A Simple, Green Method to Fabricate Composite Membranes for Effective Oil-in-Water Emulsion Separation

Most factories discharge untreated wastewater to reduce costs, causing serious environmental problems. Low-cost, biological, environmentally friendly and highly effective materials for the separation of emulsified oil/water mixtures are thus in great demand. In this study, a simple, green method was developed for separating oil-in-water emulsions. A corn straw powder (CSP)-nylon 6,6 membrane (CSPNM) was fabricated by a phase inversion process without any further chemical modification. The CSPNM showed superhydrophilic and underwater superoleophobic properties and could be used for the separation of oil-in-water emulsion with high separation efficiency and flux. The CSPNM maintained excellent separation ability after 20 cycles of separation with an oil rejection >99.60%, and the oil rejection and flux have no obvious change with an increasing number of cycles, suggesting a good antifouling property and the structural stability of CSPNM. In addition, the CSPNM exhibited excellent thermal and chemical stability under harsh conditions of high temperature and varying pH.


Introduction
Nowadays, an increasing number of industries discharge oily wastewater without any treatment, threatening human health and the aquatic ecosystem [1,2]. Emulsified oil/water mixtures generated from most industrial processes, such as petrochemistry, steel production, metal finishing, textile production, food production, and leather production, have a high percentage of oily wastewater [3]. Traditional technologies such as flotation, coalescers, depth filters, centrifugation, and oil-absorbing materials are efficient technologies for separating oil/water mixtures [4][5][6][7]; however, these technologies are ineffective for emulsified oil/water mixtures and surfactant-stabilized emulsions, particularly for emulsions with a droplet size of 20 µm [8]. Although electric field and adding chemicals can demulsify the emulsions, these methods have some disadvantages, such as higher energy consumption and secondary pollution [9]. Therefore, effective techniques to separate oil/water emulsions in wastewater are in great demand. The candidates with the most potential are membrane techniques with advantages such as recycled oil and purified water [10][11][12]. However, the biggest limitation of these materials is easy fouling, caused by the pore plugging, directly leading to a quick decline in flux [13].
Recently, membranes and films with special wettability have been used to separate oil/water emulsions and prevent membrane fouling [14][15][16]. It is usually thought that super-wettability can

Emulsion Separation Experiment
The as-prepared CSPNM was sealed within the filter apparatus. The 10 mL of feed emulsion were poured onto the membrane and then the filtrate was collected with a driving pressure of 0.01 MPa (vacuum −0.1 bar). The flux was determined by calculating the following formula [34][35][36]: where V is the volume of the solution passing through the membrane, A is the valid area of the membrane (1.77 cm 2 ), and t is the time of the solution passing through the membrane. Original emulsions and their corresponding collected filtrates after a one-time separation were measured by UV-Vis spectroscopy and total organic carbon (TOC) analyzer. The oil content in water was measured using a TOC analyzer, and the characteristic peak of the oil was measured using a UV-Vis spectroscopy. The average value of five samples was measured. In order to study the separation ability of the CSPNM deeply, the oil rejection was determined by calculating the ratio of the rejected oil content in the feed solution by the membrane to the oil content in the feed solution using the follow formula [37]: where R is the oil rejection of the CSPNM, C f1 is the oil content in feed solution, and C f2 is the oil content in filtration.

Continuous Separation
The as-prepared CSPNM was sealed between a vertical glass tube and a conical flask, and then the CSPNM was pre-wetted with water. Water (10 mL) and the H/W emulsion (10 mL) were poured into two different filter apparatuses and the filtration was driven by a vacuum pump at a pressure of 0.01 MPa (vacuum −0.1 bar). The whole feed solution permeated the membrane, and a second cycle followed. During each cycle 10 mL feed solution were added. The time it took for the solution to pass through the membrane was recorded by a timer. The flux was calculated by Formula (1) (where V is the volume of feed solution passing through the membrane, t is the time consumed by a solution of a certain volume passing through the membrane, and A is the valid area of the membrane (1.77 cm 2 )). The filtrate was measured by total organic carbon (TOC) analyzer and then the oil rejection was calculated by Formula (2), where C f1 is the oil content in the feed solution and C f2 is the oil content in filtration.

