Desalination 432 (2018) 32–39

Contents lists available at ScienceDirect

Desalination
journal homepage: www.elsevier.com/locate/desal

Polyoxadiazole hollow fibers for produced water treatment by direct contact T

membrane distillation
Jingli Xua,e,1, N.M. Srivatsa Bettahallia,1, Stefan Chiscaa, Mohammed Khalil Khalidb,
⁎
Noreddine Ghaffoura,c, Régis Vilaginesd, Suzana P. Nunesa,
a
King Abdullah University of Science and Technology (KAUST), Biological and Environmental Science and Engineering (BESE), 23955-6900 Thuwal, Saudi Arabia
b
King Abdullah University of Science and Technology (KAUST), Analytical Core Lab, 23955-6900 Thuwal, Saudi Arabia
c
King Abdullah University of Science and Technology (KAUST), Water Desalination and Reuse Center (WDRC), Biological and Environmental Science and Engineering
(BESE), 23955-6900 Thuwal, Saudi Arabia
d
Saudi Aramco, Oil & Gas Treatment Division, 31311 Dhahran, Saudi Arabia
e
Xuchang University, Chemical Engineering and Chemistry Department, China

A R T I C L E I N F O A B S T R A C T

Keywords: Treatment of produced water in the petroleum industry has been a challenge worldwide. In this study, we
Hollow fibers evaluated the use of direct contact membrane distillation (DCMD) for this purpose, removing oil and dissolved
Polyoxadiazole elements and supplying clean water from waste. We synthesized fluorinated polyoxadiazole, a highly hydro-
Produced water phobic polymer, to fabricate hollow fiber membranes, which were optimized and tested for simulated produced
Membrane distillation
water and real produced water treatment. The process performance was investigated under different operating
parameters, such as feed temperature, feed flow velocity and length of the membrane module for 4 days. The
results indicate that by increasing feed temperature and feed flow rate the vapor flux increases. The flux de-
creased with increasing the length of the module due to the decrease of the driving force along the module. The
fouling behavior, which corresponds to flux decline and cleaning efficiency of the membrane, was studied. The
performance of the fabricated hollow fiber membranes was demonstrated for the treatment of produced water,
complying with the industrial reuse and discharge limits.

1. Introduction as an opportunity for produced water treatment to provide a viable

The regular operations of gas and oil industry involve large amount
of injected water to facilitate the petroleum recovery. This water is
brought to the surface along with hydrocarbons (oil and gas), salt and
other solutes, and is commonly known as “produced water”. The
composition depends on onsite natural geological information, con-
taining various soluble mineral ions and being often acidic in nature.
Fresh water is also used for desalting crude oil, sweetening of gas and
other refinery processes. The liquid waste stream produced by the
petroleum industry is estimated to be around 250 million barrels per
day (water to oil ratio of at least 3:1) [1]. The water reuse or recycling
has become mandatory, especially in water stressed countries. Many
countries have implemented more stringent regulatory standards for
the permitted oil and grease (O&G) limits for discharging produced
water, which range from 10 mg/L, according to the China Environ-
mental Ministry, to the maximum limit of 42 mg/L, regulated by United
States Environmental Protection Agency (USEPA) [2]. This can be seen
source of water for beneficial use in many applications, for which
drinking water quality is not required. New regulations have been
promoting the development of environmentally friendly and econom-
ical disposal methods [3] to prevent serious environmental damage. As
a consequence, physical, chemical and biological methods have been
proposed to treat produced water. A single technology can hardly meet
the requirements for a clean effluent and two or more treatment sys-
tems might be used in series operation [2].
Membrane distillation (MD) was considered in this work as a po-
tential technology for produced water treatment. MD is based on
thermal gradient transport of vapor through hydrophobic porous
membranes, the driving force being the vapor pressure difference be-
tween the two sides of the membrane pore [4]. MD combines the ad-
vantages of conventional distillation and membrane technologies: high
selectivity for non-volatile compounds, meaning 100% retention of
ions, macromolecules, colloids etc. It consumes less energy, operates at
low temperatures (compared to conventional non-membrane

⁎
Corresponding author.
E-mail address: suzana.nunes@kaust.edu.sa (S.P. Nunes).
1
These authors contributed equally to this work (co–first authors).

