Accepted Manuscript

Decolour ization of Aer obically Tr eated Palm Oil M ill Effluent (AT-POM E)
using Polyvinylidene Fluor ide (PVDF) Ultr afiltr ation M embr ane I ncor po
r ated with Coupled Zinc-I r on Oxide Nanopar ticles

Y.H. Tan, P.S. Goh, A.F. Ismail, B. C, Ng, G.S. Lai

PII: S1385-8947(16)31336-5
DOI: [Link]
Reference: CEJ 15802

To appear in: Chemical Engineering Journal

Received Date: 10 July 2016

Please cite this article as: Y.H. Tan, P.S. Goh, A.F. Ismail, B. C, Ng, G.S. Lai, Decolour ization of Aer obically
Tr eated Palm Oil M ill Effluent (AT-POM E) using Polyvinylidene Fluor ide (PVDF) Ultr afiltr ation
M embr ane I ncor por ated with Coupled Zinc-I r on Oxide Nanopar ticles, Chemical Engineering Journal (2016),
doi: [Link]

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Decolourization of Aerobically Treated Palm Oil Mill Effluent (AT-POME) using

Polyvinylidene Fluoride (PVDF) Ultrafiltration Membrane Incorporated with Coupled

Zinc-Iron Oxide Nanoparticles

Y. H. Tan, P. S. Goh, A. F. Ismail*, B. C, Ng and G. S. Lai

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

UTM Skudai, Johor, Malaysia.

Faculty of Chemical and Energy Engineering (FCEE), Universiti Teknologi Malaysia, 81310 UTM

Skudai, Johor, Malaysia.

*Corresponding author. Advanced Membrane Technology Research Centre (AMTEC), Universiti

Teknologi Malaysia, 81310 Skudai, Johor, Malaysia. Email address: [Link]@[Link]

(A. F. Ismail)

Abstract

Palm oil mill effluent (POME), which is rich in organic matter, is one of the major

contributors of water pollution. To date, biological treatment has been employed before it is

released into water bodies to minimize the environmental hazards. However, the dark

brownish colour of the effluent is still a big challenge to be addressed. In regard to this

concern, membrane separation can serve as an attractive solution for this issue. This paper

aims to investigate the performance of ultrafiltration nanocomposite membrane embedded

with coupled zinc-iron oxide (ZIO) for the decolourization of POME. The ZIO was

synthesized through the solution combustion technique by employing zinc nitrate

hexahydrate and iron (III) nitrate nonahydrate as the precursors and urea as the fuel. The

1
nanocomposite flat sheet membrane was prepared via phase inversion process. The physico-

chemical properties of the MMMs were analysed using zeta potential analysis, SEM, BET

and contact angle. Continuous filtration test was carried out to evaluate the capability of the

MMMs for colour and COD removal. The results showed that the increase in the ZIO loading

has drastically increased the membrane surface negativity and led to the colour removal of

70%. In addition, with the addition of M0.5, the permeation and colour removal had

increased 25% and 17% respectively, compared to the pristine PVDF membrane. However,

the long term filtration test revealed that the structure of M0.5 collapsed after 4 cycles of

washing, but M2.0 still retained its performance. In a nutshell, this study demonstrated that

negative surface charge has improved the antifouling properties meanwhile hydrophilicity has

contributed to increased water flux during the colour removal process.

Keyword: Coupled metal oxide, negatively charged surface, decolourization, palm oil mill

effluent, ultrafiltration.

1. Introduction

Palm oil industry is one of the fast growing industries in the world where the plantation area

of oil palm tree has increased drastically over the last few years. In edible oil production, oil

palm has contributed to the highest global demand owing to its high oil yield and the

nutritional value of the palm oil. However, extraction of palm oil has resulted in the brownish

effluent which known as palm oil mill effluent (POME). POME is highly rich in organic

matter hence shows high tendency to pollute the water body through eutrophication.

Therefore, government has restricted the rule and regulation for the discharge of POME.

Conventionally, POME is treated via biological approach [1, 2]. However, one of the

2
significant drawbacks of biological treatment is the formation of dark effluent that arisen

from the biodegradation of lignocellulosic into tannin, lignin and other colour pigment. The

brownish colour effluent has raised the public concern on its impact on water body. In

addition, the colour pigment has also caused the colourization of the pipeline in industry and

reduced the possibility of the reclamation. While the world is searching for more solutions to

tackle the looming global water crisis, water reclamation from effluent for the processing

plant can definitely serve as a promising alternative water resource to address the water

shortage issues.

Decolourization of as-treated POME (AT-POME) can be accomplished via physical or

chemical method. To date, heterogeneous chemicals such as adsorbate and catalyst have been

employed to remove the colour pigment via absorption and catalysis respectively [3, 4].

However, this method requires additional operation unit to recycle the chemicals and

unfavourably increases the footprint of the processing plant. Besides, energy intensified

processes such as distillation has been used to obtain clean water through boiling and

condensation of treated POME and further promotes the carbon footprint and pollute the

environment. As such, membrane technology which is energy conservative, modular and

easily to scaled up has been introduced to counter the abovementioned shortages. Principally,

membrane separation the separate and remove the colour pigments on the molecular sieving

mechanism [5]. Colour pigments in POME can be easily removed via nanofiltration.

However, nanofiltration requires high pressure as driving force hence high energy

consumption to purify the AT-POME. Furthermore, the colour pigments tend to block the

pore of the membrane and promote the formation of cake layer on the membrane and result in

membrane fouling [6, 7]. Therefore, mixed matrix nanocomposite ultrafiltration membrane

has been developed to reduce the fouling and pressure that required in the decolourization

3
process. In general, nanoparticles such as zinc oxide and titanium dioxide are used to improve

the membrane hydrophilicity. By doing so, the membrane able to operate at low pressure and

provide a high permeation for AT-POME.

