Sie sind auf Seite 1von 6

Elsevier Editorial System(tm) for Procedia CIRP

Manuscript Draft
Manuscript Number: PROCIR-D-14-00286R1
Title: Performance study of PVDF-TiO2 membrane with different molecular weight of PVP for oily
wastewater treatment
Article Type: SI: 12th CIRP GCSM
Keywords: TiO2; polyvinylpyrrolidone; submerged membrane reactor; ultrafiltration; oily solution;
flux recovery
Corresponding Author: Dr. Lau Woei-Jye, Ph.D
Corresponding Author's Institution: Universiti Teknologi Malaysia
First Author: Lau Woei-Jye, Ph.D
Order of Authors: Lau Woei-Jye, Ph.D; Chi Siang Ong, Mr; Pei Sean Goh, Dr; Be Cheer Ng, Mr; Ahmad
Fauzi Ismail, Ph.D; Takeshi Matsuura, Ph.D

Full Paper Manuscript

Available online at

Procedia CIRP 00 (2014) 000000

12th Global Conference on Sustainable Manufacturing

Performance study of PVDF-TiO2 membrane with different molecular

weight of PVP for oily wastewater treatment
C.S. Ong a,b, W.J. Lau a,b*, P.S. Goh a,b, B.C. Ng a,b, T. Matsuura a,c, A.F. Ismail a,b

Advanced Membrane Technology Research Centre (AMTEC), Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia.

Faculty of Petroleum and Renewable Energy Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia.

Industrial Membrane Research Laboratory, Department of Chemical and Biological Engineering, University of Ottawa, 161 Louis Pasteur St, Ottawa, ON,
Canada K1N 6N5.
* Corresponding author. Tel.: +60 75535926 E-mail address:;

Ultrafiltration (UF) hollow fiber membranes consisted of polyvinylidene fluoride (PVDF), titanium dioxide (TiO2) and different molecular
weight (Mw) of polyvinylpyrrolidone (PVP) (i.e. 10, 24, 40 and 360 kDa) were prepared to treat the synthesized oily wastewater. The
membrane performances were characterized in terms of pure water flux, permeate flux and oil rejection while the membrane morphological
properties were studied using SEM and AFM. Results showed that with increasing the Mw of PVP, the composite membrane surface is tended
to become smoother while the pore size and porosity become smaller. It is also found that the composite membrane prepared from PVP40k was
the best performing membrane due to its good combination of permeability and oil rejection during filtration process. A simple backflushing
was able to retrieve approximately 60% of initial flux without affecting the membrane rejection. Based on the above findings, the PVDF-TiO2
membrane prepared from PVP40k can be potentially considered as good candidate for sustainable development of oily wastewater treatment.
2014 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of Assembly Technology and Factory Management/Technische Universitt Berlin.
Keywords: TiO2, polyvinylpyrrolidone, submerged membrane reactor, ultrafiltration, oily solution, flux recovery

1. Main text
A large volume of oil emulsion wastewater is generated from
various process industries such as metallurgy, transportation,
food processing and petrochemical industries [1]. Typical
composition ranges of oil particles in the oily wastewater
generated from oil and gas industries are reported around 1001000ppm [2]. According to Environmental Quality Act
(EQA) 1974 [3], the maximum oil discharge concentration in
Malaysia is around 10ppm. Thus, a very effective method is
required to treat the oily wastewater before it was discharged
into the natural environment. The conventional separation
processes [4] such as gravity settling (API separator),
centrifuging, electrocoagulation, advanced oxidation and
flotation process that proved to be industrially acceptable

have associated with high operating cost and incomplete

removal of oil molecules, especially when the oil droplet size
is smaller than 20 m, which is also known as stable oil-water
emulsion [5, 6]. Membrane treatment offers a potential
solution to treat the oily wastewater due to the specific
membrane pore size that capable to exclude and retain most of
oil droplets. To improve the membrane separation
performances in particular for stable oil-water emulsion,
different kinds of additives were added to the membrane
matrix by other researchers. Polyvinylpyrrolidone (PVP) has
been commonly used in membrane formation due to its pore
forming ability. Many studies focused on the addition of
different molecular weight of PVP on the membrane filtration
performance. Jung et al.[7] found that the water flux was
improved with the addition of PVP but it was decreased when

2212-8271 2014 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of Assembly Technology and Factory Management/Technische Universitt Berlin.

