Accepted Manuscript Photocatalytic nanofiber-coated alumina hollow fiber membranes for highly efficient oilfield produced water treatment Nur Hashimah Alias, Juhana Jaafssar, Sadaki Samitsu, T. Matsuura, A.F. Ismail, M.H.D. Othman, Mukhlis A. Rahman, N.H. Othman, N. Abdullah, S.H. Paiman, N. Yusof, F. Aziz PII: DOI: Reference:
S1385-8947(18)32179-X https://doi.org/10.1016/j.cej.2018.10.217 CEJ 20280
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
13 August 2018 21 October 2018 28 October 2018
Please cite this article as: N.H. Alias, J. Jaafssar, S. Samitsu, T. Matsuura, A.F. Ismail, M.H.D. Othman, M.A. Rahman, N.H. Othman, N. Abdullah, S.H. Paiman, N. Yusof, F. Aziz, Photocatalytic nanofiber-coated alumina hollow fiber membranes for highly efficient oilfield produced water treatment, Chemical Engineering Journal (2018), doi: https://doi.org/10.1016/j.cej.2018.10.217
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.
1
Photocatalytic nanofiber-coated alumina hollow fiber membranes for highly efficient oilfield
2
produced water treatment
3 4
Nur Hashimah Aliasa,b, Juhana Jaafara*, Sadaki Samitsuc, T. Matsuurad, A. F. Ismaila, M. H. D.
5
Othmana, Mukhlis A Rahmana, N. H. Othmanb, N. Abdullaha, S. H. Paimana, N. Yusofa, F. Aziza
6 7
a
8
81310 Skudai, Johor, Malaysia.
9
b
Advanced Membrane Technology Research Center (AMTEC), Universiti Teknologi Malaysia,
Department of Oil and Gas Engineering, Faculty of Chemical Engineering, Universiti Teknologi
10
MARA, 40450 Shah Alam, Selangor, Malaysia.
11
c
12
Integrated System (MaDIS), National Institute for Materials Science (NIMS), 1-2-1 Sengen,
13
Tsukuba, Ibaraki 305-0047, Japan.
14
d
15
University of Ottawa, 161 Louis Pasteur St, Ottawa, ON K1N 6N5, Canada.
Data-driven Polymer Design Group, Research and Services Division of Materials Data and
Industrial Membrane Research Laboratory, Department of Chemical and Biological Engineering,
16 17
Corresponding author:
[email protected]
18 19
Abstract
20 21
Cost-effective purification technology of oilfield produced water (OPW) is becoming a
22
global challenge for future petroleum exploration and production industry. Energy-efficient
23
operation of membrane separation is potentially promising. However, severe fouling problem of
24
oil droplets demands new robust and fouling-resistant membranes with high permeability and
25
rejection efficiency. Here, we propose a photocatalytic nanofiber-coated inorganic hollow fiber
26
membrane suitable for OPW treatment. The membrane was fabricated by coating polyacrylonitrile
27
(PAN) nanofiber incorporated with photocatalytic graphitic carbon nitride (GCN) on an alumina
28
(Al2O3) hollow fiber membrane. While the highly porous coating made of smooth hydrophilic
29
nanofibers facilitated water permeation, the coating effectively captured oil droplets in its opening,
30
resulting in a better rejection efficiency of oil contaminants. Its sparse mesh morphology prevented
31
oil contaminants to form dense fouling film on the membrane surface and maintained high
32
permeate flux even after 180 min filtration. The best permeate flux of 640 L∙m−2∙h−1 and oil
33
rejection percentage of 99% were recorded for 180 min crossflow filtration of OPW at 2 bar along
34
with the highest pure water flux of 816 L∙m−2∙h−1. The photocatalytic activity of GCN enabled the
35
coating to degrade the captured oil contaminants under UV irradiation, demonstrating permeate
36
flux of 577 L∙m−2∙h−1 and oil rejection of 97% after three cycles of 180 min filtration. The excellent
37
fouling resistance and cleaning performances of the membrane are considerably beneficial for a
38
long-term repeated filtration operation. This work will motivate researchers to develop nanofiber-
39
coated hollow fiber membranes for future membrane separation technology.
40 41
Keywords: Graphitic carbon nitride; PAN nanofiber; Alumina hollow fiber membrane; Coating;
42
Filtration; Oilfield produced water
43 44
45
1.
Introduction
46 47
A massive amount of oilfield produced water (OPW) is co-produced during petroleum
48
production and exploration, which accounts for 80%–95% of the production volume collected
49
from the production wells. Much attention has been given to reduce the pollution of OPW such as
50
aliphatic hydrocarbons, heavy aromatic compounds, alkylated phenols, and added production
51
chemicals because the contaminants are potentially hazardous to the environment [1,2]. Various
52
physical treatments such as hydrocyclone, floatation, centrifugation, evaporation, and extraction
53
have been traditionally used as a primary treatment to minimize the concentration of oil in OPW
54
[3]. Despite high efficiency of primary treatments, the resultant OPW still contains trace oil
55
contaminants, which is not allowed to be discharged according to stringent environmental
56
regulations to date [4].
57
Membrane separation is another technique that can be applied to remove trace
58
contaminants in water. Although current membrane processes generally consume more energy and
59
are more expensive compared with conventional biological and physicochemical treatment
60
processes, they have great potentials on improving their performance by developing advanced
61
membranes and filtration processes [5,6]. In principle, membrane separation systems have many
62
attractive features such as energy-efficient separation without phase change, small footprint, and
63
easy installation and maintenance [7,8]. Membrane separation for OPW treatment, however, needs
64
more improvement on robust membranes due to the complex sticky composition and high
65
temperature of OPW, which frequently degrade common polymeric membranes. Inorganic
66
membranes are the promising candidates for OPW treatment due to the excellent chemical and
67
thermal stabilities. However, the development of inorganic membranes is still challenging in terms
68
of high permeability and rejection, fouling resistance, and easy cleaning of membrane fouling.
69
Surface modification of membranes has attracted much attention for reducing membrane fouling,
70
which involves chemical modification to change surface hydrophilicity or hydrophobicity [9],
71
control surface zeta potential [10], plasma treatment [11], UV irradiation [12], surface grafting
72
[13], and surface coating [14]. Interestingly, a recent study demonstrated that electrospun
73
nanofiber coating enhanced fouling resistance and permeability for a flat sheet polymer membrane
74
[15]. The coating seems to be applicable to an inorganic hollow fiber membrane as actually
75
demonstrated in this study [16]. Hollow fiber membranes are probably more attractive for the
76
practical treatment of OPW because they are suitable for constructing a microfiltration membrane
77
module which is easily up-scaled compared with that of flat-sheet membranes [17]. Surface
78
modification of membranes has attracted much attention for reducing membrane fouling. In OPW
79
purification process using microfiltration membranes, a considerable amount of oil droplets
80
adheres on the membrane surface while water molecules pass through the membrane.
81
Conventional membranes, which have smooth surface morphology and hydrophobic property (i.e.,
82
oleophilic), suffer from severe fouling because a dense oily layer usually covers the whole
83
membrane surface. The fouling layer seriously degrades permeate flux in addition to the reduction
84
of separation efficiency, which is challenging for a long-term operation of microfiltration process.
