Customized antifouling polyacrylonitrile ultrafiltration membranes for

effective removal of organic contaminants from aqueous stream

J. S. Beril Melbiaha, J. Pramilaa, e, D. Ranab, A. Nagendranc, N. Nagendra Gandhia,

D. Mohand,*

a
Membrane Laboratory, Department of Chemical Engineering, Anna University, Chennai –
Accepted Article
600025, India
b
Department of Chemical and Biological Engineering, University of Ottawa, 161 Louis

Pasteur St., Ottawa, ON, K1N 6N5, Canada

c
Polymeric Materials Research Lab, PG & Research Department of Chemistry, Alagappa

Government Arts College, Karaikudi– 630 003, India

d
Centre for Research and Consultancy, Hindustan Institute of Technology & Science, Padur,

Chennai 600 103, Tamil Nadu, India

e
Department of Chemistry, Agni College of Technology, OMR, Thalambur , Chennai 600

130, India

*Corresponding author:

E- mail: mohantarun@gmail.com (Dr. D. Mohan); Tel.: 91-44-27474262/395

Abstract

BACKGROUND

Ultrafiltration (UF) is a promising separation technique in the removal of

macromolecular contaminants. However, the hydrophobic polymeric UF membrane

This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which
may lead to differences between this version and the Version of Record. Please cite this
article as doi: 10.1002/jctb.5833

This article is protected by copyright. All rights reserved.

 performance is suffered by fouling in the long run due to the clogging of contaminants at the

surface and pores. In this study, anti-fouling hydrophilic polyacrylonitrile (PAN) UF

membranes in the presence of an amphiphilic triblockcopolymer, Pluronic F127 (PF127)

were prepared via phase inversion technique.

Accepted Article
RESULTS

The effect of varying concentrations of PF127 on PAN UF membranes was analyzed

by ATR-FTIR, SEM and AFM. The filtration characteristics of the membranes were

measured in terms of pure water flux, membrane porosity and water content. The separation

efficiency of the membranes is explored for contaminants such as bovine serum albumin

(BSA), humic acid (HA) and oil. The results revealed that the PAN membrane with 4 wt% of

PF127 produced greatest permeates flux of 391 L/m2h with minimal fouling. A higher solute

rejection of more than 90% was observed for the tailored membranes due to the improvement

in surface properties.

CONCLUSION

The inherent hydrophilicity of the high density poly (ethylene oxide) brush-like layer

of PF127 at the membrane-water interface is utilized effectively to restrict the adsorption of

the organic contaminants onto the membrane surface. After a simple hydraulic washing of

PAN/PF127 UF membranes, the flux recovery ratio was augmented which is ascribed their

excellent antifouling property and potential use in water treatment.

Keywords: Antifouling; Ultrafiltration; Membrane; Polyacrylonitrile; Hydrophilicity;

Pluronic F127;

1. Introduction

To date, reclamation of wastewater from various industries has become crucial to

alleviate the inadequate supply of clean water. Fresh water resources are being contaminated

highly due to the exponential growth in population and industries. As a consequence, these

This article is protected by copyright. All rights reserved.

 pollutant enriched water resources have become a platform for the growth of microbes which

may deplete the oxygen level in the surface water. Currently, membrane based water

treatment technologies have emerged as an effective approach in filtering aqueous mixtures

owing to its excellent separation performance with least energy input, single phase operation,
Accepted Article
fine permeate quality and cost-effective feature.1

Ultrafiltration (UF) is an emerging low pressure membrane technique apparently

separates macromolecular proteins, organic contaminants, microorganisms and tiny particles

present in the feed streams with relatively ease of operation.2-4 In contrast, membrane fouling

that arises on continuous accumulation of solutes on the membrane surface (active layer)

limits its widespread application.5 Hence it is imperative to control fouling to maintain

excellent performance of membranes. Formulating fouling resistant membrane surface is

considered to be paramount to inhibit the non-specific adsorption of the aforementioned

organic contaminants present in the feed stream.6,7 It is well established that hydrophilic

surfaces exhibit exceptional resistance to organic fouling since the hydrophilic surfaces bind

large amount of water molecules around to configure a compact structured water layer. This

in turn defy the interactions of foulants present in the feed stream to the membrane top

surface.8,9 The hydration ability can further be intensified by altering surface properties of the

active layer by incorporating different macromolecular polymeric additives like

poly(ethylene glycol), poly(vinyl pyrrolidone) and poly(vinyl alcohol) to the casting dope

solution. Moreover, these additives are also responsible for the substantial enhancement of

hydrophilicity and porosity of asymmetric membrane during membrane formation.10

Mostly, these hydrophilic additives have been used as pore forming agents. Even

though they enhance hydrophilicity, they are unstable and tend to leach out after a prolonged

operation due to its high water solubility and poor compatibility with membrane matrix.

