Structural design of the electrospun nanofibrous membrane for membrane distillation application: a review.

Although membrane distillation (MD) is a promising technology for water desalination and industrial wastewater treatment, the MD process is not widely applied in the global water industry due to the lack of a suitable membrane for the MD process. The design and appropriate manufacture are the most important factors for MD membrane optimization. The well-designed porous structure, superhydrophobic surface, and pore-wetting prevention of the membrane are vital properties of the MD membrane. Nowadays, electrospinning that is capable of manufacturing membranes with superhydrophobic or omni phobic properties is considered a promising technology. Electrospun nanofibrous membranes (ENMs) possess the characteristics of cylindrical morphology, re-entrant structure, and easy-shaping for a specific purpose, benefiting the membrane design and modification. Based on that, this review investigates the current state and future progress of the superhydrophobic, multi-layer, and omniphobic ENMs manufactured with various structural designs for seawater desalination and wastewater purification. We expect that this paper will provide some recommendations and guidance for further fabrication research and the configuration design of ENMs in the MD process for seawater desalination and wastewater purification.

Introduction

It is anticipated that about 1.8 billion people will experience a lack of water in 2025 (Connor 2012), (WWAP (United Nations World Water Assessment Programme) 2015). Seawater occupied 70% of the water on the earth, while only 2.5% could be used by humans (Nthunya et al. 2019). Hence, freshwater production from seawater is crucial for the sustainable existence of the Earth. Many desalination technologies, such as multiple stages flash distillation (Yan et al. 2007), multiple effect distillation (Mistry et al. 2013), ion exchange (Dow et al. 2016), and membrane distillation (MD), have been developed to solve water scarcity.

Membrane technology has displayed great potential in the elimination of heavy metal ions, antibacterial applications, and selective ion removal for all types of separation, such as microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO) because of attractive advantages such as simpleness of operating conditions, no phase change, less energy consumption, compact design, and eco-friendlier. This technology is facing a significant challenge of membrane fouling. Many studies have been conducted to develop a nanocomposite membrane with improved ion removal performance and antibacterial activity as an eco-friendly membrane-based water treatment technology. By combining the inherent properties of nanomaterials (for example, graphene oxide, carboxylated-graphene oxide, silver-doped multi-walled carbon nanotube (MWCNT)) with a polymer matrix (for example, polyphenylsulfone), nanocomposite membranes improved membrane flux, antifouling, antibacterial, and solute removal performance (Mohammad Jawaid et al. 2021; Shukla et al. 2018; Shukla et al. 2017; Shukla et al. 2019).

Membrane distillation (MD) is an emerging technology that is environmentally friendly and easy to operate (Deshmukh et al. 2018). Researchers have affirmed that the MD process treats high concentrations of wastewater and brine and has the possibility of utterly using solar heat and low-grade heat energy such as industrial waste heat (Efome et al. 2016), (Joachim Koschikowski 2003), (El-Bourawi et al. 2006). Also, MD can be applied for small-scale desalination in remote areas (Lin et al. 2014). The membrane of the MD process has superhydrophobic and superoleophobic properties, so it can separate oil from wastewater solution containing oil. It has the merits of mild operating temperature, a lower operating pressure, a low cost, no requirement for extensive pre-treatment, and less energy consumption compared to traditional separation processes (Alklaibi and Lior 2004), (Su et al. 2012), (Khalifa and Alawad 2018), (Criscuoli et al. 2008), (Tun et al. 2016), (Zarebska et al. 2014), (Liu and Wang 2013), (Tomaszewska and Morawski 1995), (Feng et al. 2008). Hence, MD has been considered a next-generation promising technique for water desalination (Lalia et al. 2013), (Tijing et al. 2014a).

Figure 1 exhibits the research situation on the MD from 2011 to 2021. The rapid increasing of the number of the published paper demonstrated that the research on the MD has attracted great attention and been extensively performed in recent years.

Fig. 1

The number of the published paper on the MD in the Web of Science database

However, this technology has some main demerits, such as low flux and the membrane pores wetting and fouling or scaling. It is vital to elucidate the limiting effects of MD to improve the desalination efficiency of MD. There are many factors such as hydrophobic membrane, feed side temperature, cool side temperature, MD configuration, LEP (liquid entry pressure), and feed pre-treatment operations. The hydrophobic membrane offers a liquid–gas interface for mass and heat transfer while it does not participate in the separation, so it plays a vital role and key component in applying the MD process. Although the MD was invented in 1963 (Prince et al. 2014), the fact that MD has not yet been introduced into the water industry due to a lack of suitable membranes demonstrated that designing and manufacturing suitable membranes are of great importance.

An ideal membrane to ensure high performance in the MD process should possess good mechanical properties, low thermal conductivity, high thermal stability, hydrophobicity to prevent pore wetting, and a well-designed pore structure (high porosity, straight pore, narrow pore size distribution, appropriate pore size) to improve water flux (An et al. 2017; Bonyadi and Chung 2007; Ma et al. 2009; Shirazi et al. 2014; Woo et al. 2017). However, pore wetting is unavoidable in the MD process using these hydrophobic membranes, particularly in desalinating other wastewaters containing high levels of surfactants or low surface tension substances. The ideal membrane in the MD process should possess high LEP and low-fouling properties. Although the membrane material is hydrophobic, the membrane may be wetted because of the deposition of inorganic and organic foulants overtime on the membrane surface and internal pores (Zhou et al. 2021). It was reported that the most crucial reason for membrane wetting is because the higher transmembrane pressure than the LEP value of hydrophobic membrane in the feed channel is applied for MD because of the membrane’s organic fouling and inorganic scaling (Guo et al. 2019).

In order to develop new MD membranes that repulse wetting by both water and other low surface tension contaminants and alleviate the membrane fouling, many MD membrane designs have been proposed and developed by mimicking the hierarchical structure of ramee and louts leaves: composite hydrophilic − hydrophobic membranes, superhydrophobic and omniphobic membranes. The lotus leaf does not display omniphobic property because the surface does not exhibit a re-entrant morphology (Zheng et al. 2018). However, the long-term MD operation will cause the air and/or vapor captured under the water-membrane interface to be lost and finally result in surface wetting, even though it displays a superhydrophobic and omniphobic property. Thus, further comparisons for membrane design manufactured in various structures (lotus, pillar, acicular multilevel re-entrant, pine-needle-like hierarchical structure, and so on) must be performed for better membrane performance.

