Journal of Membrane Science 603 (2020) 118007

Contents lists available at ScienceDirect

Journal of Membrane Science

journal homepage: http://www.elsevier.com/locate/memsci

Oil-in-water separation with graphene-based nanocomposite membranes

for produced water treatment
Abdulaziz Alammar a, b, Sang-Hee Park c, Craig J. Williams d, Brian Derby d, Gyorgy Szekely a, c, *
a
School of Chemical Engineering and Analytical Science, University of Manchester, The Mill, Sackville Street, Manchester, M1 3BB, United Kingdom
b
Research & Development Center, P.O. Box 62, Saudi Aramco, Dhahran, 31311, Saudi Arabia
c
Advanced Membranes & Porous Materials Center, Physical Science Engineering Division (PSE), King Abdullah University of Science and Technology (KAUST), Thuwal,
23955-6900, Saudi Arabia
d
School of Materials, University of Manchester, Oxford Road, Manchester, M13 9PL, United Kingdom

A B S T R A C T

The treatment of wastewater from the oil and gas industry presents a very specific problem because the wastewater produced is comprised of a complex mixture of oil
and water that can be difficult to treat. In this work, polybenzimidazole (PBI), graphene oxide (GO) and reduced GO (rGO) nanocomposite membranes were
developed via the common blade coating and phase inversion technique for the treatment of produced water from the oil and gas industry. The nanocomposite
membranes were dip-coated by polydopamine (PDA), which is known for its antifouling properties. For the industrially relevant produced water, stable emulsions
with high salinity, sharp unimodal size distribution and average oil droplet size of less than 500 nm were prepared. The incorporation of just a few weight percent GO
into the PBI matrix resulted in superior oil-removal efficiency up to 99.9%, while maintaining permeance as high as 91.3 � 3.4 L m 2 h 1 bar 1. The presence of GO
also increased the mechanical stability of the membrane. The biofouling test of the nanocomposite membrane over 180 days showed remarkable improvement
compared to the pristine PBI membrane. The nanocomposite membranes described in this work demonstrated promising long-term performance for oil-in-water
emulsion separation as well as antifouling and antimicrobial properties without any alkaline or acidic cleaning. The membranes were capable of de-oiling high
salinity emulsions with excellent reusability, highlighting that these membranes are promising for produced water treatment under harsh industrial conditions.

1. Introduction
Water scarcity and the sustainable use of water has become a global
concern, as outlined by numerous studies and reports such as the United
Nations’ 2030 goals for sustainable development [1]. Industrial waste­
water management has attracted much interest in this context. Produced
water is a common expression used in the oil and gas industry to define
wastewater that is generated as a by-product during crude oil and gas
production. Produced water represents the largest waste stream during
the exploration, drilling and processing of oil and/or gas [2], with an
estimated global production of 41 million m3 day 1 [3]. The production
of one barrel of oil is estimated to produce the equivalent of around
three barrels of produced water for a typical reservoir and nine to ten
barrels for a mature reservoir [4]. New reservoirs tend to produce more
oil and/or gas than water; however, as the reservoir ages, the amount of
produced water increases dramatically and can reach up to 98% of the
produced oil/gas and water mixture [5]. Produced water is a complex
mixture with various inorganic and organic compounds, and its
composition can vary from one reservoir to another depending on the
geographical location of the well, the type of in-contact hydrocarbons
and the chemicals added throughout the drilling and production process
[6]. The rapid expansion of oil and gas production to meet increasing
energy demands means that produced water management has become a
pressing issue [7].
Produced water treatment involves de-oiling and demineralizing
processes before disposal or recycling to meet stringent environmental
regulations. However, existing conventional technologies for produced
water treatment are either susceptible to fouling or are insufficient to
separate stabilized oil-water emulsions. Industries currently employ
various technologies as stand-alone or hybrid processes, such as
adsorption, flotation, biological treatment, cyclonic and gravity sepa­
rators, coagulation-flocculation and evaporation, to treat the produced
water in order to meet stringent environmental regulations for disposal
or reuse in other processes [8]. Nonetheless, these technologies are not
effective for treating emulsified oil in water with droplet sizes smaller
than 10 μm [9,10]. Consequently, chemical dosage and the application

* Corresponding author. School of Chemical Engineering and Analytical Science, University of Manchester, The Mill, Sackville Street, Manchester, M1 3BB, United
Kingdom.
E-mail addresses: gyorgy.szekely@manchester.ac.uk, gyorgy.szekely@kaust.edu.sa (G. Szekely).
URL: http://www.szekelygroup.com (G. Szekely).

https://doi.org/10.1016/j.memsci.2020.118007
Received 4 November 2019; Received in revised form 10 February 2020; Accepted 26 February 2020
Available online 5 March 2020
0376-7388/© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
A. Alammar et al. Journal of Membrane Science 603 (2020) 118007

of an external electric field have been used to break down the oil
emulsions, but both processes tend to be energy-intensive and harmful
to the environment [11,12]. On the other hand, membrane technology
has been acknowledged as an effective method for oil emulsion sepa­
ration with lower energy consumption and a smaller environmental
impact [10]. However, membranes are susceptible to fouling due to
adsorptive surfactants or mechanical pore-plugging by the oil emul­
sions, which can cause a drastic reduction in membrane life span, flux
and removal efficiency, and may necessitate costly post-cleaning prac­
tices to achieve efficient recovery [13]. Thus, membrane fouling remains
the major challenge for the ultimate successful implementation of
membrane technology in produced water treatment.
Hydrophobic polymeric membranes such as polyvinylidene fluoride
(PVDF), polysulfone (PS), polyacrylonitrile (PAN) have been widely
studied for the treatment of oily wastewater [8]. However, these
membranes tend not to be efficient because the higher density of water
can hinder oil permeation by forming a layer that acts as a barrier on the
membrane surface [14]. However, the hydrophilicity of these mem­
branes can be increased through physical or chemical modification, as
well as additives-blending, which may improve fouling-resistance dur­
ing oil/water separation [9,15]. Graphene oxide (GO) filler is often
selected for its hydrophilicity, mechanical robustness, surface area and
antibacterial properties. Incorporating GO into the polymeric matrix is
expected to enhance the transport of water and the membrane’s ability
to withstand high operating pressures during the separation process due
to its oxygenated functional groups, inter-layer spacing, and sturdy
atom-thin structure [16,17]. GO-based membranes have been success­
fully developed for various separations, including for oil-water separa­
tion in produced water treatment [18]. Coating the membranes with a
mussel-inspired polydopamine (PDA) layer has the potential to
enhance the oil/water separation via its antifouling properties and
strong resistance to the adhesion of microorganisms [19]. Poly­
benzimidazole (PBI) has been chosen as a polymer matrix due to its
chemical, thermal and mechanical stability as well as its highly rigid
structure to withstand the harsh industrial conditions. Crosslinked GO
and hydroxylated PBI membranes were previously found to be stable
under harsh conditions, and they were successfully used for organic
solvent nanofiltration [20].
We expect that the integration of GO with a PBI nanocomposite
membrane, coated with a PDA layer, that we present in this work will
improve the permeation, antifouling and antibacterial properties, while
also providing adequate mechanical strength and thermal stability,
making the developed membranes applicable for widespread industrial
produced water treatment.
2.2. Reduction of graphene oxide
The reduction of GO was performed with ascorbic acid (AA) [21]. In
a typical experiment, 0.20 g of GO powder was added to a 250 mL
round-bottom flask, followed by the addition of water to achieve a
concentration of 0.80 mg mL 1. An aqueous GO dispersion was achieved
by vigorous stirring at 350 rpm. 1.4 g of AA was added to the GO
dispersion and left for 4 h under stirring at 90 � C.
2.3. Membrane fabrication
Dope solutions containing 16 wt% PBI and varied amounts of GO and
reduced graphene oxide (rGO) were prepared using DMAc as solvent
(Table 1, Fig. 1). The fillers were dispersed in DMAc using sonication,
followed by the addition of the suspension to the commercial PBI stock
solution (26 wt%). The dope solutions were mixed using an overhead
mechanical stirrer at room temperature and speed of 60 rpm for 6 h. The
PBI, PBI/GO and PBI/rGO membranes were prepared by wet phase
inversion, which is the most widely used technique to prepare mem­
branes for separation of oil-in-water emulsions [22]. The dope solutions
were cast on a polypropylene nonwoven support using a casting ma­
chine (Elcometer 4340) with a gate height of 250 μm. The A4 size film
was immediately immersed into a non-solvent coagulation bath, which
contained 10 L of water at 22 � C. The membranes were washed with
fresh water to remove any excess solvent.
The PDA coated membranes were prepared by the conventional dip-
coating method [23]. A dopamine monomer solution of 2 mg mL 1
concentration was prepared in water. A 100 � 100 mm membrane was
soaked in 140 mL solution at a temperature of 22 � C for 0.5 h, followed
by the rapid dissolution of NaIO4 at 5 mM concentration. The mixture
was gently shaken for 0.5 h, then washed with water and stored in 1: 9
(acetonitrile: water solution).
2.4. Emulsions preparation
Stable oil-in-water emulsions were prepared by high mechanical
shearing using the surfactant as an emulsifier. For a typical experiment,
free-salt oil-in-water emulsions were prepared by mixing 0.5 g of
vegetable oil (1 g L 1) with 20 mg of CTAB surfactant in 0.5 L water
using T-18 ULTRA-TURRAX® (IKA England Ltd) at 15,000 rpm for 5
min at 25 � C as reported elsewhere [24]. Salts were added to simulate
the salinity of produced water (Table 2) [3]. In order to maintain the size
and stability of the emulsions with salts, the turrax speed was increased
to 20,000 rpm for 12 min.