85
Membrane cleaning performance is another important issue in membrane development for
86
OPW treatment. In addition to conventional physical and chemical washing processes,
87
photodegradation of fouling contaminants has been considered as a promising technique in
88
membrane cleaning process. Various polymeric and inorganic membranes incorporating
89
photocatalyst into their matrices have demonstrated good membrane cleaning performance using
90
photodegradation under light irradiation. Unfortunately, a low amount of photocatalyst was
91
dispersed on the membrane top surface, leading to low photodegradation properties [18,19]. A
92
higher dosage of photocatalyst had caused membrane pore blockage and deterioration of
93
continuous porous structure due to the formation of agglomerates, which had seriously declined
94
the membrane permeability [20,21].
95
To address several issues on robust membranes for OPW purification by integrating these
96
preferential membrane properties, we developed a photocatalytic nanofiber-coated inorganic
97
hollow fiber membrane. Recent emerging rediscovery of graphitic carbon nitride (GCN) has
98
attracted a considerable attention for potential applications in photocatalysis [22–25]. Furthermore,
99
our previous study demonstrated an enhanced photodegradation of OPW using GCN-embedded
100
polyacrylonitrile (PAN) nanofiber mesh [26]. Therefore, in this study, we electrospun GCN-
101
embedded PAN nanofiber on top of asymmetric alumina (Al2O3) hollow fiber membrane surface
102
using direct electrospinning technique. The thermal treatment after deposition significantly
103
enhanced the adhesion of nanofiber coating on the Al2O3 membrane and successfully offered a
104
GCN-embedded nanofiber-coated Al2O3 hollow fiber membranes. The membrane demonstrated
105
high permeability and rejection, fouling resistance, and easy cleaning of membrane fouling for a
106
model example of OPW treatment.
107 108
2.
Experimental
109 110
2.1. Materials
111 112
Urea from QReC Malaysia was used as the raw material to synthesize bulk graphitic carbon
113
nitride (bGCN). The bGCN was further exfoliated to form nanosheet graphitic carbon nitride
114
(nsGCN) using isopropyl alcohol (IPA) from QReC Malaysia. Polyacrylonitrile (PAN) from
115
Sigma Aldrich and dimethylformamide (DMF) from RCI Labscan were used as polymer binder
116
and solvent, respectively, for the electrospinning suspension to fabricate nanofibers.
117
Three different sizes of alumina powder (Al2 O3) were purchased from Alfa Aesar and used
118
to fabricate hollow fiber membranes: (1) α-Al2O3 (99% metal basis, average size of 1 µm and
119
surface area of 6–8 m2/g), (2) α- and γ-Al2O3 (99.5% metal basis, average size of 0.5 µm and
120
surface area of 32–40 m2/g), and (3) α- and γ-Al2 O3 (99.8% metal basis, average size of 0.01 µm
121
and surface area of 100 m2/g). Polyethersulfone (PESf, Radal A300, Ameco Performance, USA),
122
N-methyl-2-pyrrolidone (NMP, AR grade, QRëCTM), and poly(ethylene glycol) 30-
123
dipolyhydroxystrearate (Arlacel P135 from Uniqema) were added into the spinning suspension to
124
fabricate Al2O3 hollow fiber membranes. All materials purchased in this work were used without
125
any further purification.
126
The crude oil sample used for membrane performance evaluation was supplied by Petronas
127
Refinery Malacca, Malaysia (API Grade 66). Sodium dodecyl sulfate (SDS) purchased from
128
Merck was used as surfactant during the preparation of the oilfield produced water (OPW)
129
solution. Reverse osmosis (RO) water (Millipore: ASTM Type III) was used throughout the
130
experiments and membrane filtration tests.
131 132
2.2. Preparation of asymmetric alumina (Al2O3) hollow fiber membrane
133 134
The asymmetric alumina (Al2O3) hollow fiber membrane was fabricated according to the
135
scheme established in a previous report [27]. There are mainly three steps involved in the
136
membrane preparation, which are preparation of ceramic suspension, spinning of hollow fiber
137
precursor, and sintering of the precursor. In the first step, three Al2O3 powders with different
138
average sizes of 0.01, 0.05, and 1 µm were mixed at the ratio of 1:2:7, respectively, and 106 g of
139
the powder mixture was added into 73.74 g of NMP solution containing 2.6 g of Arlacel P135.
140
The mixture was ball-milled in planetary ball machine (Magna NQM-2 Planetary Ball Mill) before
141
17.66 g of PESf was added into the Al2O3 suspension. In the second step, the prepared Al2O3
142
suspension was extruded using stainless steel spinneret to form the hollow fiber membranes. The
143
collected spun Al2O3 hollow fiber membranes were drenched into tap water to complete the phase
144
inversion process. In the third step, the collected Al2O3 hollow fiber membranes were dried at
145
room temperature before sintered at 1400 °C in a tubular furnace (Magna, XL-1700).
146 147
2.3. Fabrication of a GCN nanofiber-coated Al2O3 hollow fiber membrane
148 149
Electrospinning technique has been successfully utilized to fabricate a nanofiber mesh on
150
a metallic foil placed on a flat plate or a cylindrical drum [28]. An electrically-conductive substrate
151
with small curvature generates uniform electric field between a needle and substrate, which is
152
readily applicable for electrospinning. There are, in contrast, limited reports on nanofiber
153
electrospinning on an insulating porous substrate having large curvature like a hollow fiber
154
membrane, and therefore, electrospinning technique on such substrate has not completely
155
established yet.
156
Photocatalysts of bGCN and nsGCN were synthesized from urea using green and facile
157
template-free method according to our previous work [26]. Figure 1 illustrates a schematic design
158
of electrospinning setup for the fabrication of GCN nanofiber-coated Al2O3 membrane.
159
Commercially available nanofiber electrospinning unit (NF-1000, Progene Link Sdn. Bhd.,
160
Malaysia) and a blunt metallic needle (21G × 1", Terumo Corporation) connected to a 10 mL
161
syringe were employed in the nanofiber coating. An 8 wt% dope solution containing 7.2 wt% of
162
PAN and 0.8 wt% of GCN was electrospun on an Al2O3 hollow fiber membrane at a solution feed
163
rate of 1 mL/h and an acceleration voltage of 15 kV. The membrane of 7 cm in length was
164
immersed in advanced in IPA for 15 min and fixed on a holder directed perpendicular to the needle
165
while wrapped in cotton wadding. The holder was placed at 18 cm apart from the needle while
166
rotating at 6 rpm to homogeneously coat the nanofibers on the membrane. Electrospinning process
167
continued for nearly 3 h for each of the membrane. Al2O3 hollow fiber membranes coated with
168
PAN nanofibers only, PAN nanofibers containing bGCN, and PAN nanofiber containing nsGCN
169
were denoted as NF/Al2O3, NF-bGCN/Al2O3, NF-nsGCN/Al2O3 membranes, respectively. The
170
membranes coated with nanofibers were heated at 120 °C for 15 min, which enhanced both
171
mechanical stability of nanofiber mesh and adhesion between nanofiber layer and membrane as
172
demonstrated in the following section.
173 (d)
(a)
(b)
(e)
(c)
(f)
174
High voltage supply
175
Fig. 1. A schematic design of electrospinning setup for the fabrication of GCN nanofiber-coated
176
Al2O3 membrane: (a) syringe containing dope solution, (b) blunt metallic needle, (c) syringe pump,
177
(d) wood-based support, (e) grounded metal rotating holder, and (f) nanofiber-coated Al2O3 hollow
178
fiber membrane.