Author name / Procedia CIRP 00 (2014) 000000

further increasing Mw of PVP from 10 to 360 kDa.

Chakrabarty et al.[8] reported that increasing Mw of PVP from
24 to 360 kDa could lead to lower flux and higher hydraulic
resistance due to the reduction of pore size caused by the
swelling of PVP on the membrane surface layer. Most of
researchers reported that the use of higher Mw suppressed the
formation of macrovoids and there should be optimum values
for Mw at which thermodynamic and kinetic effects are
equally important [7-14]. There are contradictory results
found by Ochoa et al. [15] that permeabilities were increased
by adding PVP 40 and 360 kDa into PES membrane dope. To
the best of our knowledge, the effect of four types of PVP
with different Mw (i.e. 10, 24, 40 and 360 kDa) on the
morphology and performance of polyvinylidene fluoride
(PVDF) membrane blended with titanium dioxide (TiO2)
nanoparticles has not been reported. With respect to this, an
attempt is made to investigate the effect of a wide range of
Mw of PVP as additive on the properties of PVDF-TiO2
composite membrane for oily wastewater treatment under
submerged conditions. PVDF is selected as based polymer
due to its high mechanical strength, good chemical stability
and excellent membrane forming ability. In the present work,
the changes in the morphological structure and membrane
properties of PVDF-TiO2 membranes made of different M w of
PVP are reported and the comparison between each
membrane performance were made. A simple backwashing
cleaning process was performed on the optimized membrane
to investigate the membrane flux recovery and oil rejection
after cleaning process.
2. Experimental
PVDF (Kynar760) pellets purchased from Arkema Inc.,
Philadelphia, USA were used as the main membrane forming
material. N, N-dimethylacetamide (DMAc) (Merck, >99%)
was used as solvent to dissolve polymer without further
purification. Polyvinylpyrrolidone (PVP) (10, 24, 40 and 360
kDa) purchased from Sigma Aldrich and titanium dioxide
(TiO2) (Degussa P25, average particle size ~21 nm) from
Evonik were used as additives to enhance PVDF membrane
properties. The cutting oil obtained from RIDGID, Ridge
Tool Company was used to synthesize oily solution with
different oil concentrations.
2.2.1 Scanning electron microscope (SEM)
The outer surface and cross sectional morphology of
membranes was observed by tabletop SEM (Model: TM 3000,
Hitachi). Prior to the analysis, the hollow fiber was immersed
into liquid nitrogen for few minutes followed by freezefracturing to obtain perfect cut structure. The fiber was then
placed onto carbon-tape aluminum holder and coated with
gold under vacuum.
2.2.2 Atomic force microscopy (AFM)
The membrane surface roughness and images were
investigated by AFM (Model: Seiko SPA-300HV). Small
piece of fiber was cut and placed on a square paper card with
size of 1 cm2 using double-sided adhesive tape. The

membrane surface was scanned in the size of 5 m 5 m.

The surface roughness of the membrane was expressed in
terms of mean surface roughness (Ra).
2.2.3 Contact angle goniometer
The contact angle of membranes was determined by the
sessile drop technique using a contact angle goniometer
(Model: OCA 15EC, Dataphysics) with deionized water as the
liquid. At least 10 locations were arbitrarily chosen on the
membrane surface in order to yield an average value.
2.2.4 Membrane porosity
The membrane porosity, , is defined as the volume of the
pores per the total volume of the porous membrane as shown
in Equation (1).
(Wwet Wdry )

(Wwet Wdry ) Wdry



where is the membrane porosity (%), wwet is the weight of
wet membrane (g), wdry is the weight of dry membrane (g), p
is the density of the polymer (g/cm3) and w is the density of
water (g/cm3).

