Electrospinning-Based Super Liquid-Repellent Membranes for Membrane Distillation: Theory, Fabrications, Applications, and Challenges

Abstract

The potential of membrane distillation (MD) in seawater desalination and high-salt wastewater treatment makes it a highly promising application in alleviating the global water crisis. However, membrane fouling and wetting are the main obstacles to the large-scale application of MD. Bio-inspired super liquid-repellent membranes offer a viable resolution to these challenges. The rapid advancement of nanotechnology has stimulated the growing attention toward electrospun nanofiber membranes (ENMs). Electrospun fibers demonstrate excellent functionalization, controllability, and hydrophobicity. Their low energy consumption and ease of preparation promote their application prospects in the construction of super liquid-repellent membranes. This article provides a comprehensive summary of electrospinning principles and influencing factors, coupled with a detailed account of the theory and preparation of super-liquid-repellent membranes via electrospinning, thus explicating the application and challenges of these membranes in MD, facilitating a deeper understanding of the ENMs application in MD for readers.

1. Introduction

Water is an essential resource for the survival of human society. However, only 2.5% of the earth’s total water is fresh water, and most of it is deep groundwater, glaciers, and polar ice caps, rendering it unusable for humans [1]. As a result, less than 1% of the world’s fresh water resources are available for human consumption and production. With the rapid growth of industrialization and population, the tension between water supply and demand has intensified, leading to an increasingly severe water scarcity problem. This has resulted in an estimated one billion people worldwide lacking access to clean fresh water. Therefore, finding ways to alleviate the shortage of fresh water resources has become a critical issue of global concern.

Desalination can facilitate the production of freshwater and alleviate the crisis of freshwater scarcity. Its two primary existing practices include thermal treatment and membrane technology [2]. Nevertheless, thermally driven processes for seawater desalination have been observed to be susceptible to scaling, which entail high energy consumption and expenses. As a result, the membrane technology has gained attraction in the market with its associated benefits [3]. Thanks to its low energy consumption and small space requirements, reverse osmosis (RO) is currently the most widely used desalination technology worldwide. However, a significant amount of concentrated salt wastewater is produced in the RO process, adversely affecting the marine ecological environment. Furthermore, forward osmosis (FO), which utilizes natural osmosis to eliminate the need for pressurization, has become an emerging technology, but bottlenecks in high-efficiency membrane design have hindered its application.

In this context, membrane distillation (MD) has emerged as a promising seawater desalination technology due to its advantages of low energy consumption, high salt rejection and ability to handle high-concentration brine [4]. As shown in Figure 1, MD is a thermal-driven process that utilizes a hydrophobic microporous membrane as a barrier between hot-side feed solution and the cold-side permeate solution. The vapor pressure difference between the two sides of the membrane is the mass transfer driving force, which separates non-volatile components [5]. Membrane, as the core of MD, plays an important role in its operation. Traditional hydrophobic membranes are prone to fouling and wetting, which impedes the industrial application of MD. Fortunately, super liquid-repellent membranes inspired by natural surfaces, such as lotus leaves, fish scales, and cicada wings, can alleviate membrane fouling and enhance membrane wetting performance. Superhydrophobic membranes, omniphobic membranes, and underwater superoleophobic membranes fall under this category.

Figure 1.

Schematic diagram of membrane distillation process.

The fabrication of hydrophobic membranes often employs non-solvent induced phase separation (NIPS), thermally induced phase separation (TIPS), sintering, and track-etching [6]. However, electrospinning technology, which can produce nanofibrous membranes with unique structural features, is superior to these methods. Electrospinning technology can construct super liquid-repellent membranes with concave geometries and facilitate the large-scale production of nanofibrous membranes. Additionally, incorporating functional polymers or nanoparticles in spinning precursors or combining electrospinning with membrane surface modification can create super liquid-repellent membranes with different wettability.

This chapter provides a review of recent progress in the fabrication of super liquid-repellent electrospun nanofibrous membranes (ENMs) using electrospinning technology. It discusses the electrospinning mechanism and influencing parameters, followed by the theory and preparation of super liquid-repellent membranes. The section that follows illustrates the applications and challenges of these membranes in MD. Finally, the paper concludes with a brief summary and outlook.

2. Theory of membrane distillation

2.1 MD configurations

As mentioned above, MD is a thermal-driven membrane process in which the water vapors transport through a microporous hydrophobic membrane and condense on the cold permeate side [4]. Theoretically, the membrane’s inherent hydrophobicity ensures its non-wettability by the feed solution. On the thermal boundary, the hot-side feed directly engages the hydrophobic membrane surface, effectuating evaporation at the membrane interface. The motion of steam across the membrane is impelled by pressure differentials, leading to its condensation on the permeate side. The defining attributes of MD technology encompass the subsequent elements [7]: (1) the membrane is hydrophobic and microporous, (2) at least one side of the membrane surface is in contact with the feed liquid and will not be wetted, (3) conversion of volatile feed liquid constituents into steam, transiting through membrane pores, and (4) exploitation of the vapor pressure gradient across the membrane to propel the passage of volatile constituents through its pores.

To counteract the thermal losses stemming from membrane conduction during MD operations, various MD configurations have evolved. Conventionally, there are four designs for MD configurations found in the literature [8]: Direct contact membrane distillation (DCMD), air gap membrane distillation (AGMD), vacuum membrane distillation (VMD) and sweeping gas membrane distillation (SGMD). A compendium of the principal attributes, merits, and demerits of these four MD configurations is tabulated in Table 1.