Alternatively, an amphiphilic triblockcopolymer additive PF127 bearing both hydrophilic

This article is protected by copyright. All rights reserved.

 moieties and a hydrophobic backbone is more compatible with membrane matrix, tend to

linger with the membrane for much longer time imparting robust fouling resistance.11 PF127

or Pluronic is a non-ionic macromolecule holding a triblock structure of poly(ethylene

oxide)–poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO) and performs a dual

Accepted Article
function as pore former as well as surface modifying agent.12 Blending PF127 with base

polymers like poly(ether sulfone) (PES), cellulose acetate (CA), polysulfone (PSU),

poly(vinyl chloride) (PVC), poly(vinylidene fluoride) (PVDF), and poly(ether imide) (PEI)

produced a unique self-assembly pattern which improved the hydrophilicity and separation

performance of membrane.13-18 During phase inversion process, in a major portion of Pluronic

F127 content, the hydrophilic PEO moiety spontaneously tend to move upwards and

segregate at membrane-water interface imparting hydrophilic character while the counterpart,

hydrophobic PPO moiety tend to fix into membrane matrix ensuring the stability of

hydrophilic PEO moieties at membrane surface. In addition, the protruded PEO chains

increase the number of hydrophilic sites at membrane surface which in turn enhances

hydrogen bonding sites. This enables the construction of hydration layer which further

suppresses the non-specific attachment of the foulants. Further, the remaining portion of the

added PF127 self-assembles to form spherical micelles with a shell configuration and releases

out into the gelation bath. Thus, the amphiphilic additive containing hydrophobic anchoring

groups and hydrophilic segments would be a suitable candidate for membrane modification.

Herein we have reported the fabrication of antifouling PAN-UF membrane blended with

PF127 for enhancing the separation of fouling agents such as bovine serum albumin (BSA),

humic acid (HA) and oil.

2. Experimental
2.1. Materials

This article is protected by copyright. All rights reserved.

 Commercial grade PAN (150 kg/mol) was procured from Technorbital, Uttar Pradesh,

India. PF127 (12,600 g/mol) was obtained from Sigma Aldrich. BSA (69 kDa), HA, N-

methyl-2-pyrrolidone (NMP) and sodium lauryl sulfate (SLS) were purchased from SRL

Chemicals Ltd., India. NMP was sieved through molecular sieves (4A°) in order to eliminate
Accepted Article
moisture content and kept stored in dry atmosphere before usage. Sodium dodecyl sulfate

(SDS) was procured from Glaxo India Ltd. Distilled water was employed during membrane

preparation and UF experiments.

2.2. Fabrication of membranes

Dope solutions were prepared using PAN (9 wt%) with various concentration of

PF127 from 0, 1, 2, 3 and 4wt % (designated as MA, M.01, M.02, M.03, and M.04 respectively)

and blended with NMP in the pre-optimized composition as given in Table 1. To get a

homogeneous casting solution, physical blending was done for 8 h with mechanical stirrer

and then allowed to stand for the completion of degassing. Upto 4 wt% of PF127, miscibility

was found with PAN and beyond that phase separation was occurred. The obtained dope

solution was casted over a smooth and polished fine glass plate by maintaining a thickness of

0.22 mm and a relative humidity (RH) of 25%. It was then dipped carefully in the gelation

bath containing 2 % of NMP (solvent), 0.2 wt % of SLS (surfactant) in 2 L of water (non-

solvent) at 200C and kept as such overnight for the solvent and non-solvent exchange to be

completed. The membranes were then removed from the glass plate and cleaned thoroughly

with distilled water to remove the excess NMP and SLS. The prepared membranes were

stored in distilled water containing 0.1% formalin solution.

Table 1. Blend composition of the PAN and PAN/PF127 membranes

Membrane Blend composition

This article is protected by copyright. All rights reserved.

 PAN (wt%) PF127 (wt%) NMP(wt%)
MA 9 0 91
M.01 9 1 90
M.02 9 2 89
M.03 9 3 88
Accepted Article
M.04 9 4 87

2.3. Characterization of membranes

2.3.1. ATR-FTIR, porosity and water content

ATR-FTIR spectra of the membranes were recorded using Perkin-Elmer spectrometer

in the range of 4000 - 400 cm-1.

The average surface pore radius rm (m) of each membrane was calculated by the

filtration velocity method using Guerout-Elford-Ferry equation which is as follows.

( )
rm = √ (1)

where, ‘ε’ is the membrane porosity, ‘η’ is the viscosity of water at 25 °C, ‘l’ is the

membrane thickness, ‘Q’ is the water permeate flux, ‘A’ is the active membrane area and ‘P’

is the applied pressure.

Porosity and water content (WC) were computed from the wet weight and dry weight

of the membranes. After soaking the equal sized membrane samples for 24 h, the excess

moisture was removed using blotting paper and weighed then to know the wet weight.

Subsequently, these wet membranes were dried in vacuum overnight and weighed to find out

the dry weight of membranes. The porosity (ε) of the membranes was determined to know the

water uptake capacity of membranes. Membrane porosity was measured by gravimetric

method using the following equation

This article is protected by copyright. All rights reserved.

