTiO2PSF 11.1 64.7
1% NTiO2PSF 14.6 59.9
2% NTiO2PSF 22.1 55.7
increase in surface roughness could reduce membrane anti-
fouling properties following an increase in adsorption sites
[36, 37], the introduction of hydrophilic NTiO2 into PSF
membrane on the other hand will improve membrane
hydrophilicity and enhance anti-fouling. It seems there is a
competition between roughness and hydrophilicity incre-
ments upon the addition of the NTiO2. When hydrophilicity
improvement is more signicant, anti-fouling properties
increases, however, they decrease when roughness overcomes
hydrophilicity. The results of FRR that will be presented in
following sections can be justied by this hypothesis.
Contact angle measurements were also performed to
evaluate the effect of the NTiO2 on the hydrophilicity of the
membranes. The results are presented in table 1. As expected,
the contact angle decreases (hydrophilicity increases) with an
increase of the NTiO2 loading. It should be mentioned that the
contact angle is dependent not only on the surface hydro-
philicity but also on various other parameters like roughness,
5
H C Bidsorkhi et al
porosity, pore size and pore size distribution. For example,
comparing two membranes with a similar hydrophilicity, the
one which has the higher roughness shows the higher contact
angle [37]. However, the level of improvement varies
depending on the type of TiO2 (nanoporous or non-porous).
As shown in table 1, the contact angle of membrane for 0.5%
NTiO2PSF, 1% NTiO2PSF and 2% NTiO2PSF are 64.7,
59.9 and 55.7, respectively, while 83, 79 and 74 were the
contact angle values reported by Zhang et al [38] for similar
PSF composite membranes using non-porous TiO2 particles.
This signicant difference in water contact angles can be
attributed to the difference on porosity of the nanoparticles
used, i.e. one is nanoporous and the other is non-porous.
Nevertheless, we cannot rule out the effect of surface
roughness (based on Wenzel and Cassie effect) that might
play partial role in inuencing the contact angle of the
membranes incorporated with nanoporous or non-porous
particles.
According to the procedure mentioned above, calculated
porosity values for the neat PSF, 0.5% NTiO2PSF, 1%
NTiO2PSF and 2% NTiO2PSF membranes are 72%, 78%,
81% and 87%, respectively. It should be emphasized here that
these values denote to the total porosity of the membranes
including bulk porosity and skin layer one. Comparing these
membrane porosity values with the ones reported for mem-
branes containing non-porous TiO2 particles reveals that
incorporation of the NTiO2 particles increases porosity sig-
nicantly [39]. Usually, addition of an appropriate amount of
Nanotechnology 27 (2016) 415706
Figure 5. X-ray diffraction (XRD) spectrum of (a) PSF, (b) NTiO2
and (c) NTiO2PSF.
inorganic nanollers leads to an increase in the population of
pores and the total porosity of the membranes. Upon
shrinkage of the polymer after casting the polymer/inorganic
ller mixture, a mechanical stress is generated between the
polymer and ller, which becomes the driving force for
creation of pores between them. This may occur in the bulk
membrane phase as well as at the skin layer. However, it
should also be emphasized that a high loading of the nano-
particles causes creation of large macrovoids and formation of
the nanoparticle aggregates at the skin layer as well as in the
bulk [25]. This phenomenon signicantly deteriorates per-
formance of the membranes and this is why the concentration
of the NTiO2 particles was kept below 2% in this research.
XRD analysis was also carried out to evaluate dispersion
quality and possible interactions of the NTiO2. This inorganic
nanoparticle has two different structural forms named anatase
and rutile. Although the chemical structure of both forms is
similar, their dissimilar crystalline structures result in differ-
ence in angle and intensity of peaks in XRD spectrum. The
XRD spectra of PSF, NTiO2 particle and the NTiO2PSF
composite membrane are shown in gure 5. PSF shows a
broad peak at 2 of 18 and some other sharp peaks at higher
angles which are related to different crystalline structures that
exist in PSF [40, 41]. Although TiO2 shows three main peaks
at 25.32, 37.86 and 48.06, the peaks are ascribed to NTiO2
crystals with the tetragonal anatase phase [39]. The NTiO2
PSF membrane shows a broad peak at around 18, a strong
sharp peak at around 29 and one another peak at 25 which is
bolded by an arrow. These peaks and the ones at larger angles
are all originated from both NTiO2 and PSF. These obser-
vations conrm the presence of the NTiO2 in PSF-based
membrane and verify its interaction with the host polymer.