MD configuration	Main characteristics	Advantages	Disadvantages
DCMD	One side of the hydrophobic membrane is in contact with the hot side feed solution and the other side is in contact with the cold side permeate solution.	The simplest configuration.; Easy operation.; High water flux.	Lower thermal efficiency.; Highly affected by temperature polarization
AGMD	The hydrophobic membrane is only in contact with the hot side feed solution. An extra air gap compartment is introduced between the other side of the membrane and the condensation plate.	Less heat loss.	Greater mass transfer resistance
VMD	An additional vacuum pump is installed on the condensing side, and the permeated vapor is condensed in an external condensing device.	High flux.; Heat loss can be ignored.	Membrane wetting is prone to occur.; Complex configuration.
SGMD	The inert gas introduced on the permeation side will carry the vapor passing through the membrane pores to the external condensing device for condensation.	Greater mass transfer drive.; Lower mass transfer resistance compared to AGMD.	High energy consumption.; Complex configuration.

Table 1.

The main attributes, merits, and demerits of the four MD configurations.

2.2 Methods of fabrication of MD membranes

Hydrophobicity and porosity constitute pivotal attributes of MD membranes. Various techniques, such as phase inversion, stretching, track-etching, and electrospinning, are conventionally employed for the fabrication of diverse MD membranes [9]. In addition, different forms of membranes in MD such as single-layer membranes and composite membranes are used in different MD scenarios to enhance MD performance.

The phase inversion process hinges upon a transition between two phases, induced by altering the solubility of the polymer constituents [10]. Through modulation of the composition or conditions, the initially uniform polymer solution undergoes phase separation, leading the casting solution’s transition from a liquid to a solid state. The ultimate membrane is acquired by solidifying the polymer-rich phase. Versatile membrane structures are typically realized via non-solvent induced phase separation (NIPS), temperature induced phase separation (TIPS), and vapor induced phase separation (VIPS). It is worth noting, however, that the viable membrane materials are constrained due to the necessity for soluble polymers in the phase inversion process.

Stretching represents a solvent-free approach to membrane fabrication, entailing the heating of a polymer beyond its melting point and extruding it into the desired form under elevated stress, thereby engendering membrane pores through elongation [9]. This methodology finds suitability in the case of highly crystalline polymeric materials, such as Polytetrafluoroethylene (PTFE) and polyethylene (PE). Notably straightforward in terms of equipment and procedural intricacy, this technique exhibits environmental merits. Nevertheless, the resultant thin membrane is somewhat circumscribed by its substantial pore size and broad pore size distribution.

Track-etching is a widely employed means to fabricate membranes with small pore sizes by irradiating non-porous polymer membranes with high energy heavy ions to create linear damaged tracks [11]. Accurate control of the pore size distribution can be achieved using this method.

These methodologies serve the purpose of scalable MD membrane production, however, substantial technical intricacies still warrant comprehensive exploration. These challenges encompass issues such as restricted permeation flux and the propensity for facile pore wetting. These concerns emanate from the inherent deficiencies in membrane surface hydrophobicity and surface roughness, alongside the constrained porosity, diminutive pore dimensions, and the presence of intricate tortuous closed-pore configurations within the membrane matrix. Conversely, nanofiber membranes produced via the intricate art of electrospinning present a contrasting picture. These membranes feature an intricate network of interconnected pores, distinguished by elevated porosity levels, and underscored by the added advantage of precise control over fiber morphology and membrane thickness. Notably, the process for preparing these nanofiber membranes is streamlined and straightforward, further enhancing their appeal. As a result, within the realm of membrane distillation, researchers have increasingly directed their focus toward these ENMs, recognizing their potential, as evidenced by the mounting scholarly interest in recent years.

3. Theory of electrospinning process

Electrospinning is a highly versatile and cost-effective technique for fabricating nanofibrous membranes with exceptional specific surface area, pore structure, and surface morphology. This technology was first discovered in 1897 and made significant progress in the early 1990s after undergoing a series of detailed studies [12]. Nanofibrous membranes produced by electrospinning typically exhibit high porosity and interconnected pores, which reduce mass transfer resistance within the structure, making them highly suitable for membrane separation applications. In this section, we will discuss the fundamental principles of electrospinning technology and the influence of various parameters on the nanofibrous membrane during the spinning process.

3.1 Understanding electrospinning

In recent years, considerable attention has been given to the production, characterization, and application of electrospun nanofibers, as well as the enhancement of spinning devices for more precise and efficient fabrication of complex nanofibers tailored to specific applications. To this end, various sophisticated electrospinning devices have been developed, such as vibration electrospinning, magneto-electrospinning, and melt electrospinning [13]. Nevertheless, regardless of the design, every electrospinning apparatus is composed of four fundamental components: a high-voltage power supply, a syringe pump, a spinneret, and a collector (Figure 2) [14]. Generally, a high-voltage power supply refers to a DC power supply capable of generating thousands to tens of thousands of volts, which will form a high-voltage electric field to charge the fluid, causing disturbance and polarization of the fluid, and finally forming a jet. The spinneret is connected to a container containing the polymer solution, and the fluid flows through the spinneret at a constant and controllable speed driven by the syringe pump. The collector, typically constructed from metal, is either a rotating drum or aluminum foil connected to the negative terminal of the high-voltage power supply.

Figure 2.

Schematic diagram of electrospinning process.