Beside tailoring the hydrophilicity of the membrane to ease the transport of the water

molecule across the membrane and improve the membrane’s performance, modification on

the membrane surface charge would be another approach to enhance its performance by

increasing the rejection of unwanted solutes. To date, there are several methods to alter the

membrane surface charge like surface grafting, coating, use of pre-functionalized polymers as

additives and blending with hydrophilic and negatively charged NPs [8, 9]. These methods

have been attempted to alleviate the membrane fouling issue, however, most of the

techniques suffer from some shortages. For example, UV grafting will increase the

production cost due to its process complexity. Coating process may decrease membrane

performance as some of the functional pores are blocked by the coating agent and

delamination of the coating layer during the operation. Among the methods, direct blending

of polymer with NPs to produce a new class of nanocomposite membrane is considered a

promising method owing to its simplicity and the unique features of the NPs such as

photocatalytic properties can be retained in the membrane processes. A variety of organic and

inorganic NPs such as graphene oxide (GO), carbon nanotubes (CNTs), titanium dioxide

(TiO2), zirconium (ZrO2), silica (SiO2), zinc oxide (ZnO) , iron(III) oxide (Fe3O4) and

hydrous ferric oxide (HFO) are employed in the membrane matrix to enhance membrane

hydrophilicity and mitigate fouling problem [9-16]. Literature reported that 0.1 wt% of ZnO

was capable to improve the membrane performance by increasing 61% of the pure water flux

[17]. In addition, the incorporation of 1.0 wt% SAPO-14 in PVDF matrix had increased 51%

of the pure water flux [18].

4
Recently, many efforts have been devoted to the development of nanocomposite membranes

incorporated with hybrid metal oxide system to improve the performance in term of surface

charge, hydrophilicity and absorption properties. To achieve this target, the hybrid of highly

hydrophilic NPs such as GO-TiO2, TiO2-HNTs and TiO2-ZnO-CNTs are known to be

effective as they can render abundant hydroxyl group (-OH) on the membrane structure to

improve flux [9, 14, 19]. The coupling of the inorganic NPs shows promising improvement

of the membrane surface charge. From literature [20], zirconium (ZrO2) was coupled with

silica at various ratio. The result showed that the addition of silica can alter the membrane to

more negatively charged surface. The binary metal oxide of manganese and iron (ferric-

manganese-binary-oxide, FMBO) had been used to enhance adsorption capacity of the

nanocomposite membrane for arsenic removal [21]. Some studies also revealed that the

coupled metal oxide has strong surface complexation [22]. The changes in the surface

morphology can lead to the overhaul of the surface charge properties. The existence of the

ߨ-ߨ in the mixed transition metal oxide will hold the electrons on its surface longer and resist

the surface negatively charged. This will in turn result in highly negative charged surface NPs.

Zinc oxide (ZnO) and iron (III) oxide (Fe2O3) are two widely used nanofillers for polymeric

membrane matrix to increase the hydrophilicity of the membrane. In addition, these two NPs

are photocatalysts which can induce self-cleaning under UV and visible light illumination.

Rabiee et. al. [23] demonstrated the embedment of 3 wt% ZnO was capable to increase the

pure water flux from 213 kg/m2h to 402 kg/m2 h. Besides, the flux recovery was achieved up

to 91.8%. Besides, Jorge et. al. [24] showed that the incorporation of 0.05% ZnO in PES

improved the water permeability from 100 L/m2hbar to 170 L/m2hbar. Furthermore, Fe2O3

was employed as additive in ceramic membrane to induce the Fenton-type reactions to

5
achieve flux recovery of 97% [25]. In addition, AL-Hobaib et. al. [26] successfully improved

the membrane water flux from 26 L/m2h to 44 L/m2h while maintaining high salt rejection of

98% by incorporating Fe2O3.

In this study, zinc iron coupled oxide (ZIO) was developed and incorporated into

polyvinylidene fluoride (PVDF) UF membrane to augment the decolourization of AT-POME.

The effects of ZIO on the physico-chemical properties of the resultant nanocomposite

membrane was studies in the aspects of morphology, surface charge and wettability. As

surface change is one of the key factors in UF and it has not been widely discussed in the

open literature, the main aim of this study is to study the effect of surface charge on the UF

performance. Based on the findings, the coupling of ZnO and Fe2O3 have synergically

contributed to enhance surface negative charge, which favoured the membrane separation

performance in term of colour removal efficiency, flux and fouling resistance.

2. Experimental

2.1 Materials and chemicals

Iron (III) nitrate nonahydrate, zinc nitrate hexahydrate and urea were purchased from Acros

Organics for the synthesis of coupled zinc-iron oxide. Polyvinylidene fluoride, PVDF Solef

6000 Series – powder (Solvay Specialty Polymers France), polyvinylidone, PVP K28-32

(Acros Organics), lithium chloride, LiCl (Acros Organics), n-methyl pyrolidone, NMP

(Acros Organics) and coupled zinc-iron oxide were used to fabricate nanocomposite

ultrafiltration membrane. All chemicals were used without further purification. Aerobically

6
treated palm oil mill effluent, AT-POME was collected from Kahang PPNJ Palm Oil Mill,

Malaysia. The AT-POME was pre-filtered using a 0.45 µm cellulose acetate membrane prior

any testings.

2.2 Synthesis of zinc oxide, iron (III) oxide and coupled zinc-iron oxide

The ZIO was synthesized via solution combustion method [27, 28]. 0.4 mol iron (III) nitrate

nonahydrate, 0.1 mol zinc nitrate hexahydrate and 0.5 mol urea were dissolved in 200 mL

deionised water to form a mixture. After that, the solution was heated at 100°C for 10 h with

stir rate of 200 rpm to ensure the gel formation upon the evaporation of water. Then, the gel

was heated up to 500°C until the evolution white gas occurred, followed by a yellowish gas

was observed. After the completion of the reaction, the product was ground using agate

mortar and pestle. The resulted powder was calcined at 500°C for 4 h to remove any

impurities under atmospheric condition.

ZnO and Fe2O3 have been synthesized for comparison. The synthesis of single metal oxide

(i.e. ZnO and Fe2O3) were carried out by using the same solution combustion method as

mentioned above. For the preparation of ZnO, 0.1 mol zinc nitrate hexahydrate and 0.1 mol

urea were dissolved in 200 mL deionised water to form a mixture. After that, the solution was

vapourized at 100°C for 10 h and combust at 500°C until the white and yellowish gas

evolved. The resulted powder was calcined at 500°C for impurities removal. Fe2O3 was

synthesized using the same procedure with iron (III) nitrate nonahydrate replaced by zinc

nitrate hexahydrate.