Filtration experiment

Two U-shaped hollow fiber membrane modules at the bottom

of a submerged tank containing approximately 14 litres of oily
wastewater. To minimize the fouling effect, a constant air
flow rate of 5 l/min generated by an air compressor (Model:
2HP single cylinder 24L tank, Orimas) was used to generate
air bubbles within the submerged tank through air diffuser
installed underneath the membrane modules. Water permeate
was produced using peristaltic pump (Model: 77200-60,
Masterflex L/S, Cole Parmer) by creating vacuum condition
on permeate side. Both the vacuum pressure and pump flow
rate were kept constant at 15 inHg and 15 ml/min,
respectively throughout the experimental period. A 10 ml of
sample was taken from the collected permeate for sample
analysis and the remaining permeate was recycled back to the
tank. Three measurements were made for each sample and
then the average value was reported together with its standard
deviation. The membrane water flux (J) was determined
according to Equation (2).



where J is the water flux (L/m2h), Q is the quantity of

permeate (L), A is the effective membrane area (m2), and t is
time (h) to obtain the quantity of Q. The performance of
membranes was evaluated using different feed solutions by
changing the oil concentration (250 ppm, 500 ppm and 1000
ppm) in the feed. The membrane oil rejection was then
calculated using the following equation.




Author name / Procedia CIRP 00 (2014) 000000

where R is the oil rejection (%), Cp and CF are the

concentration of oil in the permeate (ppm) and the feed (ppm),
respectively. The oil concentration in permeate and feed was
determined using an UV-Vis spectrophotometer (Model:
DR5000, Hach) measured at wavelength of 294 nm at which
the maximum absorption occurs. In addition, the antifouling
property of the resulted membrane was also studied in this
work. Firstly, the best performing membrane with balance
performance of permeability and selectivity was selected to
treat the oily solution for 180 min. After 180 min of operation,
the feed solution was replaced with deionized water and the
membrane was cleaned by reversing the direction of permeate
flow for 60 min (also known as backflushing). Bubbling of air
controlled at flow rate of 5 l/min was also employed to assist
the cleaning process by removing the oil droplets attached on
the membrane surface.

3. Results and Discussion

3.1 Effect of molecular weight of PVP on membrane
structural properties
Figures 1 and 2 show the SEM cross-sectional images and 3D
AFM surface images of the hollow fiber membranes prepared
with different Mw of PVP. It was clearly observed that PVDFTiO2 membrane with the addition of PVP has larger
macrovoids in comparison to the neat PVDF-TiO2 membrane.
However, the size of macrovoids was decreased and the
sponge layer became larger with increasing Mw of PVP. The
finger-like macrovoids gradually developed at the external
and internal surface layer with increasing Mw of PVP. This
can be explained by the increases of viscosity of the
membrane dope with increasing Mw of PVP. As for the
properties of membrane surface, the mean surface roughness
increased from PVDF to PVDF-PVP10k, followed by the
decrease from PVDF-PVP10k to PVDF-PVP360k, as can be
seen from AFM surface images in Figure 2.
Table 1 shows several important properties of the membrane
prepared from different Mw of PVP. As can be seen, an abrupt
change from PVDF to PVDF-PVP10k, which is followed by a
gradual change in the reverse direction as the Mw of PVP
increases from 10k to 360k. For example, the porosity
increased from PVDF(81.95%) to PVDF-PVP10k(88.59%)
and then gradually decreased to PVDF-PVP360k(82.27%).
The same pattern was observed in the pore size as illustrated
in Table 1. This shows a parallel relationship between the
surface pore size and the bulk macrovoid size. As more nonsolvent is drawn in to the spun polymer solution, the larger
pores and the larger macrovoid are formed at the membrane
surface as well as in the membrane bulk, respectively. On the
other hand, the contact angle decreased from PVDF to PVDFPVP10k, which was likely due to the high hydrophilicity of
PVP used. Then, the contact angle increased as the PVP Mw
increased from 10k to 360k. This change is due to the effect
of decreasing roughness as the Wenzel equation predicts.






Figure 1: SEM cross-sectional images of PVDF-TiO2 membranes prepared

from different molecular weight of PVP (a) control (without additive), (b)
10k (c) 24k (d) 40k and (e) 360k Da.
Ra = 17.58 nm

Ra = 26.50 nm


Ra = 19.15 nm

Ra = 16.37 nm


Ra = 15.35 nm

Figure 2: 3D AFM images (outer surface) of PVDF-TiO2 membranes
prepared from different molecular weight of PVP, (a) control (without
additive), (b) 10k (c) 24k (d) 40k and (e) 360k Da.