 ( )

Where, ‘W0’ is the wet weight of membrane, ‘W1’ is the dry weight of membrane, is the
Accepted Article

density of water, ‘A’ is the active membrane area and ‘L’ is the membrane thickness. WC of

the membranes is related to its porosity and hydrophilicity, calculated by the following

equation

( )

2.3.2. SEM and AFM

The cross sectional morphology of the membranes was studied by Scanning Electron

Microscopy (SEM, Cam Scan). Membrane samples with a size of 1cm × 1cm were dried with

ethanol and fractured using liquid nitrogen for 60 s. These dried membranes were gold

sputtered to produce electron conductivity and the images were obtained at a high vacuum of

15 kV.

Surface roughness of the membranes was probed by Atomic Force Microscopy (AFM,

Park System, Korea XE-100). Membrane sample attached onto a glass substrate was imaged

in a scan size of 2 µm × 2 µm by tapping mode. The surface roughness of each membrane

was measured at three different locations and an average value was reported for accuracy.

2.3.3. Contact angle and work of adhesion

The wetting characteristics of membranes was tested by water contact angle

measurements by sessile drop technique using goniometer (OCA 15EC, Data Physics,

Germany). The membrane samples with a size of 3 cm × 3 cm were mopped with a blotting

This article is protected by copyright. All rights reserved.

 paper prior to analysis. Measurements were made by placing 5µl of Milli-Q on membrane

surface at four different places on the sample and the average value was considered for

accuracy. From the contact angle values, work of adhesion ( ) was also determined by the

following equation:
Accepted Article

( ) ( )

Where, ‘θ’ is the contact angle and is the interfacial surface tension of liquid and vapor

phase (72 mN/m at 25°C).

2.3.4. Pure water flux

The efficacy of the membranes was determined by performing separation experiments

using a batch type dead-end filtration cell (UF cell, Model 8400, Amicon, USA). The UF cell

with a volumetric size of 350 mL and active filtration area of 45.38 cm2 was attached to

solution reservoir. A Teflon-coated magnetic paddle was fitted inside the cell to minimize

concentration polarization.19

Before starting UF experiments, the membranes were compacted for an hour at a

specified pressure of 413.685 kPa using distilled water taken in the solution reservoir. After

that, permeated water was measured at fixed intervals until a steady state was obtained to

determine trans-membrane flux. The pure water flux ( ) was calculated based on the

following expression.

( )

Where, ‘Q’ is the volume of water permeate (L); ∆t is the water permeation time (h); ‘A’ is

the active area of membrane (cm2).

2.3.5. BSA, HA and Oil removal

This article is protected by copyright. All rights reserved.

Accepted Article

Fig. 1. Size distribution of oil

droplets in the synthesized

emulsion

Three different kinds of

model foulant solutions such as protein (500 mg/L, BSA in PBS), humic acid (500 mg/L in

DI water) and motor oil (500 mg/L in SDS emulsifier) were prepared individually to evaluate

the rejection performance of the membranes. Oil droplet size of the emulsion prepared was 5

µm as shown in Fig. 1 which was examined by Malvern Zetasizer, Nano ZS instrument.

During filtration of each model foulant solution individually, permeate was collected at a

fixed and regular interval of time to measure trans-membrane flux which is mentioned as

protein flux (Jp), humic acid flux (Jh) and oil-in-water emulsion flux (Jo). All the fouling

experiments were carried out using UF stirred cell at a constant pressure of 344.738 KPa.

The concentrations were quantified with total organic carbon analyzer (Shimadzu,

TOC-V CPH) for HA as well as oil/water emulsion and UV absorbance at 280 nm for BSA.

After each foulant solution was tested for fouling intensity, UF cell and reservoir was cleaned

This article is protected by copyright. All rights reserved.

 thoroughly with deionized (DI) water. Subsequently, both were filled again with DI water in

order to find out the reversibility of fouling and the fluxes obtained are denoted as which

is pure water flux after protein filtration, which is pure water flux after humic acid

filtration and which is pure water flux after emulsion filtration

Accepted Article

. The solute rejection (SR) was found out from the concentration values of feed

solution ( ) and permeate flux ( ) using the following equation.

( ) ( ) ( )

2.3.6. Fouling resistance

The efficiency of each membrane in terms of antifouling ability was examined from

its Flux Recovery Ratio (FRR) and Total Flux Decline (DRT) by using the following

equations.20, 21

( ) ( )

( ) ( )

( ) ( )

( ) ( )

FRR and DRT values on filtering all the three model foulant solutions are found. The

Total Flux Decline is the sum of Reversible and Irreversible Flux Decline. Moreover, higher

FRR values and lower DR value indicates good fouling resistance property of membrane.

This article is protected by copyright. All rights reserved.