2. Experimental 2.5. Membrane separation

2.1. Materials A typical cross-flow system was used for the filtrations, composed of
a membrane cell with an effective area of 52.8 cm2, a relief valve, a high-
GO powder was purchased from William Blythe Ltd., UK. S26 Cela­ pressure pump and a recirculation gear pump (Fig. 2). The flow rate of
zole® PBI dope (26.2 wt% PBI, 72.3 wt% DMAc, 1.5 wt% LiCl) was the high-pressure pump and the recirculation gear pump were kept
supplied from PBI Performance Products Inc., USA. Novatexx 2471 constant at 3 L h 1 and 100 L h 1, respectively. Samples from the
(polypropylene nonwoven support) was purchased from Freudenberg
Filtration Technologies KG, Germany. Hexadecyltrimethylammonium
Table 1
bromide (CTAB) was purchased from Acros Organics. Dopamine hy­ Membrane designations, compositions and dip-coating.
drochloride was purchased from Alfa Aesar. Sodium chloride (NaCl),
Membrane PBI (wt%) GO or rGO (wt%) DMAc (wt%) PDA dip-coating
calcium chloride (CaCl2), magnesium chloride hexahydrate
(MgCl2⋅6H2O), sodium sulfate (Na2SO4), sodium bicarbonate (NaHCO3), MPBI 16 0.00 84.00 No
N,N-dimethylacetamide (DMAc), ascorbic acid and sodium periodate MGO
PBI
​ 0:50 16 0.50 83.50 No
(NaIO4), benzimidazole were purchased from Sigma Aldrich. 2-Phenyl­ MGO
PBI
​ 0:75 16 0.75 83.25 No
benzimdazole was received from Fluorochem. HPLC grade organic sol­ MGO
PBI
​ 1:00 16 1.00 83.00 No
vents such as n-hexane and acetonitrile were obtained from Fisher MGO
PBI
​ 1:50 16 1.50 82.50 No
Scientific. Vegetable oil was obtained from a local market. Type II MGO ​ 1:00 16 1.00 83.00 Yes
PBI=PDA
deionized water (DI) was used throughout all experiments with a re­ MPBI=PDA 16 0.00 84.00 Yes
sistivity of 18.2 MΩ cm at 25 � C (Milli-Q). All chemicals and materials MrGO ​ 1:00 16 1.00 83.00 No
PBI
were used as received without any modifications.

2
A. Alammar et al. Journal of Membrane Science 603 (2020) 118007

Fig. 1. Schematic representation of the membrane fabrication steps.

(TA Instruments) with a temperature ramp rate of 20 � C min 1 from

Table 2
25 � C to 800 � C in an N2 atmosphere. Emulsion droplet size and distri­
Applied composition of salts in the tested produced water [3].
bution were analyzed based on a dynamic light scattering (DLS) method
1
Salt Concentration (g L ) using Zetasizer 3000 HSA (Malvern Instruments). Atomic force micro­
Sodium chloride, NaCl 50 scope infrared spectroscopy (AFM-IR) was performed on a MultiMode 8
Calcium chloride, CaCl2 16 H R using TESPA-V2 probe (Bruker Instruments) in the tapping mode.
Magnesium chloride, MgCl2 8
The roughness analysis of the samples was performed on NanoScope
Sodium sulfate, Na2SO4 2
Sodium bicarbonate, NaHCO3 1 software using a 10 μm � 10 μm scan area. ATR-FTIR spectra of the
membranes were measured using an Alpha-P instrument (Bruker In­
struments). An average of 32 scans was used to generate the spectra
retentate and permeate streams were collected for analysis. under air using dried samples. Visible-light microscope reflection im­
The reported filtration results are the mean values of at least two ages of the membranes were collected by B 61 optical microscope
independently prepared membranes. The oil-removal efficiency (Eq. (Olympus). The membranes were fixed on glass slides using double-
(1)) was determined as the ratio of the obtained oil concentration in the sided tape. The swelling ratio (SR, Eq. (4)) of the membranes was
permeate (Cp) and the feed (Cf) streams. The permeance (Eq. (2)) was calculated using the dry length (ld) and wet length (lw) of the mem­
calculated based on the effective membrane area (A), the volume of the branes, and the reported values were obtained as an average of 10
collected permeate (Vpermeate) over time (t), and the transmembrane measurements.
pressure (ΔP). Continuous, long-term experiments were carried out at a
different split ratio (λ, Eq. (3)), defined as the flow rate of the permeate ðlw ld Þ
SR ​ ½%� ¼ � 100 (4)
(rpermeate) over the flow rate of the retentate (rretentate) streams. The ld
desired split ratio was set by the relief valve. Field Emission Scanning Electron Microscope (SEM) images were
� � acquired using a FEI Quanta 200 with the accelerating voltage set to 20
Cp
Removal efficiency ½%� ¼ 1 � 100 (1) kV. The samples were spin-coated with a 6 nm layer of gold or palladium
Cf
using a Quorum Q150TES under an Ar atmosphere to make the samples
� � Vpermeate conductive for surface and cross-sectional analysis. The thickness of the
Permeance L m 2
h 1
bar 1
¼ (2) membranes was determined from the average of 10 measurements
ΔP A t
across the SEM cross-sectional images of each membrane using ImageJ
λ½ � ¼
rpermeate
(3) software. The Energy Dispersive X-ray (EDX) spectrums were obtained
rretentate using FEI Quanta 200 with a backscatter detector (BSE) at 15 kV in low-
vacuum mode. The membranes, GO and rGO were uncoated since the
gold and palladium can interfere with the elements of interest. Trans­
2.6. Chemical and morphological characterizations mission electron microscopy (TEM) images of GO/rGO particles and
membranes were measured using Titan Themis Z of FEI and Titan Cryo
Thermogravimetric analysis (TGA) was determined using TGA-550 Twin of FEI, respectively. The GO and rGO particle samples were
dispersed in ethanol and then the solution was dropped on a copper grid
to obtain flat sheet. The membrane samples were embedded in an epoxy
resin for 4 h and then allow to polymerize the resin for 16 h at 65 � C. A
Leica UC6 ultramicrotome was employed to obtain the ultra-thin slice
(80 nm to 110 nm thickness) of the epoxy-embedded membranes and the
prepared sample slices were transferred to a copper grid.
Raman spectra of GO and rGO particles were collected from the
Raman spectrometer (LabRAM ARAMIS, HORIBA Jobin Yvon) with
visible laser (473 nm) and grating (1800 g mm 1) at room temperature.
X-ray photoelectron spectroscopy (XPS) analysis for GO, rGO and the
membranes was performed using an Axis Supra instrument (Kratos
analytical) equipped with a monochromatic Al Kα x-ray source (hν ¼
1486.6 eV) operated at a power of 150 W and under UHV conditions in
the range of ~10 9 mbar. The high-resolution C1 peak was deconvo­
luted and the area percentage of each peak was calculated using the
XPSpeak4.1 software. The contact angle measurements were performed
Fig. 2. Schematic of the cross-flow filtration rig for produced water treatment. on a DSA 100 (Krüss Instruments) with the built-in sessile drop fitting