Author name / Procedia CIRP 00 (2014) 000000

Table 1: Effect of different Mw of PVP on the PVDF-TiO2 membrane

properties with respect to porosity, pore size and contact angle

3.3 Flux recovery of optimum membrane

PVDF (control)


Pore size

Contact angle
( o)

3.2 Effect of different molecular weight of PVP on

membrane flux and oil rejection
Figure 3 shows the results from the experiments conducted
using pure water and synthesized oily wastewater. The pure
water flux (Jw1) was reported to increase from PVDF to
PVDF-PVP10k and gradually decreased from PVDF-PVP10k
to PVDF-PVP360k. This pattern is exactly the same as those
observed in the surface pore size and the bulk porosity. It
seems natural that Jw1 increases as both surface pore size and
bulk porosity increase. The pattern observed in Jw1 is also
exactly the same as the pattern observed in hydrophilicity, i.e.
the hydrophilicity increased, shown as the decrease in contact
angle (see Table 1) from PVDF (control) to PVDF-PVP10k
membrane and decreased gradually from PVDF-PVP10k to
PVDF-PVP360k. Hence the increase in surface hydrophilicity
is also contributing to enhance of Jw1. The permeate flux of
the oil/water mixture (Jw2) was slightly lower than Jw1. This
phenomenon was attributed to the rapid formation of oil layer
on membrane surface, which has led to the increase in water
transport resistance. On the other hand, a reverse pattern was
shown in oil rejection, probably due to the increase of
membrane hydrophobicity with the increasing of PVP Mw (as
shown in Table 1), which has facilitated stronger interaction
between membrane surface and oil layer . Overall, it can be
said that PVDF-PVP40k is the best performing among the
studied membranes due to its good combination of water
permeability and oil selectivity.

Figure 4: Permeate flux before and after cleaning and flux recovery of PVDFPVP40K membrane (Operating conditions: temperature = 25 C, air bubble
flow rate = 5 l/min and vacuum pressure = -15 inHg)

Figure 5: Rejection before and after cleaning of PVDF-PVP40K membrane

(Operating conditions: temperature = 25 C, air bubble flow rate = 5 l/min
and vacuum pressure = - 15 inHg)

In order to assess the extent of fouling to the membrane

performance with respect to water flux and rejection. PVDFPVP40K was subject to the simple backflushing cleaning for
60 min after being used for treating synthetic oily wastewater
of 250 ppm. Figure 4 compares the membrane performance as
a function of time before and after cleaning process. It was
found that above 60% of initial flux could be retrieved by
simple backflushing without involving any chemical agents
and optimized the cleaning condition. Although the
membrane water flux has not been completely recovered by
the cleaning process, a promising separation efficiency was
achieved with the rejection rate against oil particles (>95%)
after 150 min of membrane filtration process.

Figure 3: Pure water flux, permeate flux and oil rejection of PVDF-TiO2
membranes prepared at different Mw of PVP (Operating conditions:
temperature = 25 C, oil concentration = 250 ppm, air bubble flow rate = 5
l/min and vacuum pressure = -15 inHg, Mw = 0 means PVDF-TiO2
membrane without PVP)

The permeation performance of PVDF-TiO2 membranes with

different Mw of PVP was evaluated for the treatment of
oil/water mixture. The results obtained showed that the
membrane properties were influenced by the Mw of PVP as
additive during membrane preparation. Based on the results
obtained from this study, the addition of PVP to PVDF
membrane could improve the hydrophilicity, pore size and
porosity of PVDF membrane. Among the membranes with
different Mw of PVP, PVDF-PVP40k is concluded as the best

