DES-12409; No of Pages 12

Desalination xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Desalination
journal homepage: www.elsevier.com/locate/desal

Novel polymer matrix composite membrane doped with fumed silica particles for
reverse osmosis desalination
Aneela Sabir a,, Atif Islam a, Muhammad Shaq a, Amir Shafeeq b, Muhammad Taqi Zahid Butt c,
Nasir M. Ahmad d, Khairuddin Sanaullah e, Tahir Jamil a
a

Department of Polymer Engineering and Technology, University of the Punjab, Lahore 54590, Pakistan
Institute of Chemical Engineering and Technology (ICET), University of the Punjab, Lahore 54590, Pakistan
Faculty of Engineering and Technology, University of the Punjab, Lahore 54590, Pakistan
d
School of Chemical and Materials Engineering (SCME), National University of Sciences and Technology (NUST), H-12 Islamabad, Pakistan
e
Department of Chemical Engineering and Energy Sustainability, Faculty of Engineering, Universiti Malaysia Sarawak (UNIMAS), 94300 Kota Samarahan, Sarawak, Malaysia
b
c

H I G H L I G H T S

Doping of FSP in polymer matrix membrane; involving dissolution casting methodology

FSP is used as an additive to inate salt rejection properties.
Mechanical properties augmented for PMCM
Uniform dispersion of FSP on PMCM posing dense structure observed in SEM micrographs

a r t i c l e

i n f o

Article history:
Received 22 July 2014
Received in revised form 25 December 2014
Accepted 28 December 2014
Available online xxxx
Keywords:
Polymer matrix composite membrane
Fumed silica particles
Reverse osmosis
Desalination

a b s t r a c t
Novel polymer matrix composite membranes of cellulose acetate (CA)/polyethylene glycol (PEG) doped with
1030 wt.% fumed silica particles (FSP) were synthesized. The dissolution casting methodology was used to construct reverse osmosis (RO) membrane which accounts the explicit application for desalination process. The interactions between polymer chains and doped FSP were conrmed by Fourier transform infrared spectroscopy
(FTIR). Differential scanning calorimetry thermograms support the physical nature of polymer matrix composite
membranes (PMCMs) and an improved glass transition temperature (Tg) from 78.3 to 92.4 C was observed. The
thermal stability of the composite membranes signicantly enhanced with the addition of FSP. The typical morphology of PMCM was observed using scanning electron microscopy (SEM) and atomic force microscopy (AFM).
The 30 wt.% of FSP lled PMCM showed substantial improvement in tensile strength (8.2 MPa) and Young's modulus (854.0 MPa) as compared to the PMCM without FSP. The percentage water content (WC) of the membrane
doped with 30 wt.% FSP absorbed more water as compared to the other membranes. It was also noticed that the
FSP doped PMCM enhanced the desalination process which was monitored in terms of permeation ux (L/m2h)
and salt rejection (%). The optimum performance of desalination process was shown by 30 wt.% FSP doped
PMCM; the permeation ux was 0.66 L/m2h and salt rejection was 98.4% for the initial feed of 0.35 wt.% NaCl
aqueous solution at 25 C and operating pressure of 4.0 bar (osmotic pressure 2.9 bar).
2014 Elsevier B.V. All rights reserved.

1. Introduction
The water scarcity around the world has become an inescapable
issue because of the negligence in wastewater management and exponential growth in world population [1,2]. Seawater desalination is considered to be an adequate and sustainable solution for the supply of
water to deal with enormous issue of supply and demand of puried
Corresponding author.
E-mail address: aneela.pet.ceet@pu.edu.pk (A. Sabir).

drinking water [3]. Desalination process is the removal of salts from

brackish, saline and seawater by RO process, which is today the leading
technology to convert salt water into clean water for human consumption, industrial and agriculture applications [4,5]. RO membrane separation process is the most promising tool for saline water treatment and
has overtaken the conventional thermal processes like multistage ashing (MSF), multieffect distillation (MED) and electrodialysis [57]. RO
technology requires less energy and endures a great potential for solving and consistently alleviating the most pervasive problem of water
scarcity. Desalination process by RO has become a viable solution with

http://dx.doi.org/10.1016/j.desal.2014.12.041
0011-9164/ 2014 Elsevier B.V. All rights reserved.