7
2.3 Characterization of coupled zinc-iron oxide

The synthesized ZIO was characterized by field emission scanning electron microscope

(FESEM), energy dispersive X-ray Spectrometer (EDX), X-ray diffraction (ERD) and

scanning transmission electron microscope (STEM). FESEM imaging analysis was

performed to examine the morphology and also the size of nanoparticles. Besides, the energy

dispersive spectrum is used to determine the element that present in the ZIO. The metal plate

was adhered with the carbon double-sided tape and the sample was poured homogeneously

onto the adhesive layer with a thin layer. After that, the sample was observed under Hitachi

SU8020 for the imaging process. EDX was performed by using the E-MaxN that

manufactured by Oxford Instruments which attached to the FESEM unit. By using coupled

software, the dimension and elemental analysis of ZIO was analysed. STEM imaging analysis

was performed to determine the dispersion of the ZIO in solvent. The ZIO was dispersed in

ethanol solution (with 0.1 wt% concentration) and under a vigorous ultrasonication. After that,

10 µL of solution was dropped onto the copper grid with mesh size of 300. The sample was

observed under Hitachi SU8020 for imaging process. XRD analysis was conducted to

examine the crystallinity of ZIO. Powder XRD diffractogram of the sample was recorded by

X-ray diffractometer using Ni filter Cu Kα radiation (λ = 1.54056 Ǻ). The scanning speed of

3°/min with step size of 0.05° from 5° to 80°. The acceleration voltage and emission current

was fixed at 50 kV and 50 mA respectively.

2.4 Preparation of composite flat sheet membranes

8
The asymmetric flat sheet composite PVDF membranes were prepared using phase inversion

induced by immersion precipitation. The components of the casting solution were PVDF

(18 wt%), PVP K28-32 (1 wt%), LiCl (0.5 wt%), measured amount of ZIO, and NMP was

used as the solvent. First, ZIO, LiCl and PVP K28-32 were dissolved into NMP by

mechanical stirrer with stirred rate of 600 rpm for 1 day to prepare homogeneous solution.

Five different concentration of ZIO were used in this study: 0.0, 0.5, 1.0, 1.5 and 2.0 wt%.

PVDF powder was dried at 60°C in an oven for overnight to remove the absorbed water

before compounding (shown in Table 1). Subsequently, PVDF (18 wt%) was dissolved in the

solution and stirred for 24 h at 60°C until the solution was completely became homogeneous.

After being placed at room temperature overnight to remove the bubbles, the solution was

cast on a glass plate using a casting roller (glass rod with diameter of 25 mm). The casting

solution was left at ambient for 10 seconds before immersed into a water coagulation bath for

2 days prior to post-treatment using methanol and n-hexane. The post-treated membranes

were dried in ambient condition prior for storage.

Table 1: Composition of dope solution

Composition (wt%)
Membrane
NMP PVDF PVP LiCl ZIO

M0 80.5 18.0 1.0 0.5 0.0

M0.1 80.4 18.0 1.0 0.5 0.1

M0.5 80.0 18.0 1.0 0.5 0.5

M1.0 79.5 18.0 1.0 0.5 1.0

M2.0 78.5 18.0 1.0 0.5 2.0

2.5 Membranes characterization

9
The morphologies of the flat sheet membrane were characterized with a field emission

scanning electron microscope (FESEM) (Zeiss Supra 35VP Field Emission Scanning

Electron Microscope, Carl Zeiss, Inc., USA). The flat sheet membrane was first immersed in

the methanol, then frozen in liquid nitrogen and fractured by using tweezers. The samples

were then placed on a metal plate with carbon double-sided tape. The samples were then

sputter coated with platinum and the accelerating voltage was 5 kV throughout the analysis.

The surface roughness on the top surface of the membrane was characterized with atomic

force microscopy (AFM) using non-contact AFM mode (Park System XE-100, Park System

Inc., Korea). The scanning area of each membrane was set at 10 µm × 10 µm. The surface

roughness was then analysed by software.

The membrane wettability was measured by conducting static contact angle (SCA) using a

contact angle ganiometer (OCA-15Pro, DataPhysics Instruments GmbH) and RO water was

used as the probe liquid. The water droplet shape was captured by the camera that mounts on

the instrument. The measured data was analysed by SCA20 contact angle software. For each

membrane sample, fifteen measurements were taken at different areas of the membrane at

room temperature and the average value was reported.

Embedment of coupled zinc-iron oxide in the membrane matrix was examined by using X-

ray diffraction (XRD). The membranes were cut into very fine pieces and put into the sample

cell of XRD. The XRD patterns of the samples were recorded by X-ray diffractometer using

Ni Filter Cu Kα radiation (λ=1.54056 Å). The scanning speed of 3°/min with step size of 0.05°

from 20° to 80°. The acceleration voltage and emission current is 50kV and 50 mA

10
respectively. The crystalline size of the coupled zinc-iron oxide is 5.7179 nm as calculated by

following equation:

0.89ߣ
‫=ܦ‬
ߚܿ‫ߠݏ݋‬

where λ is the wavelength (Cu Kα), β is the full width at the half maximum (FWHM) of the

coupled zinc-iron oxide and θ is the diffraction angle.

Pore size and pore size distribution of the flat sheet membrane were characterized using a

BET surface area analyser (Surfer Gas Adsorption Porosimeter, Thermo Scientific). The BET

study was carried by using N2 as an adsorption-desorption agent. The adsorption-desorption

analysis was carried out at -197° in automated surface area. Prior to the analysis, the samples

were degassed in vacuum condition for 4 h at 60°C. The pore size distribution was calculated

with the nitrogen desorption isotherm. The relationship between Kelvin radiuses (critical

radius) and relative pressure was represented by the classical Kelvin equation as follows:

ܲ −4.14
݈‫݃݋‬ =
ܲ଴ ‫ݎ‬௞

௉
where rk is Kelvin radius (critical radius) in Å and is the relative pressure in the BET
௉బ

surface area measurement.

The membrane surface charge was carried out by conducting streaming potential analysis

(Anton Paar SurPASS Analyzer, Anton Paar Inc., Ireland). For each measurement, two

membrane samples with area of 20 mm × 10 mm were fixed on sample holder using double

sided adhesive tape. The maximum pressure difference was set at 400 mbar to assure a

laminar flow. A 1 mM of KCl solution was used as the background electrolyte. The pH value

11
of the electrolyte was adjusted between 2 and 10 using 0.1 M hydrochloric acid or 0.1 M

sodium hydroxide solution. Rinsing and flow check steps were carefully performed before

the measurement in order to ensure reproducibility. The zeta potential was calculated based

on the Helmholts-Smoluchowski (HS) approach.