Most of the commercially applicable hydrophobic membranes usually were fabricated by thermally induced phase separation (TIPS), non-solvent induced phase separation (NIPS), phase inversion sintering, and stretching methods for microfiltration (MF) purposes, and they possess some disadvantages such as inner disconnected microporous structures and poor porosity (Deng et al. 2018; Zhao et al. 2018). Those MF membranes are usually not precisely appropriate for MD operation and tend to experience low permeability and wetting problems making them unsatisfactory in the MD process. Electrospinning (ES) technology has been recognized as a promising and indispensable technique for manufacturing highly porous and durable nanofiber membranes due to the broad selectivity of polymer materials. The advantages of the electrospinning method over conventional membrane manufacturing methods are the same as the following: firstly, it can realize the mass production of nanofibers; secondly, it has unique structural characteristics: narrow pore size distribution, large surface area per unit volume or mass, interconnectivity, low tortuosity, excellent structural stability, steerable scaffold thickness (Ahmed et al. 2015; Li et al. 2015; Tijing et al. 2014a; Wang et al. 2013), and high porosity of about 80% (conventional membranes possess a porosity of 5–35% (Tlili and Alkanhal 2019)); thirdly, significantly, the electrospinning allows the fibrous membrane to construct an omniphobic surface because the ENMs are composed of a high number of nanofibers possessing cylindrical morphology, which can feature a re-entrant geometry (Hou et al. 2019).

Many researchers reviewed the fabrication procedures and the modification strategies of the various hydrophobic and hydrophilic membranes by using electrospinning and other methods used for water treatment: MD, pressure-driven membrane process, osmotically driven membrane process, membrane bioreactor, water–oil separation, heavy metal ion adsorption, and antibacterial nanofibrous membrane (Suja et al. 2017; Saikat et al. 2016; Shengjie Peng et al. 2016; Tijing et al. 2014a; Xianfeng Wang 2016; Yuan Liao et al. 2018). Then, others have reviewed membranes materials (polymer and more functionalization of nanomaterials) and the membranes fabrication methods for specific polymers via ES (Muhammad et al. 2019), (Yao et al. 2020), (Pan et al. 2019), (Zhou et al. 2021), (Lingling Zhong et al. 2021), (Ravi et al. 2020). Regrettably, they did not systematically analyze the properties of membranes fabricated using the ES technique in terms of membrane design and structure in detail.

Despite the unparalleled advantages of ENMs, reducing the permeability of the ENMs from wetting cannot be avoided. Therefore, new materials and membrane structures are needed to address the above limiting factors. We categorized the membrane with special wettability into three subgroups according to membrane structure: superhydrophobic membranes, multi-layered membranes (composite hydrophilic–hydrophobic/hydrophobic–oleophobic underwater membranes, superhydrophobic dual-layer, triple-layer membrane), omniphobic membrane (i.e., with the surface that possesses hydrophobic and oleophobic). First, this paper discussed the recent progress and challenges of the electrospinning technology to fabricate nanofiber. Second, this paper discussed fabrication methods of representative membranes fabricated with novel structural design via ES technology or the combination of ES and several surface modification methods and characteristics of these membranes for application in the MD process: (1) functional nanomaterials and doping methods to introduce roughness structures on the membrane surface and representative grafting materials; (2) grafting methods used in the surface modification to lower the surface tension of the MD membrane for prolonged operation. Thirdly, we made a prospect in electrospinning technology and the MD process.

We hope that this paper will inspire peer researchers in this field for the unique design and fabrication of high-performance ENM-based MD membranes.

Electrospinning as promising nanomaterial manufacturing technology to fabricate the nanofiber membrane

Nanotechnology, which manufactures nanomaterials with unique physical, chemical, and biological properties compared to macroscale mass materials, is attracting great attention as a basic core technology that can solve challenges that could not be solved by conventional methods.

Numerous techniques have been used to fabricate membranes with high hydrophobicity (Ma and Hill 2006; Qian and Shen 2005): lithography, plasma treatment, sol–gel method, phase separation, and electrospinning. Among them, electrospinning technology is a versatile and indispensable technique with low cost, vast material selectivity, versatility, multifunction, and high speed. Compared to the membranes produced from these conventional fabrication methods, the membranes fabricated by electrospinning methods have higher porosity and excellent permeability owing to interconnected open-pore architecture and steerable membrane thickness. This superior ES technology has been developed via many kinds of research and efforts to attract significant attention as a powerful candidate for fabricating nanofibrous membranes (Fig. 2).

Fig. 2

Diagrammatic configuration of the single nozzle electrospinning setup (Shirazi et al. 2013)

The electrospinning process usually consists of a high voltage power supply system, a syringe pump, a syringe, a rotating drum as a collector, and a needle. It is based on the uniaxial stretching of an injection ejected from the surface of a charged long-chain polymer solution with sufficient molecular entanglements under the action of a strong electric field (usually 105–106Vm−1) between the emitter and a conductive collector. When the electrostatic power acting on the polymer solution’s surface is beyond the polymer solution’s surface tension, a charged liquid injection is attracted to the Taylor cone and spun toward the target collector. The solvent concurrently evaporates during the electrospinning process while the fluid injection travels from the spinneret to a collector (Manley and Larrondo 1981). The fiber morphology is dependent on parameters including the applied electric field voltage, the distance between the spinneret and the collector, atmosphere temperature, flow rate, humidity, properties of the feed materials such as the polymer, concentration, surface tension, conductivity, and a viscosity (Burger et al. 2006). The single nozzle electrospinning process is a simple technique. However, industrialization for this process is limited due to low productivity. The following picture shows the possible geometries of electrospinning spinneret (Persano et al. 2013) (Fig. 3).

Fig. 3

Schemes of the possible architecture of the electrospinning emitter (Persano et al. 2013)

Recently, multi-needle and needleless electrospinning have been developed for mass production (Nayak et al. 2012; Zhou et al. 2009). However, nanofiber’s morphology control in multiple-needle electrospinning requires numerous architecture optimization and modeling works due to cross-talk between neighboring needles and interactions. The mutual repulsion between charged polymeric injections and continuous production at this process may be limited due to polymer plugging at the spinneret jet. The multiple-needle electrospinning based on multi-emitter components can be employed to augment the overall throughput per setup and the thickness of resultant mats and be useful for depositing large area surfaces and producing composite fibers composed of different materials (Ding et al. 2004).

Figure 4 shows a geometric structure of the representative multiple-needle electrospinning setup (Al-Mezrakchi 2018).