https://doi.org/10.1016/j.desal.2017.12.014
Received 29 August 2017; Received in revised form 2 December 2017; Accepted 10 December 2017
0011-9164/ © 2017 Elsevier B.V. All rights reserved.
J. Xu et al.
processes), is a low hydrostatic pressure process (in comparison to
pressure driven membrane process like reverse osmosis (RO)), is com-
pact (for offshore deployment), simple, and is claimed to be less sus-
ceptible to fouling than other membrane processes [5-7]. Most of the
published MD works focus on seawater desalination. Only few groups
have explored it for the treatment of produced water by MD [8–12].
Whether desalination or treatment of produced water/wastewater, the
membranes for MD should be highly porous and hydrophobic to pre-
vent the feed solution from entering into the pores (high liquid entry
pressure). Commonly used membranes for MD are polytetra-
fluorethylene (PTFE), polypropylene (PP) and polyvinylidene fluoride
(PVDF) with typical pore sizes varying from 0.1 μm to 1 μm
[5,9,13–22]. The MD membrane performance can be enhanced by dif-
ferent means, such as incorporation of nanoparticles [23–33], graphene
oxide [34] or use of alternative solvents such as triethyl phosphate
(TEP) [35] for the membrane manufacture. Few efforts have been
dedicated to development of new polymers, tailor-made for the needs of
MD [36–37]. Our group has previously reported polyazole-based
membranes for water desalination by MD [38–40]. Polyoxadiazole
could be an excellent material comparable to ceramic membrane due to
its high thermal and oxidative stability and good mechanical strength
[41].
In this study, fluorinated polyoxadiazole hollow fiber membranes
have been fabricated, characterized and optimized for produced water
treatment. Synthetic and real produced water samples were in-
vestigated. Various parameters influencing the permeate flux and re-
jection factors for salts and organic compounds have been investigated
in DCMD operation. The performance of polyoxadiazole hollow fiber
membranes, as well as their fouling susceptibility, have been in-
vestigated in long-term separation/operation stability tests.
2. Materials and methods
2.1. Materials and chemicals
Fluorinated polyoxadiazole (f-POD) was synthesized according to
previously reported procedure [40]. N-methyl-2-pyrrolidinone (NMP,
≥ 99.5%, Merck) was used as solvent to prepare hollow fiber spinning
dope solution. Methanol (HPLC grade, Fischer scientific) and glycerol
(≥ 99%, Sigma Aldrich) were used in the process hollow fiber fabri-
cation and long periods of storage. Produced water was supplied by
Saudi Aramco, from one of the drilling locations in Saudi Arabia.
2.2. Fiber spinning
The synthesized polyoxadiazole was washed with methanol to re-
move any moisture and dried in vacuum oven at 80 °C overnight, before
using it to prepare dope solution. The dried polymer flakes were added
to NMP in small portions to avoid lump formation and stirred at
600 rpm for 24 h at 70 °C, using overhead mechanical stirrer and hot oil
bath. The prepared dope solution was loaded into the high-pressure
syringe holder of the spinning machine and degassed to remove any air
bubbles entrapped within the solution.
A set-up was built to spin hollow fibers from the laboratory-syn-
thesized polymers, with capacity for 20 to 30 g polymer. The dope
solution was loaded in a stainless steel syringe and extruded through
the spinneret using a high-pressure syringe pump (neMESYS high
pressure syringe pumps, Cetoni GmbH). The extruded polymer solution
was immersed in a water tank to form the hollow fibers by phase in-
version. The bore liquid was pumped through the needle of the spin-
neret to form the lumen of the hollow fiber, using HPLC dual head gear
pump (Lab Alliance ™).
The hollow fibers were washed by replenishing fresh water (purified
by RO) for at least 10 h (at room temperature and a few trials at 50 °C)
and stored in RO water for three days (exchanging the water every day)
to remove any residual solvent. Finally, the hollow fibers were kept in
33
Desalination 432 (2018) 32–39
25:75 (vol%) glycerol/water solution overnight, slowly air dried and
stored in dry environment. Before use, the fibers were extensively wa-
shed with water to remove the glycerol.
2.3. Membrane characterization
2.3.1. Hollow fiber membrane morphology
The membrane morphology was investigated by field emission
scanning electron microscopy (SEM) (FEI Quanta 600 FE or Nova Nano
SEM microscopes) at accelerating voltage of 5 kV. The fibers were dried
overnight and were carefully fractured in liquid nitrogen for cross
sectional SEM analysis. The fibers were sputter coated with Iridium
(∼ 3 nm-thick, Quorum Q150T ES) to make the polymer surface con-
ductive for surface analysis.
2.3.2. Pore size distribution and membrane liquid entry pressure (LEP)
The membrane pore size distribution was estimated by using
Porolux™ 1000 (Porometer.com, Belgium), at the pressure range from 0
up to 34.5 bar and using perfluoroether (Porefil) with a surface tension
of 16 dynes cm− 1 as pore filler. The same analysis supplied the liquid
entry point (LEP), which is the minimum pressure required for liquid to
pass through the membrane.
2.4. DCMD experimental set-up
The schematic of the experimental setup for cross-flow DCMD is
shown in Fig. 1.
All the experiments were carried out in modules containing 3 to
4 hollow fibers, using 6 mm polyethylene tube housings, which were
approximately 15 cm long. The fibers were carefully inserted into the
polyethylene tube and sealed with epoxy glue (Devcon, 2 components
flow mixture, 5 min epoxy, 1500 psi strength). Fig. 2 shows the hollow
fiber modules. At the beginning and end of the housing, T-junctions
were placed to circulate cold water in the shell side to condense the
permeating water vapor. Before measuring the vapor flux, the fibers
were completely dried to regain their intrinsic hydrophobicity. The
vapor flux (JV) through the membrane was measured by pumping
(peristatic pumps, Cole Parmer) hot feed (produced water) through the
lumen of the hollow fiber and cold RO water (countercurrent flow)
through the shell side of the module. The change in weight of the
permeate reservoir over time (10 min interval) at steady state was re-
corded, using the data acquisition software on to a computer. Si-
multaneously temperatures, pressures and flowrates were recorded. The
data presented in this work is an average of at least 3 hollow fiber
modules prepared from different sections of the total fiber spun.
The salt rejection (SR) and permeate vapor flux (Jv) were calcu-
lated, using the following equations:
Cp ⎞
SR% = ⎜⎛1 − ⎟ × 100
⎝ Cf ⎠ (1)
mp
JV =
At (2)
where Cp and Cf are the salt concentration of permeate and feed solu-
tions, respectively, and mp is the mass of collected permeate, A is the
effective membrane area, and t is the running time.
2.5. Produced water analysis
The total organic content was analyzed by Dohrmann Series Apollo
9000 TOC analyzer. Total dissolved solids (TDS) was analyzed, using an
OAKTON conductivity/TDS meter.
2.6. Treatment of produced water by DCMD
Real produced water supplied by Saudi Aramco was used as feed
J. Xu et al.
Fig. 1. Schematic of cross flow hollow fiber membrane based DCMD bench scale experimental set-up (C1, C2 are conductivity sensors; P1, P2 are pressure sensors; T1, T2, T3, T4 are
temperature sensors).
Fig. 2. Hollow fiber module potted inside polyethylene tube, using epoxy glue.
solution. Pre-treatment was carried out by aerating (air or nitrogen) at
approximately 0.5 bar pressure for 24 h at room temperature and rested
for 12 h. The upper clear solution was taken as feed solution. For all the
following experiments, the modules were prepared with four hollow
fibers, which correspond to a total membrane active surface area of
approximately 0.001718 m2. The effective length of the module was
approximately 15 cm.
2.7. Hollow fiber membrane fouling analysis and cleaning methods
The hollow fiber lumen was analyzed by SEM, after operation, to
visualize the fouling layer and potential salt crystal growth on the
surface. Elemental analysis of the fouling layer was conducted by
Energy dispersive X-ray spectroscopy analysis (EDS).
The cleaning efficiency of the fouled lumen/inner surface of the
hollow was evaluated by flushing the tubing of the MD set-up and the
feed channel of membrane module with water until its conductivity
was < 25 μS/cm. Multiple cycles of fouling-cleaning were conducted to
check the performance and stability of the fiber.
34
Desalination 432 (2018) 32–39
3. Results and discussion
In a previous study comparing flat-sheet membranes based on PVDF
and fluorinated polyoxadiazole [40], which were prepared under si-
milar conditions, we confirmed that the water contact angle for poly-
oxadiazole was higher (120 o) than for PVDF (88 o). Preliminary DCMD
tests [39] using pure water showed that fluorinated polyoxadiazole
hollow fibers had higher flux than those based on PVDF, prepared and
operated under similar conditions. Therefore, we focused our study on
produced water, by systematically investigating the performance of
fluorinated polyoxadiazole hollow fibers.
3.1. Preparation of fluorinated polyoxadiazole hollow fiber membranes
The fluorinated polyoxadiazole hollow fibers were fabricated by
using a method analogous to that previously reported by Maab et al.
[39], but using water as bore liquid. The morphology of selected hollow
fibers is shown in Fig. 3. In summary, the study was carried out by
varying the polymer concentration of the dope solution ranging from 18
to 20 wt% dissolved in NMP. The influence of the dope and bore liquid
flow rate, air gap and collector bath temperature was investigated. The
hollow fibers fabricated with a dope solution containing 18 wt%
polymer were mechanically not stable. The top layer of those prepared
with 20 wt% dope solution was excessively dense with scarcely dis-
tributed nanopores. Hollow fibers from 19 wt% dope solutions had the
best optimized nanopores. An improved water flux (275 kg
m− 2 h− 1 bar− 1) was observed for hollow fibers washed at elevated
temperature.
Further optimization was carried out by changing the composition
of the bore liquid from pure water to a mixture of NMP (solvent) and
water, to reduce the demixing rate and increase the porosity of the
inner surface. Data corresponding to different fabricated hollow fibers
are listed in Table 1.
The formed hollow fibers were analyzed using SEM to check the
pore distribution and interconnectivity across the wall and surface, and
the exact dimensions of the hollow fiber formed (lumen diameter and
wall thickness). Fig. 3 shows the SEM images of POD-5. The surface
pores and distinctive pore distribution across the wall of the hollow
fiber can be seen. The pore size distribution and liquid entry point were
characterized by using a Porolux 1000. The results are listed in Fig. 4.
J. Xu et al. Desalination 432 (2018) 32–39

Fig. 3. SEM images of 19 wt% polyoxadiazole dope solution, fabricated with air gap of 30 cm, bore liquid flow rate of 5 mL/min and 80:20 (vol%) NMP: water mixture as bore liquid.

Table 1
Fluorinated polyoxadiazole hollow fiber membranes fabricated with different bore liquid
compositions and dope flow rates, using a dope solution with 19 wt% polymer, air gap of
30 cm, and 5 mL/min bore liquid flow rate.

Hollow fiber Dope flow rate Bore liquid NMP:water ratio

(mL/min) (vol%)

POD-1 6 0:100
POD-2 50:50
POD-3 60:40
POD-4 70:30
POD-5 80:20
POD-6 5 80:20
Fig. 5. DCMD permeate flux and salt rejection measured for various fabricated hollow
fibers (feed solution 90 g/L NaCl; feed flow rate: 50 mL/min; permeate flow rate: 50 mL/
min; one fiber per module).

temperature at 20 °C. Feed and coolant flow rates were maintained at

1:1 ratio at 50 mL/min. All experiments were carried out for at least
4 h. The water vapor flux measured for different hollow fibers is plotted
in Fig. 5. For all tested hollow fibers the salt rejection was measured
based on the permeate conductivity, which was > 99.9%. The water
vapor flux for the POD-5 hollow fiber was higher than those of all other
fabricated fibers. The flux almost doubled, compared to the POD-1,
which was fabricated using water as bore liquid. By taking this into
consideration, the POD-5 membranes were selected for experiments of
produced water treatment. The high content of NMP in the bore fluid
delays the phase separation leading to the pore formation. An internal
more porous skin is then formed.
Fig. 4. Average pore size and liquid entry point (LEP) of various fabricated hollow fibers.

High LEP values indicate that the tendency of pore wetting would be
less pronounced and higher operational pressure difference could be
tolerated. As shown in Fig. 4, POD 4 and POD 5 had the highest LEP
values, when compared to the other fabricated membranes.
The hollow fibers were tested in the DCMD experimental set-up,
using first 90 g/L NaCl dissolved in deionized (DI) water, as test feed
solution. The feed temperature was kept at 70 °C and the coolant
3.2. Hollow fiber membrane performance for treatment of produced water
by DCMD
3.2.1. MD tests with simulated produced water
Before performing the MD experiments with fluorinated poly-
oxadiazole membranes using real samples of produced water as feed,
simulated water containing crude oil of 0.1 g/L was used. The

35
J. Xu et al.
Table 2
Composition of simulated produced water.
Compounds Concentration (g/L)
Sodium chloride (NaCl) 50
Calcium chloride dihydrate (CaCl2·2H2O) 16
Magnesium chloride hexahydrate (MgCl2·6H2O) 8 Feed T(outlet)/°C
Sodium sulfate (Na2SO4) 2 Cold permeate T(outlet)/°C
Sodium bicarbonate (NaHCO3) 1 ΔT(outlet)/°C
Crude oil (from Aramco) 0.1
Fig. 6. Long-term DCMD experiment (Feed 100 mL/min; permeate 340 mL/min; 3 fi-
bers).
composition of simulated produced water (provided by Saudi Aramco)
was listed in Table 2. The conductivity was 96.3 mS/cm. Simulated
produced water was treated by POD 5 membrane at the feed and
permeate temperature of 70 °C and 20 °C, respectively. The permeate
and feed flow rates were 340 mL/min and 100 mL/min, respectively.
The results were shown in Fig. 6. The flux decreased from 6.5 kg
m− 2 ∙h− 1 to 5 kg m− 2·h− 1 in the initial 3 h and then slightly decreased
over longer time. The sharp decrease in the initial time might be caused
by oil accumulation in the surface of the membrane. The absence of oil
in the permeate side indicates that oil was not transported through the
membrane. The conductivity of the permeate was about 200 μS/cm and
the rejection was > 99.9% after 100 h operation. The fluorinated
polyoxadiazole membranes had therefore an excellent performance in
treating produced water, at least with the composition described in
Table 2.
3.2.2. Effect of the feed inlet temperature on the permeate flux
The effect of the feed inlet temperature on the permeate flux was
investigated with a feed flow rate of 155 mL/min and a permeate flow
rate of 340 mL/min. The results are shown in Fig. 7. The temperature
change between inlet and outlet was listed in Table 3. Heat loss, which
Fig. 7. Effect of the inlet feed (produced water) temperature on the permeate flux (4
fibers per module; feed flow rate: 155 mL/min; permeate flow rate: 340 mL/min).
36
Desalination 432 (2018) 32–39
Table 3
DCMD experimental conditions for treatment of produced water.
Condition
Feed T(inlet)/°C 43.88 52.31 61.28 70.46
Cold permeate T(inlet)/°C 20.22 20.56 20.80 21.06
41.09 49.97 57.5 64.1
21.49 21.87 23.39 24.37
19.60 26.1 32.11 37.34
is caused by the fact that the membrane is not an excellent thermal
insulator, could be detected from the inlet to outlet in both sides of the
membrane. The increase of the water flux was observed with the in-
crease of the feed temperature, maintaining the constant permeate
temperature. The water flux increased from 4.4 to 17.9 kg m− 2 h− 1, by
increasing the feed inlet temperature from 40 °C to 70 °C. The driving
force for the water transport is increased, because the water vapor
pressure difference between the permeate and feed side increases, as
the feed temperature increases. In membrane distillation, the permeate
flux can be described as the function of vapor pressure gradient across
the membrane, JV = BwΔPw, where Bw is a constant relevant to the
membrane characteristics and ΔPw is the transmembrane water vapor
pressure difference [42–44].
With the assumption of liquid-vapor equilibrium at the membrane
surface, the vapor pressure of the salt solution can be described by
P = γ (1 − x ) P ∗ (3)
where P is the water vapor pressure, x is the mole fraction of the solute
in the salt solution, γ is the water activity coefficient which is function
of temperature and solute concentration and P* is the vapor pressure of
pure water which is function of temperature [45]. The composition of
the produced water was kept constant in the experiments; γ and P only
depend on the temperature. The water activity coefficient and water
vapor of pure water increase with the increase of temperature [46].
3.2.3. Effect of feed flow rate on permeate flux
The effect of feed flow rates on the permeate flux was investigated
for feed and permeate temperatures of 70 °C and 20 °C, respectively.
During this study, the cold permeate flow rate was kept constant at
340 mL/min. The investigated flow rates were 80, 120, 155, 170, 210
and 250 mL/min (which corresponds to Reynolds numbers of 990,
1485, 1918, 2103, 2599, 3093, respectively). The results are shown in
Fig. 8. The salt rejection was above 99.98%.
The water vapor flux increased with the increase of the feed flow
rate. This is because the increase of the feed flow rate reduced the
temperature polarization. The temperature and salt concentration at
the membrane surface become closer to the corresponding bulk tem-
perature and bulk concentration, resulting in higher transmembrane
Fig. 8. Effect of the feed (produced water) flow rate on the permeate flux (number of
fibers per module: 4; feed temperature: 70 °C; permeate temperature: 20 °C).
J. Xu et al.
Fig. 9. Effect of the hollow fiber module length on the permeate flux (number of fibers
per module: 4; feed flow rate: 155 mL/min; permeate flow rate: 340 mL/min; feed tem-
perature: 70 °C; permeate temperature: 20 °C).
temperature difference and greater permeate flux. The feed flow rate
must be well controlled in order to avoid membrane pore wetting, as
the transmembrane hydrostatic pressure (TMP) must be lower than the
liquid entry pressure (LEP). Higher flow rates above 250 mL/min were
not tested due to the limitation of liquid entry pressure (LEP) of the
membrane, which was shown in Fig. 4.
3.2.4. Effect of the length of membrane module on permeate flux
The effect of the module length on the performance was in-
vestigated and the results are shown in Fig. 9. The flux significantly
decreased with the increase of the length of the membrane module.
This is attributed to the increase of residence time as the length of
the module increases. We assume that the cross-sectional area of the
fluid channel is constant along the module, and then the following
equation can be used to calculate the residence time, which relates to
flow rate and membrane length (module effective length):
Effective length of module
residence time(τ ) =
fluid velocity
The increase of the effective length of module provides enough time
for the fluids to exchange mass and heat across the membrane. It causes
the decrease of temperature at the outlet of the membrane in the hot
side, which will cause the temperature difference across the membrane
(the driving force of mass transfer) to decrease. This behavior is de-
monstrated by the experimental results shown in Fig. 9.
3.2.5. Long term experiment for real produced water
After the operational conditions optimization shown above, the
long-term performance stability was tested. The experiments were
performed with POD-5 hollow fibers, using real produced water (with
84 g/L TDS, 41 mg/L TOC and 82 mS/cm conductivity). Long-term
experiments were carried out with feed and permeate inlet temperature
of 70 °C and 20 °C, respectively. The feed and permeate flow rates were
maintained at 155 mL/min and 340 mL/min, respectively. The con-
ductivity of the produced water feed increased from 82 to 129 mS/cm,
after 100 h DCMD operation, due to continuous recirculation of the
retentate. The conductivity of the permeate was about 195 μS/cm,
which corresponds to salt rejection larger than 99.7% or salt con-
centration of condensed permeate water ≈0.088 mg/L or 88 ppb (the
minimum allowable standard for discharging produced water by China
environmental ministry is 10 mg/L). As shown in Fig. 10, the permeate
flux for POD-5 membrane was stable during experiments, lasting for
100 h. The permeate flux was maintained at 12.5–11.8 kg m− 2·h− 1.
The permeate had 1.6 mg/L TOC and 197 TDS.
3.2.6. Fouling and cleaning study
A slight gradual decrease of the permeate flux occurred during the
Desalination 432 (2018) 32–39
Fig. 10. Experimental results showing hollow fibers long term performance and corre-
sponding salt rejection (number of fibers per module: 4; feed flow rate: 155 mL/min;
permeate flow rate: 340 mL/min; feed temperature: 70 °C; permeate temperature: 20 °C).
first 100 h of operation (blue line shown in Fig. 11). One reason for the
flux decline might be fouling caused by salt scale, which could be
clearly observed in SEM images (Fig. 12). Fig. 12 shows the SEM images
and corresponding elemental analysis performed by EDS, which in-
dicate the occurrence of salt scaling on the surface of the membrane.
Once scaling occurs on the membrane surface, the membrane per-
meation can be recovered by effective cleaning. Membrane cleaning is
usually achieved by physical and chemical processes. For chemical
cleaning, a chemical agent should have a good solubility for the scale,
while not causing a substantial damage to the membrane. In this study,
only a physical method, deionized (DI) water flushing, was in-
vestigated, as we didn't observe severe fouling.
The efficiency of cleaning with DI water is shown in Fig. 11. The
results clearly indicate that the flux can be recovered to the initial
value. The flux was stable (17–16 kg m− 2 h− 1) for about 90 h. Salt
rejection was also higher than 99.5% after cleaning. The SEM images
(Fig. 12) and the results of EDS show that DI water cleaning is effective
to remove salt scale formed on the inner surface (lumen) of hollow
(4) fiber.
4. Conclusion
Nanoporous hydrophobic hollow fiber membranes were fabricated
using fluorinated polyoxadiazole. Their performance was investigated
for treatment of produced water by DCMD. The influence of the bore
liquid composition during the hollow fiber fabrication on their perfor-
mance was clearly demonstrated. The vapor flux through the mem-
branes increased with the increase of the NMP concentration in the bore
fluid during the fiber fabrication. Furthermore, various operating
parameters, like feed inlet temperature, feed flow rate and module
length, were tested for the optimization. The permeate flux increased
with the increase of the feed flow rate and significantly decreased with
the increase of the length of the membrane module.
A stable water vapor flux of at least 12 kg m− 2 h− 1 and very high
total dissolved solid (TDS) rejection of > 99.5% (conductivity of
194.7 μS/cm, corresponding to a salt concentration of 88 ppb) have
been achieved during 100 h of continuous operation. Although this
would be an acceptable value of salt content for many applications, a
continuous decrease of rejection could indicate that an undesirable
wetting is taking place or the substantial scaling on the membrane