2.6 Colour Removal and antifouling performance

The filtration performance of the membranes was evaluated using an in house fabricated

cross flow flat sheet permeation system. The membrane was initially compacted at 2 bar

before lowering to 1 bar for data recording. The flux of pure water (J0) and AT-POME

sample were calculated using following equation.

ܳ
‫=ܬ‬
‫ݐ∆ܣ‬

where J is the flux (L/m2h), Q is the permeate volume (L), A is the membrane area (m2) and

∆t is the permeating time (h).

To examine the antifouling capability, the membrane was tested for 4 h continuously. The

flux recovery and flux declination were observed throughout the experiment. The membrane

was then taken out and subjected to a physical cleaning, i.e. washing the membrane surface

with running water. After that, the membrane was reinstalled into the system for subsequent 4

h of AT-POME filtration process. The process was repeated for four cycles.

2.6 Analysis

Chemical oxygen demand (COD) measurement was performed with APHA Standard method

8000 using the colorimetric method. 2 mL of centrifuged sample was added into COD vial

12
(HR 20 – 1500 mg/L) and digested for 2 h by using HACH digester (DRB200, Loveland,

CO). The COD concentration was then measured with HACH spectrophotometer (DR5000,

Loveland, CO).

Colour measurement was carried by spectrophotometric method (DR5000, HACH, Loveland,

CO). American Dye Manufacturing Index (ADMI) colour unit were used to determine the

apparent colour of AT-POME.

The removal of colour and COD were calculated using the following equations.

‫ܫܯܦܣ‬௣௘௥௠௘௔௧௘
ܴ஺஽ெூ = ቆ1 − ቇ × 100
‫ܫܯܦܣ‬௙௘௘ௗ

‫ܦܱܥ‬௣௘௥௠௘௔௧௘
ܴ஼ை஽ = ቆ1 − ቇ × 100
‫ܦܱܥ‬௙௘௘ௗ

where RADMI and RCOD are the rejection (%) of colour in ADMI unit and COD respectively,

subscript “after reaction” and “before reaction” represent the value of AT-POME after the

photocatalytic reaction and before the photocatalytic reaction for respective parameter,

respectively.

3. Results and discussion

3.1 Nanoparticles characteristic

The morphology of the synthesised ZIO was assessed using FESEM and STEM meanwhile

the elements present in the ZIO were identified by EDX to ensure the removal of impurities

13
during the solution combustion process. As shown in Figure 1a and 1b, the synthesised ZIO

was appeared in nano-spherical shape. Figure 1C shows the EDX spectrum of the synthesised

ZIO. The spectrum indicates that the nitrate group in the zinc nitrate and iron nitrate (metal

oxide precursors) were successfully eliminated through the solution combustion and

calcination processes. At the same time, the ratio of the zinc to iron was found to be 1:4.

XRD analysis was performed to determine the crystallinity and crystalline size of the

synthesised nanoparticles. In addition, the formation of ZIO was also identified through the

matching with the database. The sharp and symmetric peaks obtained from the XRD peaks as

shown in Figure 1d indicates the presence of high crystallinity of the ZIO.

Figure 1 Characteristics of the ZIO, (a) FESEM and (b) STEM reveal that the ZIO

nanoparticles exhibits in nanosperical shape, (c) EDX peaks show the ratio of zinc to iron is

1:4 and (d) XRD diffractogram proves the synthesised ZIO has lattice structure of [311] and

[440] with crystalline size of 5.7179 nm.

14
3.2 Membrane morphology

Figure 2a – 2c show the FESEM images of the nanocomposite UF membrane surface.

Despite of the different ZIO loadings, the nanocomposite membranes posed woven thread

like surface structure as shown in Figure 2a. However, the incorporation of ZIO in the

membrane had resulted in different membrane cross-section morphologies. From Figure 2b,

all the fabricated membranes show typical asymmetric porous structure that consists of a thin

upper dense layer and a finger-like porous sub-layer. Also, the membrane surfaces remained

intact, indicating that the integrity of the membrane was well conserved after the addition of

ZIO. Upon the addition of ZIO, the macrovoids that present in the sub-layer of pristine PVDF

membrane have been greatly suppressed. This structure is preferable for water separation as

the existence of macrovoids is undesirable for long term operation stability of the membrane

[18]. The alteration of the membrane morphology was due to the presence of ZIO in the

casting dope solution which has increased the dope viscosity and delayed the mixing rate [29].

In addition, ZIO also acted as the pore-forming centre and slowed down the precipitation

process, thus increased the thickness of the top dense layer. As the ZIO loading increased

from 0 wt% to 2.0 wt%, the dense layer thickness increased from 77.71 nm to 231.38 nm.

Surface properties of a membrane play an important role to alter the hydrophilicity as well as

the antifouling properties of the membrane. In this study, the membrane surface roughness

was examined using AFM. As shown in Figure 2d, the surface roughness of MMMs was

compared with the pristine PVDF membrane. As observed in Figure 2d, the roughness of the

membranes decreased with the addition of ZIO. It showed that the addition of ZIO had

suppressed the crystallization of the polymer and allowed the formation of smoother surface

15
[30, 31]. The membrane average roughness (Ra) was calculated using AFM analysis software

to indicate the arithmetic average of the absolute values of the surface height deviations

measured from the centre plane. From Figure 2d, the membrane surface roughness (Ra) has

increased from 32.646 nm to 57.745 nm when the ZIO loading was increased from 0wt% to

0.5 wt%. This observation was probably due to the hydrophilicity of the ZIO that led to the

migration of nanoparticles to the surface of the membrane during phase inversion process. As

a result, the surface roughness increased when the ZIO loading increased from 0 wt% to 0.5

wt%. However, when the ZIO loading increased from 0.5 wt% to 1.0 wt%, the surface

roughness decreased from 57.745 nm to 25.532 nm. The addition of the nanoparticles (ZIO)

has increased the viscosity of the casting dope solution and resulted the formation of dense

layer on top of the membrane. Besides, the ZIO also plays the important role in the reduction

in the membrane porosity [32]. As a result, a smoother surface was observed. It was found

that the further increase of ZIO loading up to 2 wt% has again increased the surface

roughness of the membrane. This is due to the less dispersion of ZIO on the membrane

surface which has resulted in a rougher surface. The high content of ZIO can produce the

hunks due to the decrease in the distance between nanoparticles in casting solution [33].

Nevertheless, no obvious nanoparticle agglomeration was observed from the surface

micrographs.

16
Figure 2 Membrane morphologies of ZIO/PVDF nanocomposite UF membrane. (a)

Surface morphology, (b) cross sectional view, (c) dense top layer and (d) surface roughness.

3.3 Membrane porosity and pore size

Table 2 tabulates the effective surface area, pore size and porosity of the fabricated

membranes. The membrane porosity decreased with the addition of ZIO. The results revealed

that the addition of ZIO has effectively altered the physico-chemical properties of the

fabricated membrane by influencing the thermodynamic and kinetic driving forces during the

phase inversion process. This resulted in a slow exchange rate between the solvent and non-

solvent, thus, led to the formation of dense layer and lower porosity [13, 31]. Besides, the

incorporation of ZIO into the polymeric matrix decreased the membrane pore size. However,

the change in pore size of the membrane was not significant. The pore size of the membranes
17
ranges from 18.63 nm to 29.07 nm. Based on the porosity and pore size, M0.5 showed the

lowest pore size among the fabricated membranes. Therefore, it is expected that M0.5 can

perform well in UF application owing to the size exclusion separation properties. Meanwhile,

the membrane surface roughness has a close relationship with the membrane effective surface

area. As the roughness increased, the peak and valley were formed and increased the

membrane surface area. From Figure 2d, M0.5 exhibited the roughest surface and recorded

relatively high surface area based on BET result.

Table 2 Surface area, pore size and porosity of fabricated membrane

Membrane Surface area (m2/g) Pore size (nm) Porosity (%)

M0 26.62 23.23 65.36

M0.1 27.01 29.07 47.43

M0.5 28.76 18.63 44.62

M1.0 33.17 20.90 37.50

M2.0 25.79 22.58 39.71

3.4 Membrane XRD analysis

The XRD diffractogram in Figure 3 shows the intensity of the characteristic peak at 2θ=35.4

increased with the increasing ZIO loading. However, the peak at 2θ = 20.6° and 29.5°

became less intense when the loading of coupled zinc-iron oxide increased. This indicated

that the addition of ZIO had successfully reduced the crystallinity of PVDF.

18
Figure 3 The XRD diffractogram of fabricated membranes shows intense peak at

2θ=29.5° which indicates the PVDF crystalline structure in the membrane. As ZIO loading

increased from 0 wt% (M0) to 2.0 wt% (M2.0), the peak at 2θ=35.4° became more intense.

3.5 Effect of coupled zinc-iron oxide on membrane surface charge

Figure 4 shows the surface zeta potential of the fabricated membranes over a wide range of

pH. As expected, the pristine PVDF membrane (M0) has isoelectric point (IEP) at nearly pH

4.3 as reported previously. As ZIO loading increases, the zeta potential of the membrane

became more negative (i.e. more acidic surface). This observation suggests that the ZIO was

hydrated when mixed in NMP solvent. When the hydrated ZIO in the membrane matrix

dissociated in water, it remained negatively charged. During the fabrication of MMMs, some

19
of the ZIO accumulated on the membrane surface and exposed some functional groups

(hydroxyl group) that readily to be protonated in aqueous medium, hence causing the

membrane surface to be negatively charged [34]. Theoretically, the surface charge able to

repel the charged solute in the wastewater to improve the rejection of the colour pigment.

Therefore, the colour pigment rejection increased as the membrane surface charge became

more negative. Hence, it is expected that M2.0 (ξ potential = -53.5 at pH 8) able to achieve

higher colour removal. The figure also shows that all the membranes started to deform at

operating pH less than 3 (plateau of the zeta potential in Figure 4), which suggests that the

operating condition should be controlled above pH 3 to maintain the integrity of the

membranes.

Figure 4 Zeta potential of membrane

20
3.6 Membrane wettability

Surface hydrophilicity is one of the significant properties of the membrane which can

strongly affect the flux and the antifouling ability of a membrane. In general, a lower contact

angle indicates that the membrane surface is more hydrophilic. In Figure 5, the contact angles

of membrane surface measured using pure water and AT-POME are plotted as the function of

ZIO loading. Membrane wettability in this study was evaluated based on the static contact

angle. In this study, the main factors that influenced the change in membrane wettability are

liquid surface tension, surface charge and surface roughness of the membrane. It can be

noticed that the contact angle of water and AT-POME with the membrane surface exhibited

different trend when the ZIO loading increased from 0 wt% to 0.5 wt%. This was most

probably due to the presence of colour pigments (i.e. tannin, lignin and other compounds)

which tend to reduce the liquid surface tension. Therefore, the contact angle of AT-POME on

membrane is less than that of water. However, ZIO in the nanocomposite membrane tends to

repel the colour pigments in the AT-POME hence hindered the attachment of AT-POME on

the membrane surface. As a result, the contact angle of AT-POME with membrane increased

from 0 wt% to 0.5 wt% of ZIO. On the other hand, when water act as the probe liquid, the

contact angle decreased as the ZIO loading increased from 0 wt% to 0.5 wt%. When the ZIO

loading increases from 0.5 wt% to 1.0 wt%, both water and AT-POME contact angle behave

similarly, which decreasing. During the phase inversion process, the ZIO nanoparticles

migrated to the membrane surface and resulted more oxygen-containing functional group of

ZIO were exposed on the membrane surface [29]. Thus, the hydrophilicity of the membranes

was increased when the concentration of hydrophilic nanoparticles in the membrane

increased. However, when the loading of the ZIO was further increased to 2 wt%, the

21
membrane wettability decreased (i.e. membrane became more hydrophobic). This shows that

the water contact angle is also relying on the membrane effective surface area. From Table 2

in Section 3.3, the membrane surface area of M2.0 was the lowest due to less peak and valley

structures on the membrane surface. At the same time, the less surface area of ZIO located on

the surface of the membrane due to agglomeration which led to less absorption of the water

molecules and reduced the membrane hydrophilicity [17].

Figure 5 Water and AT-POME contact angle of the nanocomposite membranes with

different ZIO loading

3.7 Membrane performance

The fabricated membranes were tested for their pure water permeabilities (PWP), AT-POME

permeabilities, colour removal efficiency and chemical oxygen demand (COD) removal

efficiency. Figure 6a shows the PWF and AT-POME flux at different loadings of ZIO. As

shown in Figure 6a, the flux of the membrane can be associated with the membrane

hydrophilicity. The increase in membrane hydrophilicity enhanced the absorption of water

22
molecules within the membrane matrix, hence promoted water permeation through the

membrane [18]. Figure 6a also demonstrates that the addition of ZIO has improved the PWP

owing to the enhancement of the membrane hydrophilicity (the neat membrane is 85.92

Lm2h-1 and M0.1 is 223.94 Lm2h-1). However, the increasing in ZIO loading has resulted in

the decreasing in flux (i.e. from 223.94 Lm2h-1 to 125.35 Lm2h-1 with the loading from 0.1 wt%

to 2.0 wt%). This can be explained by the decrease in the porosity of the membrane and the

formation of the dense layer on the top surface which restricted the water passage. On the

other hand, the AT-POME flux increased by 25% upon the addition of 1.0 wt% ZIO. The

drastic reduction in flux was due to the presence of more suspended particles in AT-POME,

which prevented the water to penetrate through the pores of the membrane. However, the flux

of the membrane decreased significantly when ZIO loading used was more than 1wt% due to

the agglomeration of ZIO in the membrane matrix.

Figure 6b reveals the membrane removal efficiency in terms of COD and apparent colour.

The graph shows that both colour removal and COD removal had increased by 17% with the

addition of ZIO. Among the fabricated membranes, M0.5 and M2.0 exhibited the best colour

removal efficiency. The membrane pore size plays an important role to separate the colour

pigment in the AT-POME. Hence, the smaller the pore size, the higher the rejection.

However, this will sacrifice the flux of the membrane and decrease the water permeability.

M2.0 performed very well owing to the thicker dense layer that facilitated the rejection. On

the other hand, M0.5 with relatively high flux also capable to remove the colour of the AT-

POME, mainly due to its strong negatively membrane surface charge. AT-POME is

negatively charged at pH 8 due to the presence of colour pigments such as tannin, lignin and

carotenoid [35]. These compounds have acidic surfaces (negatively charge surfaces) which

lead to the agglomeration of these compounds when the AT-POME is adjusted to the acidic

23
condition. Hence, one of the ideal approaches to mitigate the fouling issue on membrane

surface is to employ a membrane with negatively charged surface to repel and prevent the

deposition of foulants. By doing so, the fouling of the membrane can be mitigated. The

colour removal efficiency is in agreement with the negative surface charges that present on

the membrane surface. The results indicate that the major role of ZIO is to enhance the

membrane surface charge in order to improve the colour removal efficiency without

sacrificing the permeability.

In order to examine the colour intensity of the membrane treated AT-POME, UV-Vis

spectrophotometer was employed to check the absorbance of the solution from 700 nm to

200 nm. As shown in Figure 6c, the treated AT-POME had relatively low absorbance above

the visible region, i.e. from 700 nm to 400 nm which implies that the apparent colour had

been significantly reduced after the membrane treatment. It is clearly shown that the M2.0

showed a promising colour removal efficiency among all the fabricated membranes, which is

70.97% with flux of 31.69 Lm2h-1. However, the relatively thick dense top layer formed at

M2.0 reduced the water permeability of the membrane even though it showed high rejection

towards the colour of the AT-POME. Therefore, M0.5 is chosen as the optimum membrane

owing to its relatively high colour removal efficiency, which is 69.53% with flux of 40.14

Lm2h-1.

24
Figure 6 Membrane performance: (a) permeability, (b) removal efficiency and (c)

comparison of AT-POME before and after membrane treatment

Figure 7 illustrates the colour pigment removal based on the negative charged membrane

surface. The colour pigments such as tannin and lignin protonated and become negatively

charged with the presence of water. In addition, the hydrated ZIO in the PVDF matrix also

protonates in the water and hence, improve the negative charge of the membrane surface.

Theoretically, there is a repulsion force between two compounds which exhibit same charge.

Therefore, it is expected that the negatively charged surface of the membrane can enhance the

repulsion between the colour pigments and membrane surface, which resulting in better

colour removal.

25
Figure 7 Illustration of colour pigment in AT-POME being repelled by the negative

charged membrane surface. Tannin is protonated in water and becomes negatively charged (a)

and the hydrated ZIO on the PVDF membrane surface protonated in water and become

negatively charged (b). As a result, the negatively charged tannin and membrane surface

repel each other (c). Therefore, water molecules and some unrepelled tannins can penetrate

through the membrane pores (d).

3.8 Comparison of coupled zinc-iron oxide and single metal oxides

In order to compare the effect of ZIO to its single metal oxide, i.e. zinc (II) oxide and iron (III)

oxide, towards the membrane performance, nanocomposite PVDF membranes with addition

of 0.5 wt% of ZIO, zinc (II) oxide (ZnO) and iron (III) oxide (Fe2O3) were fabricated,

respectively. As shown in Figure 8, the addition of nanoparticles did not bring any significant
26
changes to the contact angle. The nanocomposite membranes added with ZnO and Fe2O3

exhibited extremely low pure water flux and AT-POME flux due to the formation of

relatively dense structure. Despite the dense structure, PVDF/ZnO and PVDF/Fe2O3 also

demonstrated poor colour removal efficiency. From the zeta potential analysis, PVDF/ZnO

had the almost similar surface negativity with pristine PVDF membrane which is -35.3 and -

32.6 mV respectively. However, when the iron (III) oxide was added into the PVDF matrix,

the membrane surface became more positively charge which recorded -12.2mV of zeta

potential. Based on the hypothesis, PVDF/ZnO and pristine PVDF membrane should exhibit

similar colour removal as they have similar surface negative charge. On the other hand,

PVDF/Fe2O3 might experience lower colour removal owing to the low surface negativity.

Figure 8c reveals that the PVDF/ZIO achieved the highest colour removal, which was 69.53%

while pristine PVDF and PVDF/ZnO achieved 59.25% and 59.32% of colour removal

respectively. Besides, PVDF/Fe2O3 achieved only 53.77% colour removal. This observation

has firmly proved the membrane surface negativity plays an important role in the colour

pigment removal.

27
Figure 8 Membrane performance: (a) contact angle, (b) permeability and (c) removal

efficiency

3.9 Effect of coupled zinc-iron oxide in membrane fouling mitigation

This section discusses the effect of ZIO on the membrane fouling mitigation. The long term

operating performance of the membrane is greatly dependents on the severity of fouling.