3
A. Alammar et al.
function in drop-shape analysis system, which involved injecting 1.5 μL
of deionized water on membrane surface fixed on a flat stage. The vis­
cosity of dope solutions was obtained using Elcometer 2300 R V (Elc­
ometer Inc, USA) with TL7 connection at 6 rpm speed and 25 � C.
Nanoindentation was performed in water using a Hysitron Bio­
indenter (Hysitron, Eden Prairie, Minneapolis, MN) fitted with a 50 μm
diameter conospherical diamond probe. Samples were mounted to a
glass slide using a custom made liquid cell cover. A 5 � 5 grid pattern
was used to perform 25 indents per sample using a 5 s load, 2-s hold and
a 5-s unload, loading function. Indents were performed in load control
using a maximum load of 50 μN. Analysis was carried out using the
Hysitron Triboscope software using the Oliver and Phar method to
determine the reduced elastic modulus and hardness.
2.7. Antibacterial and antifouling tests
The antibacterial properties of MPBI , MGO
PBI
​ 1:00
and MGO ​ 1:00
PBI=PDA mem­
branes were evaluated based on microscopic visualization via SEM after
long-term exposure to water. The membranes were immersed in 0.5 L
tap water at room temperature for 180 days without changing the water
or adding any chemicals. Samples were taken out of the water and dried
after 45, 90, 135 and 180 days to characterize the surface of the mem­
branes using SEM. The antifouling property was assessed by the water
flux recovery ratio (FRR, Eq. (5)) [25,26], which was calculated based
on the pure water flux before (F0) and after (Fi) the produced water
filtration, i.e., a fouling-cleaning cycle.
Fi
FRR ½%� ¼ � 100 (5)
F0
The membranes were conditioned with pure water and the flux (F0)
was measured for 10 min. After filtering 150 mL of the simulated pro­
duced water, pure water was fed to the system to wash the membranes
for 10 min and the flux was measured within this period and denoted by
Fig. 3. TEM images of (a) GO and (b) rGO, (c) Raman spectra and (d) XRD patterns of GO and rGO.
4
Journal of Membrane Science 603 (2020) 118007
Fi. The fouling-cleaning cycle was repeated three times to measure the
FRR. SEM-EDX analysis was performed on the membrane top surface to
examine the fouled layer after the filtration of 0.5 L produced water,
followed by rinsing with pure water.
2.8. Adsorption experiment
An adsorption experiment was performed to investigate the in­
teractions between GO/rGO and amine moieties of PBI, namely benz­
imidazole and 2-phenylbenzimidazole. Samples were prepared by
loading various amount of GO and rGO (0.01–0.3 g) in 1 mL DMAc
solution containing 1 g L 1 of benzimidazole or 2-phenylbenzimidazole
considered as the feed. The sample vials were placed in an IKA Control
incubator shaker at 400 rpm for 24 h. The GO and rGO particles were
filtered using PVDF syringe filter (pore size ~0.45 μm). HPLC was used
to determine the concentration of the moieties in the feed stock solution
(Cf) and after filtering the GO and rGO (Cp) to obtain the adsorption
isotherms.
3. Results and discussions
3.1. Materials characterization
rGO was synthesized by the reduction of GO to use as a filler in PBI
membranes. Fig. 3 shows the TEM images and Raman spectra of GO and
rGO. The GO showed a smooth surface (Fig. 3a), whereas the rGO pre­
sented a clumped and wrinkled structure after the reduction (Fig. 3b).
This difference in morphology is due to the decomposition of the oxygen
functional groups, which resulted in the stacking of the wrinkled gra­
phene sheets during reduction [27]. In the Raman spectra of GO and rGO
(Fig. 3c), the specific G band and 2D band were observed at 1580 cm 1
and 2700 cm 1, respectively, which corresponds to the sp2 hybridized
carbon-carbon bonds in graphene [28]. The observed D band at 1350
cm 1 for both samples indicates that there were distortions in the lattice.
A. Alammar et al.
As a result of the reduction, an increase in the intensity ratio of D to G
bands from 0.98 (GO) to 1.05 (rGO) was observed. The amount of
oxygen-containing functional groups significantly decreased, which was
also demonstrated by the FTIR and XPS analysis (Fig. S6, Table S4).
These results demonstrate the decline of the average size of the sp2
domain and the recovery of the defected hexagonal network of carbon
atoms by the reduction [28,29]. Fig. 3d shows the XRD patterns of GO
and rGO to explore their interlayer distances. GO showed a strong
characteristic diffraction peak at 9.72� , which is in good agreement with
previous studies [28,29]. The interlayer distance between the GO sheets
(d ¼ 9.09 Å) is attributed to the oxygenated functional groups. On the
other hand, the graphene peak for rGO was shifted to around 25.16� ,
indicating a decrease in the interlayer distance (d ¼ ~3.52 Å), which is a
result of restoring the conjugate structure of the sp2 carbon via the
removal of the oxygen-containing functional groups. In addition, the
recovered hexagonal structures and the narrow interlayer distance of
rGO contributed to the enhancement of thermal stability after the
reduction (Fig. S7).
ATR-FTIR spectra of the membranes are shown in Fig. 4 and Fig. S8.
For the pristine PBI, a broad absorption peak at 2000–3600 cm 1 was
observed, corresponding to stretching vibrations of the hydrogen
bonding (N–H⋅⋅⋅N) [30,31]. The specific N–H absorption peak of PBI at
~3200 cm 1 was obscured with in the wide absorption peak. The three
characteristic peaks of PBI at 1630 cm 1, 1532 cm 1 and 1438 cm 1
were observed for all the membranes, which are assigned to C¼C, C¼N
and C–N stretching bands, respectively [31]. These peaks for MPBI
coincide with those of MGO ​ 1:00
PBI=PDA as a result of the similar chemical bonds
found in both PBI and PDA. Nonetheless, the successful PDA coating of
the PBI-based membrane was confirmed by i) a distinguished peak
attributed to the α,β-unsaturated ketone observed at 1610 cm 1, and ii)
the increased intensity of the stretching vibrations of catechol –OH
groups at about 3200 cm 1. On the other hand, there was no obvious
indication of the incorporation of GO and rGO owing to the low content
(<2 wt%) (Figs. S8–S10).
To further demonstrate the presence of GO and rGO in the PBI
membrane, visible light reflection images and XPS spectra were
collected as shown in Fig. 5. Compared to the MPBI , MGO PBI
​ 1:00
and
rGO ​ 1:00
MPBI revealed the shiny flakes in the visible-light microscope im­
ages, corresponding to the incorporation of GO and rGO [20]. It was
found that MGO PBI
​ 1:00
has well-distributed GO particles appearing as
small-sized shiny flakes (1–3 μm), which is in good agreement with the
size of the GO in the inset TEM image. However, some rGO aggregation
in the MrGOPBI
​ 1:00
were found due to the hydrophobic interactions between
Fig. 4. ATR-FTIR spectra of MPBI , MGO ​ 1:00
and MGO ​ 1:00
PBI PBI=PDA membranes, as well as
GO and PDA.
5
Journal of Membrane Science 603 (2020) 118007
the rGO particles. The presence of GO and rGO in the membranes was
demonstrated by their cross-sectional TEM images, which is in a good
agreement with the literature [32,33]. In addition, the MGO PBI
​ 1:00
showed
a significant increase in the oxygen content of the chemical composition
determined by XPS, which is attributed to the oxygen-based functional
groups of GO (Fig. 5b, Fig. S11). The increased area percentage of C–O
and C¼O peaks, and the presence of COOH peak in the deconvoluted C1s
peak demonstrate the successful incorporation of the 2D materials into
the polymer matrix (Fig. 5b, Table S5).
The dope solution viscosity, membrane thickness, swelling ratio and
surface properties are summarized in Table 3, while the morphologies
and topographies of the membranes are shown in Fig. 6. The higher the
loading of the GO, the higher the viscosity of the dope solution, which
culminated in a 68% increase at 1.5% GO loading. The observed in­
crease in viscosity can be attributed to the oxygen-containing functional
groups of the GO that can form secondary interactions, such as hydrogen
bonding [34], with the amine moieties of the PBI. This hypothesis is
further supported by the approx. 10% decrease in viscosity when rGO is
used instead of GO. In order to check the affinity of the 2D materials to
PBI, adsorption isotherms were obtained, which revealed that the fillers
adsorb benzimidazole derivatives. GO was found to have a higher af­
finity than rGO, which explains the higher viscosity of the dope solution
for MGO ​ 1:00
than for MrGO ​ 1:00
. Refer to the Supplementary Information
PBI PBI
for the details.
The thickness of the membrane increased from 38.8 μm to 49.9 μm as
the GO loading content increased, which was governed by the viscosity
of the dope solution (Table 3). The increase in the viscosity of the dope
solution resulted in a longer demixing time. The diffusion rate of water
into the film of the polymer dope solution and the diffusion rate of the
DMAc into the water phase were delayed during the phase inversion
process [35,36]. The longer demixing time contributed to the formation
of less porous structures and reduced the structural contraction, result­
ing in the formation of a thicker membrane. The MGO ​ 1:00
PBI=PDA was found to
be the thickest membrane with an increase of 8 μm. This increase was
not the result of a pristine PDA layer on top of the support PBI mem­
brane. We hypothesize that the PDA coating prevented [37,38] the usual
shrinkage [39] of the membrane structure upon drying before the SEM
analysis.
Owing to the hydrophilic nature of GO, the swelling ratio of the
membranes increased with the GO loading from 5.71% for the pristine
PBI membrane to 8.17% for the MGO PBI
​ 1:00
. The enhanced hydrophilicity
of the membrane allowed water molecules to easily penetrate the
membrane pores [40] and to enable the retention of water molecules
[41], leading to an increase in the swelling ratio. A further increase of
1.5% GO loading resulted in a decrease in the swelling ratio due to the
GO agglomeration. Both the reduction of GO and the PDA coating
decreased the swelling compared to the corresponding MGO PBI
​ 1:00
membrane.
Hydrated nanoindentation analysis was carried out to assess the
mechanical properties of the membranes (Table 3). The addition of 0.5%
GO resulted in an increase of one order of magnitude in the Young’s
modulus and hardness, which is in good agreement with other GO/
polymer composites [42]. The substantial increase in the mechanical
property can be attributed to the strong hydrogen bonding between the
GO and PBI [34], which is also evident from the increase in the viscosity
of the dope solution. Although an increase of the GO content to 1%
improved the mechanical properties, a further increase to 1.5% resulted
in a moderate decrease in both Young’s modulus and the hardness,
which could indicate that GO underwent agglomeration. The PDA
deposition resulted in similar increase in the Young’s moduli and the
1:00
hardness for both MPBI=PDA and MGO PBI=PDA , which is significantly higher
1:00
than those of MPBI , but lower than those of MGO
PBI . It proves that the
PDA dominates the surface properties of the membranes. Furthermore,
1:00
MGO
PBI showed high thermal stability even at 600 � C, which is
A. Alammar et al. Journal of Membrane Science 603 (2020) 118007