 3. Results and discussions

3.1. ATR-FTIR
Accepted Article

Fig. 2. ATR- FTIR spectra

of pristine and modified

PAN membranes

Fig. 2 depicts the

ATR-FTIR spectra of PAN and PAN/PF127 membranes. The typical peak associated with

nitrile (-C N stretching) is observed at 2245 cm−1 in both PAN and PAN/PF127

membranes.22,23 After the addition of PF127, significant changes in the spectra of PAN films

can be noted. In PAN/PF127 spectra, a new peak observed at 1080 cm-1 indicates C-O-C

stretching attributed to ether group which infers the presence of the hydrophilic PEO chains

on the surface. On the other hand, the incremental dosage of PF127 content from 0 to 4 wt%,

shows a notable increase in transmittance at 1080 cm−1 which assures that the membrane

matrix contains PF127.9,15 This must be owing to the surface segregation of PF127 molecules

This article is protected by copyright. All rights reserved.

 during phase inversion process. The peaks observed at 2280 cm-1 is corresponding to the CH

vibration of methylene. In PAN/PF127, a fair increase in intensity of this peak can be seen

due to the presence of methylene group from the added PF127.14 These obtained results

evidenced the presence of PF127 in the membrane matrix as PPO block provides a good
Accepted Article
anchorage within the membrane matrix despite the highly water soluble nature of PF127.

3.2. SEM and AFM

Fig. 3. Cross-section SEM images of pure PAN (MA), PAN/ 1 wt% PF127 (M.01), PAN/ 2

wt% PF127 (M.02), PAN/ 3 wt% PF127 (M.03) and PAN/ 4 wt% PF127 (M.04)

membranes

The morphological analysis leads to predict the membrane properties such as

selectivity and permeability. Fig. 3 shows the cross-sectional SEM images of PAN/PF127

membranes with different compositions. The images demonstrate an asymmetric

morphological structure with a separating skin layer at the top, a spongy layer beneath the

skin and finger like spongy layer projecting towards the bottom with fully opened pores at the

This article is protected by copyright. All rights reserved.

 end.24 The pure PAN membrane exhibited a macrovoidal structure as seen in cross-section

compared to the spongy structure of PAN/PF127 blend membranes. The incorporation of

PF127 in the dope solution produced porous membranes with a sponge like sub-layer. During

phase inversion, the solvent/non-solvent exchange between casted dope and gelation bath
Accepted Article
lead the polymer thermodynamically unstable thus influences demixing by diminishing free

energy of mixing. Moreover, on immersion, the polymeric molecules tend to form cluster in

order to reduce the interaction with the non-solvent leading to pore formation.25, 26

As mentioned earlier, a portion of the added PF127 might have leached out into

gelation bath facilitating instantaneous demixing by fastening the passage of water through

those voids to form macrovoids.27.On further increase in Pluronic addition, the finger like

formation increased where as macrovoids are suppressed substantially forming a denser

membrane structure. This must be due to the drastic increase in viscosity that decelerates the

diffusion rate of exchange of solvent and non-solvent thus slowed down demixing to form

porous substructure. Hence, in Fig. 3, PAN/PF127 membranes shows more finger like slender

channels than control PAN membrane. In AFM micrographs, all the membranes possessed a

nodule-valley like structure which is the typical structure of asymmetric membranes prepared

by phase inversion mechanism. As shown in Fig. 4, pure PAN membrane surface contains

large nodules and as the concentration of PF127 is increased, the nodules decreased in its size

for modified membranes. In general, surface roughness increases thereby the total active

surface area enhances the filtration; however intensify the deposition and attachment of

foulants28. The pure PAN membrane has an average roughness ( ) value of 71.81nm.

Alternatively, the increase in PF127 content has led to decrease the nodules and a reduced

alue of 47.62 nm at 4 wt% PF127. From values, it is evident that on increasing PF127

content in PAN membrane, smoothness on the membrane surface is enhanced as shown in

This article is protected by copyright. All rights reserved.

 Table 2. These results confirm that the increase in smoother surface shows better antifouling

property which is reliable to the obtained results of foulant studies. 29

Accepted Article

Fig. 4.

2D and 3D AFM micrographs of PAN (MA), PAN/1 wt% PF127 (M.01),

PAN/2 wt% PF127 (M.02) and PAN/4wt % PF127 (M.04) membranes

3.3. Contact angle and work of adhesion

The wetting nature of the membrane plays an important role in evaluating its water

permeation ability. Membrane is said to be hydrophilic when a water drop positioned on

surface of membrane immediately spreads from membrane air interface. The dynamic change

in water contact angle with respect to time is due to some measures like capillary action,

adsorption, roughness factors etc. Herein, the pure PAN membrane showed the higher contact

angle of 76.2°. At the same time, the higher wt% of PF127 embedded membranes had the

This article is protected by copyright. All rights reserved.

 least contact angle of 36.8° which must be due to the high surface coverage of hydrophilic

PEO brushes. High surface energy associated with PEO is the reason for the lag or drop in

contact angle from 76.2°to 36.8° as shown in Fig. 5.

Accepted Article

Fig. 5. Water contact angle and work of adhesion of the pure and modified PAN membranes

The work of adhesion, calculates the extent of interactive forces in between

membrane surface and the liquid. has incremented from 89.17 mN/m for MA to 139.65

mN/m for M.04 membrane as shown in Table 2. As mentioned earlier, at a higher dosage of

PF127, a high density PEO brush layer may be established at the membrane water interface.