Cycle Experiment
Feed solution (10 mL) was poured into the apparatus and the filtrate was measured by total organic carbon (TOC) analyzer, while the time of solution passing through the membrane was recorded by a timer. After one cycle, the membrane was washed with ethanol (10 mL), cleaned, and dried at 70 • C (the test used the same sample to directly reflect the variation trend). The above process was repeated 20 times.

Environmental Durability Experiments
The CSPNM was heated at different temperatures for 12 h to test the thermal stability of the CSPNM. The oil contact angles (CAs) and roll-off angles of the CSPNM were measured underwater after the CSPNM stayed at 0 • C to 120 • C for 12 h. The CSPNMs were in contact with corrosive aqueous solutions (HCl and NaOH aqueous solutions, pH 1 to 14) for 30 h to study the chemical durability. Then, the oil CAs and roll-off angles of the membranes were measured underwater.

Characterizations
The morphology of the as-prepared CSPNM was investigated by scanning electron microscopy (SEM, TM3030, Tokyo, Japan). The water CAs were measured using 5 µL of deionized water droplets, and the oil CAs underwater were measured using 5 µL of DCE droplets after immersing the CSPNM into water at room temperature, which was conducted on an OCA20 instrument (Data-physics, Stuttgart, Germany). The tensile strength of the membranes was measured using a tensile testing machine (AI-7000S, Gotech testing machines, Dongwan, China). Feed solutions, filtrates, and emulsifier aqueous solutions were analyzed by optical microscopy (Eclipse 80i, Tokyo, Japan), a UV-Vis spectrometer (Cary 100, Agilent, Melbourne, Australia), and total organic carbon analysis (TOC-VCPN, Shimadzu, Kyoto, Japan), respectively.