Fig. 5. Visible-light microscope reflection images (left), cross-sectional TEM images (right), as well as XPS wide spectra and deconvoluted C1s peaks of (a) MPBI , (b)
MGO
PBI
​ 1:00
and (c) MrGO
PBI
​ 1:00
membranes; b-i) The inset TEM image of GO; b-ii) The inset high magnification images of the GO flakes.

Table 3
Dope solution viscosity, membrane thickness and swelling ratio as well as membrane surface properties including Young’s modulus, hardness contact angle and
roughness. n.a. ¼ not applicable.
Membrane Viscosity (mPas) Thickness (μm) Swelling ratio (%) Surface properties

Young’s modulus (GPa) Hardness (GPa) Water contact angle (� ) Roughness, Rq (nm)

MPBI 3976 � 9 38.8 � 0.6 5.71 � 0.32 0.1 � 0.03 0.2 � 0.06 55.5 � 1.6 16.5 � 1.0
MGO
PBI
​ 0:50 4640 � 7 47.5 � 0.9 6.82 � 0.24 1.4 � 0.3 1.4 � 0.6 51.5 � 1.7 30.1 � 2.7
MGO
PBI
​ 0:75 5082 � 13 49.5 � 0.6 7.35 � 0.28 2.1 � 0.8 1.3 � 0.6 49.1 � 1.6 32.9 � 3.8
MGO
PBI
​ 1:00 5656 � 11 45.2 � 1.0 8.17 � 0.21 3.1 � 0.9 2.1 � 0.8 45.8 � 0.8 44.2 � 2.8
MGO
PBI
​ 1:50 6684 � 17 49.9 � 0.8 7.44 � 0.26 1.9 � 0.5 1.7 � 0.6 51.0 � 0.8 30.7 � 2.5
MGO ​ 1:00
PBI=PDA
n.a. 53.3 � 1.2 7.12 � 0.24 1.6 � 0.4 1.1 � 0.4 59.5 � 1.0 23.9 � 2.9
MPBI=PDA n.a. 44.5 � 1.6 6.02 � 0.34 1.5 � 0.3 1.3 � 0.1 61.3 � 1.9 13.2 � 3.0
MrGO
PBI
​ 1:00 5128 � 15 45.3 � 1.1 6.26 � 0.22 1.7 � 0.3 1.4 � 0.4 50.0 � 1.2 17.2 � 1.2