When the deposition of foulants occurs on/in the membrane pores, it significantly reduces the

membrane permeability [6]. In this study, the surface charges rendered by the coupled metal

oxides was capable to repel the organic foulants, i.e. tannin and lignin from depositing on the

membrane surface. As such, the colour rejection was improved and at the same time, the

fouling was mitigated. Figure 9a and 9b show the variation of the permeate flux and colour

removal of AT-POME for four cycles throughout the 16 h operation. The results showed that
28
the flux of all membrane decreased as the function of time in each cycle. This was due to the

deposition of organic compounds on the membrane surface over the operating time which

formed a cake layer and increased the resistance to permeate. It is clearly shown that in the

subsequence cycles of operation, M2.0 exhibited the most stable performance. Interestingly,

the colour removal efficiency of neat PVDF membrane increased slightly during the process.

It was due to the high fouling tendency of neat PVDF membrane in which the formation of

cake layer had prevented the passage of the colour pigment, hence increased the colour

removal efficiency [8].

Figure 9a shows that M2.0 successfully remained the AT-POME flux at 20 – 25 Lm2h-1 after

4 cycles of operation. Besides, the recovery of AT-POME flux was approximately 99% after

the physical cleaning process. However, M0.5 experienced the distortion on the membrane

structure after 4 cycles of continuous operation. Therefore, the AT-POME flux fluctuated

after the physical cleaning. M0.5 recorded AT-POME flux of 21.13 Lm2h-1 at the first cycle

and the flux increased to 85.35 Lm2h-1 at the fourth cycle. Figure 9b demonstrate the colour

removal efficiency of the fabricated membrane. M2.0 recorded consistence colour removal

ranging 65.55% to 74.12%, however M0.5 experienced drastically drop in colour removal

from 59.25% to 35.31%. The decline in the performance was ascribed to the collapse of

macrovoids during the pressure driven filtration process, which has enlarged the pores and

resulted in higher flux as shown in Figure 9a. However, the enlargement of pore provided

non-selective channels for the passage of colour pigments, therefore the colour pigment

penetrated through the membrane pores and resulted in poor colour removal efficiency.

Although the initial performance of the M0.5 was comparable to M2.0 in terms of flux and

colour removal, however the performance of M0.5 deteriorated with the prolonged operating

duration. Based on the finding, it is suggested that, if long term membrane operation is

29
required, M2.0 should be considered for stable colour removal and flux for prolonged

running due to its greater stability.

Figure 9 Long term performance of membrane: (a) flux and (b) colour removal

efficiency

4. Conclusions

A novel, highly surface negative charged PVDF/ZIO UF nanocomposite membrane was

fabricated via phase inversion process. AT-POME decolourization was improved after the

incorporation of ZIO nanoparticles in PVDF polymeric matrix membrane. The addition of

ZIO nanoparticles had improved the membrane surface negativity from -32.6 mV (pristine

PVDF membrane) to -53.5 mV (with 2.0 wt% of ZIO loading). This enhancement of the

surface negativity is expected to improve the decolourization process by repelling the

negative charged colour pigment that present in the AT-POME. Besides, the hydrophilic ZIO

nanoparticles has increased the pure water permeability from 85.92 Lm2h-1 (pristine PVDF

membrane) to 223.94 Lm2h-1 (with 0.1 wt% of ZIO loading). When the ZIO loading

increased, the pure water flux declined to 125.35 Lm2h-1 (with 2.0 wt% ZIO loading) owing

30
to the delay in the demixing during the phase inversion process and resulted a denser top

layer formed on the flat sheet membrane. On the other hand, the AT-POME flux also

increased from 31.69 Lm2h-1 (pristine PVDF) to 41.41 Lm2h-1 (with 0.1 wt% ZIO loading).

From the observation, it clearly shows that the addition of ZIO in PVDF polymeric matrix

had enhanced the colour removal from 59.25% to 70.97% by 2.0 wt% of ZIO without

sacrificing the AT-POME permeability. This clearly showed that the improvement in the

membrane surface charge has tremendously increased the colour removal of the AT-POME.

This work also provides an idea to tackle the decolourization issue by alter the membrane’s

surface charge.

Acknowledgment

We gratefully thank the financial support from Ministry of Higher Education for the HiCOE

grants (4J182, 4J204 and 4J195), LIMPID FEP-7 Collaborative European Project

Nanocomposite Materials for Photocatalytic Degradation of Pollutants (Project number:

NMP3-SL-2012-310177) and Zamalah Scholarship from Universiti Teknologi Malaysia.

Besides, the authors would like to thank Pertubuhan Pertanian Negeri Johor (PPNJ) for the

supply of AT-POME throughout this study.

References

1. Liew, W.L., et al., Conventional methods and emerging wastewater polishing

31
2. Neoh, C.H., et al., Decolorization of palm oil mill effluent using growing cultures of
Curvularia clavata. Environmental Science and Pollution Research, 2014. 21(6): p.
4397-4408.
3. Said, M., et al., Removal of COD, TSS and colour from palm oil mill effluent (POME)
using montmorillonite. Desalination and Water Treatment, 2016. 57(23): p. 10490-
10497.
4. Ng, K.H., et al., Photocatalytic degradation of recalcitrant POME waste by using
silver doped titania: Photokinetics and scavenging studies. Chemical Engineering
Journal, 2016. 286: p. 282-290.
5. Nataraj, S.K., K.M. Hosamani, and T.M. Aminabhavi, Nanofiltration and reverse
osmosis thin film composite membrane module for the removal of dye and salts from
the simulated mixtures. Desalination, 2009. 249(1): p. 12-17.
6. Said, M., et al., Blocking mechanism of PES membrane during ultrafiltration of
POME. Journal of Industrial and Engineering Chemistry, 2015. 21: p. 182-188.
7. Ali Amat, N.A., et al., Tackling colour issue of anaerobically-treated palm oil mill
effluent using membrane technology. Journal of Water Process Engineering, 2015. 8:
p. 221-226.
8. Li, X., et al., Characterization and antifouling performance of negatively charged
PES/mesoporous silica ultrafiltration membrane for raw water filtration. Desalination
and Water Treatment, 2016. 57(24): p. 10980-10987.
9. Kumar, M., et al., Preparation and characterization of low fouling novel hybrid
ultrafiltration membranes based on the blends of GO−TiO2 nanocomposite and
polysulfone for humic acid removal. Journal of Membrane Science, 2016. 506: p. 38-
49.
10. Hairom, N.H.H., A.W. Mohammad, and A.A.H. Kadhum, Effect of various zinc oxide
nanoparticles in membrane photocatalytic reactor for Congo red dye treatment.
Separation and Purification Technology, 2014. 137: p. 74-81.
11. Lu, D., et al., Hydrophilic Fe2O3 dynamic membrane mitigating fouling of support
ceramic membrane in ultrafiltration of oil/water emulsion. Separation and
Purification Technology, 2016. 165: p. 1-9.
12. Shi, H., et al., A modified mussel-inspired method to fabricate TiO2 decorated
superhydrophilic PVDF membrane for oil/water separation. Journal of Membrane
Science, 2016. 506: p. 60-70.