179 180
2.4. Characterization methods
181 182
A field emission scanning electron microscope (S-4800, Hitachi High-Tech. Co.) was
183
employed to examine the as-prepared membranes morphological structures. The samples were
184
fixed with double sided carbon tape and coated with a thin platinum layer under argon pressure of
185
7 Pa for 50 s using an ion sputter (E-1030, Hitachi High-Tech. Co.). Thermal stability of the
186
nanofiber samples was determined using a thermal gravimetric analyzer (TGA4000, PerkinElmer
187
Inc.) under air flow at a heating rate of 10 °C/min. The nanofiber samples were cut into small
188
pieces and placed in platinum pans.
189 190 191
2.5. Filtration experiments
192 193
Cross-flow microfiltration is known as a suitable way to remove colloidal particles and
194
reduce fouling effect [29]. In the crossflow filtration of colloidal droplets, droplets in a stream are
195
subjected to two different forces in the direction perpendicular to the membrane surface. The
196
balance between the forces determines the formation of a filter cake on the membrane surface. Due
197
to the filtrate flow passing through the membrane surface, a drag force, Fy, moves droplets to the
198
membrane surface. Fy can be calculated using Stokes equation as follows [29]:
199
𝐹𝑦 = 3π 𝜂 𝐷 𝑣𝐹 ,
200
where η is the viscosity of the liquid medium, D is the droplet size, and vF is the permeate rate.
201
According to the equation, Fy is proportional to D. The other force acting on droplets is a lift force,
202
FL, which moves droplets away from the membrane surface. FL originates from shear flow and
203
estimated using an equation [29],
204
𝐹𝐿 = 0.761
𝜏𝑤 1.5 𝐷3 𝜌0.5 𝜂
205
where τw is the shear stress and ρ is the density of the liquid medium. According to the equation,
206
the lift force is proportional to the cube of D. As D increases, both Fy and FL increase. However,
207
due to different dependences on D, FL overcomes Fy when D becomes sufficiently large.
208
Synthetic OPW at 1000 ppm was prepared according to a previously reported method [30].
209
The total organic carbon (TOC) measurement carried out using TOC analyzer (Shimadzu Co.,
210
TOC-LPCN) on 1000 ppm of OPW solution reveals the TOC value of 531.2 ppm. Oil particle
211
sizes in feed and permeate were measured using a particle analyzer (Zetasizer Ver. 7.11, Malvern
212
Instruments Ltd.). A benchtop crossflow membrane filtration setup was used to evaluate the
213
separation performance of the synthesized membranes (Fig. 2). The membranes were wetted via 1
214
h immersion in reverse osmosis (RO) water before installation. Prior to collecting the filtration
215
results, the membranes were compacted at a pressure of 3 bar for 10 min to ensure a steady state
216
condition of permeation experiment. The filtration setup was operated using a transmembrane
217
pressure of 2 bar and a feed flow rate of 72 L/h. Pure water flux (PWF) and permeate flux, J
218
(L∙m−2∙h−1), are calculated using Eq. (1):
219
𝐽=𝐴
∆𝑉
𝑚 ∆𝑡
(1)
220
where Am is the effective membrane area (m²), Δt is the time used to collect permeate (h), and ΔV
221
is the respective collected permeate volume (L). The oil rejection (%) was assessed according to
222
our previous study [26] using Eq. (2): 𝑂𝑖𝑙 𝑟𝑒𝑗𝑒𝑐𝑡𝑖𝑜𝑛 (%) = (
223
𝐶𝑖 −𝐶𝑜 𝐶𝑖
) × 100 (2)
224
where Ci and Co are the absorbance values of feed and permeate solutions, respectively, and were
225
measured using a UV–visible spectrophotometer (DR5000, Hach, U.S.A) at the wavelength of 238
226
nm. The crude oil used in this study has large UV absorbance at the 238 nm wavelength. The OPW
227
filtration experiment was conducted for 180 min by collecting permeate solutions at every 10 min
228
to examine antifouling property. To examine the cleaning property after the 180 min filtration, the
229
contaminated membrane was taken out from the membrane module, immersed in RO water, and
230
subjected to UV irradiation (30 W UV lamp, peak wavelength of 312 nm) for 180 min. The GCNs
231
photocatalyst has strong UV absorbance in this wavelength range, which was demonstrated in our
232
previous paper [26]. The membrane was then reinstalled into the module and the OPW filtration
233
was performed for another 180 min. The filtration–cleaning cycle was repeated for three times
234
using fresh feed of OPW solution in each cycle.
Retentate
OPW feed tank Valve
Pump
235 236
Membrane module
Permeate Pressure gauge
Nanofiber coated Al2O3 hollow fiber membrane
Fig. 2. Schematic diagram of crossflow membrane filtration setup.
237 238
3.
Results and discussion
239 240
3.1. Fabrication of nanofiber-coated on alumina (Al2O3) hollow fiber membrane surface
241 242
As described in Section 2.3, the electrospinning technique was able to coat PAN
243
nanofibers homogeneously on a thin Al2O3 hollow fiber membrane by selecting appropriate
244
electrospinning parameters such as acceleration voltage, solution flow rate, and tip-to-membrane
245
distance based on our previous study [26]. The whole membrane was uniformly covered by
246
nanofiber mesh, which was at least confirmed by visual inspection of the membrane appearance.
247
In spite of the uniform nanofiber coating on the membrane surface, when the membrane
248
was used in the filtration experiment, some of the nanofibers were left in the filtrate and the coating
249
sometimes peeled off from the membrane surface due to insufficient adhesion. The as-prepared
250
coating of electrospun nanofibers suffered from low mechanical properties and weak adhesion on
251
substrates due to low electrical conductivity of Al2O3 hollow fiber membrane, which made the
252
coating impossible to be applicable to a filtration membrane. To ensure good membrane stability
253
in a prolonged filtration process, the as-prepared nanofiber-coated membranes were further heated
254
at a high temperature. When the membrane was heated at 150 °C for 15 min, the membrane was
255
lightly browned. The result indicates that the PAN nanofibers partially decomposed although the
256
thermal gravimetric analysis (TGA) exhibited no detectable nanofibers degradation at a
257
temperature below 280 °C [31] (Fig. S1) (Details in Supplementary Information). On the other
258
hand, the 15 min heat treatment at 120 °C did not cause detectable change of nanofiber coating by
259
visual inspection. In addition to no color change, the mechanical properties of the nanofiber mesh
260
were also significantly improved by the heat treatment as demonstrated in Fig. 3. When the as-
261
prepared and heat-treated NF-nsGCN meshes were immersed in RO water for 60 min, the as-
262
prepared mesh was so soft that tweezers could not hold it firmly while the heat-treated mesh
263
offered a mechanical strength enough for handling. Figure 3(a) shows the photographs of the as-
264
prepared and heat-treated meshes containing RO water. The as-prepared mesh showed many
265
wrinkles with a deformed shape while the heat-treated mesh exhibited a flat rectangular shape.
266
Furthermore, when they were subjected to 60 min ultrasonication in RO water, the heat-treated
267
mesh maintained its shape (Fig. 3c) while the as-prepared mesh lost its original shape and exhibited
268
rough appearance accompanying partial fragmentation (Fig. 3b). The results directly indicate a
269
significant improvement on the mechanical strength through the heat treatment process. Since the
270
temperature of 120 °C is higher than the glass transition temperature of PAN (97 °C) [32], the
271
high-temperature short-time treatment allows PAN polymers to plasticize and bound strongly at
272
nanofiber intersections while keeping the nanofiber shape as confirmed by SEM. The high
273
mechanical strength of nanofiber coating plays a significant role to determine the stability of the
274
membrane during prolonged filtration operation.