consistent with the pristine PBI membrane (Fig. S15). The pH stability of
the membrane, covering the wide range of pH ¼ 1–11, was also
demonstrated through a long-term stability test up to 30 days (Fig. S18).
The improved mechanical properties and the excellent thermal and pH
stability of the PBI/GO nanocomposite membrane show that these
membranes are suitable for harsh conditions, including high pressure,
high temperature and high concentrations of acids or bases.
The water contact angle of the membranes monotonously decreased
1:00
from 55.5� for the pristine PBI to 45.8� for the MGO as the GO loading
PBI
increased due to the increase in the oxygen-containing functional groups
present in the GO (Table 3, Fig. S16) [43,44]. However, a further in­
crease in the GO loading resulted in a 10% increase in the water contact
angle for MGO PBI
1:50
. This phenomenon can be attributed to the increased
hydrophobicity of the membrane due to GO aggregation at a higher
loading content [45]. The reduction of graphene oxide, MGO 1:00
to
PBI
rGO 1:00
MPBI , somewhat increased the water contact angle due to the more
1:00
hydrophobic nature of rGO than that of GO. MPBI=PDA and MGO
PBI=PDA
revealed approximately a 60� water contact angle, which is consistent
with previously reported results for the PDA coating layer [46,47].
Compared to the results of water contact angles, the inversed trend for
the crude oil contact angle of the membranes was observed, which
supports the reliability of the trend of water contact angle (Fig. S17).
The trend in the change in roughness as a result of varied membrane

6
A. Alammar et al. Journal of Membrane Science 603 (2020) 118007

1:00 1:00
Fig. 6. SEM images of (a–c) cross-section and (d–f) top surface and (h–j) AFM height images of (a,d,h) MPBI , (b,e,i) MGO
PBI and (c,f,j) MGO
PBI=PDA membranes (also refer
to Figs. S12–13).

composition was the opposite as for the water contact angle (Table 3).
The roughness of the PBI/GO membranes monotonously increased from
16.5 nm for the pristine PBI up to 44.2 nm for MGO PBI
1:00
, due to the
accelerated rate of phase inversion by the presence of the hydrophilic
GO [48,49] Nonetheless, the roughness decreased at the highest GO
loading due to its probable agglomeration. The PDA coating and the
replacement of GO with rGO both resulted in a smoother surface.
Therefore, the improved hydrophilicity of the PBI/GO membranes can
enhance the water permeance [50], because the more hydrophilic sur­
face can adsorb more water and create a hydration layer, preventing oil
interaction, which ultimately minimizes any fouling and improves
oil-in-water emulsion separation [51].
3.2. Produced water treatment
The performance of the membranes was assessed in oil-in-water
emulsions in the presence and absence of salts (Fig. 7). The permeance
and oil-removal efficiency for all the membranes showed a similar ten­
dency against both feed solutions with respect to the changes in the
composition of the materials (Fig. 7a). The permeance and oil-removal
efficiency increased monotonously with increasing GO loading up to
1 wt%, reaching approx. 91 L m 2 h 1 bar 1 and 100%, respectively.
This observed tendency is in line with previous reports on graphene-
based polymer membranes [20]. The hydrophilic surface facilitates
the transport of water into the pores, while the higher roughness
(Table 3) provides a higher surface area, which results in the enhance­
ment of water permeance. A further increase in the GO loading to 1.5 wt

Fig. 7. (a) Permeance and removal efficiency of oil-in-water emulsions with (striped bars) and without (solid bars) salt. (b) Long-term permeance and removal
1:00
efficiency for MGO
PBI during produced water treatment. The split ratio (λ) was kept constant by adjusting the pressure.

7
A. Alammar et al. Journal of Membrane Science 603 (2020) 118007

% maintained the 100% removal efficiency but resulted in an approx. the emulsions, and therefore the permeance was normalized by the
5% decrease in the permeance. average droplet size of the emulsions. The MGO1:00
PBI membrane demon­
The oil-removal efficiency of the PDA-coated membranes MPBI=PDA strated excellent oil-removal efficiency at the expense of permeance.
and MGO ​ 1:00
PBI=PDA is comparable to the corresponding uncoated membrane
Only two membranes reported in the literature showed higher per­
meance, although we note that i) they were not tested with produced
(MGOPBI
​ 1:00
). However, the permeance was reduced to approx. 50–60 L
water (i.e., oil-in-water emulsion without salt), and ii) neither the long-
m 2 h 1 bar 1 due to the PDA growth on the membrane surface. The
term stability nor the antifouling properties and cleaning were assessed.
coating layer is known to increase the hydraulic resistance, and the
membrane performance is dominantly governed by the thicker and more
compact PDA layer, which can lead to lower permeance [52]. The 3.3. Antifouling properties
reduction of the GO resulted in membranes with inferior performance
with respect to both oil-removal efficiency and permeance, since the The membrane fouling that forms on the surface during operation
rGO increased the hydrophobicity of the membrane. increases the hydraulic resistance and subsequently leads to flux decline.
The membrane with 1 wt% GO showed superior performance This fouling eventually affects the membrane life-time and influences
compared to the rest of the series, and therefore we tested this mem­ the maintenance costs and efficiency. Two methodologies were used to
brane for long-term produced water treatment up to 10 L (Fig. 7b). The investigate membrane fouling: i) FRR (Eq. (5)) was calculated for the
effect of split ratio on the membrane performance was investigated. The membranes after produced water treatment (Fig. 8), and ii) elemental
oil-removal efficiency (approx. 100%) was constant during the filtration analysis by EDX was performed on the surface of the fouled membranes
for all split ratios. However, an initial decline in the permeance, over the before and after cleaning (Fig. 10). The permeance of the pure water
first 2 L of the filtration, was observed in all cases. The lower the split gradually decreased after each fouling-cleaning filtration cycle due to
ratio, i.e., a higher relative permeate flow rate, the more pronounced the fouling. After three cycles, the FRR values for the MPBI , MGO 1:00
and
PBI
decline in permeance. At 90% and 99% permeate flow rates, the per­ MGO 1:00
membranes were found to be 60.5%, 91.8% and 72.6%,
meance decreased by 27% and 92% corresponding to steady-state values
PBI=PDA
respectively (Table S6). MGO ​ 1:00
showed the best fouling resistance
of 62 and 7 L m 2 h 1 bar 1, respectively. Consequently, the trade-off PBI
along with a high FRR, which can be explained by the presence of GO
between the steady-state permeance and the split ratio was found to
and the hydrophilic nature of the membrane (Table 3). The more hy­
be a crucial factor to consider during produced water treatment. The
1:00
stable separation performance during the long-term experiment drophilic surface of MGO
PBI can adsorb water molecules and thus forms
demonstrated the first step in the assessment of industrial applicability a hydration layer on its surface, which helps to prevent hydrophobic
of the nanocomposite membranes. foulants from attaching to and detaching from the membrane surface,
contributing to the high FRR.
The regeneration of MGO PBI
​ 1:00
was studies through filtration-cleaning
The salt scaling and descaling, which mainly consisted of sodium,
cycles (Fig. 8). The removal efficiency of oil was found to be constant at
calcium and chlorine ions, on the membrane surface before and after
100% during the experiments. The pure water permeance was measured
cleaning by the hydraulic force was confirmed through SEM images and
before each cycle, which showed a diminishing 4.9%, 2.3% and 1.2%
EDX elemental analysis (Fig. 10). After cleaning with water, some salts
decrease over the four cycles. The permeance decline during the start-up
remained on the membrane surface of MPBI , whereas the surfaces of
period was different for each cycle, whereas the same steady-state per­
1:00 1:00
meance of approx. 64 L m 2 h 1 bar 1 was achieved each time. The MGO
PBI and MGO
PBI=PDA were easily cleaned. The more hydrophilic sur­
1:00 1:00
initial decline in permeance, as well as the standard deviation of per­ face of MGO
PBI and MGO
PBI=PDA resulted in weaker interactions between
meance decreased after each cycle. the membrane surface and the salts [53], which in turn facilitated the
The best performing membrane ðMGO1:00 PBI Þ was compared with descaling process.
various published membranes for oil-in-water emulsion separations A long-term fouling study was performed by immersing the mem­
including polymeric, inorganic and composite membranes (Fig. 9, branes in tap water and taking samples for SEM analysis after 45 and 180
Table S7). The effectiveness of separation depends largely on the size of days. Bacteria growth on the MPBI membrane surface after 45 days was
observed, which resulted in extensive biofouling covering of the surface
of the membrane after 180 days (Fig. 11, Fig. S19). On the contrary,
virtually clean surfaces on MGO PBI
​ 1:00
and MGO ​ 1:00
PBI=PDA membranes were