This highly dense hydrophilic PEO moiety that stretches out into the aqueous environment

could hinder solute adsorption onto membrane surface by forming a highly hydrated compact

layer onto it.30,31 The increment in work of adhesion substantially reveals the enhancement of

adhesive force which is the indirect indication of the increased wetting property of the

respective M.04 membrane.

This article is protected by copyright. All rights reserved.

 Table. 2. Roughness parameter, contact angle, work of adhesion and pure water flux of pure

and modified PAN membranes.

Contact
Ra Work of adhesion, Pure water fiux
Membrane angle
(nm) (mN/m) (L/m²h)
(θ)
Accepted Article

MA 71.81 76.2 89.17 242

M.01 65.19 65.1 102.3 304

M.02 58.06 53.6 114.73 340

M.03 51.03 41.8 125.67 368

M.04 47.62 36.8 129.67 391

Table. 3. Membrane porosity, water content and mean pore radius of pure and modified

PAN membranes.

Porosity Water content Mean pore size

Membrane MWCO
(%) (%) (A°)
MA 72.1 64.7 33.6 < 69
M.01 79.0 70.1 35.0 < 69
M.02 82.1 71.3 35.3 < 69
M.03 84.0 73.9 36.1 < 69
M.04 86.3 75 36.5 < 69

3.4. Porosity and water content

The membrane porosity results are presented in Table 3. It is inferred that membrane

porosity has increased from 72.1% (MA) up to 86.3% (M.04). The effect of PF127 on the hike

in porosity of PAN/PF127 blend membranes might be explained by thermodynamics and

kinetics of membrane formation. The increase in the porosity with respect to the increase in

This article is protected by copyright. All rights reserved.

 PF127 addition must be owing to the availability of more PF127 content in the PAN (M.04)

for the formation of small networks. The higher porosity of M.04 in turn increased the wetting

property of the membranes.32

The WC values of the pure base polymer and Pluronic blend membranes is another
Accepted Article
indication of membrane wetting property which is shown in Table 3. The WC value has

increased from 64.7% (MA) to 75% (M.04). The augment in water content on increase in

addition of PF127 content in blend membranes is because, more porous structure can uphold

more amount of water within it. These results are consistent with the pure water flux and

water contact angle measurements.33,34

3.5. Pure water permeation

Fig. 6. Pure water flux profile of PAN control and modified PAN membranes

Pure water flux is considered to be the key factor for any UF membrane. The pure

water flux profile obtained for the various prepared membranes is displayed in Fig. 6. Here,

the control PAN membrane exhibited lower pure water flux of around 242 L/m²h whereas an

opposite trend of higher water flux of 391 L/m²h could be noticed on incremental dosage of

This article is protected by copyright. All rights reserved.

 the hydrophilic additive PF127. This linear flux escalation observed for M0.4 membrane is in

accordance with its higher hydrophilic nature. Considerable increase in hydrophilicity of the

modified membranes led to enhance the transport of water molecule through membrane

pores. Similar kind of results was also reported by Kanagaraj et al.18.

Accepted Article

Fig. 7. Schematic illustration of fouling resistant mechanism for modified PAN/PF127

membrane

This article is protected by copyright. All rights reserved.

 3.6. Antifouling studies

In general, hydrophilic surface favors in assembling a dense hydration layer barrier on

to membrane surface that possesses steric repulsion effect (fouling resistant mechanism).

Therefore the interaction between membrane surface and organic pollutants such as BSA, HA
Accepted Article
and oil is diminished. Based on the fouling nature of organic foulants on

hydrophobic/hydrophilic membrane surfaces, a fouling mechanism has been proposed and

shown in Fig. 7.

During the initial stage of UF of model foulant solutions, the adsorption of the

rejected solute molecules generates intense fouling on membrane surface. This eventually

gives rise to sharp decline in flux. On continuous filtration, equilibrium is accomplished on

further deposition of solute owing to the rigorous stirring near the membrane surface and

finally stable flux is reached.

3.6.1. BSA removal

Efficient purification and separation of macromolecular therapeutic proteins is

influential in diverse industries such as food, biomedical and pharmaceutical. The effect of

varying PF127 content on protein separation was investigated by filtering BSA as a model

protein solution through all the membranes. The BSA permeates flux values are presented in

Fig. 8. The pure PAN membrane possessed the permeate flux value of 95 L/m2h whereas the

membrane M.04 containing PF127 showed the highest permeate flux of 156 L/m²h with higher

molecule rejection of 95%.