Electrospinning is a distinctive technique of producing fibers utilizing electrostatic forces. At room temperature and atmospheric conditions, a polymer solution or molten liquid is placed under a high-voltage electric field to form a charged jet. As the jet traverses the air, solvent evaporation and fiber solidification occur, leading to the formation of polymeric nanofibers on the collecting plate [15]. The electrospinning process typically progresses through three primary phases: (1) spindle formation and jet initiation; (2) rectilinear jet development and bending instability, and (3) nanofiber solidification and collection.

Initially, a high-voltage power supply is employed to apply a high constant electric field between the polymer droplets at the tip of the spinneret and the grounded collector, charging the pendulous droplet at the spinneret. Subsequently, through the action of Coulombic repulsion force and electrostatic force, the semicircular droplet progressively changes into a cone shape, known as the Taylor cone. Above a critical electric field, the electrostatic force surpasses the surface tension of the polymer solution, resulting in the ejection of a fine jet from the tip of the Taylor cone.

During flight, the jet first moves in a straight line at a distance from the spinneret that has a critical value proportional to the applied electric field, electrical conductivity and flow velocity of the polymer solution [14]. After linear motion, electrically induced bending instabilities occurs, forming a conical unstable region, which is crucial for refining the fiber diameter from micrometers to nanometers. Ultimately, rapid solvent evaporation due to the high surface area to volume ratio leads to the solidification of the fibers and the production of the desired nanofiber membrane on the collection plate.

3.2 Different parameters affecting resultant nanofibrous membranes

Different parameters in the electrospinning process including the inherent properties of the polymeric dopes, operational conditions and surrounding environmental conditions can significantly affect the resultant nanofibrous membranes. By controlling these parameters, nanofibers with ideal morphology and diameter can be obtained.

3.2.1 Effect of polymeric dopes

It is widely acknowledged that the efficacy of electrospinning is significantly influenced by the polymer materials employed [16, 17]. In the MD process, the materials that can be used to manufacture nanofibrous membranes need to be hydrophobic and soluble. Different types of polymer materials have different molecular weights, which play a pivotal role in determining the electrical and rheological characteristics of the spinning solution. Generally, smaller molecular weights lead to lower solution viscosities, resulting in smaller fiber diameters that are beneficial for mitigating wetting but not permeate fluxes. Furthermore, the diameter of the nanofibers depends greatly on the concentration of the polymer solution. When the concentration of the solution is too low, nanofibers with a beaded structure are formed, and when the concentration is increased, the spherical nanofibrous microspheres tend to become spindle-like shapes [18].

During the spinning process, a high-voltage electric field is required for the polymer solution to be ejected from the end of the spinneret [19]. Hence, it is imperative that the solution is conductive to a certain extent to ensure that it can be ejected appropriately. If the solution’s conductivity is too strong, the ejected fiber will undergo considerable bending deformation, which is not conducive to its spreading on the collector, resulting in microbeads and a dense network [20]. Conversely, too low conductivity results in insufficient stretching of the jet, leading to non-uniform fibers. Therefore, at the initial stage of spinning, it should be ensured that the conductivity of the polymer solution is within an appropriate range. Usually, ions are introduced into the spinning solution to adjust the conductivity and increase the spinnability due to their high mobility and charge density [20].

It was revealed that the morphology and properties of electrospun nanofibers are affected by the surface tension of polymer solutions [21, 22]. A suitable surface tension contributes to the stability of the jet. By reducing the surface tension of the solution, the electrospinning process can be completed at lower electric field strength. Adding a suitable dose of surfactant in the electrospinning solution helps to generate uniform nanofibers. However, an excessive amount of surfactant can lead to self-assembly, causing colloidal aggregation and ultimately resulting in defects on the fiber surface.

The choice of solvent is crucial because its evaporation governs the solidification of nanofibers during electrospinning. The solvent type affects the solution’s viscosity, with higher viscosities making it difficult for polymer solutions to eject under appropriate electric field strength and lower viscosity producing discontinuous fibers. Currently, N, N-dimethylformamide (DMF) and N-methylpyrrolidone (NMP) are relatively common polar solvents that can dissolve polymers, while acetone is a non-solvent that can dilute polymer solutions. Usually, both solvent and non-solvent are utilized to prepare the spinning solution.

3.2.2 Effect of operational conditions and surrounding environmental conditions

In addition to the intrinsic properties of polymer solutions, extrinsic factors such as operational and environmental conditions can also exert a substantial influence on the nanostructure of ENMs. Among the operational parameters, spinning time directly dictates the thickness of the nanofibrous membrane and hence modulates the membrane permeability. The working distance, which denotes the distance between the spinneret and the collector, affects the strength of the electric field and the evaporation kinetics of the solvent. An excessively short working distance may impede solvent evaporation and lead to undesired adhesion among fibers. Another critical parameter is the applied electric field. Theoretically, a higher voltage facilitates fiber stretching and formation of smaller diameter fibers. However, as the voltage increases, a “frying” phenomenon may occur, reducing the collection efficiency of fibers [6]. In actual operation, the voltage of the spinning system should be adjusted according to the shape of the “Taylor cone.”

Apart from operational factors, the ambient environmental conditions, such as temperature and relative humidity, can also significantly affect the nanostructure of fibers [23]. Temperature has a dual effect on the viscosity and surface tension of the spinning solution and influences the evaporation rate of the solvent. A higher temperature leads to a lower viscosity and surface tension, resulting in smaller diameter nanofibers. Furthermore, it can promote solvent evaporation and facilitate fiber solidification. Conversely, relative humidity is an essential factor that can impact the evaporation kinetics of the solvent, thereby affecting the resulting nanofibrous membranes. A low relative humidity can accelerate the evaporation rate and cause premature coagulation of the spinning solution, which may clog the spinneret and interfere with the spinning process. Therefore, it is crucial to control the humidity to achieve the desired nanofiber membrane morphology.

4. Preparation of super liquid-repellent surfaces by electrospinning

In the context of mitigating membrane wetting and fouling during the MD process, the development of super-liquid repellent membranes with exceptional wetting properties has emerged as a current research focus. Compared with other fabrication methods, electrospinning is a superior technique for producing nanofibers with good functionality, high porosity, and specific surface area. ENMs have gained extensive utilization in MD due to their low energy consumption and high efficiency. By means of physical and chemical modifications, the interfacial properties of ENMs can be optimized to prepare super-liquid repellent membranes that are suitable for treating complex wastewater in MD. This section will elaborate on the preparation of super-liquid repellent membranes based on electrospinning, in accordance with the classical wetting model.

4.1 Superhydrophobic surfaces

4.1.1 Theory of superhydrophobic membranes

Highly hydrophobic membranes can effectively prevent wetting by aqueous solutions during the MD process. Superhydrophobic surfaces have good resistance to wetting, which is defined as water contact angle (WCA) greater than 150° [24]. Inspired by the “lotus leaf effect,” it was found that the self-cleaning properties of the surface were attributed to the combination of micron-structured papillae and epicuticular wax layers [25]. Subsequent studies suggested that there are nanostructures on the surface of the lotus leaf, and the combination of micro/nanostructures together promotes its superhydrophobicity and self-cleaning properties. This discovery provides a theoretical basis for the construction of superhydrophobic surfaces.

The interfacial tension between the contacting phases determines the degree of wetting of the membrane surface, which is usually characterized by the contact angle (CA), the larger the CA, the less droplets adhere to the membrane. Young’s equation is the simplest equation to describe the wettability of a surface, and is usually used to describe the state of a water droplet on an ideal flat (a smooth and chemically uniform surface) [26, 27]:

$$cos{\theta }_{\mathit{eq}}=\frac{{\gamma }_{\mathit{SG}}-{\gamma }_{\mathit{SL}}}{{\gamma }_{\mathit{LG}}}$$E1

where ${\theta }_{\mathit{eq}}$ is the CA of an ideal solid surface. ${\gamma }_{\mathit{SG}}$, ${\gamma }_{\mathit{SL}}$ and ${\gamma }_{\mathit{LG}}$ represent the interfacial tension between solid–gas, solid–liquid and liquid–gas, respectively. According to this equation, it can be known that the CA of the surface can be changed by introducing different chemical groups [28].

However, perfectly smooth solid surfaces rarely exist in practical situations. Therefore, the Wenzel is used to describe the wettability of rougher surfaces [29]:

$$cos{\theta }_W=\text{rcos}{\theta }_{\mathit{eq}}$$E2

where r is the roughness factor, defined as the ratio of the wetted surface to the projected surface. ${\theta }_W$ represents the apparent WCA of the rougher surface. According to this theory, the wettability of surfaces with greater surface roughness will be enhanced. The Wenzel model is further extended by the Cassie-Baxter theory, which assumes that a rough porous surface will not be filled with liquid due to the presence of air pockets [29, 30]:

$$cos{\theta }_{\mathit{CB}}={\mathrm{f}}_{\mathit{sl}}\left(1+cos{\theta }_{\mathit{eq}}\right)-1$$E3

where ${\theta }_{\mathit{CB}}$ is the apparent contact angle under this theoretical model and ${\mathrm{f}}_{\mathit{sl}}$ is the contact area fraction of solid–liquid. According to this theory, a superhydrophobic surface can be obtained by reducing the solid–liquid interface. The left side of Figure 3(A–C) shows the wetting behavior of water droplets in three different states.

Figure 3.

Typical wetting behavior of a droplet on the solid substrates. (A) in-air Young’s model. (B) in-air Wenzel’s model. (C) in-air Cassie’s model. (D) underwater Young’s model, Wenzel’s model and Cassie’s model.

4.1.2 Engineering in-air superhydrophobic surfaces by electrospinning

Based on the aforementioned theory, superhydrophobic surfaces are constructed by designing surfaces with a certain roughness or using low surface energy materials. So far, studies have successfully combined bionics-based electrospinning technology with various modification procedures to prepare superhydrophobic surfaces for MD, including: one-step electrospinning, surface modification and mixed matrix methods.

One-step electrospinning involves mixing low surface energy materials into the spinning solution, which is a straightforward and effective approach to generate superhydrophobic nanofiber membranes. For instance, perfluorooctyltriethoxysilane (PFOTES) has been added to polymer solutions to synthesize fluorinated silane functionalized polyvinylidene fluoride (PVDF), and the nanofiber membranes obtained by electrospinning using this solution exhibited a higher WCA than those prepared by solvent evaporation (Figure 4) [31]. Furthermore, the ENMs without PFOTES (Figure 4B) contained a large number of beaded structures, while the formation of bead-free nanofibers imparted a stronger hydrophobicity to the PVDF/PFOTES nanofiber membranes (Figure 4C). PTFE is also selected to be added to the PVDF polymer solution for electrospinning. Within a certain range, the WCA of the membrane increased with increasing PTFE concentration, gradually exhibiting superhydrophobicity [32].

Figure 4.

SEM images of membranes prepared from (A) PVDF/PFOTES using solvent evaporation, (B) PVDF, (C) PVDF/PFOTES using electrospinning. Reproduced from [31] with permission. Copyright (2009) Elsevier Inc.

Despite the simplicity of one-step electrospinning, it has certain limitations since the available polymer materials for direct electrospinning are relatively limited. Therefore, it is necessary to explore the sources of hydrophobic materials further. In recent years, many attempts have been made to directly deposit low surface energy materials on the surface of ENMs to prepare superhydrophobic membranes suitable for MD processes. Chemical vapor deposition (CVD) has been utilized to produce superhydrophobic membranes. MD membranes with a WCA of 151° can be created by depositing poly(1H,1H,2H,2H-perfluorodecyl acrylate) (PPFDA) on an ENMs substrate. In this study, a custom-built reactor was used for deposition of PPFDA [33]. In addition, low surface energy fluoroalkylsilane (FAS) was used to reduce the surface tension of the membrane. The prepared electrospun nanofibrous membrane was placed in FAS solution (2 wt. %) for hydrophobic modification process. The resulting nanofiber membrane F-PVA can reach a water contact angle of 158° [34]. In a preceding investigative endeavor from our body of work, we detailed the development of a superhydrophobic membrane designated as PVDF-S, distinguished by its heightened resistance to wetting. This innovative membrane was demonstrated to exhibit a sustained and stable performance within the context of MD, particularly when subjected to feed solutions containing DTAB (N,N,N-trimethyl-1-dodecanaminium bromide). The foundational construction of the PVDF substrate membrane itself was realized via the intricate process of electrospinning, thereafter subjected to fluorination through the application of a low surface energy FAS. The ensuing fluorination procedure, integral to our approach, effectively brought about a substantial reduction in surface energy, consequently conferring a notable enhancement in hydrophobic characteristics upon the membrane [4].

Surface modification, which reduces the surface energy of the membrane, can moderately increase hydrophobicity. However, the roughness of the nanofibers produced by electrospinning is inadequate, which limits further enhancement of the final membrane hydrophobicity. According to the theory of Wenzel and Cassie-Baxter, in addition to reducing the surface energy, another way to achieve superhydrophobic properties is the construction of micro/nanoscale rough structures. The mixed matrix method refers to combining multi-level roughness and low surface energy surface through post-treatment or doping of bulk solution to prepare superhydrophobic membranes. Typically, various nanoparticles such as silicon dioxide (SiO₂), titanium dioxide (TiO₂), and silver nanoparticles (AgNPs) are usually applied to increase the roughness of the membrane surface [5, 35, 36]. For example, hydrophobically modified SiO₂ was incorporated into the spinning solution for one-step electrospinning to construct MD membranes with outstanding properties, and its WCA reached 157°. Compared with the unmodified PVDF membrane, the water osmotic pressure of the superhydrophobic membrane doubled, and the water flux was close to 25.7 kg m⁻² h⁻¹ during 100 h of DCMD [37]. Table 2 summarizes the experimental values of water contact angle (WCA), oil contact angle (OCA), and sliding angle (SA) for superhydrophobic membranes, omniphobic membranes, and Janus membranes mentioned in this chapter.

Membrane name	Category	WCA (°)	OCA (°); In-air/underwater	SA (°)
PVDF/PFOTES	Superhydrophobic	156	/	/	/
PVDF-PTFE	Superhydrophobic	152.2 ± 2.0	/	/	/
PPFDA-iCVD	Superhydrophobic	151 ± 2	/	/	/
F-PVA	Superhydrophobic	158	/	/	4
PVDF-S	Superhydrophobic	>150	>120	<30	36
PVDF/SiO2	Superhydrophobic	157 ± 1	/	/	3.7 ± 1.2
P/CF-60	Omniphobic	160.9 ± 0.9	∼149°	/	∼51
SiNPs-PVDF-HFP/BTEAC	Omniphobic	∼150	∼139	/	/
F-POSS/PVDF-HFP	Omniphobic	154.5 ± 2.6	148.8 ± 3.7	/	/
FAS@SiNPs-SFM	Omniphobic	154.2 ± 3.6	154.2 ± 3.6	/	/
PVDF-F	Omniphobic	164.5 ± 1.7	>150	/	1.6 ± 0.1
Janus(o)	Underwater superoleophobic	<30	<60	∼149°	/
PH-7.5RC	Underwater superoleophobic	0	0	140.5 ± 2.62	/
PTFE/PAN-OH	Underwater superoleophobic	<40°	/	161.7 ± 1.8	/

Table 2.

The experimental values of water contact angle (WCA), oil contact angle (OCA), and sliding angle (SA) for superhydrophobic membranes, omniphobic membranes, and Janus membranes.

4.2 Omniphobic surfaces

4.2.1 Theory of omniphobic membranes

Despite the ability of superhydrophobic membranes to alleviate wetting and fouling, their practical operation can be hampered by complex feed solutions, which can easily induce membrane wetting and halt the MD process. Young’s equation predicts that reducing surface tension enhances wettability, whereas Wenzel’s theory proposes that increasing the surface roughness of the membrane improves its anti-wetting ability. Furthermore, Cassie-Baxter theory suggests that reducing the contact area between solid and liquid is a method to further increase liquid repellency. Therefore, omniphobic membranes with re-entrant structures and low surface energy surfaces is suitable for treating complex feed liquids. These membranes are both hydrophobic and oleophobic, allowing them to retain both polar and non-polar liquids. A good omniphobic membrane should possess a relatively high water and oil contact angle, high flux, high LEP value, and excellent long-term stability.

As illustrated in the right side of Figure 3(A–C), Young’s, Wenzel’s and Cassie-Baxter’s models are also applicable in describing the wettability of oil droplets on flat or rough solid surfaces. In this context, the liquid in the model refers to oil droplets. When in the Wenzel state, the oil droplets with low surface tension can easily wet the surface of the ridged structure. The wetting tendency increases with increasing distance between two ridges. In contrast, when in the Cassie-Baxter state, due to the presence of the re-entrant structure and the low surface energy surface, the air below the recessed structure generates a negative Laplace pressure difference, which plays a key role in preventing further penetration of the oil droplets [28].

4.2.2 Engineering in-air omniphobic surfaces by electrospinning

To construct an omniphobic membrane surface, two key factors must be considered: (1) building a re-entrant structure to enhance the surface roughness, and (2) creating a surface with low surface energy. The preparation of omniphobic membranes via electrospinning involves three main techniques. The first method involves electrospinning a polymer-only solution, followed by modifying the resulting fibers with low surface energy materials or nanoparticles. The second approach involves adding fluoride to the electrospinning solution, resulting in the formation of low surface energy nanofibers. Finally, the third method involves incorporating nanoparticles into the spinning solution to create roughened nanofibers, which are then treated with fluoride.

The first method is commonly used to achieve omniphobic membrane surfaces. Plasma treatment is an effective technique to enhance omniphobicity. For instance, Woo et al. formed new CF₂-CF₂ and CF₃ bonds on the surface of ENMs, reducing their surface energy and ensuring stable MD performance of the prepared omniphobic membranes P/CF-60 for real RO brine [38]. They modified ENMs using a plasma system equipped with parallel plate electrodes coupled with a radio frequency (RF) glow discharge system. ENMs were placed on the plates and CF₄ gas was introduced into the chamber at a certain flow rate. Plasma treatment can also generate in-plane nanopores in the nanosheets and provide active sites for constructing omniphobic membranes with stable performance [39]. Dip-coating is another method used to design ENMs with excellent wetting properties. Lee et al. added cationic surfactants to the PVDF-HFP spinning solution, making the substrate positively charged. They then grafted negatively charged SiNPs onto the surface by dip-coating and fluorinated it, obtaining omniphobic membranes with a low surface energy surface and a concave structure. The prepared membranes SiNPs-PVDF-HFP/BTEAC showed stable performance in the 8 h DCMD test [40].

The second way is to obtain omniphobic membranes by directly adding fluoride to the spinning solution. Omniphobic membranes with uniform fibrous structure can be obtained by adding fluorinated-decyl polyhedraloligomeric silsesquioxane (F-POSS) to the PVDF-HFP suspension solution, which exhibits good wetting resistance to liquids with different surface tensions and shows stable MD performance (Figure 5) [41]. Similarly, Lu et al. also used F-POSS as a fluorinating agent to prepare an omniphobic membrane with a contact angle of 120.3° to ethanol. Spinning solutions were prepared by dispersing F-POSS particles (3 wt.%) in a mixed solution containing PVDF-HFP, DMF, and acetone, followed by electrospinning [42].

Figure 5.

Schematic diagram of preparation method, wetting performance, and MD performance of the omniphobic membrane. Reproduced from [41] with permission. Copyright (2018) Elsevier Inc.

The last method involves incorporating nanoparticles into the spinning solution. Huang et al. used SiNPs to prepare a sheath solution for coaxial electrospinning to produce fibers with nanoscale roughness, and a PVA/silica solution as the core solution. After fluorination by FAS, a hydrophobic and oleophobic membrane was obtained, which exhibited resistance to surfactants in MD tests [43]. Generally, SiNPs, AgNPs, and TiO₂ are often used to increase the surface roughness of the membrane. In our previous study, a robust omniphobic membrane #PVDF-F is developed by fluorinating silver nanoparticles (AgNPs)-coated nanofibrous membrane, which has a concave structure with low surface energy [5].

4.3 Underwater superoleophobic surfaces

4.3.1 Theory of underwater superoleophobic membranes

Janus membranes with asymmetric wettability usually have underwater superoleophobic surfaces and are widely used for treating oily wastewater in MD. The term “Janus” is derived from ancient Greek and Roman mythology and refers to a god with two faces. As shown in Figure 6, Janus membranes for MD are usually composed of a hydrophilic surface layer and a substrate with different hydrophobicity. Thanks to their asymmetric properties, properly designed Janus membranes exhibit resistance to wetting and fouling when faced with different types of liquids.

Figure 6.

Schematic diagram of three common Janus membranes.

The underwater oleophobicity of the surface layer is an important factor that influences the performance of Janus membranes. Similarly, the three classical wetting models mentioned above are applicable to the wetting behavior of underwater oil droplets on solid surfaces. The underwater oil contact angle (UOCA) for the ideal case can be calculated using Eq. (4) [44]:

$$cos{\theta }_{{\mathit{eq}}^{\ast }}=\frac{{\gamma }_{\mathit{SO}}-{\gamma }_{\mathit{SW}}}{{\gamma }_{\mathit{OW}}}=\frac{{\gamma }_{\mathit{OA}}\mathit{cos}{\theta }_{\mathit{OA}}-{\gamma }_{\mathit{WA}}\mathit{cos}{\theta }_{\mathit{WA}}}{{\gamma }_{\mathit{OW}}}$$E4

where, ${\gamma }_{\mathit{SO}}$, ${\gamma }_{\mathit{SW}}$ represent the interfacial tension of solid-oil and solid-water, ${\gamma }_{\mathit{OW}}$ is the surface tension of oil–water, and ${\theta }_{{\mathit{eq}}^{\ast }}$ refers to the CA of oil droplets under water. Similarly, when medium to large-scale roughness exists on solid surfaces, oil droplets exist in the Wenzel state and the underwater equation is [44]:

$$cos{\theta }_{\mathit{UWO}}=\text{rcos}{\theta }_{{\mathit{eq}}^{\ast }}$$E5

where ${\theta }_{\mathit{UWO}}$ is the UOCA on a solid surface in the Wenzel state, and the roughness of the solid surface determines its magnitude. In contrast, in the underwater Cassie-Baxter state, the grooves of the hierarchical micro/nanostructure on the rough solid surface are filled with water, which endows it with underwater oleophobicity and effectively reduces the adhesion of oil droplets to the membrane surface during MD. The equation in this state is [44]:

$$cos{\theta }_{\mathit{UCO}}={\mathrm{f}}_{\mathrm{SO}}\left(1+cos{\theta }_{\mathit{OW}}\right)-1$$E6

where ${\mathrm{f}}_{\mathrm{SO}}$ and ${\theta }_{\mathit{UCO}}$ are the contact area fraction and UOCA of the multi-level roughness solid surface, respectively. Figure 3D shows the wetting behavior of oil droplets on different membrane surfaces under water.

4.3.2 Engineering underwater superoleophobic membranes by electrospinning

The underwater superoleophobic top layer of the Janus membrane can form a hydration layer at the solid–liquid interface, effectively inhibiting the adhesion of oil droplets on the membrane surface. Therefore, its preparation should consider the surface morphology and hydrophilic chemical functional groups of the top layer. Additionally, due to the requirement of hydrophobicity in the MD process, substrate modification should be fully considered. There are two main ways to prepare Janus membranes using electrospinning technology: the first method is to use the combination of electrospinning technology and post-modification to prepare a double-layer composite membrane; the second method is to uses electrospraying technology to introduce hydrophilic materials on the hydrophobic substrates.

At present, some researches have realized the successful preparation of Janus membrane through post-treatment modification based on electrospinning. For instance, electrospinning was used to prepare positively charged CTAB/PVDF-HFP substrates, followed by adsorption of negatively charged SiNPs on it by dip-coating method, and finally fluorination (Figure 7A). The prepared membranes had a secondary re-entrant structure (Figure 7B) and exhibit omniphobicity. Finally, SiNPs, chitosan (CTS), and perfluorooctanoate (PFO), SiNPs-CTS/PFO, was sprayed on the substrate to obtain Janus membranes Janus (O). Wettability tests revealed that the membrane Janus (O) has asymmetric wettability and its top layer is underwater oleophobic (Figure 7C) [45]. In addition, the simple stacking of two membranes with different wettability based on electrospinning can also obtain a dual layer membrane with underwater oleophobicity. Makanjuola et al. directly coated different concentrations of cellulose on PVDF-HFP membranes to obtain Janus membranes that can treat oily wastewater [46]. Vacuum filtration, UV-mediated modification strategy, multi-step coating and surface modification can usually achieve the preparation of the Janus membrane hydrophilic layer [47]. Our previous study prepared Janus membranes with hydrophobic bottom and hydrophilic top by spray-coating multi-walled carbon nanotubes (MWCNTs) on electrospun PVDF substrate, which showed improved anti-fouling properties [4].

Figure 7.

(A)Schematic diagram of fabrication procedure of the Janus membrane, (B) SEM image of the substrate and surface of Janus membrane, (C) photographic images and CA of different liquid droplets on the bottom surface and top surface of the Janus membrane. Reprinted (adapted) with permission from [45]. Copyright (2017) American Chemical Society.

Most studies use commercial membranes as substrates and use electrostatic spraying of hydrophilic materials to prepare Janus membranes. Hydrophilic materials such as CTS, polyacrylonitrile (PAN), polydopamine (PDA), polyethyleneimine (PEI) and carbon nanotubes (CNT) are usually used for the preparation of the underwater oleophobic layer of Janus membranes [48, 49, 50, 51]. For instance, Tang et al. prepared Janus membranes with asymmetric wetting characteristics by electrospraying PAN on the surface of hydrophobic PTFE membrane and hydrolyzing it with ethylenediamine (EDA) and sodium hydroxide (NaOH). The Janus membranes PTFE/PAN-OH exhibited a UOCA > 150° and displayed stable permeate flux and salt rejection in MD-treated oil-containing feed solution [48]. Overall, so far, the research on the preparation of Janus membranes by electrospinning/spraying has become increasingly mature, which is of great significance to the industrial application of MD technology.

5. Application of super liquid-repellent membrane in MD

In comparison to conventional hydrophobic membranes, superhydrophobic and omniphobic membranes offer superior performance in the MD process for the treatment of complex components and contaminants owing to their robust wetting stability. Superhydrophobic membranes are widely utilized for seawater desalination and brine treatment due to their excellent permeability. When dealing with high-concentration brine (25 wt%), the superhydrophobic membrane maintained a stable flux and salt rejection in the 180 h DCMD test [52]. Notably, superhydrophobic membranes also demonstrated remarkable long-term stability in treating industrial wastewater [53, 54, 55].