1:00
Fig. 8. Membrane cleaning study for MGO PBI . Produced water was used during Fig. 9. Comparison of separation performance and the trade-off between oil-
each cycle. Before each cycle, 0.5 L pure water was used to clean the mem­ removal efficiency and permeance over emulsion droplet size. The inset
brane. Pure water permeance and flow recovery ratio (FRR) are reported for the shows the duration of the different studies. Refer to Table S7 in the Supple­
final 0.05 L filtrate of the washing cycle. mentary Information for the data and references.

8
A. Alammar et al. Journal of Membrane Science 603 (2020) 118007

1:00 1:00
Fig. 10. SEM-EDX elemental analysis of (a) MPBI , (b) MGO
PBI and (c) MGO
PBI=PDA after filtering 0.5 L of produced water (Fouled) and after washing it with pure water
(Cleaned). The y-axis represents count of elements and the x-axis represents the energy in keV.

observed during the six-month study. In line with previous observations
on GO [54], our results further demonstrate that the incorporation of GO
and PDA in polymer membranes can effectively inhibit the initial
attachment of bacteria and suppress bacterial growth on the poly­
benzimidazole supports.
4. Conclusion
Nanocomposite membranes comprising of polybenzimidazole (PBI),
graphene oxide (GO), reduced GO (rGO), with and without polydop­
amine (PDA) coating were developed using phase inversion and dip-
coating techniques. The introduction of as low as 0.5–1.5 wt% GO
into the polymer matrix resulted in enhanced membrane performance in
multiple ways. Compared to the pristine PBI, the permeance increased
by 17%, reaching a maximum of 91 L m 2 h 1 bar 1, while the oil-
removal efficiency increased from 80% to 100%. The water flux recov­
ery (FRR) was maintained above 90% over four produced water filtra­
tion/cleaning cycles. The retentate/permeate split ratio was found to
have no effect on oil-removal, but this ratio determined the steady-state
permeance during long-term produced water treatment. Both the GO
and PDA improved the antifouling properties of the membranes, which
we evaluated through cleaning cycles and SEM-EDX analysis. The

1:00 1:00
Fig. 11. SEM images of the membrane surface for (a, d) MPBI , (b, e) MGO
PBI and (c, f) MGO
PBI=PDA at 1600x magnification. The membranes were soaked in tap water for
(a–c) 45 days and (d–f) 180 days prior to analysis.