Membrane Characterization
The CSPNM was fabricated via the phase inversion using water as a coagulation bath. Figure 1a-c show the schematic fabrication process of the membrane. As shown in Figure 1a, the CSPNFS was cast onto the glass pane and immersed into deionized water, where the solvents were immediately exchanged between the water and formic acid. During the phase-inversion process, the water will take some of the insecure nylon particles (INP) away from the surface of the CSPNM, as shown in the top of Figure 1b 1a-c show the schematic fabrication process of the membrane. As shown in Figure 1a, the CSPNFS was cast onto the glass pane and immersed into deionized water, where the solvents were immediately exchanged between the water and formic acid. During the phase-inversion process, the water will take some of the insecure nylon particles (INP) away from the surface of the CSPNM, as shown in the top of Figure 1b Figure S1 of the Supplementary Materials. Figure 2a shows many micro-flaws on the top surface of the PNM, and obvious flaws were observed, as shown in Figure 2b. The cross section of the PNM (Figure 2c) shows many particles piled together.
The addition of the CSP improved the morphology and microstructure of the membrane. We consider that the CSP plays two distinct roles in the CSPNM. First, CSP reduces the number of INPs to prevent nylon particles (NPs) from leaving, so more nylon will participate in the formation of the structure. Secondly, Figure 3d shows a rougher structure at the top surface of the CSPNM. According to the Wenzel model [39], the hydrophilicity of a hydrophilic solid substrate will be enhanced by surface roughness because of the capillary effect. The high-magnification image (Figure 2e) clearly shows the microstructure. Moreover, the microstructure changed from a particles-accumulation structure to a continuously spongy structure, as shown in Figure 2f. During the phase-inversion process, the morphology of nylon transformed with the addition of CSP, thereby compensating for the flaws.  Figure S1 of the Supplementary Materials. Figure 2a shows many micro-flaws on the top surface of the PNM, and obvious flaws were observed, as shown in Figure 2b. The cross section of the PNM (Figure 2c) shows many particles piled together.
The addition of the CSP improved the morphology and microstructure of the membrane. We consider that the CSP plays two distinct roles in the CSPNM. First, CSP reduces the number of INPs to prevent nylon particles (NPs) from leaving, so more nylon will participate in the formation of the structure. Secondly, Figure 3d shows a rougher structure at the top surface of the CSPNM. According to the Wenzel model [39], the hydrophilicity of a hydrophilic solid substrate will be enhanced by surface roughness because of the capillary effect. The high-magnification image (Figure 2e) clearly shows the microstructure. Moreover, the microstructure changed from a particles-accumulation structure to a continuously spongy structure, as shown in Figure 2f. During the phase-inversion process, the morphology of nylon transformed with the addition of CSP, thereby compensating for the flaws. To test the flexibility of CSPNM, the membrane was rolled up and released over 500 times. As shown in Figure 3a, there were no cracks after the bending cycles, indicating the superior mechanical performance and flexibility of the CSPNM. The PNM and the CSPNM were cut into 30 × 10 mm 2 pieces to determine the tensile strength, as shown in Figure 3b. The PNM has a tensile strength of 0.377 MPa and an elongation of 2.2%, respectively. The CSPNM showed higher mechanical strength with a tensile strength of 0.689 MPa and lower elongation of 1.4%, attributed to the structural transformation from the particles-accumulation structure to the sponge-like structure [40]. The binding strength between nylon particles (NPs) in the PNM is weaker, thus leading to inferior mechanical performance. The stronger binding strength between CSP and nylon improves the mechanical performance of the CSPNM, suggesting that the sponge-like structure of the CSPNM is To test the flexibility of CSPNM, the membrane was rolled up and released over 500 times. As shown in Figure 3a, there were no cracks after the bending cycles, indicating