Fig. 4

Geometric structure of the representative multiple-needle electrospinning spinneret: a) geometric design of the reservoirs with needles and b) Taylor cone angles, radius, and near field calculation according to Taylor cone height (Al-Mezrakchi 2018)

The needleless electrospinning process to eject multi-polymeric jets from polymer solution’s surface concurrently has attracted significant attention due to these disadvantages of the multi-needle electrospinning plants. Figure 4 shows the geometric architecture of the representative needleless electrospinning plants (Fig. 5).

Fig. 5

Schematic configuration representative needleless electrospinning unit: a) rotating-disk electrospinning unit (Ng and Supaphol 2018); b) the needleless linear electrospinning unit (Li et al. 2018)

The electrospinning methods can be divided into melt spinning, solution spinning, and emulsion spinning according to the material’s state to eject. Melt electrospinning does not require the utilization of solvents compared to traditional solution electrospinning. It can avoid the residual, recovery, and treatment of toxic solvents. It is considered the development direction of green manufacturing of polymer nanofibers. However, due to the complicated equipment and process delaying, melt electrospinning has been facing difficulties in realizing fiber refinement and low preparation efficiency. In the past period, solution electrospinning technology has successfully manufactured nanofibers of hundreds of materials and has initially realized a commercial application (Persano et al. 2013). However, the solvents used for solution electrospinning, such as chloroform and methylene chloride, are mostly toxic. Therefore, in the mass production process, the preparation, removal, and recovery of poisonous solvents increase equipment costs. Thermoplastic materials do not have suitable solvents at room temperature, making it challenging to prepare nanofibers by solution electrospinning. A non-melting and insoluble compound is usually spun by using emulsion spinning. This method is employed to produce fibers consisting of fluorocarbons with a high melting point, inorganic materials like ceramics, and blends possessing flame-retardant properties (Persano et al. 2013).

Most of the nanofibers are fabricated from the electrospinning (ES) technique. However, this method also has got a few drawbacks, such as dependency on electrical characteristics of dope solutions, many other safety issues, including electrical shock risks and high electric field requirements, and low productivity per spinneret. Among many methods, centrifugal spinning (CS) uses centrifugal force instead of electrostatic power. It is considered a nanofiber fabrication strategy, ES’s promising candidate, and it has received great attracts because of its excellent products, which are almost 50 times higher than the conventional ES technique (Ren et al. 2014). CS process can be employed to fabricate nanofibers by using polymer solutions or polymer melts without dependency on the dielectric constant and the requirement of a high voltage electric field (Saleem et al. 2020). The following picture shows a schematic representation of the experimental plant of a centrifugal spinning method (Gundogdu et al. 2018) (Fig. 6).

Fig. 6

Schematic portrayal of the experimental centrifugal spinning plant (Gundogdu et al. 2018)

In addition to the CS method, new methods such as plasma-induced synthesis (Hu et al. 2014) and solution blow spinning (SBS) (Polat et al. 2016) to overcome several restrictions, including the requirement of higher voltage, solvent recovery problems, and low productivity of the ES processes, have been developed for efficient nanofiber fabrication in the recent years.

The ES technology has demonstrated excellent potential in many applications requiring high porosity, high specific surface area, and easy functionalization, especially in the MD process.

Configuration of MD setup

Without consuming much energy to evaporate water, the MD process can treat brine at comparatively low temperatures and pressure using solar and low-grade heat energy such as industrial waste heat. The MD process’s driving force is not an absolute pressure difference, a concentration gradient, or an electrical potential gradient. A vapor pressure difference across the hydrophobic porous membrane as a barrier between the feed and permeate sides plays a vital role (Lawson and Lloyd 1997; Schneider and van Gassel 1984). MD configurations usually are composed of DCMD (direct contact membrane distillation), AGMD (air gap membrane distillation), VMD (vacuum membrane distillation), and SGMD (sweeping gas membrane distillation) (Dong et al. 2015a; Efome et al. 2016; Prince et al. 2012) (Fig. 7).

Fig. 7

Different membrane distillation process configurations (Alkhudhiri et al. 2013). a) DCMD; b) AGMD; c) SGMD; d) VMD

DCMD

The membrane in the DCMD process is positioned between the feed and permeate sides, and the membrane gets in contact with both the hot feed solution and cold permeate stream. The driving force, the vapor pressure of the DCMD process, is generated from the heat difference between both sides of the membrane. The water vapor in the DCMD process passes through the hydrophobic membrane pores by diffusion or convection. DCMD process of high water flux is usually considered the most straightforward process and is used mostly. A heat loss is higher than the other configurations because hot streams are in direct contact with cold streams (Jafari et al. 2018).

AGMD

A gap is placed between the permeate side and membrane for the AGMD process. The membrane gets in direct contact with the hot feed solution, and the water vapor passes through the hydrophobic membrane and is condensed on the cold condenser surface via an air gap. The filtrated liquid is achieved between hot feed and permeate solution. The critical advantage of AGMD over DCMD is the lessening of the heat loss from the membrane of the condensing surface because of the introduction of the stagnant air layer between the membrane and condensate side, which makes AGMD possess a higher potential for commercialization (Attia et al. 2017a; Jafari et al. 2018). Also, it is more available than VMD owing to an internal condenser (air gap) (Alklaibi and Lior 2007). However, AGMD experiences low permeate flux than DCMD owing to the augmentation of mass transfer resistance by the separating air gap.

VMD

The VMD process does not have a cooling system. The use of a vacuum pump endows a lower vacuum pressure on the permeate side. The feed side has higher saturation pressure, resulting in the separation of the volatile molecules from the hot feed solution (Yalcinkaya 2019). The condensation is performed outside of the MD module. The condensation process enables the heat conduction loss to reduce on the feed solution side (Dao et al. 2013; Sarbatly and Chiam 2013; Wang and Chung 2013). A higher permeate flux can be achieved due to a higher pressure difference between the two sides of the membrane than other configurations (Wang and Chung 2013). However, in the VMD process, the relatively high trans-membranous hydrostatic pressure is inadequate for some membrane systems possessing low feed LEP (Zhou et al. 2014).

SGMD

SGMD is relatively less investigated in the MD process because the system is intricate. A cold inert gas sweeps in the hydrophobic membrane’s permeate side instead of a cold feed solution in the SGMD process. The volatile liquid molecules pass through to hydrophobic membrane pores toward the cold inert gas side, and then condensation is performed outside of the MD module (Yalcinkaya 2019). SGMD and VMD require additional condensers to convert the permeated vapor to liquid, increasing the cost and complicating these processes. However, SGMD was proved to possess the merits of a comparatively low conductive heat loss and a decreased mass transfer resistance (Zhang et al. 2017).

Despite the MD process’s superb characteristics, the MD process still has not been commercialized because of wetting, low permeability, and poor mechanical properties of the membrane.

Membrane structural designs to be adopted in the MD process

Current commercially available membranes inevitably experience membrane wetting over operation duration in the MD process. Therefore, it is imperative to develop a membrane designed so as to satisfy the MD process: superhydrophobicity or superomniphobicity, high porosity, good thermal stability, excellent chemical resistance, and good mechanical strength.

The main factors influencing the hydrophobicity of membrane, the most important factor in the MD process, can be classified as polymer, fillers, and the method used for membrane fabrication. These porous hydrophobic membranes have been usually produced from the hydrophilic materials (poly(vinyl alcohol) (PVA) (Dong et al. 2015b), polyacrylonitrile (PAN) (Cai et al. 2018), and polysulfone (PSF) (Khayet et al. 2019)) and the hydrophobic materials (polypropylene (PP), poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP), polyethylene (PE), polytetrafluoroethylene (PTFE), and polyvinylidene fluoride (PVDF) (Ahmed et al. 2015; El-Bourawi et al. 2006; Elmarghany et al. 2020)). The additives can be classified as one-dimensional (for example, nanotubes and fibers), two-dimensional (clay), or three-dimensional (spherical particles) (Bhattacharya 2016) and could be organic (for example, carbon black, cellulosic nanofiller) and inorganic (SiO2, Al2O3, TiO2, ZnO) in nature.

According to the membrane’s design method, the available ENMs for MD application could be classified as a superhydrophobic single layer, multi-layer (composite hydrophilic–hydrophobic/hydrophobic–oleophobic underwater dual-layer, superhydrophobic dual-layer, triple-layer nanofiber membrane), and omniphobic nanofiber membrane.

Superhydrophobic membranes

The first experiment in applying electrospun nanofiber membrane in MD processes (i.e., AGMD) was reported by Matsuura (Feng et al. 2008). The hydrophobicity is usually estimated according to the water contact angle (WCA). Sixty-five degrees has been defined as the boundary angle between hydrophilicity (WCA < 65°) and hydrophobicity (WCA > 65°) based on the difference in the architecture of interfacial water (Tian and Jiang 2013; Vogler 1998). Figure 8 shows schematic portrayals of different superhydrophilic, hydrophilic, hydrophobic, and superhydrophobic surfaces according to the WCA (left to right, respectively) (Tlili and Alkanhal 2019).

Fig. 8

Graphic portrayals of different superhydrophobic, hydrophobic, hydrophilic, and superhydrophilic surfaces according to water contact angle (Tlili and Alkanhal 2019)

From the above picture, a superhydrophobic surface displays a contact angle greater than 150°, and a superhydrophilic surface exhibits a contact angle less than 5°. The highly water-repellent superhydrophobic surfaces possessing water contact angles bigger than 150° have been investigated and designed by simulating the self-cleaning property of lotus leaves (Zhang et al. 2008). The hydrophobicity of any solid surface depends on its surface energy and geometrical architecture, such as porosity and roughness. Therefore, superhydrophobic surfaces could be achieved by introducing a rough surface on hydrophobic materials and modifying a rough surface by utilizing a low surface energy material.

Producing a rough surface on hydrophobic materials

Although ENMs have outstanding properties such as excellent hydrophobicity, good porosity, and interconnected structure, the smooth ENMs manufactured only by hydrophobic polymers generally exhibit strong water adhesion despite their high contact angles (Liao et al. 2013). The superhydrophobic membranes have been developed to alleviate membrane wetting in the MD process, and the superhydrophobicity can impart less pore-wetting problems, improved LEP and enhanced water vapor flux, and excellent salt rejection than hydrophobic membranes (Lee et al. 2016a; Tijing et al. 2014a). Based on the Cassie–Baxter model, constructing a rough membrane surface is vital in achieving a superhydrophobic membrane (Cassie and Baxter 1944). Recently, various methods have mainly been adopted to introduce a rough surface on the membrane.

The first method is incorporating nanoparticles into the previously prepared solutions to prepare membranes directly (Dong et al. 2015a; Moradi et al. 2018). It is well known that inorganic nanoparticles, including Al2O3, SiO2, and TiO2, enhance the electrical conductivity of polyelectrolyte, especially the photocatalytic activity of TiO2 has the potential to improve the membrane fouling properties (Chung et al. 2001; Madaeni and Ghaemi 2007). Numerous superhydrophobic membranes by using unique properties of nanoparticles such as PVDF-functionalized SiO2 (Nthunya et al. 2019), PVDF-clay (Prince et al. 2012), PVDF-HFP-functionalized TiO2 (Lee et al. 2016a), PVDF-graphene quantum dots (GQDs) (Jafari et al. 2018), PVDF-functionalized Al2O3 (Attia et al. 2017a), and PVDF-HFP-graphene (Woo et al. 2016) were prepared via one-step electrospinning to display excellent performances in the MD process.

The second method is to fasten inorganic nanoparticles such as TiO2, Al2O3, SiO2, and Ag, on the membrane surface, followed by post-treatment (Boo et al. 2016; Lee et al. 2016a; Shan et al. 2018). A superhydrophobic membrane with a hierarchical structure (Attia et al. 2018b) was fabricated by spraying a mixture of non-fluorinated alumina (Al2O3) NPs on an electrospun base membrane produced from PVDF via electrospinning. Alumina (Al2O3) NPs were functionalized with isostearyl acids and PVDF. The choreographed membranes properties: 87.5% of porosity, 0.39 μm of mean pore size, 25psi of LEP, and WCA of 154°. Then, many superhydrophobic membranes such as PTFE-PVA (Zhou et al. 2014), PVDF-PTFE (Li et al. 2019a), PVDF-TBAC (tetrabutylammonium chloride)-FA (fluorinated acrylate copolymer) (Li et al. 2019c), and PTFE-PVA (Zhou et al. 2020) were fabricated by combining electrospinning and post-treatment such as sintering, heat-press, ultrasonic treatment, and vacuum sintering to demonstrate applicable properties for MD process.

The third method is to manufacture a superhydrophobic membrane possessing a hierarchical structure via colloid or emulsion electrospinning followed by post-treatment or without it, such as sintering like PVDF- hydrophobic silica (Li et al. 2015) and PTFE-PEO (Su et al. 2019).

Many ENMs with superhydrophobicity, like PDMS (polydimethylsiloxane)-PMMA (polymethyl methacrylate) (Ren et al. 2017b), a copolymer of tetrafluoroethylene, hexafluoropropylene, and vinylidene fluoride (THV) (Zhang et al. 2017) and PVDF-HFP-PET (polyethylene terephthalate) (An et al. 2020), were fabricated from various kinds of polymers on the base of the unique hydrophobic properties of these polymers (Fig. 9).

Fig. 9

Diagrammatic electrospinning process of iHFC (interweaving hierarchical fibrous composite) membrane preparation from PET and PVDF-HFP solutions (An et al. 2020)

The effects of additives affecting the superhydrophobic ENMs performance, such as PVDF-LiCl (Liao et al. 2013) and PVDF-PTFE- IL (Li et al. 2020b) (ionic liquid (BMIMPF6: 1-butyl-3-methylimidazolium hexafluorophosphate)) were investigated in the MD process.

The effects of pre-treatment on the ENMs performance possessing superhydrophobicity during the MD process were also studied. A superhydrophobic TiO2 electrospun membrane (E- TiO2) (Guo et al. 2019) was developed via electrospinning by using PVDF-HFP, followed by coating with TiO2 nanoparticles using electrospraying. TiO2 nanoparticles were functionalized with a 1H, 1H, 2H, 2H-perfluorooctyltriethoxysilane (FTES). The membrane antifouling control was achieved by combining a ferrate pre-treatment and subsequently dissolved air floatation (DAF) in the DCMD process. The membrane regeneration was performed with easy physical water flushing.

Also, many studies on the mathematical modeling of the MD process were performed. A superhydrophobic ENM (Attia et al. 2017b) was fabricated from PVDF polymer and eco-friendly superhydrophobic alumina functionalized with isostearyl acids via electrospinning to remove heavy metal lead from the feed side of AGMD. Then, they used various models to calculate membrane flux on the base of mass and heat transfer balance in the case of AGMD. The experimental results demonstrated that membrane flow rate increased exponentially according to augmentation of feed solution temperature and slightly with an enlargement of feed flux. However, the flow rate decreased according to increasing feed solution concentration and coolant temperature. In Fig. 10, T, Tmf, Tmp, Tcd, Tca, Tcp, and T are the feed side inlet temperature, the feed side membrane temperature, the air gap side membrane’s temperature, the thin film condensate temperature, the cooling plate (permeate side) temperature, the cooling plate (coolant side) temperature, and the coolant temperature respectively.

Fig. 10

The vapor pressure and temperature diagram in the AGMD process (Attia et al. 2017b)

The purpose of forming a rough surface on the hydrophobic material by incorporating nanomaterials is to produce a Cassie–Baxter state, which creates an air gap between the water droplet and the superhydrophobic surface, thereby increasing the allowable pore diameter of the membrane and improving the permeate flow (Dong et al. 2015a). The purpose of introducing a rough surface on the hydrophobic material by incorporating post-treated nanomaterials is to enhance the stability of the resultant membrane by strengthening the adhesion between the nanomaterial and the substrate and its overall performance of it (Ren et al. 2017a) (Fig. 11).

Fig. 11

Schematic of a) Wenzel's state, b) Cassie–Baxter state (Zhou et al. 2021)

However, the existence of nanoparticles inlaid into the membrane substrate useless for roughness structure adversely may influence the prepared membrane’s physical and chemical properties. The low adhesion force between these inorganic constituents and the membrane matrix makes the nanoparticles easily separate from the membrane surface (Zhong et al. 2019).

Then, post-processing results in undesirable effects like pore-clogging and the possibility of destroying the nanofiber architecture. It is difficult to steer the thickness and amount of functional groups on the nanofiber surface in this process (Khayet et al. 2019).

Modification of a rough surface by utilizing a material with low surface energy

A pore wetting of the membranes can be alleviated by enhancing surface hydrophobicity or reducing pore diameter (Gryta and Barancewicz 2010): broader pore diameter results in higher vapor flux. However, the allowable pore sizes of existing membranes are limited to the range of 50 to 200 nm, so the membrane hydrophobicity by reducing pore diameter is restricted (Ma et al. 2009). As compared to reducing the pore diameter of the membrane, it can be considered a rational way to enhance surface hydrophobicity to abate pore wetting because decreased pore size would lower permeability.

The as-prepared nanofiber layers via electrospinning possess excellent porosity and uniform pore size distribution. The interconnected porous architecture of electrospun membranes as the path to transfer water vapors enhances the membrane permeability. It is susceptible to functionalizing the membrane due to the high surface area. ENMs have excellent advantages such as multiple-scale roughnesses (Zhou et al. 2014), high hydrophobicity, interconnected structure, and high porosity (An et al. 2017; Dong et al. 2015b; Feng et al. 2008; Guo et al. 2015a; Tijing et al. 2014b) compared to traditional membranes (Tadanaga et al. 2003) to generate superhydrophobic surfaces for MD process. Many methods have been developed for surface modification, such as CF4 plasma (Tian et al. 2015), fluorosilanization (Fang et al. 2012), and hydrophobic nanoparticles (Wang et al. 2011).

Many studies for preparing superhydrophobic ENMs were performed via fabrication of functional polymers with low surface energy functional groups using various methods such as free radical polymerization and atom transfer radical polymerization followed by one-step electrospinning of them (Guo et al. 2015b).

The surface modifications were achieved via electrospinning of hydrophilic (PAN, PSF) or hydrophobic (PVDF) polymers followed by simple dip-coating by using materials with low surface energy such as 1H,1H,2H,2H-perfluorodecyl acrylate (PFDA) monomer, and PDMS and post-treatment such as cold-press post-treatment or without it (Cai et al. 2018; Jiao et al. 2020; Li et al. 2017).

Then, the surface modifications were performed by hydrophobization with materials such as (heptadecafluoro-1,1,2,2-tetradecyl) trimethoxysilane (FAS-17); PDMS; and 1H, 1H, 2H, 2H-perfluorododecyl-trichlorosilane (FTCS) fluorination (Dong et al. 2015b; Ren et al. 2017a; Zhong et al. 2019).

Novel superhydrophobic ENMs (Dong et al. 2015b) were fabricated via the electrospinning of hydrophilic PVA followed by chemical cross-linking using glutaraldehyde. Finally, surface modification through fluoroalkylsilane ((heptadecafluoro-1,1,2,2-tetradecyl) trimethoxysilane (FAS-17)) with low surface energy. The resultant membrane displayed excellent properties like the following: WCA of 158°, SA of 4°, stable permeate flux of 25.2 kg/m2·h, and excellent permeate conductivity (< 5 μs/cm).