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

the advent of membrane technology due to its high efciency, low energy consumption and operational simplicity [6]. However, a lot of potential in membrane technology entails to be explored and optimized.
The polymers withstand as the prime choice for [7] membrane technology but due to varied chemical and physical properties only a few selected polymers can be employed [8]. Cellulose acetate (CA) is the rst
polymer specically adequate for desalination process among other
polymeric materials like polyvinylpyrrolidone, polyvinyl alcohol, and
polyamide [9,10]. CA membranes have prominent desalination efciency, ease of availability, relatively low cost, nontoxicity, biocompatibility,
hydrophilicity, lm forming ability, good solubility in the organic solvents and high ux potential [1113]. CA has some demerits like low
oxidation, chemical and thermal resistance and poor mechanical
strength [10,1416]. In order to delineate these issues, CA is doped
with other polymers which improved the membrane performance, permeability and permselectivity as compared to pristine polymer [17].
Polyethylene glycol (PEG 200) can enhance the hydrophilicity and
thermal stability while promoting the environmental friendly chemical
reactions in the membrane [18,19]. It has been investigated that the
polymeric membranes incorporated with silica (SiO2), titania (TiO2),
alumina (Al2O3) and zirconia (ZrO2) overcome the trade-off phenomenon and inveterate the improved mechanical properties, fouling resistance and water permeability [2022]. Doping of FSP functionally
affects the surface chemistry and transport properties of the polymer
chains like polyvinylpyrrolidone, and polyvinyl alcohol and enhances
the selectivity and permeability of the membrane [23,24]. The study of
silica in low molecular weight PEG showed that silica interacts with
PEG through hydrogen-bonding between oxygen in PEG chains and
silanol groups (\Si\OH) in the silica [25].
In this study, novel CA/PEG 200 polymer matrix composite membrane (PMCM) doped with FSP was synthesized for RO desalination application. Employing the doping of FSP (1030 wt.%) in membranes was
characterized by Fourier transform infrared spectroscopy (FTIR), differential scanning calorimetry/thermogravimetric (DSC/TGA), scanning
electron microscopy (SEM), atomic force microscopy (AFM) and mechanical testing. Assessment of PMCM for desalination application in
terms of permeation ux and salt rejection was carried out in indigenously fabricated RO pilot plant using 0.35 wt.% NaCl solution at 25 C
temperature and operating pressure of 4.0 bar (osmotic pressure
2.9 bar).
2. Experimental
2.1. Materials
Analytical grade cellulose acetate (CA) was obtained from Fluka
(USA) and fumed silica (FS) powder of around 1 m size was obtained
from Sigma Aldrich. Polyethylene glycol (PEG 200) and N,N-di methyl
formamide (DMF) were obtained from Fluka and Riedel-de Ha n, respectively. All the chemicals were used without further purication.
2.2. Membrane preparation
2.2.1. Preparation of CA/PEG 200 polymer matrix membrane
Polymer matrix membrane solutions were prepared by using dissolution casting method in which 5 g of CA was dissolved in 40 mL of DMF
solvent with constant stirring at 80 C for each solution for 6 h. The varying amount (w/w) of CA/PEG 200 (50/50, 60/40, 70/30 and 80/20) was
used. The viscous and clear solution obtained which was termed as
polymer matrix membrane solutions. The solutions were cooled at
room temperature for 2 h in an air tight condition. The stay time was
given in order to remove bubbles.
2.2.2. Membrane casting
The polymer matrix membrane solutions were spread slowly in
glass sheet by maintaining uniform thickness. The membranes were

then placed in oven at 60 C which accelerates the evaporation of solvent and clear membranes were obtained. The polymer matrix membranes were carefully removed from glass sheet by using sharp blades.
On the basis of salt rejection membrane with the ratio of CA/PEG 200
(80/20) was selected for further treatment.
2.2.3. Doping of FSP in polymer matrix membrane
The different concentrations of FSP (1030 wt.%) were doped in
polymer matrix membrane (CA/PEG 200) solution with 80/20 (w/w)
and magnetically stirred for 12 h at 55 C until homogeneous PMCM solutions were formed. The solutions were casted on a clean dried glass
plate with doctor's blade. The excess amount of solvent was removed
at 40 C in an oven. Finally, the dried polymer matrix based composite
membranes were removed from the glass plates and the membrane
thickness ranges from 25 to 35 m was measured by screw gauge.
3. Characterization of membranes
3.1. Fourier transform infrared spectroscopy
FTIR spectra of PMCMs were recorded by IR Prestige-21 (Shimadzu)
using the attenuated total reectance (ATR) technique. The air background of the instrument was run before each test of membrane. The
wavelength range was from 4000 to 600 cm1 at 60 scans per
spectrum.
3.2. Differential scanning calorimetry/thermogravimetric analysis
The differential scanning calorimetry (DSC)/thermogravimetric
analysis (TGA) measurement of the PMCM was carried out using a TA
instrument SDT Q600 simultaneous differential scanning calorimetry/
thermogravimetric analyzer. The thermal analysis of the membranes
was carried out from an ambient temperature to 1200 C at a ramp
rate of 20 C/min under the nitrogen atmosphere (15 mL/min). Also,
the glass transition temperature (Tg) of the membranes was determined by DSC.
3.3. Scanning electron microscopy
Scanning electron micrographs of FSP doped PMCM were obtained
using a JEOL (JSM-6480LV) microscope to analyze the morphology
and properties of the membranes. The PMCM was examined on a
conducting tape. The low vacuum mode was used to operate PMCM at
20 kV while the morphology was observed at 2 and 10 m
magnications.
3.4. Atomic force microscope
The topographical images were obtained using a scanning probe microscope (SPM 9500J3, Shimadzu) with tapping mode at room temperature. The scanning area was 5 m2. The values of root mean square
(RMS) roughness were derived from AFM images, which were obtained
from the average of the values measured in random areas. The membrane surface morphology was expressed in terms of various roughness
parameters, like mean roughness (Ra), the parameter represents the
mean value of the surface relative to the center plane and the plane
for which the volumes enclosed by the image above and below this
plane are equal. It is calculated by Eq. (1).
Z
Ra 1=LxLy

Lx
0

Ly
0

j f x; yjdxdy

where f(x, y) is the surface relative to the center of plane while Lx & Ly
represent the dimensions of surface in x and y directions respectively.
Whereas, the root mean square average (RMS) of the measured
height deviations from the mean surface is taken within the evaluation

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

weighed [26]. The water content % was measured by Eq. (3).

area and calculated by Eq. (2).


Z
RMS 1=Ae

Lx

Ly
0

1=2
2
Z x; ydxdy
:

WC% wet sample weight dry sample weight= wet sample weight  100

3.7. Reverse osmosis performance test

3.5. Mechanical testing
The mechanical properties such as tensile strength and Young's
modulus for PMCM were evaluated at room temperature using a mechanical tester M350-10CT ROCHOALL, England (Testometric) having
a load cell of 10 kN. The dumbbell shape of the membrane was used
as per ASTM D-638 standard with gauge length and width of 50 and
10 mm, respectively. The experiments were carried out at room temperature with 25 2% relative humidity. Three specimens were tested for
each membrane and the average value was recorded at a crosshead
speed of 10 mm/min.

3.6. Water content

Dried membrane (1 g) was placed in vial lled with distilled water
(100 mL) and the vials were set at room temperature. The water content
(%) of the membrane was attained for 24 h and weighed. The wet membranes were kept for 48 h in an oven under vacuum at 70 C and

RO experiments were performed on the plate and frame membrane

module as shown in Fig. 1 (process ow diagram) [7]. The feed of
known conductance was 0.35 wt.% NaCl in water at 25 C. The effective
area of membrane in contact with the continuous feed ow was
154 cm2. Permeate was collected for 8 h in a continuous RO membrane
unit operation. The provided feed tank capacity was 10 L and feed
inlet along with circulation pump of 1 kW followed by a rotameter.
The RO performance tests of the membranes were evaluated on basis
of salt rejection SR (%), salt passage SP (%) and permeation ux J (L/
m2h) using Eqs. (4), (5) and (6). The permeate and retentate were collected during operation from the sampling point, respectively. The operating pressure during the process was maintained at 4 bar (osmotic
pressure 2.9 bar). The conductance of permeate acquired by using
Cyber Scan Waterproof PC 300 Series (EUTECH).
Salt rejection
h
 %SR

1 Permeate conductance Cp =Feed conductance C f

Salt passage % SP 1 Salt rejection %

i

 100 4
5

Fig. 1. Process ow diagram of the reverse osmosis/pervaporation unit.

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

Scheme 1. FTIR spectra of FSP, M and doped (10-30 wt.%) PMCMs membranes.

h
 
i
2
Permeation flux PermeateL=membrane area m  Timeh 6

4. Results and discussion

4.1. Fourier transform infrared spectroscopy
Spectroscopic analysis plays a vital role in the characterization of
polymers and the incorporated components [2729]. The possible
inter and intramolecular hydrogen bonding between FSP with CA/PEG
200 polymer matrix membrane is shown in the following Scheme 1.
FTIR analysis was conducted to elucidate the nature and correlation
between the polymer matrix chains and the doped FSP. Fig. 2 shows the
FTIR spectrum of M, PMCM and FSP. Generally, FSP bands were found at
1074 cm1 (Si\O\Si asymmetric stretching) [30] and 797 cm 1
(Si\O\Si) symmetric stretching [31]. M membrane (CA/PEG 200) exhibited bands at 3435 cm1 (hydrogen bonded hydroxyl group) [32],
2889 cm1 (C\H stretching vibration) [32,33],1742 cm1 (stretching
vibration of C_O of aldehyde CA) [32,33], 1369 cm1 (\OH bending

vibration) [34], 1225 cm1 (C\O single bond stretching modes) [33],
1067 cm1 (C\O\C ether linkage from the glycosidic unit) [34] and
909 cm 1 (pyranose ring). On the other hand, spectra for PMCM
doped with FSP (1030 wt.%) which showed bands at 3742 (MFS 1
MFS 4) and 3739 cm1 (MFS 5) conrmed the presence of \OH polar
groups in the silica particles (Si\OH). The bands at 2883 and
2922 cm1 showed C\H stretching vibrations of MFS 1 and MFS 3, respectively while MFS 2, MFS 4 and MFS 5 C\H stretching vibration
bands were observed at 2924 cm1.
For all PMCM, bands at 1371 and 1454 cm1 were considered due to
symmetric and asymmetric vibrations of \CH3, respectively [35], bands
at 1739 cm1 were due to C_O stretching vibrations, pyranose ring
stretching vibrations at 904 cm 1 [36] and the characteristic bands
around 690 cm1 were due to C\H bending vibrations [7]. Hydrogen
bonding was developed by the addition of FSP resultantly weakened
the force constant of \OH bond which was implied for the shifting of
peak towards lower wavenumber. The intensity of carbonyl band
(1742 cm 1) was decreased with FSP addition in polymer matrix
which conrmed the hydrogen bonding between carbonyl and hydroxyl groups. Generally, the carbonyl group interaction through hydrogen

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

Fig. 3. DSC of FSP, M and doped (1030 wt.%) PMCM membranes.

4.3. Scanning electron microscopy

Fig. 2. FTIR spectra of FSP, M and doped (1030 wt.%) PMCM membranes.

bonding with FSP can also be conrmed by peak shifting towards lower
wavenumber. These results indicated the strong interactions of doped
FSP in PMCM [37].

4.2. Differential scanning calorimetry/thermogravimetric analysis

4.2.1. Differential scanning calorimetry
The DSC thermograms of PMCM (M-MFS 5) with and without different FSPs (030 wt.%) are shown in Fig. 3. Polymeric membrane nature
and its structure were interpreted by employing glass transition temperature (Tg) [38] as stronger the Tg, the membrane has stable structure.
It was shown that M without FSP exhibited Tg at 78.3 C while with the
addition of FSP in MFS 1MFS 5, it was increased up to 87.4, 88.2, 89,
89.8, and 92.4 C respectively. This showed that FSP has developed the
strong interactions with CA/PEG 200 as the chains become wellordered and more packed in nature.

4.2.2. Thermogravimetric analysis

Thermal properties of the pure CA/PEG 200 and FSP doped PMCM
were analyzed by TGA. Fig. 4 showed the decomposition of FSP and it
was observed that the degradation was completed mainly in 4 steps;
in the rst step, decomposition was started at about 30 C and 0.43%
degradation was observed, in the second, third and fourth steps, degradation was occurred at 127.37, 516.03 and 984.65 C with degradation
of about 0.69, 0.75 and 0.053%, respectively. Thermograms of FS showed
slight degradation which indicated its high degradation temperature.
Thermograms of PMCM (M-MFS 5) showed that the rst step
weight loss from 30 to 180 C was attributed to moisture removal, dehydration and loss of bound water. The second step showed that onset
deacetylation of polymer chains started degradation at 250 C [3941]
with almost 90% of polymer weight loss up to offset temperature (450
C). The nal step indicated the temperature range above 380 C indicating carbonization of the thermally decomposed product to ash [41]. The
TGA curve of MFS 5 was thermally stable as compared to the other FS
particle contained PMCMs, whereas, 10 and 70 wt.% loss of M membrane was at 190.08 and 375.1 C which increased up to 277.6 and
612 C in MFS 5, respectively. It was cleared from the graph that the
thermal stability of the FS doped PMCM was higher than that of M
membrane exhibiting the improvement in the thermal properties.

A better performance of the morphological structure of the membranes has been attained by top surface and cross sectional views.
Fig. 5 FSP (a, b) showed the morphology of pure FSP and M-MFS 5 (a,
b) showed the top and cross-section views of the membranes. The top
and cross section views of the membrane M showed without FSP, the
dense membrane structure with smooth surface was attained and similar observation was showed by CA and Pluronic F127 membranes [42].
Fig. 5 MFS 1MFS 5 (a, b) showed the uniform distribution and dispersion of the doped FSP. The protrusions of the PMCM were observed all
over the surface which comprised of FSP or their aggregate. Apparently,
PMCM with high FSP (30 wt.%) showed an aggregate behavior. Moreover, the FSP doped PMCM led to the formation of packed chain structure by the formation of no spongy like holes as observed in the crosssectional images. The cross-sectional layer of the PMCM without FSP
(M) had a fused nodule structure, whereas, the PMCM lled with FSP
had apparently loose nodule and macro void like structure (maximum
at 30 wt.%). This showed the existence of interfacial stresses between
FSP and polymer matrix. Similar results were shown with PVDF membrane containing lithium chloride [43]. It showed that with the increase
in the concentration of FSP, asymmetric structure of MFS 5 membrane
was obtained and it became faint as shown in CA blended membranes

Fig. 4. TGA of FSP, M and doped (1030 wt.%) PMCM membranes.

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

Fig. 5. Surface image of FSP (a, b). Surface and cross-sectional images of FSP doped PMCM (M-MFS 5) (a, b).

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

Fig. 6. AFM 3-dimensional image of M and PMCM (MFS 1MFS 5).

A. Sabir et al. / Desalination xxx (2014) xxxxxx

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

Table 1
Sample name of with and without doped FSP (1030 wt.%) PMCM.
Membrane sample

Weight fraction of FSP (wt.%)

Water content (%)

M
MFS 1
MFS 2
MFS 3
MFS 4
MFS 5

0
10
15
20
25
30

77.91
79.28
81.52
83.41
85.15
86.24

with ZnO [7]. The pores in a RO dense membrane skin can be envisaged
as voids in the membrane material and depends strongly on the method
of membrane fabrication [44]. In dense membrane, the pores or density
defects are caused by the irregular packing of almost randomly kinked
stiff-chain molecules. The pores may be circular and non-circular in
shape, open or closed, and may or may not form a continuous and
inter-connecting network [45].
SEM images showed compatibility between CA/PEG 200 polymer
matrix and FSP. The images exhibited that agglomeration of FSP increased with the increase in concentration of FS which was due to physical association of chains of polymer matrix and FSP.
FTIR analysis also conrmed the connection between FSP and polymer matrix chains of CA/PEG 200 [37,46].

4.4. Atomic force microscope

The atomic force microscopy was employed for surface topography
of the membranes and the AFM images are shown in Fig. 6. The CA/
PEG 200 polymer matrix membrane without FSP showed consistent valley and ridge like structure [47]. However, the average and root mean
square roughness (Ra and Rms) of surface increased with increase in
doped FSP as given in Table 2. The elevated surface roughness presented
better condition for the adsorption on the membranes [48]. However,
the membrane with more roughness provided greater membrane surface area and hence more active sites were available for adsorption.
The reactive functionalities of \OH group of polyethylene glycol in
membrane structure have generated the morphological improvement
in the membrane. The doping of different amounts of FSP ensured the
roughness in the valleys and ridges of the RO membrane. The AFM images clearly show the remarkable differences between the surface morphologies of the membranes. The FSP doped composite membrane
exhibits large-scale surface roughness of ridge-and valley structure
compared to the undoped polymer membrane surface which is relatively smooth.
There is a strong correlation between the membrane fouling and the
surface roughness. It was shown that RO hydrophilic membranes with
smooth membrane surfaces are less prone to colloidal fouling compared
to the relatively more hydrophobic and rougher membranes. The particles preferentially accumulate in the valleys of rough membranes,
resulting in valley clogging which causes more severe ux decline
than in smooth membranes [49,50]. This is probably due to a signicant
decrease in the electrostatic repulsion between the FSP and the surface
of membrane, resulting in the experience of a greater adhesion force in
the valley regions [51].

Fig. 7. Effect of doped FSP (1030 wt.%) on the tensile strength of M and PMCM (MFS 1
MFS 5) membranes.

This showed the important implications for the desalination of salt

water throughout the membrane; with the increase in the roughness,
the surface area increased which ultimately enhanced salt rejection %,
as shown in the salt rejection analysis.
4.5. Mechanical testing
The tensile strength and Young's modulus of the M-MFS 5 PMCM are
shown in Figs. 7 and 8, respectively. It can be seen that with an increase
in FSP, both the tensile strength and the Young's modulus of the PMCM
were increased. This tendency was because of the addition of FSP which
in turn increased the polymer matrix chain mobility and the mechanical
strength of the membranes. It can be concluded that the interaction between polymer matrix and FSP was strong enough which mechanically
strengthened the membrane. Figs. 7 and 8 showed that the tensile
strength (2.1 MPa) and Young's modulus (380 MPa) of M membrane
were lower when compared with FSP doped PMCM (MFS 5) having tensile strength (8.2 MPa) and Young's modulus (854 MPa).
4.6. Water content
The water content (WC) % of the membrane exhibited the hydrophilic nature of the membrane [10]. WC analysis of the membranes is

Table 2
Roughness values with and without FSP doped PMCM.
Membrane type

Average roughness
Ra (nm)

Root mean square roughness

Rms (nm)

M
MFS 1
MFS 2
MFS 3
MFS 4
MFS 5

21.76
56.61
58.84
50.59
57.69
98.92

27.61
69.97
69.29
64.34
73.59
116.01

Fig. 8. Effect of FSP (1030 wt.%) on the Young's modulus of M and PMCM membranes.

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

study the hydratability of the PMCM membranes that exhibits a rapid

increase in absorbed water content due to the hydrophilic nature of
the membranes.
Eq. (3) used to calculate the water content (%) for M membrane
(without FSP) i.e. 77.91% as shown in Table 1. In the case of 10 wt.% of
FSP, the water content % showed the value of 79.28%. When FS contents
reached to 30 wt.%, the water content % was increased up to 86.24%.
Similar increase in water content % was shown [42] in which Pluronic
(F127) was added in the CA membranes.
4.7. Reverse osmosis performance test

Fig. 9. Effect of time on permeate ux for M and PMCM membranes.

attributed to the absorption mechanism caused by the diffusion process.

This diffusion process is controlled by the afnity between the polymer
chains and water (external media). The hydrophilic nature of PEG 200
in the membranes acted as a driving force for sorption of water into
the membrane. PEG 200 can absorb more water molecules inside the
polymer membrane matrix [55]. These absorption properties of the
membranes play an important role in RO performance processes,
which affect the membrane permeation ux. WC analysis was to

4.7.1. Analysis of membrane fouling during permeation ux

The doping of membrane by FSP results in surface modication with
increased roughness resulting in more prone to fouling phenomena and
effecting permeation ux. Fig. 9. shows the relationship of permeation
ux with respect to time for M and PMCM membranes.
Compared to M polymer matrix membrane, the surface roughness of
MFS 15 was gradually increasing as exhibited in AFM images which
show remarkable differences between the membrane surface morphologies. The results show that proper concentration of FSP in the membrane was benecial for salt rejection but unfavorable to greater
permeation ux. The decline in membrane uxes was observed with
the time due to the membrane fouling. The reason was that a signicant
decrease in the electrostatic repulsion between the FSP and surface results in greater adhesion force in the membrane regions [51]. The particles agglomerate in the membranes, resulting in clogging which causes
permeation ux decline than that in smooth membranes [50].

Fig. 10. Effect of CA/PEG 200 concentration on permeation ux (A), salt rejection (B) and salt passage (C).

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

10

A. Sabir et al. / Desalination xxx (2014) xxxxxx

Fig. 13. Effect of FSP (1030 wt.%) doped in PMCM on permeation ux and salt rejection.

Fig. 11. Comparison of permeation ux of FSP doped (1030 wt.%) PMCM with M polymer
matrix without FSP.

4.7.2. Transport properties of membranes

The salt water feed containing 0.35 wt.% NaCl in water having osmotic pressure 2.9 bar with pH = 7.4 i.e. slightly alkaline was pumped
into a vessel in indigenously fabricated RO pilot plant. The operating
pressure of 4.0 bar with 25 L/h (feed ow rate) was applied to the RO
membranes. The operating pressure of 4.0 bar which was required in
the process enabled the pure water to pass through the membranes
leaving salts.
The results showed that M 3 membrane has a maximum ux of 1.83
(L/m2h) while it has a minimum salt rejection capacity of 48%. As the
content of cellulose acetate increases, the ux slightly decreased while
salt rejection capacity increased remarkably with decrease in salt passage. M membrane showed 78% salt rejection and its ux was 1.312
(L/m2h) with maximum CA content and minimum PEG 200 content.
The permeation ux, salt rejection and salt passage were calculated
and represented in Fig. 10.

Fig. 12. Comparison of salt rejection of FSP doped (1030 wt.%) PMCM with M polymer
matrix without FSP.

The transport through reverse osmosis membrane was described in

terms of diffusive ow in which separation occurs in membrane when
both solute and solvent permeate by solution and diffusion process.
The transport mechanism best explained on the basis of solution diffusion model. According to this model, transport process within membrane involves sorption at the surface of membrane, diffusion into
dense membrane under pressure and then desorption. The hydrophilic
nature of PEG 200 acts as driving force for sorption of water on membrane [5254].
CA desalting nature caused removal of salt [55]. Furthermore, salt rejection occurs on the basis of electrostatic repulsion known as the
Donnan effect [53]. The ions excluded by the Coulomb forces which
arise from charges residing on surface of membrane.
CA showed desalting property and improves the salt rejection efciency. This is a reason behind the maximum salt rejection ability of M
membrane.
The PEG 200 depicts hydrophilic nature and acts as a porogen [54].
M 3 polymer matrix membrane had a maximum quantity of PEG 200
giving more water ux but at the same time salt rejection was compromised. This may be attributed to the formation of macrovoids [55] on
membrane which allows the passage of salt along with water and
resulting in raised ux with increased salt passage but reduced salt rejection. The greater the PEG 200 content the higher will be the ux.
The diffusion rate of water is accelerated by the presence of PEG 200 attributed to its hydrophilic nature. PEG 200 proliferates the tendency of
porogen and as a consequence the permeation ux increased [56]
which is shown in Fig. 10. When salt rejection efciency of membrane
was optimum the salt passage % reduced depending upon the concentration of CA/PEG 200.
On the basis of salt rejection, M polymer matrix membrane was selected for doping with FSP. After doping these PMCMs were characterized and compared with the M polymer matrix membrane. The
permeation ux, salt rejection and salt passage of PMCM using various
concentrations of FSP are given in Figs. 13 and 14.
The water salinity was decreased with the increment of FSP concentration in the PMCM. The permeation ux salt rejection % and salt passage were signicantly affected by the integration of FSP. MFS 1
membrane showed a maximum ux of 1.09 L/m2h but as doping of
FSP increases the ux gradually declines. The results show that permeation ux declines at higher doping concentration of FSP as shown in
Fig. 13. The presence of FSP particles on the surface of the membrane reduces the effective/exposed membrane surface area accessible to permeation. The rate of convective particle transport (i.e., the product of
permeate ux and particle concentration) towards the membrane surface increases which lead to some agglomeration at surface of the membrane, which can be seen from SEM micrographs and consequently

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

A. Sabir et al. / Desalination xxx (2014) xxxxxx

11

(UNIMAS), which supported this work through grant DPI [Grant ref. no.
02 (DPI23)/ 999/2013 (06)].
References

Fig. 14. Effect of FSP (1030 wt.%) doped in PMCM on salt rejection and salt passage.

lowering the ux [57]. These PMCM reduced ux as compared to M

polymer matrix membrane. The comparison between the M and
PMCM membrane is given in Fig. 11. It is evident that the permeation
ux of PMCM is reduced. MFS 5 gives the lowest ux of 0.66 L/m2h.
The salt rejection capacity of PMCM is enhanced when compared
with M membrane as indicated in Fig. 12. MFS 5 showed the highest
salt rejection capacity of 98.4% attributed to FSP 30 wt.% doped in
PMCM as shown in Figs. 13 and 14.
The effect on permeation ux was due to doping of FSP. This PMCM
possessed a hydrophilic nature as its driving force. Resultantly, the salt
rejected through the hydrophilic membranes because of its ability for
hydrogen bonding and smaller molecular sizes. The doping of FSP
changed the permeate mobility and also the polymer chain segmental
motion which increased the free volume of PMCM. The electrostatic interactions develop due to FSP and dense membrane formation with no
voids which resultantly increases the chances of salt exclusion.
5. Conclusion
The doping of FSP (1030 wt.%) on the structure and performance of
PMCM for desalination has been studied and conrmed by FTIR analysis.
The SEM images showed uniform dispersion of FSP on PMCM having inuence on membrane morphology and a dense membrane structure.
Generally, doping of FSP led to the increase in the thermal stability of
polymer matrix membrane. The mechanical properties have been increased with the addition of FSP as compared to the membrane without
FSP. The water content % showed that the MFS 5 composite membrane
has the highest water content as compared to other membranes. The
systematic study in this work for desalination of salt water revealed
its feasibility for RO process and has been studied by permeation ux
and salt rejection. It showed that 1030 wt.% doping of FSP into CA/
PEG 200 PMCM distinctly increased the salt rejection % at operating
pressure of 4.0 bar (osmotic pressure 2.9 bar) and at 25 C. Higher content of FSP 30 wt.% (MFS 5) led to the signicant increase in salt rejection of 98.4% and the decrease in permeation ux of 0.66 L/m2h
compared to M polymer matrix membrane. The results have shown a
new approach for RO process and the developed membranes doped
with FSP have been exible enough for the desalination of salt water.
Acknowledgments
The authors express their cordial gratitude to the team of the Department of Polymer Engineering and Technology, University of the
Punjab, for their co-operation during the execution of this research project. The authors are also thankful to the University of Malaysia Sarawak

[1] O.A. Bamaga, A. Yokochi, B. Zabara, A.S. Babaqi, Hybrid FO/RO desalination system:
preliminary assessment of osmotic energy recovery and designs of new FO membrane module congurations, Desalination 268 (13) (2011) 163169.
[2] A. AlTaee, A.O. Sharif, Alternative design to dual stage NF seawater desalination
using high rejection brackish water membranes, Desalination 273 (23) (2011)
391397.
[3] W. Den, C.-J. Wang, Removal of silica from brackish water by electrocoagulation pretreatment to prevent fouling of reverse osmosis membranes, J. Sep. Purif. Technol.
59 (2008) 318325.
[4] H.L. Yang, C. Huang, J.C.-T. Lin, Seasonal fouling on seawater desalination RO membrane, Desalination 250 (2010) 548552.
[5] P.H. Gleick, The World's Water 20082009 the Biennal Report on Freshwater Resources, Island Press, Chicago, 2008.
[6] A. Malas, P. Parthajit, S. Giri, A. Mandal, C.K. Das, Synthesis and characterizations of
modied expanded graphite/emulsion styrene butadiene rubber nanocomposites:
mechanical, dynamic mechanical and morphological properties, Compos. Part B 58
(2014) 267274.
[7] M. Ali, M. Zafar, T. Jamil, M.T.Z. Butt, Inuence of glycol additives on the structure
and performance of cellulose acetate/zinc oxide blend membranes, Desalination
270 (2011) 98104.
[8] M. Mulder, Basic Principles of Membrane Technology, Second Edition Kluwer Academic Pub., 1996
[9] H. Bai, Y. Zhou, X. Wang, L. Zhang, The permeability and mechanical properties of
cellulose acetate membranes blended with polyethylene glycol 600 for treatment
of municipal sewage, Procedia Environ. Sci. 16 (2012) 346351.
[10] M. Sivakumar, D.R. Mohan, R. Rangarajan, Studies on cellulose acetatepolysulfone
ultraltration membranes: II. Effect of additive concentration, J. Membr. Sci. 268
(2006) 208219.
[11] E. Saljoughi, M. Sadrzadeh, T. Mohammadi, Effect of preparation variables on morphology and pure water permeation ux through asymmetric cellulose acetate
membranes, J. Membr. Sci. 326 (2009) 627634.
[12] J.-J. Qin, Y. Li, L-S., H. Lee, Cellulose acetate hollow ber ultraltration membranes
made from CA/PVP 360K/NMP/water, J. Membr. Sci. 218 (2003) 173183.
[13] D.Y. koseoglu-Imer, N. Dizge, I. Koyuncu, Enzymatic activation of cellulose acetate
membrane for reducing of protein fouling, Colloids Surf. B: Biointerfaces 92
(2012) 334339.
[14] D. Zavastin, I. Crestescu, M. Bezdadea, M. Bourceanu, M. Drgan, G. Lisa, I.
Mangalagiu, V. Vasi, J. Savic, Preparation, characterization and applicability of cellulose acetatepolyurethane blend membrane in separation techniques, Colloids
Surf. A Physicochem. Eng. Asp. 370 (2010) 120128.
[15] O. Kutowy, S. Sourirajan, Cellulose acetate ultraltration membranes, J. Appl. Polym.
Sci. 19 (1975) 14491460.
[16] M. Sivakumar, R. Malaisamy, C.J. Sajitha, D. Mohan, V. Mohan, R. Rangarajan, Ultraltration application of cellulose acetatepolyurethane blend membranes, Eur.
Polym. J. 35 (1999) 16471651.
[17] M. Sivakumar, R. Malaisamy, C.J. Sajitha, D. Mohan, V. Mohan, R. Rangarajan, Preparation and performance of cellulose acetatepolyurethane blend membranes and
their applicationsII, J. Membr. Sci. 169 (2000) 215228.
[18] H. Firouzabadi, N. Iranpoor, M. Gholinejad, A. Samadi, Copper(I) iodide catalyzes
odorless thioarylation of phenolic esters with alkyl derivatives using thiourea in
wet polyethylene glycol (PEG 200), J. Mol. Catal. A Chem. 377 (2013) 190196.
[19] J. Chen, S.K. Spear, J.G. huddleston, R.D. Rogers, Polyethylene glycol and solutions of
polyethylene glycol as green reaction media, Green Chem. 7 (2005) 6482.
[20] Y. Yang, H. Zhang, P. Wang, Q. Zheng, J. Li, The inuence of nano-sized TiO2 llers on
the morphologies and properties of PSF UF membrane, J. Membr. Sci. 288 (2007)
231238.
[21] N.M. Wara, L.F. Francis, B.V. Velamakanni, Addition of alumina to cellulose acetate
membranes, J. Membr. Sci. 104 (1995) 4349.
[22] R.P. Castro, Y. Chohen, H.G. Monbouquette, Silica-supported polyvinylpyrrolidone
ltration membranes, J. Membr. Sci. 115 (1996) 179190.
[23] J. Fan, S.R. Raghavan, X.-Y. Yu, S.A. Khan, P.S. Fedkiw, J. Hou, G.L. Baker, Composite
polymer electrolytes using surface-modied fumed silicas: conductivity and rheology, Solid State Ionics 111 (1998) 117123.
[24] M.G.D. Angelis, G.C. Sarti, Solubility and diffusivity of gases in mixed matrix membranes containing hydrophobic fumed silica: correlations and predictions based
on the NELF model, Ind. Eng. Chem. Res. 47 (2008) 52145226.
[25] J. Lindn, Surface Modied Silica Nanoparticles as Emulsier, 2012.
[26] M. Sivakumar, A.K. Mohanasuundaram, D. Mohan, K. Balu, R. Rangarajan, Modication of cellulose acetate: its characterization and application as an ultraltration
membrane, J. Appl. Polym. Sci. 67 (1998) 19391946.
[27] E.G. Brame, Uncommon methods for spectroscopic analyses of polymers,
Makromolekulare Chemie. Macromolecular Symposia, Wiley Online Library, 1993.
[28] P.R. Grifths, M.W. Urban, Fourier transform IR and Raman spectroscopy of polymers, Multidimensional Spectroscopy of Polymers, American Chemical Soc., 1995,
pp. 27.
[29] J.M. Chalmers, N.J. Everall, The bole of vibrational spectroscopymicroscopy techniques in polymer characterisation, Macromol. Symp. 94 (1995) 3349.
[30] J. Liu, S. Wu, M. Zou, X. Zheng, Z. Cai, Surface modication of silica and its
compounding with polydimethylsiloxane matrix: interaction of modied silica ller
with PDMS, Iran. Polym. J. 21 (2012) 583589.

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041

12

A. Sabir et al. / Desalination xxx (2014) xxxxxx

[31] M.A. Semsarzadeh, B. Ghalei, Preparation, characterization and gas permeation

properties of polyurethanesilica/polyvinyl alcohol mixed matrix membranes, J.
Membr. Sci. 432 (2013) 115125.
[32] G.-Q. Qi, C.-L. Liang, R.-Y. Bao, Z.-Y. Liu, W. Yang, B.-H. Xie, M.-B. Yang, Polyethylene
glycol based shape-stabilized phase change material for thermal energy storage
with ultra-low content of graphene oxide, Sol. Energy Mater. Sol. Cells 123 (2014)
171177.
[33] D. Murphy, M.N. de Pinho, An ATR-FTIR study of water in cellulose acetate membranes prepared by phase inversion, J. Membr. Sci. 106 (1995) 245257.
[34] H. Kamal, F.M. Abd-Elrahim, S. Lotfy, Characterization and some properties of cellulose acetate-co-polyethylene oxide blends prepared by the use of gamma irradiation, J. Radiat. Res. Appl. Sci. 7 (2014) 146153.
[35] W. Zhou, J. He, S. Cui, W. Gao, Studies of electrospun cellulose acetate nanobrous
membranes, Open Mater. Sci. J. 5 (2011) 5155.
[36] A. Islam, T. Yasin, Controlled delivery of drug from pH sensitive chitosan/poly (vinyl
alcohol) blend, Carbohydr. Polym. 88 (2012) 10551060.
[37] M. Naghsh, M. Sadeghi, A. Moheb, M.P. Chenar, M. Mohagheghian, Separation of
ethylene/ethane and propylene/propane by cellulose acetatesilica nanocomposite
membranes, J. Membr. Sci. 423 (2012) 97106.
[38] A.k. Fritzsche, C.A. Cruse, R.E. Kesting, M.K. Murphy, Polysulfone hollow ber membranes spun from Lewis acid: base complexes. II. The effect of Lewis acid to base
ratio on membrane structure, J. Appl. Polym. Sci. 39 (1990) 19491956.
[39] J.-J. Shieh, T.S. Chung, Effect of liquidliquid demixing on the membrane morphology, gas permeation, thermal and mechanical properties of cellulose acetate hollow
bers, J. Membr. Sci. 140 (1998) 6779.
[40] M.D.C.C. Lucena, A.E.V.D. Alencar, S.E. Mazzeto, S.D.A. Soares, The effect of additives
on the thermal degradation of cellulose acetate, Polym. Degrad. Stab. 80 (2003)
149155.
[41] P.K. Chatterjee, C.M. Conrad, Thermogravimetric analysis of cellulose, J. Polym. Sci. A
1 Polym. Chem. 6 (1968) 32173233.
[42] C. Lv, Y. Su, Y. Wang, X. Ma, Q. Sun, Z. Jiang, Enhanced permeation performance of
cellulose acetate ultraltration membrane by incorporation of Pluronic F127, J.
Membr. Sci. 294 (2007) 6874.
[43] D.-J. Lin, C.-L. Chang, F.-M. Huang, L.-P. Cheng, Effect of salt additive on the formation
of microporous poly (vinylidene uoride) membranes by phase inversion from
LiClO4/water/DMF/PVDF system, Polymer 44 (2003) 413422.

[44] K. Kouti, L. Katelan-Kunst, B. Kunst, Porosity of some commercial reverse osmosis

and nanoltration polyamide thin-lm composite membranes, J. Membr. Sci. 168
(2000) 101108.
[45] K. Kamide, H. Iijima, Recent advances in cellulose membranes, in: R.D. Gilbert (Ed.),
Polymers, Blends and Composites, Carl Hanser Verlag, Munich, 1994, pp. 189207.
[46] H. Wu, J. Mansouri, V. Chen, Silica nanoparticles as carriers of antifouling ligands for
PVDF ultraltration membranes, J. Membr. Sci. 433 (2013) 135151.
[47] Y. Zhao, W. Xing, N. Xu, F.-K. Wong, Effects of inorganic salt on ceramic membrane
microltration of titanium dioxide suspension, J. Membr. Sci. 254 (2005) 8188.
[48] A. Ghaee, M. Shariaty-Niasaar, J. Barzin, T. Matsuura, Effects of chitosan membrane
morphology on copper ion adsorption, Chem. Eng. J. 165 (2010) 4655.
[49] M. Elimelech, X.H. Zhu, A.E. Childress, S.K. Hong, Role of membrane surface morphology in colloidal fouling of cellulose acetate and composite aromatic polyamide
reverse osmosis membranes, J. Membr. Sci. 127 (1997) 101109.
[50] E.M. Vrijenhoek, S. Hong, M. Elimelech, Inuence of membrane surface properties
on initial rate of colloidal fouling of reverse osmosis and nanoltration membranes,
J. Membr. Sci. 188 (2001) 115128.
[51] W.R. Bowen, T.A. Doneva, Atomic force microscopy studies of membranes: effect of
surface roughness on double-layer interactions and particle adhesion, J. Colloid Interface Sci. 229 (2000) 544549.
[52] M. Mazid, Mechanisms of transport through reverse osmosis membranes, Sep. Sci.
Technol. 19 (1984) 357373.
[53] H.K. Lonsdale, U. Marten, R.L. Riley, Transport properties of cellulose acetate osmotic
membranes, J. Appl. Polym. Sci. 9 (1965) 13411362.
[54] H.-G. Burghoff, K.L. Lee, W. Pusch, Characterization of transport across cellulose acetate membranes in the presence of strong solutemembrane interactions, J. Appl.
Polym. Sci. 25 (1980) 323347.
[55] R. Malaisamy, R. Mahendran, D. Mohan, M. Rajendran, V. Mohan, Cellulose acetate
and sulfonated polysulfone blend ultraltration membranes. I. Preparation and
characterization, J. Appl. Polym. Sci. 86 (2002) 17491761.
[56] Y. Kurokawa, K. Ueno, Reverse osmosis rejection by hydrous inorganic precipitate
cellulose composite membrane, J. Appl. Polym. Sci. 27 (1982) 621630.
[57] X. Zhu, M. Elimelech, Colloidal fouling of reverse osmosis membranes: measurements and fouling mechanisms, Environ. Sci. Technol. 31 (1997) 36543662.

Please cite this article as: A. Sabir, et al., Novel polymer matrix composite membrane doped with fumed silica particles for reverse osmosis
desalination, Desalination (2014), http://dx.doi.org/10.1016/j.desal.2014.12.041