(a) (i)
(b)
(ii)
(c)
275 276
Fig. 3. Photographs of NF-nsGCN meshes (a) immersed in RO water and ultrasonicated for 60
277
min: (b) as-prepared and (c) heat-treated. The samples initially had a rectangular shape of 12
278
(width) × 17 (length) × 0.05 (thickness) mm.
279 280
3.2. Characterization of membranes
281 282
The SEM images show the surface morphologies of bare Al2O3 membrane (Fig. 4a) and
283
NF-coating layers (Fig. 4b–d). The Al2O3 membrane exhibited smooth surface having many small
284
pores with a diameter range of 175–375 nm. All the NF-coating layers formed sparse mesh of
285
straight nanofibers with an average diameter of 250 nm. The opening sizes of all NF layers are
286
similar and approximately several micrometers, which are almost ten times larger than the pore
287
size of the Al2O3 membrane. In contrast to smooth straight shape of NF, NF-bGCN had several
288
bulges on smooth nanofibers. The spherical bulges with a few micrometers in size correspond to
289
the bGCN particles embedded on the nanofibers. Due to the successful exfoliation of nsGCN, NF-
290
nsGCN had a less numbers of thin bulges compared with those in NF-bGCN. It is difficult to
291
accurately measure the thickness of the coating layer because the coating layer did not result in a
292
clear cross section on an SEM image even using a freeze fracture technique. In our previous work,
293
the obtained FTIR spectra confirmed that GCNs were successfully embedded into the PAN
294
polymer matrix. The well dispersion of GCN into PAN polymer matrix promotes a vast number
295
of active sites for effective interactions between the reactant and photocatalyst, thus enhancing the
296
photocatalytic activity [26]. Meanwhile, in comparison with NF-bGCN (mean nanofiber diameter
297
of 207 ± 2 nm), the mean nanofibers diameter of NF-nsGCN is 262 ± 6 nm with a smooth and
298
straight infinite length structure [26].
(a)
(b)
10 µm
10 µm (c)
(d)
10 µm
10 µm
299 300
Fig. 4. SEM micrographs of (a) bare Al2O3, (b) NF/Al2O3, (c) NF-bGCN/Al2O3, and (d)
301
NF-nsGCN/Al2O3 membranes.
302 303
The water affinity of the nanofiber meshes was assessed on the basis of water absorption
304
capacity. The water absorption capacity for NF and NF-nsGCN meshes were measured by
305
immersing them into RO water for 60 min. The excess water on the surface was removed using a
306
filter paper and the wetted nanofiber meshes were weighed. The water absorption capacity is
307
calculated using Eq. (3):
308
𝐴𝑏𝑠𝑜𝑟𝑝𝑡𝑖𝑜𝑛 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 =
309
where ww and wd are the masses of wetted and dry nanofibers, respectively. The NF and NF-nsGCN
310
meshes both have high water absorption capacity of 34 ± 8 and 27 ± 4, respectively. The water
𝑤𝑤 −𝑤𝑑 𝑤𝑑
(3)
311
absorption capacities correspond to 96%–97% volume of water being retained in wetted nanofiber
312
meshes, demonstrating a high water affinity and large porosity of nanofiber meshes. The slightly
313
higher absorption of NF-nsGCN suggests that the incorporation of nsGCN enhanced water affinity
314
probably due to the high water affinity of GCN. In fact, Hansen solubility parameter (HSP)
315
expressing chemical affinity between molecules and colloids [33] suggested better water
316
miscibility of GCN: HSP of GCN (δD ~ 17.8 MPa1/2, δP ~ 10.8 MPa1/2, and δH ~ 15.4 MPa1/2)
317
[34] is much closer to that of water (δD ~ 15.5 MPa1/2, δP ~ 16.0 MPa1/2, and δH ~ 42.3 MPa1/2)
318
than that of PAN (δD ~ 22.4 MPa1/2, δP ~ 14.1 MPa1/2, and δH ~ 9.1 MPa1/2). Note that the contact
319
angle measurement is not adopted as an informative characterization in this study. Cassie effect
320
resulting from large air pocket and micron-range roughness in nanofiber has a significant
321
contribution on water contact angle of nanofiber meshes [35]. As a result, a contact angle of a
322
nanofiber mesh does not indicate hydrophilicity or hydrophobicity of the nanofibers. In fact, the
323
contact angles of the nanofiber-coated membranes were relatively high (135.1°, 121.5°, and 114.9°
324
for NF/Al2O3, NF-bGCN/Al2O3, and NF-nsGCN/Al2O3 membranes, respectively) in spite of the
325
high water absorption capacity.
326 327
3.3.
Pure water flux of membranes
328 329
The PWF of bare Al2O3, NF/Al2O3, NF-bGCN/Al2O3, and NF-nsGCN/Al2O3 membranes
330
were assessed in a crossflow filtration system (Fig. 5). In spite of additional coating on bare Al2 O3
331
membrane, the average PWF of the NF/Al2O3 membrane (514 L∙m−2∙h−1) is 20% higher than that
332
of the bare membrane (421 L∙m−2∙h−1). Both NF-bGCN/Al2O3 and NFnsGCN/Al2O3 membranes
333
exhibited a much higher PWF than the bare membrane, the deviations of which were clearly
334
beyond the experimental error range. This finding is consistent with several previous studies
335
addressing that a surface coating of electrospun nanofibers is capable of effectively improving flux
336
of polymeric flat sheet membranes [15,36,37]. Although a definite mechanism of the high PWF
337
has yet to be revealed, we speculate that a coating layer of a sparse nanofiber mesh on a hollow
338
fiber membrane disturbs the laminar flow of water along the membrane surface, which results in
339
the enhanced permeation of water in a transmembrane direction [38]. The sparse mesh structure
340
of nanofibers resemble the effect of a mesh spacer placed in a spiral module of a flat sheet
341
membrane [39].
342
The PWF of the NF-bGCN/Al2O3 and NF-nsGCN/Al2O3 membranes were 728 and 816
343
L∙m−2∙h−1, respectively, both of which are much higher than that of the NF/Al2O3 membrane
344
because of the enhanced water affinity by the incorporation of GCN into the PAN nanofiber. This
345
result is consistent with the larger water capacity for NF-nsGCN as described in the previous
346
section. Furthermore, previous studies have reported that electrospun membranes improve water
347
flux by introducing hydrophilic organic nanofillers due to the enhancement of water affinity
348
[40,41]. Compared with the NF-bGCN/Al2O3 membrane, the NF-nsGCN/Al2O3 membrane gave
349
10% higher PWF, which is possibly due to the smooth surface morphology of nanofibers as shown
350
in Fig. 4(d). Among all the membranes fabricated, the NF-nsGCN/Al2O3 membrane showed the
351
highest PWF, which is nearly twice the PWF of bare membrane without the NF coating. Since the
352
nanofibers-coated membrane were completely wetted, the high porosity of nanofiber meshes
353
contributes to the large flow rates [42]. Three key parameters on the enhanced PWF were assigned:
354
(1) sparse mesh structure disturbing laminar water flow, (2) better affinity to water, and (3) smooth
355
nanofiber morphology.
356
357 358
Fig. 5. Pure water flux for bare Al2O3, NF/Al2 O3, NF-bGCN/Al2O3, and NF-nsGCN/Al2O3
359
membranes for 90 min crossflow filtration at transmembrane pressure of 2 bar.
360 361
3.4.
Crossflow filtration of oilfield produced water
362 363
In the PWF measurement, the fabricated membranes were subjected to the crossflow
364
filtration of OPW, which were evaluated in terms of permeate flux and oil rejection efficiency.
365
Fig. 6 shows the photographs of OPW feed and permeate solutions obtained by the crossflow
366
filtration. In contrast with the opaque appearance of the feed solution containing 1000 ppm oil
367
droplet, all the permeate solutions were transparent, directly indicating that most of the oil droplets
368
were successfully removed by the filtration operation.
OPW
369
(a)
(b)
(c)
(d)
370
Fig. 6. OPW feed and permeate solutions obtained by crossflow filtration using hollow fiber
371
membranes: (a) bare Al2O3, (b) NF/Al2O3, (c) NF-bGCN/Al2O3, and (d) NF-nsGCN/Al2 O3.
372 373
Figure 7(a) shows the permeate flux of bare Al2O3, NF/Al2O3, NF-bGCN/Al2O3, and
374
NF-nsGCN/Al2O3 membranes assessed in the crossflow filtration system. After 180 min, the bare
375
membrane showed a permeate flux of 236 L∙m−2∙h−1, which was only 56% of its PWF. After the
376
filtration, the SEM images demonstrated that the surface of the membrane was almost completely
377
covered by oily component (Fig. 8a). Such a dense fouling layer significantly reduced the permeate
378
flux of OPW due to the severe blockage of the open pores on the surface of the membrane. The
379
NF/Al2O3 membrane exhibited a permeate flux of 386 L∙m−2∙h−1, which was higher than that of
380
the bare membrane. The permeate flux of the NF/Al2O3 membrane corresponds to 75% of its PWF,
381
maintaining a higher permeate flux of OPW compared with that of the bare membrane. An SEM
382
image of the NF/Al2O3 membrane after filtration displayed some amount of oily component
383
captured on the opening of the nanofiber mesh (Fig. 8b). However, individual nanofibers and the
384
openings between nanofibers can be still observed because the fouling layer made of oily
385
component covered only a small portion of the nanofiber coating layer. The large opening size of
386
the nanofiber mesh prevented the oily component to form dense continuous fouling layer,
387
maintaining the permeation pathway of water to a large extent. As a result, the nanofiber coating
388
layer has the capability of preserving high permeate flux of OPW. In fact, the NF-bGCN/Al2O3
389
and NF-nsGCN/Al2O3 membranes gave permeate fluxes of 577 and 640 L∙m−2∙h−1, respectively,
390
which are much higher than that of the bare membrane. Compared with the NF/Al2O3 membrane,
391
the SEM images exhibited a less amount of oily component on the surface of the NF-bGCN/Al2 O3
392
and NF-nsGCN/Al2O3 membranes (Fig. 8c–d), which agrees with the higher permeate flux of
393
OPW for the membranes. As demonstrated in the PWF experiment, the NF–nsGCN/Al2O3
394
membrane gave a higher permeate flux of OPW than that of the NF–bGCN/Al2O3 membrane due
395
to the smooth surface morphology of the nanofibers resulting from the uniform distribution of thin
396
nsGCN. (a)
(b)
397 398
Fig. 7. (a) Permeate flux and (b) oil rejection percentage in 180 min crossflow filtration of OPW
399
using bare Al2O3, NF/Al2O3, NF-bGCN/Al2O3, and NF-nsGCN/Al2O3 membranes at a
400
transmembrane pressure of 2 bar.
401
(b)
(a)
10 µm
10 µm (d)
(c)
10 µm
10 µm
402 403
Fig. 8. Surface SEM micrographs of (a) bare Al2 O3, (b) NF/Al2 O3, (c) NF-bGCN/Al2O3, and (d)
404
NF-nsGCN/Al2O3 membranes after 180 min OPW filtration.
405 406
3.5.
Oil rejection of membranes
407 408
Oilfield produced water generally contains a complex mixture of organic and inorganic
409
materials similar to those found in crude oil, and the composition varies with the location and the
410
life of a producing field [43]. Therefore, we measured the organic species in the crude oil using
411
gas chromatography–mass spectroscopy (GC–MS) (Fig. S2) to identify the components in OPW
412
solution (Details in Supplementary Information).
413
To assess the trace amount of oil contaminants in the transparent permeates, we
414
quantitatively determined the rejection efficiency of OPW based on UV absorbance. All
415
membranes showed a high oil rejection above 94% even after 180 min filtration as shown in Fig.
416
7(b). Obviously, the nanofiber-coated membranes resulted in a better oil rejection than that of the
417
bare membrane. The NF-nsGCN/Al2O3 membrane exhibited an excellent oil rejection percentage
418
of 99% along with the highest permeate flux, demonstrating the best filtration performance among
419
the membranes examined in this study. Furthermore, oil rejection percentage of this permeate flux
420
was matched with the result obtained from TOC analysis at 8.601 ppm (98.3% oil rejection). To characterize the dispersion state of oil contaminants in the permeates, the size
422
distribution of oil droplets in the OPW feed and permeate solutions were measured using a particle
423
size analyzer, which is suitable for understanding the separation mechanism of the fabricated
424
membranes (Fig. 9).
Number weighted frequency (%)
421
425 426
Fig. 9. Particle size distribution of oil droplets in OPW feed and permeate solutions produced by
427
bare Al2O3 and NF-nsGCN/Al2O3 membranes after 180 min.
428
429
The feed solution contained a large number of oil droplets with the sizes between 0.57 and
430
1.3 μm. These size oil droplets strongly scatter light, which is consistent with the white color of
431
the feed solution. The permeate solution obtained by crossflow filtration using the bare Al2O3
432
membrane also contained oil droplets sizing from 0.24 to 0.37 µm. In addition to the low
433
concentration of oil droplets suggested from the oil rejection efficiency, the small droplet size
434
agrees with the transparent appearance of the permeate solution. The average size of oil droplets
435
significantly reduced compared with that in the feed solution.
436
The theoretical analysis as discussed in Section 2.5 expresses that crossflow microfiltration
437
has a tendency to accumulate the small droplets near the membrane surface and keep the large
438
droplets away from the surface. In addition to the spontaneous size-selectivity crossflow
439
microfiltration, membrane pore size plays a dominant role in the rejection of oil droplets. In fact,
440
the maximum size of oil droplet agrees with the maximum pore size on the surface of the bare
441
Al2O3 membrane (Fig. 4a). This indicates that the size-selectivity permeation of oil droplet through
442
the bare membrane was governed by the surface pore size of the membrane.
443
The permeate collected by crossflow filtration using the NF-nsGCN/Al2O3 membrane
444
contained oil droplets sizing from 0.07 to 0.3 μm, the average size of which is much smaller than
445
that obtained using the bare membrane. It is interesting because the opening size of the nanofiber
446
coating is much larger than the surface pore size of the bare membrane. This indicates that the
447
droplet size distribution is not directly determined by the opening size of the nanofiber coating.
448
This is probably because the nanofiber coating does not separate oil droplets based on size-
449
selectivity of the openings but absorbs them in the openings due to a large adhesive energy of the
450
droplets as illustrated in Fig.10. Indeed, our previous paper has demonstrated that the electrospun
451
PAN nanofibers incorporating bGCN or nsGCN have a good ability to absorb oil droplets
452
dispersed in OPW in a floating manner [26]. According to the evidence, we can expect the
453
nanofiber coating absorbs most of the droplets during the permeation of the droplets through the
454
coating. The SEM images proved that the nanofiber diameter definitely increased when the
455
nanofiber coating was subjected to crossflow filtration of OPW (Fig. 8). The nanofiber coating
456
captured the larger droplets more effectively due to the larger collision cross section of droplets,
457
which resulted in the permeation of smaller droplets. In addition of the effective absorption in the
458
nanofiber coating, the small pores of the bare membrane can remove large oil droplets that pass
459
through the nanofiber coating. As shown in Fig. 7(b), the ranking of oil rejection performance
460
coincides with that of oil absorption performance previously determined [26]: NF-nsGCN > NF-
461
bGCN > NF. The agreement completely supports our expectation that the absorption performance
462
of the nanofiber coating gave a better oil rejection efficiency and smaller size of oil droplets
463
remaining in the permeates. The effective absorption of oil droplets in the nanofiber coating
464
prevented the formation of fouling layer on the surface of the bare membrane. The fouling
465
suppression preserves a high flux of OPW even after 180 min operation.
466
467 468
Fig. 10. Crossflow filtration of OPW using NF-nsGCN/Al2O3 membrane.
469 470
3.6
Cycle operation of crossflow filtration
471 472
As discussed in the previous section, the nanofiber coatings on the Al2O3 hollow fiber
473
membrane was able to prevent dense fouling layer from covering the whole membrane surface
474
because of the rough surface morphology and high water affinity. However, a relative amount of
475
oil adhered on a part of the membrane surface after 180 min filtration. Thus, cleaning process is
476
necessary to recover the filtration performance for a long-term operation. To introduce good
477
cleaning properties, we incorporate GCN, an efficient photocatalyst, into the electrospun PAN
478
nanofiber. The photocatalyst-embedded nanofiber captured the oil droplets in water and exhibited
479
an efficient photodegradation property against the oil contaminants under UV and visible light
480
irradiations, which was demonstrated in our previous study [26]. On the basis of the experimental
481
evidence, the photodegradation process is expected to be applicable as a cleaning process of oil
482
contaminants absorbed by the nanofiber.
483
Figure 11 shows the permeation flux of OPW for 180 min crossflow filtration using the
484
fabricated membranes after three cycles of operation. There is a slight difference between Fig. 7
485
and 11 because the former is based on an averaged value and the latter represents one of the
486
experiments. After each interval, the membrane was taken out from the membrane module,
487
immersed in pure water, and irradiated with UV light. In the case of all filtration operation
488
independent on membrane types, the permeate flux exponentially decreased with the elapsed time
489
and fell down to 40%–50% of the initial flux after 180 min. Such reduction in the permeate flux
490
was recovered by the cleaning process, which strongly depends on membrane types. The bare
491
Al2O3 and NF/Al2O3 membranes, both of which have no photodegradation capability, exhibited a
492
recovery ratio of less than 90% of the initial flux compared with the previous operation. In contrast,
493
both NF-bGCN/Al2O3 and NF-nsGCN/Al2O3 membranes with photodegradation ability recovered
494
more than 90% of the initial flux. The results demonstrate that the photodegradation ability of the
495
nanofiber was suitable for cleaning membranes after filtration and maintained a high permeate flux
496
for repeated filtration cycles.
497
Figure 12 shows the photographs of NF-nsGCN membrane coated with nanofiber mesh.
498
While the as-fabricated membrane looks white (Fig. 12a), the membrane used for crossflow
499
filtration of OPW turned to pale orange, indicating an amount of oil component adhered on the
500
surface (Fig. 12b). By irradiating UV light in pure water, the membrane color faded away
501
considerably, suggesting an effective photodegradation of the oil component. The NF-
502
nsGCN/Al2O3 membranes gave the highest permeate flux of 577 L∙m−2∙h−1 after three cycles of
503
180 min filtration. In the third filtration cycle, the membrane maintained 87% of the permeate flux
504
after 180 min filtration of the first cycle while the fluxes of the other membranes reduced to 58%–
505
75% of their first cycle. The results agree with the highest photodegradation performance of
506
nanofiber containing nsGCN as confirmed in our previous paper [26] and provide an evidence that
507
the tphotodegradation ability of the nanofibers offers a better cleaning performance of the
508
membrane.
509
510 511
Fig. 11. Permeation fluxes of OPW filtration for bare Al2O3 (grey), NF/Al2O3 (green),
512
NF-bGCN/Al2O3 (blue), and NF-nsGCN/Al2O3 (red) at a pressure of 2 bar for 180 min in three
513
cycles operation.
514
(a)
(b)
(c)
515 516
Fig. 12. Photographs of GCN nanofiber-coated Al2O3 hollow fiber membranes (a) before OPW
517
filtration, (b) after 180 min filtration at a pressure of 2 bar, and (c) after 180 min irradiation under
518
UV light.
519
520
Figure 13 shows the oil rejection percentage for 180 min cross-flow filtration of OPW
521
using the fabricated membranes under three cycles of operation. The bare Al2O3 membrane
522
showed a decreased oil rejection percentage of around 89% after three operation cycles, which
523
corresponds to 8% reduction from the initial value. The NF/Al2O3 membrane oil rejection
524
degraded by 5%, while both NF-bGCN/Al2O3 and NF-nsGCN/Al2O3 membranes demonstrated
525
only 2% reduction due to the photodegradation of oil component by UV irradiation. The NF-
526
nsGCN/Al2O3 membranes gave the highest rejection percentage of 97% after three operation
527
cycles. The results indicate that the nanofiber mesh with photodegradation ability is a promising
528
coating layer for crossflow microfiltration membrane that maintains the high rejection
529
performance.
530
531 532
Fig. 13. Oil rejection percentage for bare Al2O3 (grey), NF/Al2O3 (green), NF-bGCN/Al2O3 (blue),
533
and NF-nsGCN/Al2O3 (red) at a pressure of 2 bar for 180 min in three cycles operation.
534
535
In comparison to other advance membrane modifications used to treat oily wastewater,
536
Yang and co-workers have investigated the efficiency of ZrO2/α-Al2O3 microfiltration membrane
537
with average pore size of 0.2 µm showed 99.8% of oil rejection from 5000 ppm of vegetable and
538
mineral oils solutions [44]. Graphene modified Al2O3 ceramic microfiltration membrane
539
fabricated by Hu et al., [45], revealed that flux increased 27.8% increased as compared to
540
unmodified Al2O3 membrane with 98.7% oil rejection. Lastly, Zhou et al., [46] reported that
541
zirconia modified membrane that have been fabricated obtained 88% initial flux and 97.8% oil
542
rejection using 1000 ppm engine-oil water emulsion as a feed solution. Nevertheless, we believe
543
that NF-nsGCN/Al2O3 membrane is a promising candidate to treat OPW based on revealed
544
excellent membrane properties, separation performances and long-term stability in repeating cycle
545
filtration.
546 547
4.
Conclusion
548 549
Photocatalytic nanofiber-coated hollow fiber membranes were successfully fabricated
550
using a newly design electrospinning technique. Polyacrylonitrile nanofibers incorporating
551
graphitic carbon nitride (GCN) were coated on alumina (Al2O3) hollow fiber membranes.
552
Crossflow filtration using the fabricated membranes exhibited a significant improvement in pure
553
water flux, OPW permeate flux, and oil rejection percentage compared with those of the bare Al2O3
554
membrane. The NF-nsGCN/Al2O3 membrane showed the highest pure water flux, OPW permeate
555
flux, and oil rejection at 816 L∙m−2∙h−1, 640 L∙m−2∙h−1, and 99%, respectively. Sparse mesh
556
structure, high water affinity, and smooth nanofiber morphology were found as key parameters for
557
nanofiber coatings that significantly improved membrane performances. Unlike the conventional
558
ultrafiltration membranes, nanofiber coating was able to prevent uniform fouling layer covering
559
the whole hollow fiber membrane surface, resulting in a high OPW permeate flux and oil rejection.