However, due to limited hydrophobicity, superhydrophobic membranes are not suitable for wastewater containing a variety of low surface tension substances. Omniphobic membranes, on the other hand, possess stronger hydrophobicity, which imparts exceptional repellency to various low surface tension liquids and enables them to resist wetting more effectively than superhydrophobic membranes. Therefore, many studies have highlighted the effectiveness of omniphobic membranes in treating low surface tension wastewater [5, 56, 57]. The modified omniphobic membrane exhibited excellent wetting resistance to 0.6 mM sodium dodecyl sulfate (SDS) [58]. In addition, the interactions between surfactants with different properties, such as anionic surfactant SDS, dodecyltrimethylammonium bromide (DTAB), and nonionic surfactant polyoxyethylenesorbitan monolaurate (Tween-20), and omniphobic membranes were elucidated [4]. Interestingly, MD membranes with omniphobicity can also treat wastewater containing emulsified oil. Specifically, the omniphobic membranes showed stable performance in treating Oil-emulsion containing 0.005% v/v [59].

Janus membranes, which feature an underwater oleophobic layer, have emerged as a promising candidate for the treatment of wastewater containing low surface tension chemicals and oily wastewater. Our group investigated the MD performance of Janus membranes with different substrates (hydrophobic, superhydrophobic, and omniphobic), and found that Janus membranes with omniphobic substrates had the best stability for the treatment of surfactants and oil-in-water emulsions [4]. The hydrophilic surface layer of the Janus membrane can form a hydration layer at the solid–liquid interface to prevent the adhesion of oil droplets to the membrane surface. In addition, the size exclusion of the hydrophilic layer proved to be effective for surfactants [60].

In conclusion, super liquid-repellent membranes have broad application prospects in the fields of desalination, wastewater treatment, and oil–water separation due to their excellent wettability, which is beneficial to promote the industrial application of MD technology.

6. Challenges facing super liquid-repellent membrane based on electrospinning in MD

Compared with traditional hydrophobic membranes, although super liquid-repellent membranes prepared by electrospinning have significant advantages, they also face many challenges that hinder their commercial application. Firstly, the fabrication process of the super liquid-repellent membrane remains intricate, and the membrane modification materials employed, such as fluorinated solvents, are both expensive and potentially toxic, which could pose hazards to drinking water production. Secondly, despite the laboratory research indicating the efficacy of super liquid-repellent membranes, there remains a paucity of practical application in large-scale production. Finally, the balance between modified membrane permeability and porosity must be attained, and the mechanical robustness of the Janus membrane requires continuous improvement. As such, it is imperative to conduct research on more eco-friendly, stable, and producible super liquid-repellent membranes and to investigate the fouling and wetting mechanisms of real wastewater in MD to expand their application range.

7. Conclusions and outlooks

ENMs present promising prospects for future MD applications as the next generation of MD membranes, owing to their exceptional porosity, hydrophobicity, and feasibility in constructing multilayered membranes. The utilization of electrospinning-based super liquid-repellent membranes in MD for anti-wetting and anti-fouling purposes has garnered significant interest. Over the past few years, the amalgamation of innovative nanomaterials and various modification strategies with electrospinning has yielded super liquid-repellent membranes with specific morphology and higher permeation flux, significantly enhancing the competitiveness of ENMs in treating low surface tension wastewater. Recent findings suggest that improved nanofibrous membranes are promising options for applications in seawater desalination, brine treatment, and oily wastewater treatment. However, further exploration and improvement of ENMs are necessary, with a focus on enhancing their mechanical properties and minimizing preparation costs.

In summary, the MD technology has emerged as a highly competitive solution for treating complex hypersaline wastewater, recovering valuable salts, and obtaining distilled freshwater from wastewater. Ongoing efforts by scientists and engineers to develop super liquid-repellent membranes based on electrospinning for MD offer great promise for achieving large-scale applications of this technology in the near future.

Acknowledgments

This work was supported by Nankai University and Cangzhou Bohai New Area Institute of Green Chemical Engineering Fund (20220142), National One Thousand Talents Foreign Experts Program of the Ministry of Science and Technology of China and Tianjin Government (040-BE044741, 040-C021801601).

Appendices and nomenclature

RO	reverse osmosis
MD	membrane distillation
FO	forward osmosis
PVDF	polyvinylidene fluoride
PTFE	polytetrafluoroethylene
DMF	N,N-dimethylformamide
SDS	sodium dodecyl sulfonate
DTAB	dodecyltrimethylammonium bromide
Tween-20	polyoxyethylenesorbitan monolaurate
NIPS	non-solvent induced phase separation
TIPS	thermally induced phase separation
ENMs	electrospun nanofibrous membranes
NMP	N-methylpyrrolidone
CAH	contact angle hysteresis
SA	sliding angle
PFOTES	perfluorooctyltriethoxysilane
CVD	chemical vapor deposition
FAS	fluoroalkylsilane
SiO2	silicon dioxide
SiNPs	silicon nanoparticle
UOCA	underwater oil contact angle
TiO2	titanium dioxide
AgNPs	silver nanoparticles
CTS	chitosan
PFO	perfluorooctanoate
WCA	water contact angle
PAN	polyacrylonitrile
PDA	polydopamine
PEI	polyethyleneimine
EDA	ethylenediamine
CNT	carbon nanotubes