In less hydrophilic membranes, protein adsorptions at membrane water interface tend

to interrupt the well-structured water molecule arrangement which leads to the disintegration

of bound water molecule. This further causes the membrane surface to interact more with

protein foulants. Moreover, protein adsorption hinders the mobility of polymeric chain by

compressing hydrophilic polymeric brushes leading to entropy loss.35 At the same time, in

This article is protected by copyright. All rights reserved.

 highly hydrophilic membrane (M.04), more hydrogen polar groups existing at membrane

surface easily bind with surrounding water molecules through electrostatic interactions or

hydrogen bonding. Therefore, the adhesion force which is responsible for protein interaction

is suppressed by the well-structured hydration layer which is attributed to the higher density
Accepted Article
of PEO coverage. Moreover, PEO groups render neutral charge to the membrane surfaces

owing to the lack of dissociable functional groups in PEO moiety. Therefore, the interaction

between the modified membrane with high density PEO (M.04) and BSA is free from

electrostatic interactions. In addition, neutrally charged polymeric membrane surface which

forms hydrogen bonding with water can engender an energetic barrier and regulates the

adsorption of biomolecules.36 After hydraulic washing, M.04 membrane showed a higher FRR

value of 93.42% whereas it possessed lesser DRirr value of 6.58 %. From this study, it is

concluded that PF127 modified membranes showed higher FRR value and lower DRirr value

due to the presence of hydrated PEO segments on the membrane surface which effectively

inhibited protein adsorption and showed good antifouling nature.37 These results showed that

M.04 membrane has an excellent antifouling property which is reliable for protein

separation13.

3.6.2. HA removal

A predominant constituent available in naturally originated surface water is natural

organic matter (NOM) which is produced when natural debris and animal tissues are getting

decayed. UF of HA was accomplished in order to speculate the performance of membranes

on NOM removal. In general, HA molecules are negatively charged at pH 6. At this pH, the

carboxylic and phenolic hydroxyl groups are ionized to a greater extent leading to higher

repulsion between HA molecules which reduces the size of HA aggregates further. At the

same time, the bivalent calcium ions present in the hard water can easily complex with HA

This article is protected by copyright. All rights reserved.

 resulting in larger aggregated forms. Therefore CaCl2 was added to HA solution to retain the

calcium composition which is present in the surface water.

Accepted Article

Fig. 8. Time-dependent permeate flux variation of the pure PAN, PAN/PF127 membranes
(left)and a summary of the corresponding DRr, DRirr, FRR, DRT values (right) during
the filtration of (a) BSA solution, (b) HA solution and (c) oil emulsion. The entire
filtration process consisted of threesteps: purewater flux (0-3.5 h), flux on model
foulant solution filtration (3.5-5.5 h) and pure water flux after simple hydraulic
washing (5.5-8 h)

This article is protected by copyright. All rights reserved.

 The effect of varying PF127 concentration on HA separation was analyzed and

depicted in Fig. 8. From HA filtration experiment, it is observed that M.04 membrane showed

the highest permeate flux value of 165 L/m²h with a rejection of 90 % whereas pure PAN

membrane showed flux value of 81 L/m²h with a rejection of 84 %. The cause for the higher
Accepted Article
flux and higher rejection of M.04 membrane is the combined effect of appropriate pore size

and wetting property of modified membranes. On the other hand, the unmodified bare PAN

membrane showed a lesser permeate flux value due to the associated hydrophobic nature that

exacerbated organic fouling by favoring the interaction of foulants to membrane pores as well

as interior walls.

From Fig. 8, it can be concluded that both filtration of HA and BSA permeate flux has

similar trend. At the same time, the effect of membrane hydrophilicity was not much

effective on HA rejection when compared to BSA rejection. This must be due to the

difference in solution chemistry between HA and protein solution.38,39 HA molecule consists

of both hydrophilic and hydrophobic bearing groups which are a mixture of many molecules

including catechol, quinone, phenol, and sugar moieties. Therefore, there is a possibility of

emergence of a very thin layer of HA on the membrane surface by means of adsorption of

hydrophilic species of HA that may further alleviate the hydrophilicity at the modified

membrane surface. Simultaneously, the membrane surface top enveloped with PEO molecule

comprised of only hydrophilic CH2CH2O group possesses higher hydrophilicity led to

decrease in DRirr value. From Fig. 8, it can also be noted that when the PF127 content is 4

wt%, the FRR value is 83.8%. On the other hand, DRirr value which is highly responsible for

fouling is as low as 16.2%. The reason is, in PF127 modified membranes, presence of higher

density of PEO which provides surface resistance property due to the anchorage of PF127

showed a better antifouling property against HA foulant.

This article is protected by copyright. All rights reserved.

 3.6.3. Oil removal

Oil is considered as a common spreadable foulant which is unstable when compared

to protein foulant. The effect of addition of PF127 on rejection of oil for the PAN/PF127

blend membranes is shown in Fig. 8. The study on the oil separation efficiency of pure PAN
Accepted Article
membrane (MA) showed a flux value of 91 L/m²h with an oil removal efficiency of 89%.