Author name / Procedia CIRP 00 (2014) 000000

performing membrane due to its high permeability and

selectivity achieved during filtration process. A stable oil
separation efficiency of PVDF-PVP40k coupled with good
water flux production renders the membrane a promising
candidate for sustainable development of oily wastewater
[1] M. Abbasi, A. Salahi, M. Mirfendereski, T. Mohammadi,
A. Pak, Dimensional analysis of permeation flux for
microfiltration of oily wastewaters using mullite ceramic
membranes, Desalination 2010; 252: 113-119.
[2] B. Chakrabarty, A.K. Ghoshal, M.K. Purkait, Cross-flow
ultrafiltration of stable oil-in-water emulsion using
polysulfone membranes, Chem. Eng. J., 2010; 165: 447-456.
[3] Malaysia Sewage and Industrial Effluent Discharged
Standards, http://www.water
(accessed 29.03.2014)
[4] M. Karhu, V. Kuokkanen, T. Kuokkanen, J. Rm, Bench
scale electrocoagulation studies of bio oil-in-water and
synthetic oil-in-water emulsions, Sep. Purifi. Technol., 2012;
96: 296-305.
[5] S.M. King, P.A. Leaf, A.C. Olson, P.Z. Ray, M.A. Tarr,
Photolytic and photocatalytic degradation of surface oil from
the Deepwater Horizon spill, Chemosphere, 2014; 95: 415422.
[6] F. Shahrezaei, Y. Mansouri, A.A.L. Zinatizadeh, A.
Akhbari, Process modeling and kinetic evaluation of
petroleum refinery wastewater treatment in a photocatalytic
reactor using TiO2 nanoparticles, Powder Technol., 2012;
221: 203-212.
[7] A. Bunturngpratoomrat, O. Pornsunthorntawee, S.
Nitivattananon, J. Chavadej, S. Chavadej, Cutting oil removal
by continuous froth flotation with packing media under low
interfacial tension conditions, Sep. Purif. Technol., 2013; 107:
[8] P. Painmanakul, P. Sastaravet, S. Lersjintanakarn, S.
Khaodhiar, Effect of bubble hydrodynamic and chemical
dosage on treatment of oily wastewater by Induced Air
Flotation (IAF) process, Chem. Eng. Res. Des., 2010; 88:
[9] B. Chakrabarty, A.K. Ghoshal, M.K. Purkait,
Ultrafiltration of stable oil-in-water emulsion by polysulfone
membrane, J. Membr. Sci., 2008; 325: 427-437.

[10] Y.-R. Qiu, H. Zhong, Q.-X. Zhang, Treatment of stable

oil/water emulsion by novel felt-metal supported PVA
composite hydrophilic membrane using cross flow
ultrafiltration, T. Nonferr. Metal. Soc., 2009; 19: 773-777.
[11] B. Jung, J.K. Yoon, B. Kim, H.-W. Rhee, Effect of
molecular weight of polymeric additives on formation,
permeation properties and hypochlorite treatment of
asymmetric polyacrylonitrile membranes, J. Membr. Sci.,
2004; 243: 45-57.
[12] B. Chakrabarty, A.K. Ghoshal, M.K. Purkait,
Preparation, characterization and performance studies of
polysulfone membranes using PVP as an additive, J. Membr.
Sci., 2008; 315: 36-47.
[13] H. Basri, A.F. Ismail, M. Aziz, Polyethersulfone (PES)
silver composite UF membrane: Effect of silver loading and
PVP molecular weight on membrane morphology and
antibacterial activity, Desalination, 2011; 273: 72-80.
[14] J.S. Kang, Y.M. Lee, Effects of molecular weight of
polyvinylpyrrolidone on precipitation kinetics during the
formation of asymmetric polyacrylonitrile membrane, J. Appl.
Polym. Sci., 2002; 85: 57-68.
[15] W.-Z. Lang, J.-P. Shen, Y.-T. Wei, Q.-Y. Wu, J. Wang,
Y.-J. Guo, Precipitation kinetics, morphologies, and
properties of poly(vinyl butyral) hollow fiber ultrafiltration
membranes with respect to polyvinylpyrrolidone molecular
weight, Chem. Eng. J., 2013; 225: 25-33.
[16] H. Matsuyama, T. Maki, M. Teramoto, K. Kobayashi,
Effect of PVP Additive on Porous Polysulfone Membrane
Formation by Immersion Precipitation Method, Sep. Sci.
Technol., 2003; 38: 3449-3458.
[17] M. Sadrzadeh, S. Bhattacharjee, Rational design of phase
inversion membranes by tailoring thermodynamics and
kinetics of casting solution using polymer additives, J.
Membr. Sci., 2013; 441: 31-44.
[18] Z. Zhang, Q. An, Y. Ji, J. Qian, C. Gao, Effect of zero
shear viscosity of the casting solution on the morphology and
permeability of polysulfone membrane prepared via the
phase-inversion process, Desalination, 2010; 260: 43-50.
[19] N.A. Ochoa, P. Prdanos, L. Palacio, C. Pagliero, J.
Marchese, A. Hernndez, Pore size distributions based on
AFM imaging and retention of multidisperse polymer solutes:
Characterisation of polyethersulfone UF membranes with
dopes containing different PVP, J. Membr. Sci., 2001; 187: