Development of a composite membrane with underwater-oleophobic fibrous surface for robust anti-oil-fouling membrane distillation

Accepted Manuscript Development of a composite membrane with underwater-oleophobic fibrous surface for robust anti-oil-fouling membrane distillation Kunpeng Wang, Deyin Hou, Peng Qi, Kuiling Li, Ziyi Yuan, Jun Wang PII: DOI: Reference:

S0021-9797(18)31354-7 https://doi.org/10.1016/j.jcis.2018.11.040 YJCIS 24305

To appear in:

Journal of Colloid and Interface Science

Received Date: Revised Date: Accepted Date:

23 July 2018 9 November 2018 12 November 2018

Please cite this article as: K. Wang, D. Hou, P. Qi, K. Li, Z. Yuan, J. Wang, Development of a composite membrane with underwater-oleophobic fibrous surface for robust anti-oil-fouling membrane distillation, Journal of Colloid and Interface Science (2018), doi: https://doi.org/10.1016/j.jcis.2018.11.040

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.

Development of a composite membrane with underwater-oleophobic fibrous surface for robust anti-oil-fouling membrane distillation

Revised Manuscript JCIS-18-3926 Submitted to Journal of Colloid and Interface Science Nov. 9th. 2018

Kunpeng Wanga,b, Deyin Houa,b*, Peng Qic, Kuiling Lia,b, Ziyi Yuand, Jun Wanga,b a

Key Laboratory of Drinking Water Science and Technology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China b

College of Resources and Environment, University of Chinese Academy of Sciences, Beijing 100190, China c

School of Chemical Engineering and Technology, Hebei University of Technology, Tianjin 300130, China d

School of Resources Environmental and Chemical Engineering, Nanchang University, Nanchang 330031, China

*Corresponding author: Deyin Hou E-mail: [email protected] Tel.: +86 10 62917207

Abstract: Membrane fouling caused by non-polar foulants is a challenging problem for hydrophobic membranes, which hinders the industrial implementation of membrane distillation (MD). The hydrophilic coating can create a hydration layer at solid-water interface, thereby the hydrophilic surfaces are expected to supply a barrier inhibiting adhesion of hydrophobic foulants. Hence, it should be possible to develop anti-fouling composite membranes through constructing a hydrophilic skin layer onto hydrophobic MD membranes. Herein, we fabricated a novel composite membrane for excellent

anti-oil-fouling

performance

in

MD

process

by

electrospinning

polyetherimide (PEI) nanofibers on the hydrophobic polyvinylidene fluoride (PVDF) membrane surface, followed by cross-linking with ethanediamine (EDA). The membrane morphology and structure properties, surface zeta potential and wettability, thermal stability were all systematically characterized, and force spectroscopy was used to quasi-quantitatively evaluate oil-membrane adhesion force. Compared with the PVDF membrane, the PVDF/PEI-EDA composite membrane exhibited strong resistance to crude oil with underwater oil contact angle of about 145º and low oil-membrane adhesion force, which contributed to the stable performance during MD desalinating an oily and saline solution. The fabricated composite membrane with underwater-oleophobic fibrous surface can effectively mitigate oil-fouling in MD and promote MD to treat highly saline wastewater with high concentration of hydrophobic foulants. Keywords: Membrane fouling; Oleophobic surface; Membrane distillation; Composite membrane; Electrospinning

1

1. Introduction Membrane-based technologies are considered as the most effective choice for water purification with little use of chemicals and high water quality, which increase water supplies through reusing of wastewater and extracting freshwater from seawater or brackish water [1-3]. However, the long term stability and operation performance of membrane separation process was severely affected by membrane fouling resulting from the formation of fouling layer on membrane surface or the membrane pores blocking, and that would decrease membrane flux, deteriorate the quality of product water and shorten the service life of membranes [4-6]. Scientists have developed some technologies to mitigate membrane fouling, such as pretreatment of feed water, optimizing operational conditions, designing membrane properties and back flushing [7, 8]. Among these efforts, developing membranes with robust resistance to foulants is an effective approach for wide application of membrane-based technologies in water treatment. Industry wastewater is a complicated aqueous solution with diverse chemical components that has varying degrees fouling propensity on different membrane processes [9-11]. Thus, fabricating membranes based on properties of feed water and membrane process is imperative [12]. Membrane distillation (MD), using a microporous hydrophobic membrane serving as a barrier to prevent direct permeation of feed solution and supplying a porous channel for water vapor transport, is a promising membrane-based thermal desalination technology for which can utilize low-grade or waste heat resources and treat highly saline water [13-15]. In a MD process, a transmembrane water vapor pressure difference generating from the temperature difference between the hot feed-side and the cold permeate-side drives the water vapor to transport from feed-side to the permeate-side, thereby desalinating the feed solution [16]. Compared with the pressure driven membrane process such as nanofiltration and reverse osmosis, MD has lower fouling trendency when treating feed water containing natural organic materials because of the larger membrane pore size and lower operation pressure [15]. In some extent, MD is unlimited by osmotic

2

pressure, therefore it has been considered as an economical and viable candidate technology to deal with the shale oil and gas wastewater, a hypersaline wastewater with high concentration of crude oil [17]. However, the hydrophobic membrane employed in MD is prone to attracting non-polar contaminants such as oil via strong hydrophobic-hydrophobic interactions, resulting in the blocking of membrane pores and thus reducing permeate flux [18, 19]. Fabricating membranes with an underwater oleophobic property which can resist oil adhesion is an useful method for alleviating oil fouling in MD process [20]. Recently, a mount of artificial underwater anti-adhesion surfaces have been developed via biomimetic methods by designing both hydrophilic surface chemistry and rough surface morphology for oil-water separations and marine biofouling control [21-26]. This result is due to the hydration layer formed on hydrophilic surface which serves as an energetic barrier for non-polar contaminants to adhere on substrate surface [27]. Following this principle, designing a thin hydrophilic coating on the top of a hydrophobic membrane is a feasible approach to prevent the adhesion of non-polar foulants to the hydrophobic membrane surface. With respect to constructing a hydrophilic layer on the hydrophobic membrane for MD, the selected surface modification method must not destroy the properties of the substrate, such as the surface hydrophobicity and porosity, which are important to the quality and quantity of product water. However, traditional modified techniques for fabrication of composite membrane with a hydrophilic surface such as surface grafting, dip-coating, spin coating or spray coating, may introduce wetting defects and destroy the hydrophobicity of the pristine membrane due to the hydrophilization of membrane pores [18, 28, 29]. Electrospinning technique is a versatile and flexible approach to fabricate nanometer or sub-micrometer scales fibrous membrane with high porosity, interconnected open structure and re-entrant structure, which is suitable for MD to realize considerable permeability, high energy efficiency and excellent anti-wetting performance [30-32]. During electrospinning, continuous fibers are produced through elongating and stretching polymer solution under high static electric field, and a dual- or multilayer membrane can be easily formed through spinning 3

fibrous layer on extra membrane surface without deteriorating the substrate properties because the electrospun fibers with ultrahigh aspect ratio can not penetrate into the substrate membrane pores. In this study, a novel composite membrane for robust anti-oil-fouling performance in MD process was fabricated by electrospinning a hydrophilic layer on the

commercial

hydrophobic

polyvinylidene

fluoride

(PVDF)

membrane.

Polyetherimide (PEI), a common microfltration and ultrafltration membrane material, was selected to construct the anti-oil-fouling layer due to its excellent membrane-forming capacity, thermal stability, mechanical strength and chemical resistance [33-36]. Direct contact MD (DCMD) experiments were conducted to evaluate the anti-oil-fouling performance of the modified membranes using a simulated saline wastewater with the oil concentration of 1000 mg/L. To elucidate the difference of anti-oil-fouling performance in DCMD between the hydrophobic PVDF membrane and the modified composite membranes, membrane morphology, surface chemical components, surface wettability, surface zeta potential were all characterized. To evaluate thermal stability of the composite membranes, thermo gravimetric analysis were also conducted. Additionally, the oil-membrane interfacial interactions were measured using a force tensiometer to compare the adhesion tendencies of oil on different membrane surfaces.

2. Materials and methods 2.1 Materials and chemicals The hydrophobic PVDF membranes with a nominal pore size of 0.80 μm were provided by Millipore Co., Ltd. PEI polymer (Ultem 1000) was produced by General Electric Co. Dimethylacetamide (DMAc, 99%), ethylenediamineand (EDA, 99%), sodium chloride (NaCl, 99.5%) and potassium chloride (KCl, 99.5%) were purchased from Sigma-Aldrich. The crude oil was provided by Daqing Oil Field of China National Petroleum Corporation. 2.2 Preparation of underwater oleophobic membrane The composite membrane with asymmetric wettability was fabricated by

4

electrospinning PEI fibrous network on a hydrophobic PVDF substrate and then cross-linking with EDA, and the preparation procedure is illustrated in Fig. 1. Before electrospinning, a 12 wt% PEI solution was prepared through adding dried PEI polymer (12 g) to 88 g DMAc solvent at 60°C under vigorous stirring using a cantilever mixer (Eurostar 20, Germany) for 24 h to dissolve the PEI polymer. The obtained homogeneous solution was left overnight to remove the air bubbles, and then poured into a syringe with a spinneret of 18-gauge needle. A commercial hydrophobic PVDF membrane was attached onto the rotating drum using two-side tape. During 6.0 h electrospinning duration, the applied voltage was 20 kV with the positive voltage of 15 kV and the negative voltage of -5 kV, and the tip-to-collector distance was fixed at 15 cm. The feed flow rate and drum speed kept at 1.5 mL/h and 100 rpm, respectively. The chamber humidity maintained 30% and the chamber temperature was 30°C. The electrospun composite membrane was dried in air at room temperature to remove the residual solvents and named as PVDF/PEI. Cross-linking post-treatment is essential to improve the hydrophilicity of the electrospinning layer. The PVDF/PEI membrane was placed above a beaker, which contained EDA solution (3 wt%) of 100°C.

The

cross-linked

membrane

was

natural

dried

and

denoted

as

PVDF/PEI-EDA.

Fig. 1. Schematic illustration of the procedure for fabricating the composite membrane.

2.3 Membrane characterizations The cross-section and surface morphologies of the commercial PVDF membrane 5

and modified membranes were examined at different magnification using a Hitachi SU-70 scanning electron microscopy (SEM). Fourier transform infrared spectroscopy (FTIR) analysis was conducted to analyze the chemical groups of the fabricated membranes, using a FTIR Spectrometer (Perkin Elmer Spectrum One, Perkin Elmer, America) fitted with an attenuated total reflectance (ATR) sampling accessory. Thermo gravimetric analysis (Diamond TG/DTA, Perkin Elmer, America) was conducted under nitrogen atmosphere from 30 to 600°C with a heating rate of 10 °C/min. Membrane surface zeta potential was measured using Zetasizer nano ZS90 (Malvern instruments Inc., UK) under the test mode of surface zeta potential. The pore size, pore size distribution and liquid entry pressure of water (LEPw) were investigated through a Capillary Flow Porometer (Porolux 1000, IB-FT GmbH, Germany). Thickness gauge (Exploit, China) was carried out to measure the membrane thickness. A moisture analyzer (Mettler Toledo HR83, Switzerland) equipped with a low surface tension solution was conducted to characterize the membrane porosity by a gravimetric method. All membrane samples were measured more than 3 times. The in-air water contact angles (WCA) and underwater oil contact angles (OCA) of different membranes were determined using a contact angle meter (OCA20, DataPhysics, Germany), to evaluate the surface wettability of membranes with DI water and crude oil, respectively. Considering the crude oil would float in water for its low density, the membrane mounted on a glass slide should be placed face down to contact with the floating oil droplet for the measurement of OCA, just similar to the captive bubble technique. The final contact angle results of different membranes were obtained from more than 5 measurements. 2.4 Measurement of underwater oil-membrane adhesion force To quantitatively clarify the anti-fouling behaviors of different membranes when contacting with crude oil, the underwater oil-membrane adhesion force was measured by a force tensiometer (DCAT11, DataPhysics, Germany). In the force spectroscopy experiments, the membrane sample was glued on the bottom of a special beaker. A certain amount of DI water was poured into the beaker and then put the beaker on a 6

vertical mobile platform. A small crude oil droplet (10 μL) was carefully fixed on the adhesion force probe which was immerged in DI water. For obtaining the force curve, the platform moved upward toward the immobilized adhesion force probe at a set speed of 0.01 mm/s until the membrane contacted with the crude oil droplet, then continued to move upward for 1.5 mm to compress the oil droplet, after that moved downward at the same speed to the original position. The interaction force between the crude oil and membrane surface at different displacement was recorded through a microelectro-mechanical sensor. 2.5 Membrane performance in MD The commercial PVDF membrane and modified membranes were tested in DCMD experiments to evaluate their desalination performance and anti-oil-fouling behavior. The DCMD experimental setup is schematically as shown in Fig. 2. During the DCMD experiment, the modified side of the membrane faced to the feed-side and the total permeate area was about 74.7 cm2. The feed and distillate solution circulated co-currently in the membrane module with a flow velocity of 0.42 m/s and 0.36 m/s, and the temperature was maintained at 55°C and 20°C controlled by water bath, respectively. The synthesized feed solutions, 35 g/L NaCl and 35 g/L NaCl with 1000 mg/L crude oil, were used to evaluate membrane permeability and anti-oil-fouling performance, respectively. The permeate weight and conductivity were recorded automatically at set intervals with a computer software system. The production of pure water and salt rejection rate can be calculated according to the change of the permeate weight and the conductivity of feed and permeate solutions.

Fig. 2. Schematic of the experimental setup for DCMD. 7

3. Results and discussion 3.1 Membrane morphology Fig. 3 shows the surface and cross-section SEM images of the commercial PVDF membrane and the fabricated composite membranes. It can be observed that the composite membranes had a typical double-layer structure, and the fibrous coatings seemed much looser than the PVDF substrate. The SEM images of the cross-section of the membranes also showed that there was no obvious delamination between the electrospun fibrous layer and the PVDF substrate, which indicated that the fibrous layers adhered well to the PVDF substrate surface for the composite membranes. In addition, compared with the PVDF/PEI composite membrane, the fibrous layer of the PVDF/PEI-EDA composite membrane became denser because of the cross-linking treatment. The commercial PVDF membrane exhibited a porous and rough surface (Fig. 3A1), and we can find some polymer crystalline particles embedded in the PVDF matrix as shown in Fig. 3A2 which would improve the surface roughness and enhance the intrinsic hydrophobicity of PVDF according to the Wenzel model [37]. The modified membrane with an electrospun fibrous layer exhibited absolutely different membrane surface morphology. The electrospun fibrous layer resulting from the continuous overlapping of the nanofibers represented a beads-on-string structure with interconnected pores and large porosity (Fig. 3B1 and C1). The “three dimension (3D)” fibrous interconnected structure can be obviously observed in the higher magnification image (Fig. 3B2 and C2), which endowed the electrospun layers with a rough membrane surface. After cross-linking treatment, although there was no obviously difference between the PEI nanofibers (Fig. 3B3) and the PEI-EDA nanofibers (Fig. 3C3), the membrane structure become more compacted, and the shape of the beads have translated to spindle shape as shown in Fig. 3C2. The slender and interlaced nanofibers in the electrospun layers led to a multilevel re-entrant structure with high surface roughness and can enhance surface wettability according to our previous study [38].

8

Fig. 3. SEM images of cross-section and surface of the commercial PVDF membrane and the fabricated composite membranes. A, B, C represent the commercial PVDF membrane, the PVDF/PEI membrane and the PVDF/PEI-EDA membrane, respectively.

3.2 Surface chemical features and membrane thermal stability To confirm the successful cross-linking between PEI and EDA, the ART-FTIR analysis for the electrospun PEI layer and PEI-EDA layer was carried out, and the ART-FTIR spectra are displayed in Fig. 4. For the pure PEI fibrous layer, there were three characteristic absorption bands. The obvious peaks appearing at 1778 cm-1 and 1720 cm-1 were attributed to the asymmetric and symmetric stretching vibration of C=O group, respectively. The absorption bands at 1360 cm-1 can be assigned to vibration of tertiary amine groups (C-N). After cross-linking with EDA, all these three characteristic absorption peaks became weak. In the case of the PEI-EDA fibrous layer, a broad band was observed around 3350 cm-1 due to the presence of the secondary amine group (N-H). Another two weak but identifiable peaks appearing at 1648 cm-1 and 1538 cm-1 were correspond to the vibration of C=O group in amide bond and the bending vibration of N-H group. The chemical groups of PEI and PEI cross-linked with EDA can be found in Fig. 1. These results demonstrated that the electrospun PEI nanofibers have successfully cross-linked with EDA molecular.

9

1648 1538

Transmittance（ a.u.）

3350

PEI-EDA

PEI 1778 1720

3500

3000

2500 2000 -1 Wavenumber (cm )

1360

1500

1000

Fig. 4. ART-FTIR spectra of the top surface of PEI fibrous layer and PEI-EDA fibrous layer.

In MD application, to exploit the anti-fouling performance, the modified side of the composite membrane faced to the hot-feed solution. Generally, MD is operated with the feed temperature below 100°C. Therefore, it is necessary to investigate the thermal stability of the modified membrane. We conducted thermo gravimetric analysis and the results can be found in Fig. 5. The commercial PVDF membrane and the PEI fibrous layer exhibited excellent thermal stability with the decomposition temperature both above 400°C. As for the PEI-EDA fibrous layer, the weight exhibited an obviously decline at about 120°C because of the decomposion of the amido bond forming in the crosslinking process. However, all these three membranes showed excellent thermal stability when the temperature is below 100°C, so it can be concluded that the modified membrane PVDF/PEI-EDA can also be successfully applied in MD experiment with good thermal tolerance.

10

Weight (%)

100

80 PVDF PVDF/PEI PVDF/PEI-EDA

60

40

20 100

200

300 400 Temperature (°C)

500

600

Fig. 5. Thermo gravimetric analysis of the commercial PVDF membrane, the PEI layer and the PEI-EDA layer.

3.3 Membrane surface wettability and structural properties Membrane surface wettability is a vital parameter to determine the adhesion of crude oil on membrane surface. It is already a consensus that the underwater oleophobic is usually hydrophilic in-air. Contact angle measurement results, including in-air WCA and underwater OCA, are shown in Fig. 6a. The commercial PVDF membrane and the fabricated composite membranes exhibited different wetting properties. For the commercial PVDF membrane, the in-air WCA was 127.2±0.5º and the underwater OCA was 40.5±0.7º, which suggested that the commercial PVDF membrane was hydrophobic in air and oleophilic under water, and prone to be fouled by the oil. The fabricated PVDF/PEI membrane showed similar wettability with the PVDF membrane, which may be due to the weak hydrophilicity of the PEI polymer. In addition, large amount of air entrapped in the gaps among the electrospun PEI fibers also prevented the water permeating into the fibrous layer. In contrast, for the PVDF/PEI-EDA membrane, the in-air WCA was 38.5±1.6º and the underwater OCA was 145.3±0.8º. The outstanding underwater oleophobic properties of the PVDF/PEI-EDA membrane was beneficial from the cross-linking treatment which introduced lots of secondary amine groups into the PEI polymer and thus enhanced the bonding ability between the polymer and surrounding water, resulting in a strong 11

hydration layer to resist the adhesion of the oil. The bulk porosities of the fabricated composite membranes were much larger than that of the commercial PVDF membrane as listed in Table 1. This was mainly because the electrospun fibrous layer was more porous than the substrate membrane (with a porosity of about 80.2±3.5% for PEI layer and about 85.6±2.8% for PEI-EDA layer). Meanwhile, the fibrous electrospun layers have little effect on the pore size distribution. In Fig. 6b, we can see that the commercial PVDF membrane have a relative narrow pore size distribution with an average diameter of 0.78±0.03 µm and the average pore size of the two fabricated membranes slightly decreased to 0.68±0.04 µm. There is no doubt that the overall thickness of the composite membranes increased with the electrospun layer on the substrate membrane surface. Compared with the PVDF/PEI membrane, the overall thickness of PVDF/PEI-EDA membrane decreased slightly which can be explained by the fact that cross-linking made the PEI-EDA layer more compacted than PEI layer. However, for the PVDF/PEI-EDA membrane, a successful modified membrane with a hydrophilic layer, water can easily permeate through the electrospun fibrous layer and reach the surface of the substrate PVDF membrane, which meant the channel for vapor transfer in MD was not really prolonged with the membrane thickness increase. If the feed solution permeated into the hydrophobic membrane, the vapor diffusion channel will be blocked resulting in the decline of permeate flux or the feed solution permeating through the membrane with deteriorating the quality of the permeate water. LEPw is an important parameter affected by membrane surface wettability, pore size and membrane thickness, which reflects the ability of membrane to resist wetting in long-term MD application [39]. Compared with the commercial PVDF membrane with a LEPw of 2.38±0.05 bar, the LEPw value of the PVDF/PEI membrane was improved due to the additional PEI fibrous layer. For the PVDF/PEI-EDA membrane, the presence of electrospun fibrous PEI-EDA layer hardly affected the LEPw value in some extent because the PEI-EDA layer is strong hydrophilic. Although the hydrophilic PEI-EDA fibrous coating can be easily wetted by water, the surface of the substrate PVDF membrane can still maintain its intrinsic 12

hydrophobicity because the electrospun PEI-EDA nanofibers with ultrahigh aspect ratio can not penetrate into the tortuous membrane pores.

b) 70 60

PVDF PVDF/PEI PVDF/PEI-EDA

Percentage (%)

50 40 30 20 10 0 0.2

0.4

0.6 0.8 Pore size (μm)

1.0

1.2

Fig. 6. (a) In-air WCA of (A) the commercial PVDF membrane, (B) the PVDF/PEI composite membrane and (C) the PVDF/PEI-EDA composite membrane; Underwater OCA of (D) the commercial PVDF membrane, (E) the PVDF/PEI composite membrane and (F) the PVDF/PEI-EDA composite membrane. (b) Pore size distribution of the commercial PVDF membrane, the PVDF/PEI composite membrane and the PVDF/PEI-EDA composite membrane. Table 1 Membrane structural properties and surface wettability of the commercial PVDF membrane and the fabricated composite membranes. Membrane

PVDF

PVDF/PEI

PVDF/PEI-EDA

Mean pore diameter (μm)

0.78±0.03

0.68±0.04

0.68±0.03

Bulk thickness (μm)

125±1.7

236±3.2

220±2.6

Fibrous layer thickness (μm)

--

111±4.5

95±3.8

Bulk porosity (%)

61.3±2.5

70.1±3.8

72.4±3.3

Fibrous layer porosity (%)

--

80.2±3.5

85.6±2.8

Liquid entry pressure of water (bar)

2.38±0.05

2.96±0.04

2.36±0.06

Water contact angle (In-air)

127.2±0.5º

133.6±0.4º

38.5±1.6º

Oil contact angle (Underwater)

40.5±0.7º

32.8±1.5º

145.3±0.8º

Peel load of fibrous layer (N)

--

0.95±0.04

1.18±0.03

3.4 Membrane surface potential Except for membrane surface wettability, membrane surface potential is another important parameter affecting the anti-fouling performance of the membranes for which governed electrostatic interaction between membrane surface and foulants. Therefore, membrane fouling could be mitigated through manipulating electrostatic interaction. With respect to the crude oil, a negative charged foulant, constructing a negative charged membrane surface is a useful strategy to alleviate the adhesion of 13

crude oil, and this had been successfully realized in oil/water separation field [40]. In this study, surface zeta potential of each membrane was measured at different pH using 1 mM KCl solution and experimental solution, the results are shown in Fig. 7. Membrane surface potential is related to the chemical composition of the membrane materials, and the value varies with pH because of protonation or deprotonation of surface chemical functional groups. These three membranes were all negative charged with the pH range from 3 to 9 in 1 mM KCl. Compared with PEI layer, the PVDF substrate exhibited more negative surface zeta potential. This can be attributed to the fact that the PVDF polymer contains a large number of fluorine atoms which are the most electronegative. After cross-linking between PEI and EDA, a mount of secondary amine groups were introduced into the PEI polymer chain. The lone pair electrons on the nitrogen atom of the secondary amine group can bind with the ionized H+ of water to enhance the positive electricity of the membrane surface. Thus, the surface zeta potential of PEI-EDA layer was higher than PEI layer. However, when the 1 mM KCl solution was replaced by experimental solution (35 g/L NaCl solution), the membrane surface zeta potential of the PVDF membrane, the PVDF/PEI membrane and the PVDF/PEI-EDA membrane turned to -7.7±0.8 mV, -6.7±0.7 mV, -6.4±1.1 mV, respectively. There was no obvious difference in surface zeta potential among these three membranes in experimental solution. This phenomenon can be explained by two consequences of the high salt concentration. One is the counter-ions concentration in bulk solution becoming close to that near the membrane surface, decreasing the driving force of counter-ions to be adsorbed on membrane surface. The other is the growing motion resistance of the counter-ions, making them harder to be absorbed. Thus, the surface zeta potential decreased significantly and the difference in surface zeta potential among the three membranes shrank. It can be concluded that the surface zeta potential had little effect on membrane fouling mitigation during MD for these three membranes. The similar conclusion had been drawn in previous study that the effect of electrostatic interactions between membrane surface and foulants became negligible and the surface zeta potential might not play a significant role in high-salinity applications [17]. 14

0

Zeta potential (mV)

-10

-20

-30 PVDF PVDF/PEI PVDF/PEI-EDA

-40

-50 3

4

5

6

7

8

9

pH

Fig. 7. Surface zeta potentials of the commercial PVDF membrane, the PVDF/PEI composite membrane and the PVDF/PEI-EDA composite membrane as a function of pH in background electrolyte concentration of 1 mM KCl.

3.5 Force spectroscopy with oil droplet probe To further understand the interaction between oil droplet and membrane surface, underwater dynamic oil-membrane interaction force was measured using an oil-probe force spectroscopy [22, 41]. The changes of the underwater oil-membrane interaction force with the movement of membrane sample are shown in Fig. 8, and the shape changes of the oil droplet during the measurement process can be found in Fig. S1. The measurement process contained two main stages, the advancing stage and the receding stage. In the advancing stage, the membrane sample was moved upward to the oil droplet until contacted with the oil droplet, then continued to move upward to a set position. In the receding stage, the membrane moved downward at a constant speed until completely separated with the oil droplet. The differences of recorded force among the three membranes at different stage can be used to evaluate the anti-oil-adhesion properties of the membrane. It should be noted that the value of underwater oil adhesion force was set to zero before membrane sample contacting with the oil droplet. If the oil droplet was compressed by the membrane sample, it would push the probe upward, and we got a negative adhesion force value, while a positive value was observed if the oil droplet was attracted by the membrane sample and pulled the probe downward. 15

For the commercial PVDF membrane, the force curve exhibited a significantly rise once the membrane contacted with the oil droplet on account of the strong hydrophobic-hydrophobic attraction force between the hydrophobic membrane surface and the oil droplet. With the membrane moving upward, the oil droplet was compressed and spread out on membrane surface. Thus, the force became negative gradually. In the receding stage, the force curve increased rapidly in the positive direction, which was because the oil droplet, adhering on the membrane surface, dragged the probe downward. When the oil droplet split with the membrane, there was a sharp decrease in force curve because of the adhesion interaction between membrane surface and oil droplet. The baseline of force curve can not return to zero after split event, which can be attributed to some oil still left on the membrane surface and thus decreased the floating force of oil droplet. The PVDF/PEI membrane showed similar force curve with the PVDF membrane for these two membranes had similar surface wettability that was shown in Fig. 6a. However, it was very interesting that the force curve of the PVDF/PEI membrane presented a second rise in the advancing stage. The first rise in the force curve can be put down to the hydrophobic-hydrophobic interaction between the fibrous PEI layer and the oil droplet. With this composite membrane moving upward, the oil droplet was spread out on the membrane surface. Inevitably, the oil would permeate through the loose fibrous PEI layer and contact with the hydrophobic PVDF substrate, as a result, the force curve exhibited a second rise. In strong contrast, the force curve of the PVDF/PEI-EDA membrane was significantly different with the commercial PVDF membrane and the PVDF/PEI membrane. When the membrane contacted with the oil droplet, a negative force was recorded because the hydration force, resulted from the hydrophilic fibrous layer, was repulsive to the oil droplet. There was a biggest repulsive force was recorded in the compression process with the oil droplets transforming from spherical to elliptical rather than spreading out on the membrane surface. When the oil droplet split from the membrane in the receding stage, the baseline is much lower than that of the commercial PVDF membrane and PVDF/PEI membrane, which meant considerably 16

little oil remained on the PVDF/PEI-EDA membrane surface. These results indicated that the PVDF/PEI-EDA membrane with an underwater oleophobic surface exhibited strong resistance to oil adhesion than the hydrophobic PVDF membrane and PVDF/PEI membrane, and the fabricated PVDF/PEI-EDA composite membrane may be suitable for dealing with wastewater containing high concentration of oil.

Underwater oil adhesive force (μN)

300 PVDF PVDF/PEI PVDF/PEI-EDA

200

100

0

-100 -0.5

0.0

0.5

1.0

1.5

Position (mm)

Fig. 8. Force-distance curves of the commercial PVDF membrane, the PVDF/PEI composite membrane and the PVDF/PEI-EDA composite membrane recorded the changes of oil-membrane adhesion force with displacement.

3.6 Membrane performance in MD To evaluate the desalination performance of the three membranes, DCMD experiments were carried out with salty water (35 g/L NaCl solution), and the results can be found in Fig. 9a. Compared with the commercial PVDF membrane, the additional electrospun layer of the composite membranes had a negative effect on the permeate flux. For the PVDF/PEI composite membrane, the permeate flux was about 4.5 L/m2h, which was lower than that of the commercial PVDF membrane (above 7.0 L/m2h). This was mainly because that the hydrophobic PEI layer increased the diffusion distance of the water vapor, thus enhanced the mass transfer resistance. As for the PVDF/PEI-EDA membrane, although water can easily permeate through the hydrophilic PEI-EDA fibrous layer and reached the surface of hydrophobic substrate, the presence of the fibrous layer decreased the mix of feed liquid near hydrophobic

17

substrate surface with the bulk solution, aggravating temperature polarization in some extent. As a result, the enhanced temperature polarization lowered the driving force of vapor transporting through the hydrophobic membrane, and the permeate flux declined compared with that of the commercial PVDF membrane. Fortunately, the permeate conductivity of the three kinds of membranes was all below 10 μS/cm in MD process, which meant the salt rejection was nearly 100%, suggesting that the extra electrospun layers would not cause wetting defects for the hydrophobic PVDF substrate. To compare the anti-oil-fouling performance of the commercial PVDF membrane and the fabricated composite membranes, the salty feed solution was replaced with oily and saline feed containing 35 g/L NaCl and 1000 mg/L crude oil in DCMD experiments. The PVDF membrane and PVDF/PEI membrane were severely fouled by the crude oil, as indicated by the rapidly decline of permeate flux in a short operation time in Fig. 9b. The PVDF membrane and PVDF/PEI membrane were underwater oleophilic as shown in Fig. 6a, which caused the adhesion of crude oil on the membrane surface via strong hydrophobic-hydrophobic interaction. Once the oil droplets attached on membrane surface, they would result in membrane fouling and form a physical barrier to impede the permeation of vapor as confirmed by Fig. S2, thus induced the decline of permeate flux. In contrast, the PVDF/PEI-EDA membrane with a hydrophilic fibrous layer showed a stable permeate flux with a normalized water flux above 0.95 when facing oily feed solution. The robust anti-oil-fouling performance was attribute to the presence of hydrophilic fibrous layer that has a strong resistance to the adhesion of crude oil. Speaking in detail, water can easily permeate into the gaps between the fibers because of the strong hydrophilicity of the PEI-EDA fibers, forming an underwater oil/water/solid interface (similar to the water/air/solid interface in-air). According to Cassie model, the presence of water in the fibers gap can decrease the contact area between oil droplets and the fibrous surface, coupled with that the oil droplets were subjected to a horizontal shear force resulting from the flow of feed solution, so the oil droplets can hardly attach on the PVDF/PEI-EDA composite 18

membrane surface. The photographic images of the membranes shown in Fig. S2 offered intuitive evidence that the PVDF/PEI-EDA membrane had the strongest anti-oil fouling resistance compared with the other two membranes. After rinsing with DI water, the oil on the PVDF/PEI-EDA membrane surface can be completely removed, and the rinsed membrane looks exactly the same as the original. Whereas, it was clearly visible that the entire surfaces of the PVDF commercial membrane and the PVDF/PEI composite membrane were covered by the oil foulant, and it is very difficult to flush oil off from membrane surface which indicated that oil fouling on hydrophobic

6

1000

4

PVDF PVDF/PEI PVDF/PEI-EDA

2

100

10

b) 8 Original Membranes

Rinsed Membranes

PVDF PVDF/PEI PVDF/PEI-EDA

6

4

Normalized permeate flux (J/JO)

10000

Permeate flux (L/m2h)

Permeate flux (L/m2h)

a) 8

Permeate conductivity (μS/cm)

membrane surfaces was almost irreversible.

2

1.00

0.75

0.50

0.25

0.00 0

0 0

5

10

15

20

25

10

30

0

Operation time (h)

20

30

40

50

60

Operation time (h)

0

1

10

20

30

40

50

60

Operation time (h)

Fig. 9. (a) Permeate fluxes from DCMD tests of the commercial PVDF membrane, the PVDF/PEI composite membrane and the PVDF/PEI-EDA composite membrane with salty water. (b) Anti-fouling performance of the commercial PVDF membrane, the PVDF/PEI composite membrane and the PVDF/PEI-EDA composite membrane in DCMD experiments under 1000 mg/L crude oil, and the insert is the normalized permeate flux with time.

Flux recovery rate of the fouled membrane after rinsing is an important parameter to evaluate the reusability of the fabricated membranes, the DCMD experiments were conducted again using the rinsed membrane with oily and saline solution. As shown in Fig. 9b, there was little difference between the rinsed PVDF/PEI-EDA membrane and the original membrane in permeate flux, which indicated that the PVDF/PEI-EDA composite membrane had robust anti-oil-fouling performance and desirable reusability. By contrast, the flux recovery rate of rinsed PVDF and PVDF/PEI membranes was less than 15% because oil fouling on these two membrane surfaces was almost irreversible as shown in Fig. S2, and most surface 19

pores were still blocked by oil even after rinsing. All the experimental findings suggested that the fabricated PVDF/PEI-EDA composite membrane can be potentially used in MD process to treat refractory wastewater with high concentration of salt and hydrophobic foulants.

4. Conclusion On the basis of the previously reported approaches about fabricating membrane with asymmetric wettability for anti-oil-fouling MD process [18, 28, 29], this study developed a novel composite membrane without destroying the substrate hydrophobicity that is crucial for MD process. The composite membrane was prepared via electrospinning PEI fibrous layer on a hydrophobic PVDF substrate membrane surface, and the PEI nanofibers need to be cross-linked with EDA to improve their hydrophilicity through introducing secondary amino groups which can interact with water to provide a hydration layer to resist oil adhesion. Compared with the commercial PVDF membrane and the PVDF/PEI membrane, the PVDF/PEI-EDA membrane exhibited excellent underwater oleophobic property and showed fairly less attractive to oil droplet. We also measured the surface zeta potential of the three membranes and found that electrostatic repulsive force had little effect on anti-fouling performance in MD process, similar to a previous study [17]. In the DCMD tests with an oily and saline feed solution, both the hydrophobic PVDF membrane and the PVDF/PEI membrane were severely fouled by the crude oil. In contrast, the PVDF/PEI-EDA membrane presented a stable performance in the long-term operation with robust resistance to oil-fouling. The

facile

but

feasible

surface

modification

strategy

to

yield

an

underwater-oleophobic fibrous coating has tremendous potential to promote MD dealing with salty water with high concentration of hydrophobic foulants, and the specific modification technique also can be utilized to relieve non-polar organic fouling in pressure driven membrane processes. Further study is under way to reduce the

permeate

flux

loss

of the

composite

membrane,

and

whether

the

underwater-oleophobic fibrous layer can resist other foulants such as colloids and

20

amphiphilic organics still remains to be experimentally confirmed.

Acknowledgements Financial support provided by the National Natural Science Foundation of China (51678555 and 51478454) and the National Key R&D Program of China (No. 2016YFC0400500) are gratefully acknowledged.

Reference [1] K.R. Zodrow, Q. Li, R.M. Buono, W. Chen, G. Daigger, L. Duenas-Osorio, M. Elimelech, X. Huang, G. Jiang, J.H. Kim, B.E. Logan, D.L. Sedlak, P. Westerhoff, P.J.J. Alvarez, Advanced materials, technologies, and complex systems analyses: emerging opportunities to enhance urban water security, Environ. Sci. Technol. 51 (2017) 10274-10281. [2] X. Yi, W. Shi, Membrane science and technology: leader in water treatment industry, Rev. Environ. Sci. Bio. 11 (2012) 227-229. [3] P.S. Goh, T. Matsuura, A.F. Ismail, N. Hilal, Recent trends in membranes and membrane processes for desalination, Desalination 391 (2016) 43-60. [4] R. Zhang, Y. Liu, M. He, Y. Su, X. Zhao, M. Elimelech, Z. Jiang, Antifouling membranes for sustainable water purification: strategies and mechanisms, Chem. Soc. Rev. 45 (2016) 5888-5924. [5] W. Guo, H.H. Ngo, J. Li, A mini-review on membrane fouling, Bioresour. Technol. 122 (2012) 27-34. [6] D.M. Warsinger, J. Swaminathan, E. Guillen-Burrieza, H.A. Arafat, J.H. Lienhard V, Scaling and fouling in membrane distillation for desalination applications: a review, Desalination 356 (2015) 294-313. [7] M. Racar, D. Dolar, K. Košutić, Chemical cleaning of flat sheet ultrafiltration membranes fouled by effluent organic matter, Sep. Purif. Technol. 188 (2017) 140-146. [8] F. Meng, S.R. Chae, A. Drews, M. Kraume, H.S. Shin, F. Yang, Recent advances in membrane bioreactors (MBRs): membrane fouling and membrane material, Water Res. 43 (2009) 1489-1512.

21

[9] A. Asatekin, A.M. Mayes, Oil industry wastewater treatment with fouling resistant membranes containing amphiphilic comb copolymers, Environ. Sci. Technol. 43 (2009) 4487-4492. [10] B. Van der Bruggen, G. Cornelis, C. Vandecasteele, I. Devreese, Fouling of nanofiltration and ultrafiltration membranes applied for wastewater regeneration in the textile industry, Desalination 175 (2005) 111-119. [11] X. Yang, M. Crespi, V. Lopez-Grimau, A review on the present situation of wastewater treatment in textile industry with membrane bioreactor and moving bed biofilm reactor, Desalin. Water Treat. 103 (2018) 315-322. [12] S.A. Deowan, F. Galiano, J. Hoinkis, D. Johnson, S.A. Altinkaya, B. Gabriele, N. Hilal, E. Drioli, A. Figoli, Novel low-fouling membrane bioreactor (MBR) for industrial wastewater treatment, J. Membr. Sci. 510 (2016) 524-532. [13] D. Hou, J. Wang, X. Sun, Z. Ji, Z. Luan, Preparation and properties of PVDF composite hollow fiber membranes for desalination through direct contact membrane distillation, J. Membr. Sci. 405-406 (2012) 185-200. [14] N. Subramanian, A. Qamar, A. Alsaadi, A.G. Jr, M.G. Ridwan, J.G. Lee, S. Pillai, S. Arunachalam, D. Anjum, F. Sharipov, N. Ghaffour, H. Mishra, Evaluating the potential of superhydrophobic nanoporous alumina membranes for direct contact membrane distillation, J. Colloid Interface Sci. 533 (2019) 723-732. [15] A. Deshmukh, C. Boo, V. Karanikola, S. Lin, A.P. Straub, T. Tong, D.M. Warsinger, M. Elimelech, Membrane distillation at the water-energy nexus: limits, opportunities, and challenges, Energ. Environ. Sci. 11 (2018) 1177-1196. [16] M. Qtaishat, T. Matsuura, B. Kruczek, M. Khayet, Heat and mass transfer analysis in direct contact membrane distillation, Desalination 219 (2008) 272-292. [17] D.L. Shaffer, L.H. Arias Chavez, M. Ben-Sasson, S.R.V. Castrillón, N.Y. Yip, M. Elimelech, Desalination and reuse of high-salinity shale gas produced water: drivers, technologies, and future directions, Environ. Sci. Technol. 47 (2013) 9569-9583. [18] G. Zuo, R. Wang, Novel membrane surface modification to enhance anti-oil fouling property for membrane distillation application, J. Membr. Sci. 447 (2013) 26-35. 22

[19] L.D. Tijing, Y.C. Woo, J.-S. Choi, S. Lee, S.-H. Kim, H.K. Shon, Fouling and its control in membrane distillation-a review, J. Membr. Sci. 475 (2015) 215-244. [20] N.G.P. Chew, S. Zhao, C. Malde, R. Wang, Superoleophobic surface modification for robust membrane distillation performance, J. Membr. Sci. 541 (2017) 162-173. [21] H. Liu, Y. Wang, J. Huang, Z. Chen, G. Chen, Y. Lai, Bioinspired surfaces with superamphiphobic properties: concepts, synthesis, and applications, Adv. Funct. Mater. 28 (2018) 1707415. [22] F. Zhang, W. Zhang, Z. Shi, D. Wang, J. Jin, L. Jiang, Nanowire-haired inorganic membranes

with

superhydrophilicity

and

underwater

ultralow

adhesive

superoleophobicity for high-efficiency oil/water separation, Adv. Mater. 25 (2013) 4192-4198. [23] M. Obaid, N.A.M. Barakat, O.A. Fadali, M. Motlak, A.A. Almajid, K.A. Khalil, Effective and reusable oil/water separation membranes based on modified polysulfone electrospun nanofiber mats, Chem. Eng. J. 259 (2015) 449-456. [24] F. Li, Z. Wang, S. Huang, Y. Pan, X. Zhao, Flexible, durable, and unconditioned superoleophobic/superhydrophilic surfaces for controllable transport and oil-water separation, Adv. Funct. Mater. 28 (2018) 1706867. [25] B. Su, Y. Tian, L. Jiang, Bioinspired interfaces with superwettability: from materials to chemistry, J. Am. Chem. Soc. 138 (2016) 1727-1748. [26] Y. Shen, P.O. Saboe, I.T. Sines, M. Erbakan, M. Kumar, Biomimetic membranes: a review, J. Membr. Sci. 454 (2014) 359-381. [27] J. Cui, A. Xie, S. Zhou, S. Liu, Q. Wang, Y. Wu, M. Meng, J. Lang, Z. Zhou, Y. Yan, Development of composite membranes with irregular rod-like structure via atom transfer radical polymerization for efficient oil-water emulsion separation, J. Colloid Interface Sci. 533 (2019) 278-286. [28] Z. Wang, D. Hou, S. Lin, Composite membrane with underwater-oleophobic surface for anti-oil-fouling membrane distillation, Environ. Sci. Technol. 50 (2016) 3866-3874. [29] Z. Wang, J. Jin, D. Hou, S. Lin, Tailoring surface charge and wetting property for robust oil-fouling mitigation in membrane distillation, J. Membr. Sci. 516 (2016) 23

113-122. [30] Y. Liao, C. H. Loh, M. Tian, R. Wang, A.G. Fane, Progress in electrospun polymeric nanofibrous membranes for water treatment: fabrication, modification and applications, Prog. Polym. Sci. 77 (2018) 69-94. [31] Z. Zhu, Y. Liu, H. Hou, W. Shi, F. Qu, F. Cui, W. Wang, Dual-bioinspired design for constructing membranes with superhydrophobicity for direct contact membrane distillation, Environ. Sci. Technol. 52 (2018) 3027-3036. [32] H. Wang, H. Zhou, H. Niu, J. Zhang, Y. Du, T. Lin, Dual-layer superamphiphobic/superhydrophobic-oleophilic

nanofibrous

membranes

with

unidirectional oil-transport ability and strengthened oil-water separation performance, Adv. Mater. Interf. 2 (2015) 1400506. [33] Y. Zhang, M. Zhong, B. Luo, J. Li, Q. Yuan, X.J. Yang, The performance of integrally skinned polyetherimide asymmetric nanofiltration membranes with organic solvents, J. Membr. Sci. 544 (2017) 119-125. [34] R.S. Hebbar, A.M. Isloor, B. Prabhu, Inamuddin, A.M. Asiri, A.F. Ismail, Removal of metal ions and humic acids through polyetherimide membrane with grafted bentonite clay, Sci. Rep. 8 (2018) 4665. [35] M. Namvar-Mahboub, M. Pakizeh, Development of a novel thin film composite membrane by interfacial polymerization on polyetherimide/modified SiO2 support for organic solvent nanofiltration, Sep. Purif. Technol. 119 (2013) 35-45. [36] M. Qtaishat, D. Rana, M. Khayet, T. Matsuura, Preparation and characterization of novel hydrophobic/hydrophilic polyetherimide composite membranes for desalination by direct contact membrane distillation, J. Membr. Sci. 327 (2009) 264-273. [37] R.N. Wenzel, Resistance of solid surfaces to wetting by water, Ind. Eng. Chem. 28 (1936) 988-994. [38] Y.X. Huang, Z. Wang, D. Hou, S. Lin, Coaxially electrospun super-amphiphobic silica-based membrane for anti-surfactant-wetting membrane distillation, J. Membr. Sci. 531 (2017) 122-128. [39] K. Wang, D. Hou, J. Wang, Z. Wang, B. Tian, P. Liang, Hydrophilic surface 24

coating on hydrophobic PTFE membrane for robust anti-oil-fouling membrane distillation, Appl. Surf. Sci. 450 (2018) 57-65. [40] Y. Liu, Y. Su, J. Cao, J. Guan, R. Zhang, M. He, L. Fan, Q. Zhang, Z. Jiang, Antifouling, high-flux oil/water separation carbon nanotube membranes by polymer-mediated surface charging and hydrophilization, J. Membr. Sci. 542 (2017) 254-263. [41] X. Yao, J. Gao, Y. Song, L. Jiang, Superoleophobic surfaces with controllable oil adhesion and their application in oil transportation, Adv. Funct. Mater. 21 (2011) 4270-4276.

25

Graphical Abstract

26