Other researchers who studied the PSF/TiO2 membrane also
used XRD analysis to evaluate existence of interaction
between the inorganic nanoparticles and the host polymer
[40]. The researchers concluded the establishment of physical
interactions between TiO2 and PSF by observing some
changes in the location of the aforementioned peaks. As both
6
H C Bidsorkhi et al
XRD and FTIR results show displacement in peak locations,
the presence of physical interactions in the NTiO2PSF
membrane is conrmed.
Figure 6 shows the SEM images of the neat PSF mem-
brane and the membrane incorporated with different loadings
of NTiO2. As can be seen, presence of the pores at the
membranes surface is clearly observed at scale bar of 20 mm.
White spots in the images denote the NTiO2 particles.
Although the surface of the TiO2 particles was modied,
some small aggregates are still observable particularly in the
case where NTiO2 loading was at the highest. The size of the
aggregates increases with an increase in the ller loading. It is
often reported that addition of nanoparticles causes increase in
population of pores at the membrane surface [18].
In gure 7, SEM cross-sectional micrographs of the
membranes are presented. The upper and lower rows show
the low (scale bar: 200 mm) and high magnications (scale
bar: 50 mm) of the image, respectively. The bright spots in the
cross-section are the NTiO2 particles. Similar to the surface
images, small aggregates are observable in the 2.0%NTiO2
PSF membrane. It seems that the cross-section consists of
three layers, a top skin layer, an intermediate layer of nger-
like structure and a layer of large macrovoids. As the ller
loading increases, the skin layer becomes thinner, the nger-
like voids turn into macrovoids, and nally the size of mac-
rovoids increases. These results can be explained as follows.
With an increase in the NTiO2 loading, the casting dope
becomes more hydrophilic and viscous. The increase in
hydrophilicity enhances the solvent (NMP)/nonsolvent
(water) exchange rate and more nonsolvent is drawn into the
cast lm, whereas viscosity increase slows down the solvent/
nonsolvent exchange. These two opposing effects are super-
imposed in controlling the membrane morphology. At the low
NTiO2 loading, the hydrophilicity effect is predominant and
the pore size and porosity increase with an increase in the
NTiO2 loading, whereas at higher NTiO2 loading, the visc-
osity effect becomes predominant resulting in smaller pore
size and less porosity. It seems the hydrophilicity effect is
predominant at the NTiO2 concentration below 2% and the
viscosity effect may become observable at the NTiO2 con-
centration above 2%. In fact, it has been reported that the
macrovoids disappear at the NTiO2 concentration of 3% [37].
Subsequently, the membrane becomes homogeneous and
dense at higher NTiO2 loadings. Finally, the NTiO2 nano-
particles aggregate when their loading level is high and this
causes the surface pores blocking and membrane ux
decreasing. Thus, a great attention should be paid to the
amount of the incorporated NTiO2.
The presence of NTiO2 nanoparticles at the surface of the
membrane was examined by performing an energy-dispersive
x-ray spectroscopy (EDX) test on the 0.5% NTiO2PSF
membrane. In gure 8, the peaks at 4.5 and 2.5 KeV are
attributed to Ti and Si elements, respectively. The presence of
these two important peaks suggest the successful embedment
of NTiO2 in the PSF membrane matrix [33]. Other NTiO2
PSF nanocomposite membranes exhibit similar spectra, but
with slight difference in the peaks intensity (results not
shown here).
Nanotechnology 27 (2016) 415706 H C Bidsorkhi et al

Figure 6. SEM micrographs of the surface morphology of the (a) pure PSF, (b) 0.5% NTiO2PSF, (c) 1% NTiOPSF and (d) 2%
NTiO2PSF.

Figure 7. SEM micrographs of cross section of membranes containing 0.5% NTiO2 (a), (d), 1% NTiO2 (b), (e) and 2% NTiO2 (c), (d).

The pure water uxes of the membranes containing ux decreases with time and the rate of the ux decline
various amounts of the NTiO2 are displayed in gure 9. The depends on the ller loading. The ux of neat PSF and 0.5%
gure shows that the pure water ux increases with an NTiO2PSF membranes decreases only little with time while
increase in the amount of the NTiO2. However, the pure water the decrease is more signicant for 1% NTiO2PSF and 2%

7
Nanotechnology 27 (2016) 415706 H C Bidsorkhi et al

Figure 8. EDX spectrum of PSF membrane containing 0.5% NTiO2.

Figure 9. Pure water ux of membranes versus ltration time. Figure 11. Rejection of membranes against BSA foulant at
different time.

NTiO2PSF membranes. As mentioned earlier, the effects of

Figure 10. Permeate ux of the membranes in ltrating fouling
aqueous solution.
hydrophilicity and viscosity on the macrovoid formation are
opposite, among which the effect of hydrophilicity is stronger
as the NTiO2 concentration falls behind 2%, resulting in
increase in pore size and porosity. This is the main reason for
the increase in pure water ux with an increase in NTiO2
loading.
The results of the UF experiments with the BSA feed
solution are presented in gures 10 and 11. Similar to pure
water ux, the ux in the presence of foulant increases as the
NTiO2 loading increases, and a more signicant ux reduc-
tion is observed at higher NTiO2 loadings. Especially for 2%
NTiO2PSF membrane, the ux falls behind the ux of 1%
NTiO2PSF membrane after 50 min. With respect to foulant
rejection, it is found that membrane separation efciency
tends to decrease with the NTiO2 loading. This is in

8
Nanotechnology 27 (2016) 415706
Figure 12. Fouling experiments on different membranes.
Table 2. Flux recovery ratio (FRR), total fouling (Rt), reversible
fouling (Rr) and irreversible fouling (Rir) of various membranes.
Membrane FRR Rt Rr Rir
PSF 95.8 33.3 29.1 4.2
0.5% NTiO2PSF 100.0 33.3 33.3 0.009
1% NTiO2PSF 82.7 60.9 43.6 17.3
2% NTiO2PSF 82.7 72.4 55.1 17.3
accordance with other work where the negative effects of
inorganic nanoparticles loading on the solute rejection were
reported [30]. The BSA foulant separation of the neat PSF
and 0.5% NTiO2PSF membranes are nearly equal to 100%,
but those of 1% NTiO2PSF and 2% NTiO2PSF are lower
than 100%. This is because of the pore size increases together
with the enhancement of the macrovoid formation, as men-
tioned when SEM cross-sectional images were discussed.
In comparison with the work of Yang et al [25] who used
non-porous surfactant-modied TiO2, the membranes pre-
pared in this work with nanoporous TiO2 showed higher BSA
foulant rejections. The BSA foulant rejections of NTiO2
containing membranes are generally very high. This is
because of the negative charge of TiO2 at a pH above 5. Since
foulant is also a negatively charged molecule, it is rejected
from the membrane surface due to the electrostatic repulsive
forces. In addition, high hydrophilicity of the surface of the
membranes caused by incorporation of the NTiO2 leads to
formation of a water layer on the surface of the membranes.
Adhesion of organic molecules to this water layer is rather
difcult. This phenomenon along with the electrostatic
repulsion effect is therefore the main reason contributing to
improved anti-fouling resistance [42].
In gure 12, results of the antifouling experiments are
presented. For the initial 60 min of the experiment, pure water
was passed through the membranes while during the second
hour BSA foulant solution was ultraltered. Finally, the pure
water was passed again after 120 min to test the recovery of
the pure water ux. The results clearly show that the neat PSF
and 0.5% NTiO2PSF membranes recover their pure water
ux after being polluted by BSA molecules, however, 1%
9
H C Bidsorkhi et al
NTiO2PSF and 2% NTiO2PSF membranes do not fully
regain this ability. In other words, FRR is more pronounced
for the former samples in comparison with the latter ones. The
trend in FRR is shown in table 2 in details. Referring to the
table, FRR increases by addition of 0.5% NTiO2, however, its
further addition causes reduction in FRR. Thus, the addition
of NTiO2 can affect FRR both positively and negatively.
Similar effects were also reported elsewhere [41, 43].
Therefore, it can be said that a small amount of NTiO2
enhances FRR but a large amount of NTiO2 causes the
decrease in FRR as a result of poor particle dispersion in
membrane matrix. In table 2, total, reversible and irreversible
fouling dened by equations (5)(7) are also shown. Severe
irreversible fouling is observed for 1% NTiO2PSF and 2%
NTiO2PSF and this is the reason why these membranes
cannot recover their pure water ux after being polluted by
foulant. Large pores can accommodate the molecules, causing
the pore blocking. A similar explanation was given by other
researchers for the increase of irreversible fouling due to the
non-porous TiO2 addition [43].
4. Conclusions
PSF membranes containing different amounts of NTiO2 (zero,
0.5, 1 and 2 wt%) nanoparticles were prepared by the phase
inversion method. The amount of the nanoparticle was limited
to 2 wt% in order to prevent the formation of the nanoparticle
aggregates. The nanoporous NTiO2 synthesized by sol-
vothermal method is very potential to be used as nanoller for
the fabrication of UF membrane with high water ux, high
BSA foulant rejection and near zero irreversible fouling. The
FTIR and XRD analyses conrmed the formation of strong
interactions between the silane modied nanoparticles and the
host polymer matrix. The uniform dispersion of the NTiO2
observed in both top-surface and cross-sectional SEM images
of the membranes conrmed the appropriateness of the silane-
based modier for surface treatment of the nanoparticles.
Incorporation of the nanoporous particles also resulted in low
contact angle and high porosity. The comparison between
nanoporous particles and its non-porous counterparts stipu-
lated the advantages of the former over latter.
References
[1] Cooper A R 1980 Ultraltration membranes and applications
Polymer Science and Technology vol 13 (New York:
Plenum Press)
[2] Jnsson A S and Trgrdh G 1990 Fundamental principles of
ultraltration Chem. Eng. Process.: Process Intensication
27 6781
[3] Kazemimoghadam M and Mohammadi T 2007 Chemical
cleaning of ultraltration membranes in the milk industry
Desalination 204 2138
[4] Wu T, Mohammad A, Jahim J M and Anuar N 2007 Palm oil
mill efuent (POME) treatment and bioresources recovery
using ultraltration membrane: effect of pressure on
membrane fouling Biochem. Eng. J. 35 30917
Nanotechnology 27 (2016) 415706
[5] Chew C M, Aroua M, Hussain M and Ismail W W 2015
Practical performance analysis of an industrial-scale
ultraltration membrane water treatment plant J. Taiwan
Inst. Chem. Eng. 46 1329
[6] Chen W, Peng J, Su Y, Zheng L, Wang L and Jiang Z 2009
Separation of oil/water emulsion using Pluronic F127
modied polyethersulfone ultraltration membranes
Separation Purication Technol. 66 5917
[7] Radovanovic P, Thiel S W and Hwang S-T 1992 Formation of
asymmetric polysulfone membranes by immersion
precipitation: I. Modelling mass transport during gelation
J. Membr. Sci. 65 21329
[8] Shukla R and Cheryan M 2002 Performance of ultraltration
membranes in ethanolwater solutions: effect of membrane
conditioning J. Membr. Sci. 198 7585
[9] Emadzadeh D, Lau W and Ismail A 2013 Synthesis of thin lm
nanocomposite forward osmosis membrane with
enhancement in water ux without sacricing salt rejection
Desalination 330 909
[10] Shi X, Tal G, Hankins N P and Gitis V 2014 Fouling and
cleaning of ultraltration membranes: a review J. Water
Process Eng. 1 12138
[11] Chakrabarty B, Ghoshal A and Purkait M 2008 Ultraltration
of stable oil-in-water emulsion by polysulfone membrane
J. Membr. Sci. 325 42737
[12] Ochoa N, Masuelli M and Marchese J 2003 Effect of
hydrophilicity on fouling of an emulsied oil wastewater
with PVDF/PMMA membranes J. Membr. Sci. 226 20311
[13] Adelnia H, Bidsorkhi H C, Ismail A and Matsuura T 2015 Gas
permeability and permselectivity properties of ethylene vinyl
acetate/sepiolite mixed matrix membranes Separation
Purication Technol. 146 3517
[14] Bidsorkhi H C, Adelnia H, Pour R H and Soheilmoghaddam M
2015 Preparation and characterization of ethylene-vinyl
acetate/halloysite nanotube nanocomposites J. Mater. Sci.
50 323745
[15] Ghanbari M, Emadzadeh D, Lau W, Lai S, Matsuura T and
Ismail A 2015 Synthesis and characterization of novel thin
lm nanocomposite (TFN) membranes embedded with
halloysite nanotubes (HNTs) for water desalination
Desalination 358 3341
[16] Ghanbari M, Emadzadeh D, Lau W, Matsuura T and Ismail A
2015 Synthesis and characterization of novel thin lm
nanocomposite reverse osmosis membranes with improved
organic fouling properties for water desalination RSC Adv. 5
2126876
[17] Emadzadeh D, Lau W J, Matsuura T, Ismail A F and
Rahbari-Sisakht M 2014 Synthesis and characterization of
thin lm nanocomposite forward osmosis membrane with
hydrophilic nanocomposite support to reduce internal
concentration polarization J. Membr. Sci. 449 7485
[18] Ng L Y, Mohammad A W, Leo C P and Hilal N 2013
Polymeric membranes incorporated with metal/metal oxide
nanoparticles: a comprehensive review Desalination 308
1533
[19] Zodrow K, Brunet L, Mahendra S, Li D, Zhang A, Li Q and
Alvarez P J 2009 Polysulfone ultraltration membranes
impregnated with silver nanoparticles show improved
biofouling resistance and virus removal Water Res. 43
71523
[20] Jian P, Yahui H, Yang W and Linlin L 2006 Preparation of
polysulfoneFe3O4 composite ultraltration membrane and
its behavior in magnetic eld J. Membr. Sci. 284 916
[21] Lau W J, Gray S, Matsuura T, Emadzadeh D, Paul Chen J and
Ismail A F 2015 A review on polyamide thin lm
nanocomposite (TFN) membranes: history, applications,
challenges and approaches Water Res. 80 30624
[22] Ghanbari M, Emadzadeh D, Lau W, Matsuura T,
Davoody M and Ismail A 2015 Super hydrophilic TiO2/
10
H C Bidsorkhi et al
HNT nanocomposites as a new approach for fabrication of
high performance thin lm nanocomposite membranes for
FO application Desalination 371 10414
[23] Emadzadeh D, Lau W, Matsuura T, Rahbari-Sisakht M and
Ismail A 2014 A novel thin lm composite forward osmosis
membrane prepared from PSfTiO2 nanocomposite
substrate for water desalination Chem. Eng. J. 237 7080
[24] Ghanbari M, Emadzadeh D, Lau W, Riazi H, Almasi D and
Ismail A 2016 Minimizing structural parameter of thin lm
composite forward osmosis membranes using polysulfone/
halloysite nanotubes as membrane substrates Desalination
377 15262
[25] Yang Y, Wang P and Zheng Q 2006 Preparation and properties
of polysulfone/TiO2 composite ultraltration membranes
J. Polym. Sci. B 44 87987
[26] Mingliang L, Qingzhi W, Jialin L, Hongjian L and Zilong J
2011 Fabrication of SPES/nano-TiO2 composite
ultraltration membrane and its anti-fouling mechanism
Chin. J. Chem. Eng. 19 4551
[27] Ye J, Liu W, Cai J, Chen S, Zhao X, Zhou H and Qi L 2010
Nanoporous anatase TiO2 mesocrystals: additive-free
synthesis, remarkable crystalline-phase stability, and
improved lithium insertion behavior J. Am. Chem. Soc. 133
93340
[28] Emadzadeh D, Lau W, Rahbari-Sisakht M, Daneshfar A,
Ghanbari M, Mayahi A, Matsuura T and Ismail A 2015 A
novel thin lm nanocomposite reverse osmosis membrane
with superior anti-organic fouling afnity for water
desalination Desalination 368 10613
[29] Emadzadeh D, Lau W, Rahbari-Sisakht M, Ilbeygi H, Rana D,
Matsuura T and Ismail A 2015 Synthesis, modication and
optimization of titanate nanotubes-polyamide thin lm
nanocomposite (TFN) membrane for forward osmosis (FO)
application Chem. Eng. J. 281 24351
[30] Chen D, Cao L, Huang F, Imperia P, Cheng Y-B and
Caruso R A 2010 Synthesis of monodisperse mesoporous
titania beads with controllable diameter, high surface areas,
and variable pore diameters (1423 nm) J. Am. Chem. Soc.
132 443844
[31] Riazi H, Mohammadi N and Mohammadi H 2013 Emulsion
copolymerization of methyl methacrylate/butyl acrylate/
iodine system to monosize rubbery nanoparticles containing
iodine and triiodide mixture Ind. Eng. Chem. Res. 52
244956
[32] Emadzadeh D, Ghanbari M, Lau W, Rahbari-Sisakht M,
Matsuura T, Ismail A and Kruczek B 2016 Solvothermal
synthesis of nanoporous TiO2: the impact on thin-lm
composite membranes for engineered osmosis application
Nanotechnology 27 345702
[33] Razmjou A, Resosudarmo A, Holmes R L, Li H,
Mansouri J and Chen V 2012 The effect of modied TiO2
nanoparticles on the polyethersulfone ultraltration hollow
ber membranes Desalination 287 27180
[34] Ukaji E, Furusawa T, Sato M and Suzuki N 2007 The effect of
surface modication with silane coupling agent on
suppressing the photo-catalytic activity of ne TiO2 particles
as inorganic UV lter Appl. Surf. Sci. 254 5639
[35] Li J B, Zhu J W and Zheng M S 2007 Morphologies and
properties of poly (phthalazinone ether sulfone ketone)
matrix ultraltration membranes with entrapped TiO2
nanoparticles J. Appl. Polym. Sci. 103 36239
[36] Razmjou A, Mansouri J, Chen V, Lim M and Amal R 2011
Titania nanocomposite polyethersulfone ultraltration
membranes fabricated using a low temperature hydrothermal
coating process J. Membr. Sci. 380 98113
[37] Rana D and Matsuura T 2010 Surface modications for
antifouling membranes Chem. Rev. 110 2448
[38] Zhang G, Lu S, Zhang L, Meng Q, Shen C and Zhang J 2013
Novel polysulfone hybrid ultraltration membrane prepared
Nanotechnology 27 (2016) 415706
with TiO2-g-HEMA and its antifouling characteristics
J. Membr. Sci. 436 16373
[39] Cao X, Ma J, Shi X and Ren Z 2006 Effect of TiO2
nanoparticle size on the performance of PVDF membrane
Appl. Surf. Sci. 253 200310
[40] Yang Y, Zhang H, Wang P, Zheng Q and Li J 2007 The
inuence of nano-sized TiO2 llers on the morphologies
and properties of PSF UF membrane J. Membr. Sci. 288
2318
[41] Padaki M, Emadzadeh D, Masturra T and Ismail A 2015
Antifouling properties of novel PSf and TNT composite
11
H C Bidsorkhi et al
membrane and study of effect of the ow direction on
membrane washing Desalination 362 14150
[42] Emadzadeh D, Lau W, Matsuura T, Hilal N and Ismail A 2014
The potential of thin lm nanocomposite membrane in
reducing organic fouling in forward osmosis process
Desalination 348 828
[43] Vatanpour V, Madaeni S S, Khataee A R, Salehi E,
Zinadini S and Monfared H A 2012 TiO2 embedded mixed
matrix PES nanocomposite membranes: inuence of
different sizes and types of nanoparticles on antifouling and
performance Desalination 292 1929