A novel superhydrophobic surface-modified nanofiber (Khayet et al. 2019) was prepared from blends of a hydrophilic PSF as host polymer, and small quantities of a fluorinated polyurethane (FPA) as an additive via one-step electrospinning: high permeate fluxes of 53.8 kg/m2·h and stable low permeate conductivities (< 5.1μS/cm).

A novel superhydrophobic PVDF membrane with a microsphere/fiber porous structure (Su et al. 2020) prepared from premixed solutions composed of the polymer (PVDF), a suitable solvent (NMP), and a nonsolvent (H2O) by combining NIPS and electrospinning followed by simple submerging in fresh FDTS (1H,1H,2H,2H-perfluorodecyltrichlorosilane) solution for the MD application. The experimental results revealed that adding a small quantity of water into the NMP + PVDF casting solution made phase separation happen earlier the formation of more nanoscale pores and microscale spheres, and the surface would become rougher than the absence of water to make the surface possess more hydrophobic. Then, the resultant membrane exhibited a high contact angle of 152.4° and an applicable permeability.

Recently, the primary methods for surface modification of electrospun membranes are to adopt fluorinated additives that can move to the nanofiber’s surface to be separated spontaneously from the surface, and it makes the surface possess lower surface energy. The main disadvantage of surface modification using materials with low surface energy like fluoride, silane, and silicon compounds is to use silane and fluorinated materials with potential environmental hazards.

Composite hydrophilic–hydrophobic/hydrophobic–oleophobic underwater and multi-layer nanofiber membranes have been developed to enhance the MD membrane design for better MD performance (for example, high permeability, high mass transfer, low thermal conductivity, mechanical durability) (Table 1).

Table 1

The characters of representative superhydrophobic single-layer nanofiber membrane and the performance in the MD process

Polymer	Solvent	Additives	Feed side temperature (°C)	LEP (kPa	Operation time (h)	Mean pore size (μm)	WCA (°)	Feed solution	MD type/ reference
PVDF	Acetone; N, N-dimethyl acetamide (DMAc)	Clay	80	200 ± 6	8	0.64 ± 0.22	154.20 ± 3.04°	3.5wt% NaCl	DCMD (Prince et al. 2012)
PVDF	DMAc; N, N-dimetylformamide (DMF); acetone	Lithium chloride (LiCl)	50	350	15	0.3	> 135°	3.5wt% NaCl	DCMD (Liao et al. 2013)
PTFE emulsion (FR301B); PVA			80	165	10		> 150°	3.5wt% NaCl	VMD(Zhou et al. 2014)
PVDF	DMF	SiO2	60	240 ± 10	24	0.1479–1.9339	155.6	3.5wt% NaCl	DCMD (Li et al. 2015)
PVDF-HFP	DMF; acetone	TiO2, LiCl	60–61	194.5	48	0.3–0.6	149	7.0 wt% NaCl	DCMD (Lee et al. 2016a)
PVDF	DMF; acetone	Al2O3, hexadecyl trimethyl ammonium bromide (HTAB)		186.2	5	0.47–0.715	150	Pb concentrations (500, 1000, 1500, 2000 mg/l)	AGMD (Attia et al. 2017a)
PVDF	DMF	Al2O3	70	186.2	5	0.37–0.467	150	Pb concentrations (1000 mg/l)	AGMD (Attia et al. 2017b)
PDMS; PMMA	DMF; tetrahydrofuran (THF);		70	13	24	9.15–13.43	163	3.5wt% NaCl	DCMD (Ren et al. 2017b)
PVDF	DMAc; acetone	SiO2 NPs	60	72.3 ± 2.3–84.2 ± 2.8	150	1.24–1.41	150	3wt% NaCl	DCMD (Nthunya et al. 2019)
PVDF; fluorinated acrylate copolymer (FA)	DMF; acetone	Tetrabutylammonium chloride (TBAC)	60	200	9	0.21–0.26	137 ± 3	3.5wt% NaCl	DCMD (Li et al.,, 2019c)
PVDF-HFP	DMF; acetone	TiO2; LiCl	65		24	0.52	157 ± 1.6	A synthetic AOM feed solution (algal organic matter: AOM)	DCMD (Guo et al. 2019)
PTFE emulsion; PVA			60	205 ± 15	10	0.55 ± 0.05	161.5 ± 2.4	3.5 wt% NaCl; 0.1 mM sodium dodecyl sulfate (SDS)	VMD (Zhou et al. 2020)
PVDF-HFP; PET	DMAc; acetone; trifluoracetic acid (TFA)		65	38.3	60	1–1.5	146	3.5wt% NaCl	DCMD (An et al. 2020)
PVDF	DMF	Graphene quantum dots (GQDs)	60	> 135	60	4900	> 125	3.5wt% NaCl	AGMD (Jafari et al. 2018)
PVDF-HFP	DMF; acetone	Graphene	60	> 186	60	0.86 ± 0.02	> 162	3.5wt% NaCl	AGMD (Woo et al. 2016)
PVDF	DMF; acetone	Al2O3; HTAB	60	106.9	30	0.507	> 154	3.5wt% of mixed heavy metals	AGMD (Attia et al. 2018b)
PTFE; PEO	Distilled water		65	245 ± 7	16.6	0.41	> 164	3.5–15wt% NaCl	DCMD (Su et al. 2019)
THV	DMF		80	118.0	6	0.79	130.7	3.5wt% NaCl	SGMD (Zhang et al. 2017)
PVDF; PTFE	DMAc; acetone	Ionic liquid (BMIMPF6)	60	83	26	0.93	146.4 ± 1.8	3.5wt% NaCl	DCMD (Li et al. 2020b)

Multi-layer nanofiber membranes

Composite hydrophilic–hydrophobic/oleophobic underwater-hydrophobic membrane

High mass fluxes and low heat transfer in DCMD are considered preferable properties to improve the vapor permeation flux and sustain the driving force occurring from the temperature difference (Yang et al. 2017). Therefore, thin membranes with suitable porous are preferred to achieve high mass transport coefficients. Then, thick membranes usually warrant excellent thermal efficiency, mechanical strength, and physical robustness. To resolve the issues mentioned above, reducing the vapor transport distance through a possibly thin hydrophobic layer is preferable (Liu et al. 2018), while sustaining the whole membrane thickness through a thick hydrophilic layer to decrease the temperature polarization effect (Qtaishat et al. 2009).

The ideal design of the composite hydrophilic − hydrophobic is to ensure the hydrophobic layer keeps dry to enable desalination while the hydrophilic layer is completely wetted to strengthen the flux (Jiaming Zhu 2015). A dual-layer membrane with hydrophobicity and hydrophilicity can be considered a Janus membrane. This dual-layer membrane’s main feature is to possess the opposite properties of both hydrophilicity and hydrophobicity, or negative and positive charges (Yang et al. 2016). In the 1980s, Cheng and Wiersma developed the dual-layer membrane patent composed of a top layer with hydrophobicity and a sublayer with hydrophilicity (Cheng and Wiersma 1982; Cheng and Wiersma 1983). They first discovered the concept of Janus membrane to MD in 1982 (Cheng and Wiersma 1982). A composite hydrophilic–hydrophobic membrane with asymmetric wettability has received significant attention in the application of MD.

At least one layer among the MD membrane should possess hydrophobicity or possibly superhydrophobicity to avoid pore wetting of the membrane (Liao et al. 2014b). While the hydrophilic layer in the composite hydrophilic–hydrophobic membrane is wetted with liquid (usually water), vapor transmission occurs in the hydrophobic layer. The dual-layer membrane with asymmetric wettability in the MD process results in higher mass flux because of the vapor’s shorter moving path between the liquid/vapor interfaces. The dual-layer membrane with asymmetric wettability was considered a more suitable way to solve the membrane thickness issue and displayed positive enhancement because the thicker hydrophilic layer as a support layer can mechanically hold the weak, thin hydrophobic layer. In contrast, the weak, thin hydrophobic layer effectively decreases the mass transfer resistance and results in a higher vapor flux (Bonyadi and Chung 2007). In this membrane, vapor condensation occurs at the interface of hydrophobic and hydrophilic layers. The mass transfer resistance between the evaporation and the condensation side may be significantly decreased due to the hydrophilic layer’s existence compared to the neat hydrophobic membrane with the same thickness (Zhao et al. 2020). Furthermore, a hydration layer is created on the hydrophilic skin layer of the hydrophobic/hydrophilic membranes due to the robust interaction between the water and the hydrophilic layer through hydrogen bonding, resulting in powerful underwater fouling resistance (Chen et al. 2010).

Many dual-layer composite nanofiber membranes were fabricated for the MD process via electrospinning of hydrophobic polymer (for example, PVDF-HFP, PTFE) or hydrophobic polymer-hydrophobic additives (for example, activated carbon (AC)) blend solution as the top active layer and hydrophilic polymer (for example, PVA, nylon-6 (N6), PAN, PSF) or hydrophilic polymer-hydrophilic additives (for example, cellulose acetate (CA), silica NPs) dope solution as bottom supporting layer followed by post-treatment or without it (Hou et al. 2018; Khayet et al. 2018; Woo et al. 2017; Zhao et al. 2018).

Figure 12 schematically demonstrates how AC nanoparticles affect the water contact angles of the membranes. AC enhanced hydrophobic/hydrophilic dual-layer nanofiber composite membranes were fabricated for DCMD (Zhao et al. 2018). Firstly, the commercial hydrophilic nylon 6,6 membrane was fixed onto the surface of the rotating drum collector. Then, a doping solution composed of the hydrophobic PVDF-HFP and different concentrations (0–3.0 wt%) of AC nanoparticles was electrospun on top of the nylon 6,6 substrate. All of the AC-incorporated membranes in the experiments exhibited higher water contact angles as the micro-wrinkles and AC protrusions enhanced the surface hydrophobicity than the membrane without AC addition. It is expected that the PVDF-HFP/AC nanofibers with the hierarchical roughness structure composed of micro-wrinkles and AC prominences would decrease the actual contact area between the water droplet and the hydrophobic surface to result in enhanced hydrophobicity of the membranes.

Fig. 12

Schematic diagram of the impact of AC nanoparticles on the water contact angles (Zhao et al. 2018)

A novel nanofibers-covered hollow fiber membrane (N-HFM) (Su et al. 2018) was fabricated from a commercial support woven tube and PVDF-HFP via non-rotational electrospinning for the DCMD process to combine the unique advantages of electrospun nanofibers and hollow fiber membrane (Fig. 13). The nanofibers film covered on the supporting woven tube via electrospinning acts as the skin layer of the resultant N-HFM. The fabricated hollow fiber membrane’s mechanical strength was then enhanced by a solvent vapor welding that fuses the nanofibers at their cross-points to maintain high porosity. The experimental results demonstrated that after a 40-min fusing, the strain at break, tensile strength, and Young’s modulus of the welded N-HFM increased by 79%, 90%, and 117%, respectively, compared to the pristine N-HFM.

Fig. 13

Diagrammatic portrayal of the process for electrospinning and post-treatment of N-HFMs (Su et al. 2018)

A novel mathematical model (Zhao et al. 2020) on the heat and mass transfer mechanism was developed for the membrane distillation with hydrophobic/hydrophilic dual-layer composite membranes in DCMD mode. The primary purpose of this work is to explore the effect of the hydrophilic layer on the flux rate of the hydrophobic/hydrophilic dual-layer composite membranes. The parameters are the same as following: total thickness (δ), hydrophobic layer (δ) and hydrophilic layer (δ) thickness, hydrophobic layer mean pore diameter (d), hydrophobic layer (ε) and hydrophilic layer (ε) porosity, and hydrophobic surface water contact angle (θ). They fabricated a new single and dual-layer hydrophobic/hydrophilic membrane with total thicknesses (240 μm, 260 μm, and 285 μm) and different thickness ratios (δ/δ = 3.69, 2.48, and 2.19) through electrospinning of hydrophobic PVDF-HFP solution on the commercial hydrophilic nylon 6,6 membrane substrate. Then, by using the proposed simultaneous heat and mass transfer theoretical model, they derived a standard parameter (β) to determine the augmented degree of the flux for the resultant hydrophobic/hydrophilic composite membranes. Furthermore, they theoretically derived a flux ratio model (DL=SL) and a critical value of the thickness ratio (χ = 1 + 1/β) to systematically investigate the flux performance of a hydrophobic/hydrophilic dual-layer composite membrane to the single-layer hydrophobic membrane with equal total thickness and validated through experiments.

Then, the pores of the hydrophobic membranes adopted in the MD process are susceptible to being wetted or fouled by oil because of long-range hydrophobic–hydrophobic interactions that result in fouling or a loss of selectivity of the membranes (Olawale Makanjuola 2019). This prevents the application of the MD in the treatment of oil-containing wastewater. It is very crucial to develop a membrane with repellency for both water and oil to utilize MD in the treatment of oil-containing wastewater. The composite hydrophilic–hydrophobic dual layer and omniphobic membranes have been recognized as promising methods. In the composite hydrophilic–hydrophobic membrane, the hydrophilic surface, which becomes oleophobic underwater, interacts with the oily feed solution and inhibits the passage of oil to the hydrophobic sublayer.