560
Furthermore, the NF-nsGCN/Al2O3 membrane also maintained the highest permeate flux (577
561
L∙m−2∙h−1) and oil rejection (97%) after three operation cycles, demonstrated an excellent cleaning
562
performance for a long-term membrane operation. The photodegradation ability of the NF-nsGCN
563
nanofiber enabled the coating to degrade the captured oil contaminants under UV irradiation,
564
which is beneficial to maintain the high permeate flux and rejection in repeated filtration system.
565
These findings indicate the potential application of the NF-nsGCN-coated Al2O3 hollow fiber
566
membrane in the industrial OPW treatment. They also provide useful information for further
567
research to develop nanofiber-coated hollow fiber membranes for future membrane separation
568
technology.
569 570
Acknowledgments
571 572
The authors gratefully acknowledge PETRONAS Penapisan (Melaka) Sdn. Bhd. for the
573
supply of crude oil sample. N.H.A. and S.S. would like to thank Dr. Masanobu Naito at NIMS for
574
his kind support. N.H.A, J.J, and N.Y. would like to express their sincere gratitude towards the
575
Malaysia Ministry of Higher Education for the research funds provided under UTM–HiCOE
576
Research Grants (R.J090301.7846.4J184) and (R.J090301.7846.4J185), UTM for the financial
577
support under Research University Grant (GUP) Tier 1 (Q.J130000.254616H43) and Japan
578
government for the Kurita Water and Environmental Foundation (KWEF) research grant
579
(18P001). N.H.A. would like to thank Universiti Teknologi Malaysia (UTM)–National Institute
580
for Materials Science (NIMS) Cooperative Graduate School Program (ICGP) 2017/18 for the
581
graduate fellowship awarded. A part of this work was supported by "Nanotechnology Platform"
582
(project No. A-17-NM-0208) of the Ministry of Education, Culture, Sports, Science and
583
Technology (MEXT), Japan.
584 585
Appendix A. Supplementary data
586 587
Supplementary data associated with this article can be found, in the online version, at
588 589
References
590 591
[1]
T.D. Kusworo, N. Aryanti, Qudratun, D.P. Utomo, Oilfield produced water treatment to
592
clean water using integrated activated carbon-bentonite adsorbent and double stages
593
membrane process, Chem. Eng. J. 347 (2018) 462–471.
594
[2]
M.Z. Shahruddin, N.H. Othman, N.H. Alias, S.N.A. Ghani, Desalination of produced water
595
using bentonite as pre-treatment and membrane separation as main treatment, Procedia -
596
Soc. Behav. Sci. 195 (2015) 2094–2100.
597
[3]
A.M. a Pintor, V.J.P. Vilar, C.M.S. Botelho, R. a R. Boaventura, Oil and grease removal
598
from wastewaters: Sorption treatment as an alternative to state-of-the-art technologies. A
599
critical review, Chem. Eng. J. 297 (2016) 229–255.
600
[4]
R. Zhang, Y. Liu, M. He, Y. Su, X. Zhao, M. Elimelech, Z. Jiang, Antifouling membranes
601
for sustainable water purification: strategies and mechanisms, Chem. Soc. Rev. 45 (2016)
602
5888–5924.
603
[5]
C.K. Pooi, H.Y. Ng, Review of low-cost point-of-use water treatment systems for
604 605
developing communities, Npj Clean Water. 1 (2018) 11. [6]
D.M. Warsinger, S. Chakraborty, E.W. Tow, M.H. Plumlee, C. Bellona, S. Loutatidou, L.
606
Karimi, A.M. Mikelonis, A. Achilli, A. Ghassemi, L.P. Padhye, S.A. Snyder, S. Curcio,
607
C.D. Vecitis, H.A. Arafat, J.H.L. V, Progress in Polymer Science A review of polymeric
608
membranes and processes for potable water reuse, Prog. Polym. Sci. 81 (2018) 209–237.
609
[7]
610 611
S. Munirasu, M.A. Haija, F. Banat, Use of Membrane technology for oil field and refinery produced water treatment–A review, Process Saf. Environ. Prot. 100 (2016) 183–202.
[8]
S.E. Weschenfelder, M.J.C. Fonseca, C.P. Borges, J.C. Campos, Application of ceramic
612
membranes for water management in offshore oil production platforms: Process design and
613
economics, Sep. Purif. Technol. 171 (2016) 214–220.
614
[9]
A.L. Ahmad, A.A. Abdulkarim, B.S. Ooi, S. Ismail, Recent development in additives
615
modifications of polyethersulfone membrane for flux enhancement, Chem. Eng. J. 223
616
(2013) 246–267.
617
[10] D. Breite, M. Went, A. Prager, A. Schulze, The critical zeta potential of polymer
618
membranes: how electrolytes impact membrane fouling, RSC Adv. 6 (2016) 98180–98189.
619
[11] H. Chen, G. Belfort, Surface modification of poly(ether sulfone) ultrafiltration membranes
620
by low‐ temperature plasma‐ induced graft polymerization, J. Appl. Polym. Sci. 72 (1999)
621
1699–1711.
622
[12] W. Yu, L.C. Campos, N. Graham, Application of pulsed UV-irradiation and pre-coagulation
623
to control ultrafiltration membrane fouling in the treatment of micro-polluted surface water,
624
Water Res. 107 (2016) 83–92.
625 626
[13] S. Leong, A. Razmjou, K. Wang, K. Hapgood, X. Zhang, H. Wang, TiO2 based photocatalytic membranes: A review, J. Memb. Sci. 472 (2014) 167–184.
627 628
[14] W. Yu, M. Brown, N.J.D. Graham, Prevention of PVDF ultrafiltration membrane fouling by coating MnO 2 nanoparticles with ozonation, Sci. Rep. 6 (2016) 1–12.
629
[15] K.M. Dobosz, C.A. Kuo-leblanc, T.J. Martin, J.D. Schi, Ultrafiltration membranes
630
enhanced with electrospun nanofibers exhibit improved Flux and fouling resistance, Ind.
631
Eng. Chem. Res. 56 (2017) 5724–5733.
632 633 634 635
[16] Y. Fang, Z. Xu, J. Wu, Surface Modification of Membranes, 3rd ed., John Wiley & Sons, Inc, 2013. [17] C.F. Wan, T. Yang, G.G. Lipscomb, D.J. Stookey, T.S. Chung, Design and fabrication of hollow fiber membrane modules, J. Memb. Sci. 538 (2017) 96–107.
636
[18] N.A.M. Nor, J. Jaafar, A.F. Ismail, M.A. Rahman, M.H.D. Othman, T. Matsuura, F. Aziz,
637
N. Yusof, W.N.W. Salleh, M.N. Subramaniam, Effects of heat treatment of TiO2 nanofibers
638
on the morphological structure of PVDF nanocomposite membrane under UV irradiation,
639
J. Water Process Eng. 20 (2017) 193–200.
640 641
[19] M. Nan, B. Jin, C.W.K. Chow, C. Saint, Recent developments in photocatalytic water treatment technology : A review, Water Res. 4 (2010) 2997–3027.
642
[20] M.S. Jyothi, M. Padaki, R.G. Balakrishna, R.K. Pai, Synthesis and design of PSf/TiO 2
643
composite membranes for reduction of chromium (VI): Stability and reuse of the product
644
and the process, J. Mater. Res. 29 (2014) 1537–1545.