the superior mechanical performance and flexibility of the CSPNM. The PNM and the CSPNM were cut into 30 × 10 mm 2 pieces to determine the tensile strength, as shown in Figure 3b. The PNM has a tensile strength of 0.377 MPa and an elongation of 2.2%, respectively. The CSPNM showed higher mechanical strength with a tensile strength of 0.689 MPa and lower elongation of 1.4%, attributed to the structural transformation from the particles-accumulation structure to the sponge-like structure [40]. The binding strength between nylon particles (NPs) in the PNM is weaker, thus leading to inferior mechanical performance. The stronger binding strength between CSP and nylon improves the mechanical performance of the CSPNM, suggesting that the sponge-like structure of the CSPNM is more stable than the particles-accumulation structure of the PNM. The lower elongation of the CSPNM showed that the CSPNM has better dimensional stability, suitable for repeated use. These results indicate the excellent mechanical performance and flexibility of the CSPNM.  Figure 4a shows the wetting behavior of water and oil on the top surface of the as-prepared CSPNM. A water droplet (5 μL) spreads out and permeates into the CSPNM within 0.78 s when it comes into contact with the membrane (Movie S1), and a water CA of nearly 0° in the air is observed (Figure 4a, left), indicating the high hydrophilicity of the CSPNM. The oil droplet attained a quasisphere on the underwater surface of the membrane with a CA of 157°, showing the underwater superoleophobicity of the membrane (Figure 4a, right). When the underwater CSPNM comes into contact with oil, the water trapped in the rough structures will significantly decrease the contact area between the surface of the membrane and oil droplet, showing a large oil CA underwater (underwater superoleophobicity). Thus, the CSPNM shows superhydrophilic and underwater superoleophobic properties.
To better assess the membrane's oil repellency underwater, its adhesion behavior was measured during the process of oil contact with the membrane surface underwater. Before the beginning of the experimentation, an oil droplet (5 μL) was squeezed on the CSPNM surface underwater under a preload, and then the oil droplet was released. In the relaxing process, the oil droplet maintains the spherical shape underwater, as shown in Figure 4b, indicating that CSPNM displays an ultralow adhesion with the oil droplet underwater. The results demonstrate that the CSPNM has a much better anti-adhesion performance to oil underwater, probably attributed to the formation of hydrogen bonds between the water and hydrophilic components in the CSPNM [35].  Figure 4a shows the wetting behavior of water and oil on the top surface of the as-prepared CSPNM. A water droplet (5 µL) spreads out and permeates into the CSPNM within 0.78 s when it comes into contact with the membrane (Movie S1), and a water CA of nearly 0 • in the air is observed (Figure 4a, left), indicating the high hydrophilicity of the CSPNM. The oil droplet attained a quasi-sphere on the underwater surface of the membrane with a CA of 157 • , showing the underwater superoleophobicity of the membrane (Figure 4a, right). When the underwater CSPNM comes into contact with oil, the water trapped in the rough structures will significantly decrease the contact area between the surface of the membrane and oil droplet, showing a large oil CA underwater (underwater superoleophobicity). Thus, the CSPNM shows superhydrophilic and underwater superoleophobic properties.
To better assess the membrane's oil repellency underwater, its adhesion behavior was measured during the process of oil contact with the membrane surface underwater. Before the beginning of the experimentation, an oil droplet (5 µL) was squeezed on the CSPNM surface underwater under a preload, and then the oil droplet was released. In the relaxing process, the oil droplet maintains the spherical shape underwater, as shown in Figure 4b, indicating that CSPNM displays an ultralow adhesion with the oil droplet underwater. The results demonstrate that the CSPNM has a much better anti-adhesion performance to oil underwater, probably attributed to the formation of hydrogen bonds between the water and hydrophilic components in the CSPNM [35].