Whereas, PAN/PF127 modified membranes exhibited an increased flux value of 183 L/m²h

with an oil rejection of 96.1% on oil filtration which indicates the better efficacy of

separation. When a hydrophobic membrane surface is immersed in water, the hydrophobic oil

molecules present in the emulsified oil in the feed stream adheres onto membrane surface and

deform by coalescing with each other and tend to spread over the solid surface in the

meantime of filtration.40 In this context, in pure PAN membrane (MA), the hydrophobic

interaction drives the oil molecules to deposit directly over the hydrophobic surface that leads

to a sharp flux decline over time. In contrast, in the case of PF127 added membranes the

presence hydrophilic PEO moieties make hydrophilic surface in aqueous media which helps

to reduce the hydrophobic interactions and adhesion of the oil droplets on the membrane

surface.41,42

The PAN/PF127 modified membrane with higher concentration of PF127 (M.04)

exhibited increased level of flux of 183 L/m²h with a oil rejection of 96.1% on oil filtration

which indicates the better efficacy of separation. It can be noted that DRirr value reduced

progressively with increasing PF127 content. Hence it is concluded that oil resistant property

is probably enhanced. This was attained by (i) The high surface energy PEO domains that

hampered the relocation of oil droplets at the membrane surface43 and (ii) The appropriate

pore size that arrested the squeezing in of oil molecules.44

This article is protected by copyright. All rights reserved.

Accepted Article

Fig. 9 Percentage rejection of BSA, Oil and HA for all the prepared membranes

3.6.4. Percentage rejection of BSA, HA and Oil

The percentage rejection of foulants for all the membranes is depicted in Fig. 9.

According to the above results, the irreversible fouling (DRirr) is highly depended on the

hydrophilic characteristics at membrane surface. Among all the prepared membranes,

Pluronic added membrane with a higher dosage (M.04) possessing denser PEO brushes at

membrane surface rendered high hydrophilicity end, large excluded volume as well as unique

coordination with water molecules around, endow these membranes with extraordinary

ability to resist BSA, HA and oil molecule adsorption.45 On filtering these three model

foulants, the FRR value was in the order of MA<M.01 <M.02<M.03<M.04. On increasing PF127

concentration, the decrease in mass transfer resistance is affirmed to the high surface energy

created by PEO segments at membrane surface.

This article is protected by copyright. All rights reserved.

 4. Conclusions

Hydrophobic PAN membrane surface is modified with highly hydrophilic property to

retard the non-specific adsorption of foulants such as protein, NOM and hydrocarbon by

embedding an amphiphilic macromolecule PF127 via phase inversion technique. The surface
Accepted Article
modification done led a significant enhancement of pure water flux which in turn supported

to enhance its fouling resistant property. From performance studies, best results were

obtained for M.04 membrane on adding the maximum Pluronic content (4wt%). A highly

hydrated compact layer which formed owing to the dense PEO polymer layer at membrane

surface prevented the surface from getting fouled by protein molecules, HA and oil molecules

at the same time increasing its selectivity. With the combined stability of PAN membrane and

relatively hydrophilic chemistry of PF127, it can be concluded that the incorporation of

PF127 with PEO side chains and PPO backbone greatly improves the hydrophilicity to

minimize the interactions of foulants present in feed onto membrane surface. The excellent

antifouling property endows PAN/PF127 membranes with potential applications of UF

processes.

References
[1] Xie, M., Shon, H. K., Gray, S. R., Elimelech, M. WaterRes. 2016, 89, 210-221.
[2] Ding, L., Zhang, W., Dris, A. O., Jaffrin, M. Y., Tang, B. Ind. Eng. Chem. Res. 2016, 42
11130–11138.
[3] Shamsuddin, N., Das, D. B., Starov, V. M. Chem. Eng. J, 2015, 276, 331–339.
[4] Hagen, K. Desalination 1998, 119, 85-91.
[5] Guo, W., Ngo, H. H., Li, J. Bioresour. Technol. 2012, 122, 27-34.
[6] Rana, D., Matsuura, T. Chem. Rev. 2010,110, 2448–2471.
[7] Zhang, R., Liu, Y., He, M., Su, Y., Zhao, X., Elimelech, M., Jiang, Z. Chem. Soc. Rev.
2016, 45, 5888-5924.
[8] Miller, D. J., Kasemset, S., Wang, L., Paul, D. R., Freeman, B. D. J. Membr. Sci. 2014,
452, 171–183.

This article is protected by copyright. All rights reserved.