surface is locally increasing the salt concentration much higher levels.
Both factors could compromise extended long-term operation. Taking
this concern into consideration, mild cleaning tests, consisting of
washing the membranes with DI water, were performed, and the flux
could be recovered to the initial value.
The achieved results indicate that the DCMD process using fluori-
nated polyoxadiazole hollow fibers could be a viable solution for
treatment of produced water with compositions similar to those
37
J. Xu et al.
investigated here. The MD process could operate with 2 identical
membrane modules, exchangeable in sequences of 4 days, to allow re-
generation and performance restoration. In this form, a scalable mod-
ular MD system could be considered for treatment of produced water,
industrial waste-water and other types of contaminated water.
Acknowledgement
The research reported in this publication was financially supported
by Saudi Aramco (Reference – RGC/3/1582-01-01) and King Abdullah
University of Science and Technology (KAUST). The authors thank
colleagues of the Nanostructured Polymeric Membrane Lab (npm.kaust.
edu.sa) and Water Desalination and Reuse Center (WDRC) for their
valuable scientific inputs and data analysis.
38
Desalination 432 (2018) 32–39
Fig. 11. Results illustrating multi stage cleaning efficiency with DI
water (number of fibers per module: 4; feed flow rate:155 mL/min;
permeate flow rate: 340 mL/min; feed temperature: 70 °C;
permeate temperature: 20 °C). (For interpretation of the references
to color in this figure, the reader is referred to the web version of
this article.)
Fig. 12. SEM image and corresponding EDS elemental analysis of
fouled and cleaned fluorinated polyoxadiazole hollow fiber
membranes.
References
[1] Z. Khatib, P. Verbeek, Water to Value - Produced Water Management for
Sustainable Field Development of Mature and Green Fields, in Society of Petroleum
Engineers, (2003), http://dx.doi.org/10.2118/73853-MS.
[2] A. Fakhru'l-Razi, A. Pendashteh, L.C. Abdullah, D.R.A. Biak, S.S. Madaeni,
Z.Z. Abidin, Review of technologies for oil and gas produced water treatment, J.
Hazard. Mater. 170 (2009) 530–551.
[3] S. Mondal, S.R. Wickramasinghe, Produced water treatment by nanofiltration and
reverse osmosis membranes, J. Membr. Sci. 322 (2008) 162–170.
[4] M. Khayet, T. Matsuura, Membrane Distillation Principles and Applications
(Chapter 1), 1st ed., Elsevier, 2011.
[5] R. Tian, H. Gao, X.H. Yang, S.Y. Yan, S. Li, A new enhancement technique on air gap
membrane distillation, Desalination 332 (2014) 52–59.
[6] M. Qtaishat, M. Khayet, T. Matsuura, Novel porous composite hydrophobic/hy-
drophilic polysulfone membranes for desalination by direct contact membrane
distillation, J. Membr. Sci. 341 (2009) 139–148.
[7] A. Politano, P. Argurio, G. Di Profio, V. Sanna, A. Cupolillo, S. Chakraborty,
H.A. Arafat, E. Curci, Photothermal membrane distillation for seawater desalina-
tion, Adv. Mater. 29 (2017) 1603504.
J. Xu et al.
[8] A. Alkhudhiri, N. Darwish, N. Hilal, Produced water treatment: application of air
gap membrane distillation, Desalination 309 (2013) 46–51.
[9] D. Singh, K.K. Sirkar, Desalination of brine and produced water by direct contact
membrane distillation at high temperatures and pressures, J. Membr. Sci. 389
(2012) 380–388.
[10] F. Macedonio, A. Ali, T. Poerio, E. El-Sayed, E. Drioli, M. Abdel-Jawad, Direct
contact membrane distillation for treatment of oilfield produced water, Sep. Purif.
Technol. 126 (2014) 69–81.
[11] N.A. Elsayed, M.A. Barrufet, M.M. El-Halwagi, An integrated approach for in-
corporating thermal membrane distillation in treating water in heavy oil recovery
using SAGD, J. Unconv. Oil Gas Resour. 12 (2015) 6–14.
[12] A. Hussain, J.M. Matar, A. Janson, S. Adham, Treatment of produced water from oil
& gas operations by Membrane Distillation, Proceedings of the 4th International Gas
Processing Symposium, Elsevier, Doha, Qatar, 2014, pp. 285–292.
[13] S. Adnan, M. Hoang, H. Wang, Z. Xie, Commercial PTFE membranes for membrane
distillation application: effect of microstructure and support material, Desalination
284 (2012) 297–308.
[14] H.J. Hwang, K. He, S. Gray, J. Zhang, I.S. Moon, Direct contact membrane dis-
tillation (DCMD): experimental study on the commercial PTFE membrane and
modeling, J. Membr. Sci. 371 (2011) 90–98.
[15] P. Jacob, P. Phungsai, K. Fukushi, C. Visvanathan, Direct contact membrane dis-
tillation for anaerobic effluent treatment, J. Membr. Sci. 475 (2015) 330–339.
[16] K. Kezia, J. Lee, M. Weeks, S. Kentish, Direct contact membrane distillation for the
concentration of saline dairy effluent, Water Res. 81 (2015) 167–177.
[17] M.M.A. Shirazi, A. Kargari, M. Tabatabaei, Evaluation of commercial PTFE mem-
branes in desalination by direct contact membrane distillation, Chem. Eng. Process.
Process Intensif. 76 (2014) 16–25.
[18] T.Y. Cath, V.D. Adams, A.E. Childress, Experimental study of desalination using
direct contact membrane distillation: a new approach to flux enhancement, J.
Membr. Sci. 228 (2004) 5–16.
[19] M.A. Izquierdo-Gil, M.C. Garíca-Payo, C. Fernández-Pineda, Air gap membrane