Filtration
The filtration system is shown in Figure 5a. The as-prepared CSPNM was sealed between a vertical glass tube and a conical flask. After pre-wetting the CSPNM with water, the oil-in-water emulsion was poured into the glass tube and the emulsion was separated at a low pressure (0.1 bar) by a vacuum pump. The separation process is shown in Movie S2. A schematic is shown in Figure 5b for a clear understanding of this separation process. The emulsion comes into contact with the top surface of the CSPNM and will demulsify once the emulsion droplets touch the membrane surface. The water phase passes through the CSPNM and meanwhile the oil droplets block and coalesce at the surface of the membrane, as shown in Figure 5b. The wetting behavior of the membrane transformed from superhydrophilicity in the air to superoleophobicity underwater after being prewetted by water. The water layer formed at the interface of oil and the membrane, which repelled the oil.

Filtration
The filtration system is shown in Figure 5a. The as-prepared CSPNM was sealed between a vertical glass tube and a conical flask. After pre-wetting the CSPNM with water, the oil-in-water emulsion was poured into the glass tube and the emulsion was separated at a low pressure (0.1 bar) by a vacuum pump. The separation process is shown in Movie S2. A schematic is shown in Figure 5b for a clear understanding of this separation process. The emulsion comes into contact with the top surface of the CSPNM and will demulsify once the emulsion droplets touch the membrane surface. The water phase passes through the CSPNM and meanwhile the oil droplets block and coalesce at the surface of the membrane, as shown in Figure 5b. The wetting behavior of the membrane transformed from superhydrophilicity in the air to superoleophobicity underwater after being prewetted by water. The water layer formed at the interface of oil and the membrane, which repelled the oil.
To better understand the separation capability of the CSPNM, a series of oil-in-water emulsions including surfactant-free/surfactant-stabilized emulsions were used for this separation. Optical microscopy was used for observing different feed solutions and their corresponding permeate filtrates. Figure 6a shows the separation results of the H/W emulsion as an example. In the middle photograph, the transparent collected filtrate (right) contrasts with the original white feed solution (left). The emulsion droplets of micrometer size were observed in the feed solution, and no emulsion droplets were observed in the image of the collected filtrate. For the D/W emulsion, the entire view shows densely packed droplets (Figure 6b), whereas no droplets were observed in the photo of the collected filtrate, indicating the successful removal of diesel. The separate results of other emulsions are shown in Figure S2, and none of the images of the filtrates show any droplet, implying the good separation capability of the CSPNM for various emulsions. pump. The separation process is shown in Movie S2. A schematic is shown in Figure 5b for a clear understanding of this separation process. The emulsion comes into contact with the top surface of the CSPNM and will demulsify once the emulsion droplets touch the membrane surface. The water phase passes through the CSPNM and meanwhile the oil droplets block and coalesce at the surface of the membrane, as shown in Figure 5b. The wetting behavior of the membrane transformed from superhydrophilicity in the air to superoleophobicity underwater after being prewetted by water. The water layer formed at the interface of oil and the membrane, which repelled the oil.  To better understand the separation capability of the CSPNM, a series of oil-in-water emulsions including surfactant-free/surfactant-stabilized emulsions were used for this separation. Optical microscopy was used for observing different feed solutions and their corresponding permeate filtrates. Figure 6a shows the separation results of the H/W emulsion as an example. In the middle photograph, the transparent collected filtrate (right) contrasts with the original white feed solution (left). The emulsion droplets of micrometer size were observed in the feed solution, and no emulsion droplets were observed in the image of the collected filtrate. For the D/W emulsion, the entire view shows densely packed droplets (Figure 6b), whereas no droplets were observed in the photo of the collected filtrate, indicating the successful removal of diesel. The separate results of other emulsions are shown in Figure S2, and none of the images of the filtrates show any droplet, implying the good separation capability of the CSPNM for various emulsions. To further study the separation ability of the CSPNM, the collected filtrates were analyzed by UV-Vis spectroscopy and TOC analysis. For each surfactant-stabilized emulsion, the oil plus the surfactant residues in the filtrate is the TOC value. As shown in Figure 7a, the oil contents of the other emulsions were <45 ppm except for the D/W emulsion, indicating the high separation efficiency of the CSPNM. The D/W emulsion showed higher oil content, probably because of the densely packed emulsion droplets flooding the feed solution (Figure 6b, left).
The oil rejection of the CSPNM directly reflects the separation effect of the membrane. The oil contents of the original emulsions and the oil rejection of the CSPNM for various oils are listed in Table S1. The oil concentration in the original emulsions ranges from 12,200 ppm to 33,360 ppm, and their corresponding oil rejections are >99.60%; some of them are even up to 99.85%, illustrating the high separation efficiency of the CSPNM.
UV-Vis spectrometry was used to analyze the emulsions and their corresponding collected filtrates. As shown in Figure 7b, no obvious characteristic peaks for toluene are observed in the spectrum of the filtrate. Similar results are also obtained for the H/W and D/W emulsions ( Figure S3). To further study the separation ability of the CSPNM, the collected filtrates were analyzed by UV-Vis spectroscopy and TOC analysis. For each surfactant-stabilized emulsion, the oil plus the surfactant residues in the filtrate is the TOC value. As shown in Figure 7a, the oil contents of the other emulsions were <45 ppm except for the D/W emulsion, indicating the high separation efficiency of the CSPNM. The D/W emulsion showed higher oil content, probably because of the densely packed emulsion droplets flooding the feed solution (Figure 6b, left).
The oil rejection of the CSPNM directly reflects the separation effect of the membrane. The oil contents of the original emulsions and the oil rejection of the CSPNM for various oils are listed in Table S1. The oil concentration in the original emulsions ranges from 12,200 ppm to 33,360 ppm, and UV-Vis spectrometry was used to analyze the emulsions and their corresponding collected filtrates. As shown in Figure 7b, no obvious characteristic peaks for toluene are observed in the spectrum of the filtrate. Similar results are also obtained for the H/W and D/W emulsions ( Figure S3).

Recyclability and Chemical Durability Test Experiments
To test the recyclability and the antifouling property of the CSPNM, a cyclic experiment was conducted, using 10 mL of feed emulsion (H/W) in every cycle. Figure 9 shows the flux and oil content in the filtrates within 20 cycles. In the whole testing process, the flux decreased by 19.54% from the first cycle (about 1940.17 L·m −2 ·h −1 ) to the 20th cycle (about 1561.09 L·m −2 ·h −1 ); no obvious flux decrease

Recyclability and Chemical Durability Test Experiments
To test the recyclability and the antifouling property of the CSPNM, a cyclic experiment was conducted, using 10 mL of feed emulsion (H/W) in every cycle. Figure 9 shows the flux and oil content in the filtrates within 20 cycles. In the whole testing process, the flux decreased by 19.54% from the first cycle (about 1940.17 L·m −2 ·h −1 ) to the 20th cycle (about 1561.09 L·m −2 ·h −1 ); no obvious flux decrease