9
A. Alammar et al.
observed changes in membrane performance was attributed to the
changes in hydrophilicity and surface topology of the membranes. The
results successfully demonstrate the potential of both PDA-coated and
uncoated PBI-GO nanocomposite membranes in the treatment of oil-in-
water emulsions and produced water.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
CRediT authorship contribution statement
Abdulaziz Alammar: Conceptualization, Methodology, Validation,
Formal analysis, Investigation, Data curation, Writing - original draft,
Visualization. Sang-Hee Park: Conceptualization, Data curation,
Writing - review & editing, Visualization. Craig J. Williams: Data
curation. Brian Derby: Supervision, Resources. Gyorgy Szekely:
Conceptualization, Resources, Methodology, Writing - review & editing,
Visualization, Supervision, Project administration, Funding acquisition.
Acknowledgment
The graphical abstract, Figs. 1 and 2 were created by Heno Hwang,
scientific illustrator at King Abdullah University of Science and Tech­
nology (KAUST). The authors would also like to thank Rachid Sougrat
(KAUST) and Hai Anh Le Phuong (University of Manchester) for their
assistance with the TEM and AFM measurements, respectively. The PhD
scholarship from Saudi Aramco is gratefully acknowledged (AA). The
authors wish to acknowledge the support provided by the Henry Royce
Institute in the use of the nanoindentation equipment (Grant ref EP/
R010145/1).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.memsci.2020.118007.
References
[1] SDG 6 Synthesis Report on Water and Sanitation, 2018. New York, https://sustaina
bledevelopment.un.org/content/documents/19901SDG6_SR2018_web_3.pdf.
[2] M.A. Al-Ghouti, M.A. Al-Kaabi, M.Y. Ashfaq, D.A. Da’na, Produced water
characteristics, treatment and reuse: a review, J. Water Process Eng. 28 (2019)
222–239, https://doi.org/10.1016/j.jwpe.2019.02.001.
[3] J. Xu, N.M. Srivatsa Bettahalli, S. Chisca, M.K. Khalid, N. Ghaffour, R. Vilagines, S.
P. Nunes, Polyoxadiazole hollow fibers for produced water treatment by direct
contact membrane distillation, Desalination 432 (2018) 32–39, https://doi.org/
10.1016/j.desal.2017.12.014.
[4] M. Ebrahimi, S. Kerker, O. Schmitz, A.A. Schmidt, P. Czermak, Evaluation of the
fouling potential of ceramic membrane configurations designed for the treatment
of oilfield produced water, Separ. Sci. Technol. 53 (2018) 349–363, https://doi.
org/10.1080/01496395.2017.1386217.
[5] R.J.R. Veil, A. John, Puder G. Markus, Elcock Deobrah, A White Paper Describing
Produced Water from Production of Crude Oil, Natural Gas, and Coal Bed Methane,
2004. https://publications.anl.gov/anlpubs/2004/02/49109.pdf.
[6] M. Nasiri, I. Jafari, Produced water from oil-gas plants: a short review on
challenges and opportunities, Period. Polytech. - Chem. Eng. 61 (2017) 73–81,
https://doi.org/10.3311/PPch.8786.
[7] Bp, Bp Statistical Review of World Energy, 2018. https://www.bp.com/content/
dam/bp/business-sites/en/global/corporate/pdfs/energy-economics/statistical-re
view/bp-stats-review-2018-full-report.pdf.
[8] X. Wei, S. Zhang, Y. Han, F.A. Wolfe, Treatment of petrochemical wastewater and
produced water from oil and gas, Water Environ. Res. (2019) 1–9, https://doi.org/
10.1002/wer.1172.
[9] Y. Zhu, D. Wang, L. Jiang, J. Jin, Recent progress in developing advanced
membranes for emulsified oil/water separation, NPG Asia Mater. 6 (2014), e101,
https://doi.org/10.1038/am.2014.23.
[10] J.M. Dickhout, J. Moreno, P.M. Biesheuvel, L. Boels, R.G.H. Lammertink, W.M. de
Vos, Produced water treatment by membranes: a review from a colloidal
perspective, J. Colloid Interface Sci. 487 (2017) 523–534, https://doi.org/
10.1016/j.jcis.2016.10.013.
10
Journal of Membrane Science 603 (2020) 118007
[11] G. Ríos, C. Pazos, J. Coca, Destabilization of cutting oil emulsions using inorganic
salts as coagulants, Colloids Surfaces A Physicochem. Eng. Asp. 138 (1998)
383–389, https://doi.org/10.1016/S0927-7757(97)00083-6.
[12] C.O. Igwe, A. Al Saadi, S.E. Ngene, Optimal options for treatment of produced
water in offshore petroleum platforms, J. Pollut. Eff. Control. 1 (2013) 1–5,
https://doi.org/10.4172/jpe.1000102.
[13] J. Zhu, J. Wang, A.A. Uliana, M. Tian, Y. Zhang, Y. Zhang, A. Volodin, K. Simoens,
S. Yuan, J. Li, J. Lin, K. Bernaerts, B. Van Der Bruggen, Mussel-inspired
architecture of high-flux loose nanofiltration membrane functionalized with
antibacterial reduced graphene oxide-copper nanocomposites, ACS Appl. Mater.
Interfaces 9 (2017) 28990–29001, https://doi.org/10.1021/acsami.7b05930.
[14] A.K. Kota, G. Kwon, W. Choi, J.M. Mabry, A. Tuteja, Hygro-responsive membranes
for effective oil-water separation, Nat. Commun. 3 (2012) 1025, https://doi.org/
10.1038/ncomms2027.
[15] M. Padaki, R. Surya Murali, M.S. Abdullah, N. Misdan, A. Moslehyani, M.
A. Kassim, N. Hilal, A.F. Ismail, Membrane technology enhancement in oil-water
separation. A review, Desalination 357 (2015) 197–207, https://doi.org/10.1016/
j.desal.2014.11.023.
[16] R.R. Nair, H.A. Wu, P.N. Jayaram, I.V. Grigorieva, A.K. Geim, Unimpeded
permeation of water through helium-leak-tight graphene-based membranes,
Science 335 (2012) 442–444, https://doi.org/10.1126/science.1211694.
[17] H. Huang, Z. Song, N. Wei, L. Shi, Y. Mao, Y. Ying, L. Sun, Z. Xu, X. Peng, Ultrafast
viscous water flow through nanostrand-channelled graphene oxide membranes,
Nat. Commun. 4 (2013) 1–9, https://doi.org/10.1038/ncomms3979.
[18] R.K. Gupta, G.J. Dunderdale, M.W. England, A. Hozumi, Oil/water separation
techniques: a review of recent progresses and future directions, J. Mater. Chem. A.
5 (2017) 16025–16058, https://doi.org/10.1039/C7TA02070H.
[19] Z. Liu, W. Wu, Y. Liu, C. Qin, M. Meng, Y. Jiang, J. Qiu, A mussel inspired highly
stable graphene oxide membrane for efficient oil- in-water emulsions separation,
Separ. Purif. Technol. 199 (2018) 37–46, https://doi.org/10.1016/j.
seppur.2018.01.041.
[20] F. Fei, L. Cseri, G. Szekely, C.F. Blanford, Robust covalently cross-linked
polybenzimidazole/graphene oxide membranes for high-flux organic solvent
nanofiltration, ACS Appl. Mater. Interfaces 10 (2018) 16140–16147, https://doi.
org/10.1021/acsami.8b03591.
[21] K.K.H. De Silva, H.H. Huang, M. Yoshimura, Progress of reduction of graphene
oxide by ascorbic acid, Appl. Surf. Sci. 447 (2018) 338–346, https://doi.org/
10.1016/j.apsusc.2018.03.243.
[22] Y. Peng, Z. Guo, Recent advances in biomimetic thin membranes applied in
emulsified oil/water separation, J. Mater. Chem. A. 4 (2016) 15749–15770,
https://doi.org/10.1039/c6ta06922c.
[23] F. Fei, H.A. Le Phuong, C.F. Blanford, G. Szekely, Tailoring the performance of
organic solvent nanofiltration membranes with biophenol coatings, ACS Appl.
Polym. Mater. 1 (2019) 452–460, https://doi.org/10.1021/acsapm.8b00161.
[24] X. Zhu, A. Dudchenko, X. Gu, D. Jassby, Surfactant-stabilized oil separation from
water using ultrafiltration and nanofiltration, J. Membr. Sci. 529 (2017) 159–169,
https://doi.org/10.1016/j.memsci.2017.02.004.
[25] L. Zhang, L. Cheng, H. Wu, T. Yoshioka, H. Matsuyama, One-step fabrication of
robust and anti-oil-fouling aliphatic polyketone composite membranes for
sustainable and efficient filtration of oil-in-water emulsions, J. Mater. Chem. A. 6
(2018) 24641–24650, https://doi.org/10.1039/c8ta10071c.