distillation of sucrose aqueous solutions, J. Membr. Sci. 155 (1999) 291–307.
[20] A.C. Sun, W. Kosar, Y. Zhang, X. Feng, Vacuum membrane distillation for desali-
nation of water using hollow fiber membranes, J. Membr. Sci. 455 (2014) 131–142.
[21] L. Basini, G. D'Angelo, M. Gobbi, G.C. Sarti, C. Gostoli, A desalination process
through sweeping gas membrane distillation, Desalination 64 (1987) 245–257.
[22] L. Francis, N. Ghaffour, A.S. Alsaadi, G.L. Amy, Fabrication and characterization of
functionally graded poly(vinylidine fluoride)-silver nanocomposite hollow fibers
for sustainable water recovery, Sci. Adv. Mater. 6 (2014) 2659–2665.
[23] M. Bhadra, S. Roy, S. Mitra, Desalination across a graphene oxide membrane via
direct contact membrane distillation, Desalination 378 (2016) 37–43.
[24] J.E. Efome, M. Baghbanzadeh, D. Rana, T. Matsuura, C.Q. Lan, Effects of super-
hydrophobic SiO2 nanoparticles on the performance of PVDF flat sheet membranes
for vacuum membrane distillation, Desalination 373 (2015) 47–57.
[25] E. Drioli, A. Ali, S. Simone, F. Macedonio, S.A. Al-Jlil, F.S. Al Shabonah, H.S. Al-
Romaih, O. Al-Harbi, A. Figoli, A. Criscuoli, Novel PVDF hollow fiber membranes
for vacuum and direct contact membrane distillation applications, Sep. Purif.
Technol. 115 (2013) 27–38.
[26] B. Li, K.K. Sirkar, Novel membrane and device for vacuum membrane distillation-
based desalination process, J. Membr. Sci. 257 (2005) 60–75.
[27] A. Rastegarpanah, H.R. Mortaheb, Surface treatment of polyethersulfone mem-
branes for applying in desalination by direct contact membrane distillation,
39
Desalination 432 (2018) 32–39
Desalination 377 (2016) 99–107.
[28] Z.D. Hendren, J. Brant, M.R. Wiesner, Surface modification of nanostructured
ceramic membranes for direct contact membrane distillation, J. Membr. Sci. 331
(2009) 1–10.
[29] A.K. An, E.-J. Lee, J. Guo, S. Jeong, J.-G. Lee, N. Ghaffour, Enhanced vapor
transport in membrane distillation via functionalized carbon nanotubes anchored
into electrospun nanofibers, Sci. Rep. 7 (2017) 41562–41572.
[30] C. Boo, J. Lee, M. Elimelech, Engineering surface energy and nanostructure of
micorporous films for expanded membrane distillation applications, Environ. Sci.
Technol. 50 (2016) 8112–8119.
[31] E.J. Lee, A.K. An, T. He, Y.C. Woo, H.K. Shon, Electrospun nanofiber membranes
incorporating fluorosilane-coated TiO2 nanocomposite for direct contact mem-
brfane distillation, J. Membr. Sci. 520 (2016) 145–154.
[32] K.Y. Wang, S.W. Foo, T.S. Chung, Mixed matrix PVDF hollow fiber membranes with
nanoscale pores for desalination through direct contact membrane distillation, Ind.
Eng. Chem. Res. 48 (2009) 4474–4483.
[33] M. Su, M.M. Teoh, K.Y. Wang, J. Su, T.S. Chung, Effect of inner-layer thermal
conductivity on flux enhancement of dual-layer hollow fiber membranes in direct
contact membrane distillation, J. Membr. Sci. 364 (2010) 278–289.
[34] K.J. Lu, J. Zuo, T.S. Chung, Novel PVDF membranes comprising n-butylamine
functionalized graphene oxide for direct contact membrane distillation, J. Membr.
Sci. 539 (2017) 34–42.
[35] J. Chang, J. Zuo, L. Zhang, G.S. O'Brien, T.S. Chung, Using green solvent, triethyl
phosphate (TEP), to fabricate highly porous PVDF hollow fiber membranes for
membrane distillation, J. Membr. Sci. 539 (2017) 295–304.
[36] M. Qtaishat, D. Rana, M. Khayet, T. Matsuura, Preparation and characterization of
novel hydrophobic/hydrophilic polyetherimide composite membranes for desali-
nation by direct contact membrane distillation, J. Membr. Sci. 327 (2009) 264–273.
[37] M. Essalhi, M. Khayet, Application of a porous composite hydrophobic/hydrophilic
membrane in desalination by air gap and liquid gap membrane distillation: a
comparative study, Sep. Purif. Technol. 133 (2014) 176–186.
[38] S.P. Nunes, H. Maab, L. Francis, Polyazole membrane for water purification, U.S.
Patent App. 13/765,228, 2013.
[39] H. Maab, A. Al Saadi, L. Francis, S. Livazovic, N. Ghafour, G.L. Amy, S.P. Nunes,
Polyazole hollow fiber membranes for direct contact membrane distillation, Ind.
Eng. Chem. Res. 52 (2013) 10425–10429.
[40] H. Maab, L. Francis, A. Al-saadi, C. Aubry, N. Ghaffour, G. Amy, S.P. Nunes,
Synthesis and fabrication of nanostructured hydrophobic polyazole membranes for
low-energy water recovery, J. Membr. Sci. 423–424 (2012) 11–19.
[41] H. Maab, S. Pereira Nunes, Porous polyoxadiazole membranes for harsh environ-
ment, J. Membr. Sci. 445 (2013) 127–134.
[42] K.W. Lawson, D.R. Lloyd, Membrane distillation, J. Membr. Sci. 124 (1997) 1–25.
[43] E. Curcio, E. Drioli, Membrane distillation and related operations—a review, Sep.
Purif. Rev. 34 (2005) 35–86.
[44] M.S. El-Bourawi, Z. Ding, R. Ma, M. Khayet, A framework for better understanding
membrane distillation separation process, J. Membr. Sci. 285 (2006) 4–29.
[45] J. Xu, Y.B. Singh, G.L. Amy, N. Ghaffour, Effect of operating parameters and
membrane characteristics on air gap membrane distillation performance for the
treatment of highly saline water, J. Membr. Sci. 512 (2016) 73–82.
[46] L.A. Bromley, D. Singh, P. Ray, S. Sridhar, S.M. Read, Thermodynamic properties of
sea salt solutions, AICHE J. 20 (1974) 326–335.