32
13. Hwang, T., et al., Ultrafiltration using graphene oxide surface-embedded polysulfone
membranes. Separation and Purification Technology, 2016. 166: p. 41-47.
14. Zeng, G., et al., Preparation and characterization of a novel PVDF ultrafiltration
membrane by blending with TiO2-HNTs nanocomposites. Applied Surface Science,
2016. 371: p. 624-632.
15. Lee, J., et al., High flux and high selectivity carbon nanotube composite membranes
for natural organic matter removal. Separation and Purification Technology, 2016.
163: p. 109-119.
16. Abdullah, N., et al., Polysulfone/hydrous ferric oxide ultrafiltration mixed matrix
membrane: Preparation, characterization and its adsorptive removal of lead (II) from
aqueous solution. Chemical Engineering Journal, 2016. 289: p. 28-37.
17. Rajabi, H., et al., Nano-ZnO embedded mixed matrix polyethersulfone (PES)
membrane: Influence of nanofiller shape on characterization and fouling resistance.
Applied Surface Science, 2015. 349: p. 66-77.
18. Vatanpour, V., M.E. Yekavalangi, and M. Safarpour, Preparation and
characterization of nanocomposite PVDF ultrafiltration membrane embedded with
nanoporous SAPO-34 to improve permeability and antifouling performance.
Separation and Purification Technology, 2016. 163: p. 300-309.
19. Bai, H., et al., Multi-functional CNT/ZnO/TiO2 nanocomposite membrane for
concurrent filtration and photocatalytic degradation. Separation and Purification
Technology, 2015. 156, Part 3: p. 922-930.
20. Colilla, M., et al., Advanced Drug Delivery Vectors with Tailored Surface Properties
Made of Mesoporous Binary Oxides Submicronic Spheres. Chemistry of Materials,
2010. 22(5): p. 1821-1830.
21. Jamshidi Gohari, R., et al., Fabrication and characterization of novel PES/Fe–Mn
binary oxide UF mixed matrix membrane for adsorptive removal of As(III) from
contaminated water solution. Separation and Purification Technology, 2013. 118: p.
64-72.
22. Taujale, S. and H. Zhang, Impact of Interactions between Metal Oxides to Oxidative
Reactivity of Manganese Dioxide. Environmental Science & Technology, 2012. 46(5):
p. 2764-2771.
23. Rabiee, H., et al., Improvement in flux and antifouling properties of PVC
ultrafiltration membranes by incorporation of zinc oxide (ZnO) nanoparticles.
Separation and Purification Technology, 2015. 156, Part 2: p. 299-310.
33
24. Garcia-Ivars, J., et al., Comparison between hydrophilic and hydrophobic metal
nanoparticles on the phase separation phenomena during formation of asymmetric
polyethersulphone membranes. Journal of Membrane Science, 2015. 493: p. 709-722.
25. De Angelis, L. and M.M.F. de Cortalezzi, Improved membrane flux recovery by
Fenton-type reactions. Journal of Membrane Science, 2016. 500: p. 255-264.
26. Al-Hobaib, A.S., K.M. Al-Sheetan, and L. El Mir, Effect of iron oxide nanoparticles
on the performance of polyamide membrane for ground water purification. Materials
Science in Semiconductor Processing, 2016. 42, Part 1: p. 107-110.
27. Tan, Y.H., P.S. Goh, and A.F. Ismail, Development of photocatalytic coupled zinc–
iron oxide nanoparticles via solution combustion for bisphenol-A removal.
International Biodeterioration & Biodegradation, 2015. 102: p. 346-352.
28. Pradhan, G.K., S. Martha, and K.M. Parida, Synthesis of Multifunctional
Nanostructured Zinc–Iron Mixed Oxide Photocatalyst by a Simple Solution-
Combustion Technique. ACS Applied Materials & Interfaces, 2012. 4(2): p. 707-713.
29. Alpatova, A., et al., Composite polyvinylidene fluoride (PVDF) membrane
impregnated with Fe2O3 nanoparticles and multiwalled carbon nanotubes for
catalytic degradation of organic contaminants. Journal of Membrane Science, 2015.
490: p. 227-235.
30. Wang, J., et al., Fabrication of hydrophobic flat sheet and hollow fiber membranes
from PVDF and PVDF-CTFE for membrane distillation. Journal of Membrane
Science, 2016. 497: p. 183-193.
31. Pezeshk, N., et al., Novel modified PVDF ultrafiltration flat-sheet membranes. Journal
of Membrane Science, 2012. 389: p. 280-286.
32. Zhao, J.Q., et al., Development of mesoporous titanium dioxide hybrid poly(vinylidene
fluoride) ultrafiltration membranes with photocatalytic properties. Journal of Applied
Polymer Science, 2016. 133(19).
33. Daraei, P., et al., Novel polyethersulfone nanocomposite membrane prepared by
PANI/Fe3O4 nanoparticles with enhanced performance for Cu(II) removal from
water. Journal of Membrane Science, 2012. 415: p. 250-259.
34. Mukherjee, R., P. Bhunia, and S. De, Impact of graphene oxide on removal of heavy
metals using mixed matrix membrane. Chemical Engineering Journal, 2016. 292: p.
284-297.

34
35. Mohammed, R.R. and M.F. Chong, Treatment and decolorization of biologically
treated Palm Oil Mill Effluent (POME) using banana peel as novel biosorbent.
Journal of Environmental Management, 2014. 132: p. 237-249.

35
Highlights

 M0.5 recorded high colour removal (70%) and AT-POME flux (40.14 LMH).

 ZIO has improve the membrane hydrophilicity and membrane surface negativity

simultaneously.

 Surface negativity plays important role in the colour removal of AT-POME.