645
[21] C.S. Ong, W.J. Lau, B. Al-anzi, A.F. Ismail, Photodegradation stability study of PVDF- and
646
PEI-based membranes for oily wastewater treatment process, Membr. Water Treat. 8 (2017)
647
211–223.
648 649
[22] S. Cao, J. Low, J. Yu, M. Jaroniec, Polymeric photocatalysts based on graphitic carbon nitride, Adv. Mater. 27 (2015) 2150–2176.
650 651
[23] J. Liu, H. Wang, M. Antonietti, Graphitic carbon nitride “reloaded”: emerging applications beyond (photo)catalysis, Chem. Soc. Rev. 45 (2016) 2308–2326.
652
[24] Q. Zheng, D.P. Durkin, J.E. Elenewski, Y. Sun, N.A. Banek, L. Hua, H. Chen, M.J. Wagner,
653
W. Zhang, D. Shuai, Visible-light-responsive graphitic carbon nitride: rational design and
654
photocatalytic applications for water treatment, Environ. Sci. Technol. 50 (2016) 12938–
655
12948.
656
[25] M.A. Mohamed, J. Jaafar, M.F.. M. Zain, L.J. Minggu, M.B. Kassim, M.S. Rosmi, N.H.
657
Alias, N.A.M. Nor, W.N.W. Salleh, M.H.D. Othman, In-depth understanding of core-shell
658
nanoarchitecture evolution of g-C3N4@C,N co-doped anatase/rutile:Efficient charge
659
separation and enhanced visible-light photocatalytic performance, Appl. Surf. Sci. 436
660
(2018) 302–318.
661
[26] N.H. Alias, J. Jaafar, S. Samitsu, N. Yusof, M.H.D. Othman, M.A. Rahman, A.F. Ismail, F.
662
Aziz, W.N.W. Salleh, N.H. Othman, Photocatalytic degradation of oilfield produced water
663
using graphitic carbon nitride embedded in electrospun polyacrylonitrile nanofibers,
664
Chemosphere. 204 (2018) 79–86.
665
[27] N. Abdullah, M.A. Rahman, M.H.D. Othman, A.F. Ismail, J. Jaafar, A.A. Aziz, Preparation
666
and characterization of self-cleaning alumina hollow fiber membrane using the phase
667
inversion and sintering technique, Ceram. Int. 42 (2016) 12312–12322.
668 669 670 671 672
[28] F. Rezaei, A. Nikiforov, R. Morent, N. De Geyter, Plasma modification of polylactic acid solutions to generate high quality electrospun PLA nanofibers, Sci. Rep. 8 (2018) 1–14. [29] J. Altmann, S. Ripperger, Particle deposition and layer formation at the crossflow microfiltration, J. Memb. Sci. 124 (1997) 119–128. [30] C.S. Ong, W.J. Lau, P.S. Goh, B.C. Ng, A.F. Ismail, Investigation of submerged membrane
673
photocatalytic reactor (sMPR) operating parameters during oily wastewater treatment
674
process, Desalination. 353 (2014) 48–56.
675 676 677 678 679 680
[31] M.S.A. Rahaman, A.F. Ismail, A. Mustafa, A review of heat treatment on polyacrylonitrile fiber, Polym. Degrad. Stab. 92 (2007) 1421–1432. [32] J. Brandrup, E.H. Immergut, E.A. Grulke, eds., Polymer Handbook, Fourth Edi, John Wiley & Sons Inc., 1999. [33] C.M. Hansen, Hansen solubility parameter-a user’s handbook, CRC Press: Boca Raton, FL, 2007.
681
[34] M. Ayan-Varela, S. Villar-Rodil, J.I. Paredes, J.M. Munuera, A. Pagan, A.A. Lozano-Perez,
682
J.L. Cenis, A. Martines-Alonso, J.M.D. Tascon, Investigating the dispersion behavior in
683
solvents , biocompatibility , and use as support for highly efficient metal catalysts of
684
exfoliated graphitic carbon nitride, Appl. Mater. Interfaces. 7 (2015) 24032–24045.
685
[35] A.B.D. Cassie, S. Baxter, Wettability of porous surfaces, Trans. Faraday Soc. 40 (1944)
686 687 688 689 690
546. [36] X. Wang, J. Yu, G. Sun, B. Ding, Electrospun nanofibrous materials: a versatile medium for effective oil/water separation, Mater. Today. 19 (2016) 403–414. [37] R. Wang, Y. Liu, B. Li, B.S. Hsiao, B. Chu, Electrospun nanofibrous membranes for high flux microfiltration, J. Memb. Sci. 392–393 (2012) 167–174.
691
[38] V. Gekas, B. Hallstrom, Mass transfer in the membrane concentration polarization layer
692
under turbulent cross flow. I. Critical literature review and adaptation of existing sherwood
693
correlations to membrane operation, J. Memb. Sci. 30 (1987) 153–170.
694 695
[39] Z.F. Cui, H.S. Muralidhara, Membrane technology: A practical guide to membrane technology and applications in food and bioprocessing, Elsevier Ltd., 2010.
696 697
[40] S. Homaeigohar, M. Elbahri, Nanocomposite electrospun nanofiber membranes for environmental remediation, Materials (Basel). 7 (2014) 1017–1045.
698
[41] M. Makaremi, R.T. De Silva, P. Pasbakhsh, R.T. De Silva, P. Pasbakhsh, Electrospun
699
nanofibrous membranes of polyacrylonitrile/halloysite with superior water filtration ability,
700
J. Phys. Chem. C. 119 (2015) 7949–7958.
701
[42] S. Kaur, S. Sundarrajan, D. Rana, R. Sridhar, R. Gopal, T. Matsuura, S. Ramakrishna,
702
Review: The characterization of electrospun nanofibrous liquid filtration membranes, J.
703
Mater. Sci. 49 (2014) 6143–6159.
704
[43] G. Li, T. An, J. Chen, G. Sheng, J. Fu, F. Chen, S. Zhang, H. Zhao, Photoelectrocatalytic
705
decontamination of oilfield produced wastewater containing refractory organic pollutants
706
in the presence of high concentration of chloride ions, J. Hazard. Mater. 138 (2006) 392–
707
400.
708 709
[44] C. Yang, G. Zhang, N. Xu, J. Shi, Preparation and application in oil ± water separation of ZrO2 /α-Al2O3 MF membrane, J. Memb. Sci. 142 (1998) 235–243.
710
[45] X. Hu, Y. Yu, J. Zhou, Y. Wang, J. Liang, X. Zhang, Q. Chang, L. Song, The improved
711
oil/water separation performance of graphene oxide modified Al2O3 microfiltration
712
membrane, J. Memb. Sci. 476 (2015) 200–204.
713
[46] J. Zhou, Q. Chang, Y. Wang, J. Wang, G. Meng, Separation of stable oil – water emulsion
714
by the hydrophilic nano-sized ZrO2 modified Al2O3 microfiltration membrane, Sep. Purif.
715
Technol. 75 (2010) 243–248.
716
717 718
719
Highlights
720
Graphitic carbon nitride (GCN) photocatalyst was incorporated in nanofibers
721
Photocatalytic nanofibers were electrospun on alumina hollow fiber membranes
722
Cross-flow microfiltration demonstrated purification of oilfield produced water
723
Hydrophilic, highly-porous nanofiber coating exhibited excellent fouling resistance
724
Photodegradation ability of GCN offered cleaning performance of robust membranes
725