[26] L. Zhang, Y. Lin, L. Cheng, Z. Yang, H. Matsuyama, A comprehensively fouling- and
solvent-resistant aliphatic polyketone membrane for high-flux filtration of difficult
oil-in-water micro- and nanoemulsions, J. Membr. Sci. 582 (2019) 48–58, https://
doi.org/10.1016/j.memsci.2019.03.090.
[27] Y.Z. Liu, C.M. Chen, Y.F. Li, X.M. Li, Q.Q. Kong, M.Z. Wang, Crumpled reduced
graphene oxide by flame-induced reduction of graphite oxide for supercapacitive
energy storage, J. Mater. Chem. A. 2 (2014) 5730–5737, https://doi.org/10.1039/
c3ta15082h.
[28] P. Cui, J. Lee, E. Hwang, H. Lee, One-pot reduction of graphene oxide at subzero
temperatures, Chem. Commun. 47 (2011) 12370–12372, https://doi.org/10.1039/
c1cc15569e.
[29] I.K. Moon, J. Lee, R.S. Ruoff, H. Lee, Reduced graphene oxide by chemical
graphitization, Nat. Commun. 1 (2010) 73–79, https://doi.org/10.1038/
ncomms1067.
[30] F. Chu, B. Lin, B. Qiu, Z. Si, L. Qiu, Z. Gu, J. Ding, F. Yan, J. Lu, Polybenzimidazole/
zwitterion-coated silica nanoparticle hybrid proton conducting membranes for
anhydrous proton exchange membrane application, J. Mater. Chem. 22 (2012)
18411–18417, https://doi.org/10.1039/c2jm32787b.
[31] J.W. Jung, S.K. Kim, J.C. Lee, Preparation of polybenzimidazole/lithium
hydrazinium sulfate composite membranes for high-temperature fuel cell
applications, Macromol. Chem. Phys. 211 (2010) 1322–1329, https://doi.org/
10.1002/macp.200900712.
[32] R. Kumar, C. Xu, K. Scott, Graphite oxide/Nafion composite membranes for
polymer electrolyte fuel cells, RSC Adv. 2 (2012) 8777–8782, https://doi.org/
10.1039/c2ra20225e.
[33] G.S. Lai, W.J. Lau, P.S. Goh, A.F. Ismail, Y.H. Tan, C.Y. Chong, R. Krause-Rehberg,
S. Awad, Tailor-made thin film nanocomposite membrane incorporated with
graphene oxide using novel interfacial polymerization technique for enhanced
water separation, Chem. Eng. J. 344 (2018) 524–534, https://doi.org/10.1016/j.
cej.2018.03.116.
[34] Y. Wang, Z. Shi, J. Fang, H. Xu, J. Yin, Graphene oxide/polybenzimidazole
composites fabricated by a solvent-exchange method, Carbon N. Y. 49 (2011)
1199–1207, https://doi.org/10.1016/j.carbon.2010.11.036.
[35] S. Darvishmanesh, J.C. Jansen, F. Tasselli, E. Tocci, P. Luis, J. Degr�eve, E. Drioli,
B. Van der Bruggen, Novel polyphenylsulfone membrane for potential use in
A. Alammar et al.
solvent nanofiltration, J. Membr. Sci. 379 (2011) 60–68, https://doi.org/10.1016/
j.memsci.2011.05.045.
[36] A.K. Hołda, B. Aernouts, W. Saeys, I.F.J. Vankelecom, Study of polymer
concentration and evaporation time as phase inversion parameters for polysulfone-
based SRNF membranes, J. Membr. Sci. 442 (2013) 196–205, https://doi.org/
10.1016/j.memsci.2013.04.017.
[37] M. Xie, J. Wang, X. Wang, M. Yin, C. Wang, D. Chao, X. Liu, The high performance
of polydopamine-coated electrospun poly(ether sulfone) nanofibrous separator for
lithium-ion batteries, Macromol. Res. 24 (2016) 965–972, https://doi.org/
10.1007/s13233-016-4140-3.
[38] C. Cao, L. Tan, W. Liu, J. Ma, L. Li, Polydopamine coated electrospun poly
(vinyldiene fluoride) nanofibrous membrane as separator for lithium-ion batteries,
J. Power Sources 248 (2014) 224–229, https://doi.org/10.1016/j.
jpowsour.2013.09.027.
[39] W. Lu, L. Qiao, Q. Dai, H. Zhang, X. Li, Solvent treatment: the formation
mechanism of advanced porous membranes for flow batteries, J. Mater. Chem. A. 6
(2018) 15569–15576, https://doi.org/10.1039/c8ta05349a.
[40] K. Divya, D. Rana, S. Alwarappan, M.S.S. Abirami Saraswathi, A. Nagendran,
Investigating the usefulness of chitosan based proton exchange membranes tailored
with exfoliated molybdenum disulfide nanosheets for clean energy applications,
Carbohydr. Polym. 208 (2019) 504–512, https://doi.org/10.1016/j.
carbpol.2018.12.092.
[41] W. Dai, Y. Shen, Z. Li, L. Yu, J. Xi, X. Qiu, SPEEK/Graphene oxide nanocomposite
membranes with superior cyclability for highly efficient vanadium redox flow
battery, J. Mater. Chem. A. 2 (2014) 12423–12432, https://doi.org/10.1039/
c4ta02124j.
[42] S. Lee, B.G. Choi, D. Choi, H.S. Park, Nanoindentation of annealed Nafion/
sulfonated graphene oxide nanocomposite membranes for the measurement of
mechanical properties, J. Membr. Sci. 451 (2014) 40–45, https://doi.org/10.1016/
j.memsci.2013.09.038.
[43] F. Jin, W. Lv, C. Zhang, Z. Li, R. Su, W. Qi, Q.H. Yang, Z. He, High-performance
ultrafiltration membranes based on polyethersulfone- graphene oxide composites,
RSC Adv. 3 (2013) 21394–21397, https://doi.org/10.1039/c3ra42908c.
[44] M.J. Park, S. Phuntsho, T. He, G.M. Nisola, L.D. Tijing, X.M. Li, G. Chen, W.
J. Chung, H.K. Shon, Graphene oxide incorporated polysulfone substrate for the
fabrication of flat-sheet thin-film composite forward osmosis membranes,
J. Membr. Sci. 493 (2015) 496–507, https://doi.org/10.1016/j.
memsci.2015.06.053.
11
Journal of Membrane Science 603 (2020) 118007
[45] A.K. Shukla, J. Alam, M. Alhoshan, L.A. Dass, M.R. Muthumareeswaran,
Development of a nanocomposite ultrafiltration membrane based on
polyphenylsulfone blended with graphene oxide, Sci. Rep. 7 (2017) 41976,
https://doi.org/10.1038/srep41976.
[46] C. Cheng, S. Li, W. Zhao, Q. Wei, S. Nie, S. Sun, C. Zhao, The hydrodynamic
permeability and surface property of polyethersulfone ultrafiltration membranes
with mussel-inspired polydopamine coatings, J. Membr. Sci. 417–418 (2012)
228–236, https://doi.org/10.1016/j.memsci.2012.06.045.
[47] Q. Wei, F. Zhang, J. Li, B. Li, C. Zhao, Oxidant-induced dopamine polymerization
for multifunctional coatings, Polym. Chem. 1 (2010) 1430–1433, https://doi.org/
10.1039/c0py00215a.
[48] X. Chang, Z. Wang, S. Quan, Y. Xu, Z. Jiang, L. Shao, Exploring the synergetic
effects of graphene oxide (GO) and polyvinylpyrrodione (PVP) on poly
(vinylylidenefluoride) (PVDF) ultrafiltration Membrane performance, Appl. Surf.
Sci. 316 (2014) 537–548, https://doi.org/10.1016/j.apsusc.2014.07.202.
[49] S. Zinadini, A.A. Zinatizadeh, M. Rahimi, V. Vatanpour, H. Zangeneh, Preparation
of a novel antifouling mixed matrix PES membrane by embedding graphene oxide
nanoplates, J. Membr. Sci. 453 (2014) 292–301, https://doi.org/10.1016/j.
memsci.2013.10.070.
[50] S. Xia, M. Ni, Preparation of poly(vinylidene fluoride) membranes with graphene
oxide addition for natural organic matter removal, J. Membr. Sci. 473 (2015)
54–62, https://doi.org/10.1016/j.memsci.2014.09.018.
[51] S.J. Maguire-Boyle, A.R. Barron, A new functionalization strategy for oil/water
separation membranes, J. Membr. Sci. 382 (2011) 107–115, https://doi.org/
10.1016/J.MEMSCI.2011.07.046.
[52] Y. Li, Y. Su, X. Zhao, X. He, R. Zhang, J. Zhao, X. Fan, Z. Jiang, Antifouling, high-
flux nanofiltration membranes enabled by dual functional polydopamine, ACS
Appl. Mater. Interfaces 6 (2014) 5548–5557, https://doi.org/10.1021/
am405990g.
[53] D. Emadzadeh, W.J. Lau, T. Matsuura, N. Hilal, A.F. Ismail, The potential of thin
film nanocomposite membrane in reducing organic fouling in forward osmosis
process, Desalination 348 (2014) 82–88, https://doi.org/10.1016/j.
desal.2014.06.008.
[54] X. Huang, K.L. Marsh, B.T. McVerry, E.M.V. Hoek, R.B. Kaner, Low-fouling
antibacterial reverse osmosis membranes via surface grafting of graphene oxide,
ACS Appl. Mater. Interfaces 8 (2016) 14334–14338, https://doi.org/10.1021/
acsami.6b05293.