Recyclability and Chemical Durability Test Experiments
To test the recyclability and the antifouling property of the CSPNM, a cyclic experiment was conducted, using 10 mL of feed emulsion (H/W) in every cycle. Figure 9 shows the flux and oil content in the filtrates within 20 cycles. In the whole testing process, the flux decreased by 19.54% from the first cycle (about 1940.17 L·m −2 ·h −1 ) to the 20th cycle (about 1561.09 L·m −2 ·h −1 ); no obvious flux decrease or oil content decrease were observed at the near cycle, indicating the superior recyclability and good antifouling property of the membrane.  To study the thermal stability of the CSPNM, the underwater oil CAs and roll-off angles of the CSPNM were measured after heating at different temperatures for 12 h. Figure 10 shows that the CSPNM maintains its superwetting property after the heating processes with the oil CAs >150° and roll-off angles <15°. These results indicate the superior thermal stability of the CSPNM. The chemical durability of the CSPNM was tested by measuring the underwater oil CAs and roll-off angles of the membrane after immersing CSPNMs in solutions with different pH values (Figure 11a,b) for 30 h. Figure 11c shows the CSPNM with an underwater oil CA >150° and roll-off angles <15° after being in contact with the corrosive aqueous solutions (1 to 14 pH values), showing the superior environmental stability of the CSPNMs under harsh conditions. To study the thermal stability of the CSPNM, the underwater oil CAs and roll-off angles of the CSPNM were measured after heating at different temperatures for 12 h. Figure 10 shows that the CSPNM maintains its superwetting property after the heating processes with the oil CAs >150 • and roll-off angles <15 • . These results indicate the superior thermal stability of the CSPNM.  To study the thermal stability of the CSPNM, the underwater oil CAs and roll-off angles of the CSPNM were measured after heating at different temperatures for 12 h. Figure 10 shows that the CSPNM maintains its superwetting property after the heating processes with the oil CAs >150° and roll-off angles <15°. These results indicate the superior thermal stability of the CSPNM. The chemical durability of the CSPNM was tested by measuring the underwater oil CAs and roll-off angles of the membrane after immersing CSPNMs in solutions with different pH values (Figure 11a,b) for 30 h. Figure 11c shows the CSPNM with an underwater oil CA >150° and roll-off angles <15° after being in contact with the corrosive aqueous solutions (1 to 14 pH values), showing the superior environmental stability of the CSPNMs under harsh conditions.

Conclusions
In summary, superhydrophilic and underwater superolephobic CSPNMs were fabricated by a simple and environmentally friendly phase-inversion method. The addition of CSP significantly improved the structure and mechanical strength of the membrane. In addition, oil-in-water emulsions were separated with high oil rejection (>99.60%) and flux >660.00 L·m −2 ·h −1 . After 20 separation cycles of the H/W emulsion, the oil rejection was still >99.50% and no obvious decline in the flux was observed, indicating superior recyclability. Moreover, the CSPNM maintained underwater superolephobicity after being immersed in various corrosive aqueous solutions and exposed to different temperatures, exhibiting excellent environmental stability. We believe that the utilization of CSP and emulsion separation would be of great significance to sustainable development.
Supplementary Materials: The following information is available online at www.mdpi.com/xxx/s1, Table S1: the oil content in the feed solutions and corresponding oil rejection, Figure S1: the details of the cross section, top surface, and bottom surface of the CSPNM, Figures S2 and S3: the separation results for various oil-in-water emulsions. Supplementary video: a water droplet permeates into the CSPNM (Video S1) and the separation process (Video S2).

Conclusions
In summary, superhydrophilic and underwater superolephobic CSPNMs were fabricated by a simple and environmentally friendly phase-inversion method. The addition of CSP significantly improved the structure and mechanical strength of the membrane. In addition, oil-in-water emulsions were separated with high oil rejection (>99.60%) and flux >660.00 L·m −2 ·h −1 . After 20 separation cycles of the H/W emulsion, the oil rejection was still >99.50% and no obvious decline in the flux was observed, indicating superior recyclability. Moreover, the CSPNM maintained underwater superolephobicity after being immersed in various corrosive aqueous solutions and exposed to different temperatures, exhibiting excellent environmental stability. We believe that the utilization of CSP and emulsion separation would be of great significance to sustainable development.
Supplementary Materials: The following information is available online at http://www.mdpi.com/2073-4360/ 10/3/323/s1, Table S1: the oil content in the feed solutions and corresponding oil rejection, Figure S1: the details of the cross section, top surface, and bottom surface of the CSPNM, Figures S2 and S3: the separation results for various oil-in-water emulsions. Supplementary video: a water droplet permeates into the CSPNM (Video S1) and the separation process (Video S2).