 [9] Chen, G. E., Sun, Li., Xu, Z. L., Yang, H., Huang, H. H., Liu, Y. J. Korean J. Chem. Eng.
2015, 32 (12), 2492-2500.
[10] Nazri, N .A . M., Lau, W. J., Ismail, A. F. Korean J. Chem. Eng. 2015, 32(9), 1853-1863
[11] Susanto, H., Ulbricht, M. J. Membr. Sci. 2009, 327, 125–135.
[12] Zhao, W., Su, Y., Li, C., Shi, Q., Ning, X., Jiang, Z. J. Membr. Sci. 2008, 318, 405–412.
Accepted Article
[13] Wang, Y. Q., Wang, T., Su, Y., L., Peng, F. B., Wu, H., Jiang, Z. Y. Langmuir. 2005,
21, 11856-11862.
[14] Lv, C., Su, Y., Wang, Y., Ma, X., Sun, Q., Jiang, Z. J. Membr. Sci. 2007, 294, 68–74.
[15] Raslan, R., Mohammad, A.W. J. App. Sci. 2010, 10, 2628-2632.
[16] Liu, B., Chen, C., Zhang, W., Crittenden, J., Chen, Y. Desalination. 2012, 307, 26–33.
[17]Loh, C -H., Wang, R., Chin. J. Chem. Eng. 2012, 20, 71-79.
[18] Kanagaraj, P., Neelakandan, S., Nagendran, A., Rana, D., Matsuura, T., Shalini, M. Ind.
Eng. Chem. Res. 2015, 54, 11628−11634.
[19] Jayalakshmi, A., Rajesh, S., Mohan,D. App. Surf. Sci. 2012, 258, 9770– 9781.
[20] Sri Abirami Saraswathi, M., Mahalakshmi, S., Vetrivel, S., Divya, K., Rana, D.,
Nagendran, A., J. Environ. Chem. Eng. 2018, 6, 1912–1917.
[21] Zhao, W., Su, Y., Li, C., Shi, Q., Ning, X., Jiang, Z., J. Membr. Sci. 2008, 318,
405−412.
[22] Zhi, S.H., Deng, R., Xu, J., Wan, L .S., Xu, Z .K. React.Funct.Polym. 2015, 86, 184–
190.
[23] Majeed, S., Fierro, D., Buhr, K., Wind, J., Du, B., Fi, A. B. D., Abetz, V. J. Membr. Sci.
2012, 403– 404, 101– 109.
[24] Kanagaraj, P., Neelakandan, S., Nagendran, A. Korean J. Chem. Eng. 2014, 31(6), 1057-
1064.
[25]Guillen, G. R., Pan, Y., Li, M., Eric, M. V .H. Ind. Eng. Chem. Res. 2011, 50, 3798–3817
[26] Kim, J. H., Min, B.R., Won, J., .Park, H.C, Kang, Y.S., J. Membr. Sci. 2001, 187, 47–55.
[27]Sun, Z., Chen, F. Int. J.Bio. Macr. 2016, 91, 143–150.
[28] Vatanpour, V., Madaeni, S. S., Rajabi, L., Zinadini, S., Derakhshan, A. A. J. Membr.
Sci. 2012, 401– 402,132– 143.
[29] Mokhtari, S., Rahimpour, A., Shamsabadi, A. A., Habibzadeh, S., Soroush, M. App.
Surf. Sci. 2017, 393, 93–102.
[30] Chen, X., Luo, J., Qi, B., Cao, W., Wan, Y. J. Water Process Eng. 2015, 7, 1–10.
[31]Zhao, X., Su, Y., Li, Y., Zhang, R., Zhao, J., Jiang, Z. J. Membr. Sci. 2014, 450, 111–
123.

This article is protected by copyright. All rights reserved.

 [32] Sumisha, A., Arthanareeswaran, G., Thuyavan, Y. L., Ismail, A.F., Chakraborty, S.
Ecotoxi. Env.Saf. 2015, 121, 174-179.
[33] Sajitha, C.J., Mahendran, R., Mohan, D. Europ. Pol. J. 2002, 38, 2507–2511.
[34] Yin, J., Zhou, J. Desalination.2015, 365, 46–56.
[35] Yu, L., Zhang, Y., Zhang, B., Liu, J., Zhang, H., Song, C. J. Membr. Sci. 2013, 447,
Accepted Article
452–462.
[36] Kang, S., Asatekin, A., Mayes, A. M., Elimelech, M. J. Membr. Sci.2017, 296, 42–50.
[37] Wang, X. F., Fang, D. F., Yoon, K., Hsiao, B.S., Chu, B. J. Membr.Sci. 2006, 278, 261-
268.
[38] Zularisam, A.W., Ismail, A.F., Salim, R. Desalination 2006, 194, 211–231.
[39] Yuan, W., Zydney, A. L. Envorn. Sci. Technol. 2000, 23,5043–5050.
[40] Zhang, G., Jiang, J., Zhang, Q., Gao, F., Zhan, X., Chen, F. Langmuir. 2016, 32,
1380−1388.
[41] Melbiah, J. S. B., Nithya, D., Mohan, D. Coll. and Surf. A: Physi. and Eng. Asp.2017,
516,147–160.
[42] Arumugham, T., Kaleekkal, N. J., Rana, D., Doraiswamy, M., J. Appl. Polym. Sci. 2015,
42848.
[43]Chen, W., Peng, J., Su, Y., Zheng, L., Wang, L., Jiang, Sep. Purif. Technol. Z. 2009, 66,
591–597.
[44]Solomon, B. R., Hyder, M. N., Varanasi, K. K. Sci. Rep.2014, 4, 5504.
[45] Hou, S., Zheng, J., Zhang, S., Li, S. Polymer 2015, 77, 48-54.

This article is protected by copyright. All rights reserved.