The dual-layered membrane has been fabricated via electrospinning for oily wastewater treatment, which is consisted of a hydrophobic electrospun PVDF-HFP membrane as a bottom layer and a hydrophilic electrospun PVDF-HFP membrane modified by regenerated cellulose as a top layer. The experimental results exhibited that the resultant membrane produced a freshwater flux of 12.8 kg m−2 h−1 for saline feed solution containing 1000 ppm of oil while completely rejecting salt in the DCMD process (Olawale Makanjuola 2019).

The purpose of adding hydrophilic additives into a hydrophilic polymer of composite membranes is to improve the surface roughness for hydrophilicity and to ensure that the hydrophilic layer adheres well to the hydrophobic layer.

One challenge that the composite hydrophilic–hydrophobic membranes face is the possibility that wetting faults between hydrophilic skin and the hydrophobic substrate can be introduced into the inner pores due to the coating of a hydrophilic layer onto a hydrophobic matrix. Then, a conductive heat loss is increased because of the hydrophobic layer’s introduction with a thinner thickness, and the hydrophilic layer produces a further heat transfer resistance by increasing the undesired temperature polarization, resulting in the reduction of the flux.

Superhydrophobic dual-layer nanofiber membranes have been developed to improve wetting repellent characteristics, surface’s interface stability, and the mechanical rigidity of ENMs (Table 2).

Table 2

The characters of representative composite hydrophilic–hydrophobic dual-layer membrane and the performance in MD process

Polymer (hydrophilic layer)	Polymer (hydrophobic layer)	Flux, L/m2·h	Feed side temperature (°C)	LEP (kPa)	Mean pore size (μm)	WCA (°)	Feed solution	MD type/reference
Additives	Additives
(PAN); PVA; nylon-6 (N6)	PVDF-HFP	109–15.5	60	185	< 0.27	148.9 ± 1.0	3.5wt.% NaCl	AGMD (Woo et al. 2017)
	LiCl
PSF	PVDF	47.7 kg/ m2·h	80	98 ± 1.4	-	140.5	3wt.% NaCl	DCMD (Khayet et al. 2018)
Cellulose acetate (CA)	PTFE	19.92	53		PTFE/CA: 0.45 PTFE/CA-SiNPs: 0.47	154.2 ± 0.3	600 mM NaCl and 1000 mg/L crude oil	DCMD (Hou et al. 2018)
Silica NPs
Commercial hydrophilic nylon 6,6 membrane	PVDF-HFP	45.6	60	136 ± 4	0.7 ~ 0.9	142.7	3.5wt.% NaCl	DCMD (Zhao et al. 2018)
	AC; LiCl
Commercial supporting woven tube	PVDF-HFP	13.2	50	181 ± 13	1.91 ~ 1.10		3.5wt.% NaCl	DCMD (Su et al. 2018)

Robust superhydrophobic dual-layer membrane

Inspired by the appealing self-cleaning characters of the silver ragwort leaf, lotus leaf, and water strider leg in the natural world, these prominent superhydrophobic surfaces have attracted significant attention. They have been widely investigated because of their promising applications in water harvesting, antifogging surfaces, and self-cleaning materials, especially desalination (Celia et al. 2013).

A three-dimensional (3D) superhydrophobic materials possess specific properties to maintain the air at both the surface and within the material’s bulk. A new interface between water, air, and material can be continuously provided as water penetrates it (Liao et al. 2014a). Two kinds of superhydrophobic dual-layer membranes (Liao et al. 2014a) were prepared via electrospinning for the MD process: (1) to form an ultrathin 3D superhydrophobic selective skin through electrospinning of FS10 (α, ω-triethoxysilane-terminated perfluoropolyether ((EtO)3Si-PFPE-Si(OEt)3))-modified silica and PVDF dope solutions on a porous PVDF nanofibrous support; (2) to form a thicker 3D superhydrophobic selective skin by electrospinning FS10-modified silica and PVDF dope solutions on commercial PET nonwoven support. The results showed that the water contact angles of the superhydrophobic dual-layer are above 150°. So, they exhibited excellent superhydrophobicity against numerous kinds of liquid, including coffee, juice, milk, and oil-in-water emulsion as well as water.

The superhydrophobic dual-layer membranes were fabricated from a hydrophobic polymer or hydrophilic polymer-hydrophobic additive (for example, modified SiO2 NPs, functional alumina NPs, functionalized GO) blend solution as a skin layer on a commercial hydrophobic membrane substrate or hydrophobic ENMs or hydrophobic membrane prepared by various methods such as phase inversion method as supporting layer via electrospinning (Attia et al. 2018a; Du et al. 2019; Efome et al. 2016; Li et al. 2020a; Liao et al. 2014b).

A novel composite membrane (Li et al. 2019b) was fabricated through PVDF fibrous coatings as the hydrophobic functional layer via electrospinning. They used two kinds of nonwoven fabrics and two kinds of spacers as the substrate to achieve high mechanical strength and excellent permeate flux of the membrane. The experimental data displayed that at the feed temperature of 80 °C, the permeate flux reached 49.3 kg/m2·h during the DCMD test. The spacer fabrics as substrates had minor less resistance to mass transfer than the nonwoven fabrics.

Many kinds of research for a triple-layer nanofiber membrane have been conducted to improve further the performance of the single-layer and dual-layer nanofiber membranes in MD application (Table 3).

Table 3

The characters of representative superhydrophobic dual-layer nanofiber membrane and the performance in the MD process

Polymer (bottom layer)	Polymer (top layer)	Flux, kg/m2·h	Feed side temperature (°C)	LEP (kPa)	Operation time (h)	WCA (°)	Feed solution	MD type/reference
Additives	Additives
PVDF	PVDF	3.2	27.5	379–565	6	139 ± 5	3.5wt.% NaCl	DCMD and VMD (Efome et al. 2016)
	Silica particles;
PVDF	PVDF	18.9	60	179 ± 7	50	> 153.9	3.5wt.% NaCl	DCMD (Liao et al. 2014b)
	Silica particles
PET	PVDF	24.6 ± 1.2	60	150	40	> 150	3.5wt.% NaCl	DCMD (Liao et al. 2014a)
	Silica particles
PVDF	PVDF	23	30–80	144.79	5	150 ± 0.3	2.5wt.% metal concentration (Zn, Cd, Pb, Cu, Ni)	AGMD (Attia et al. 2018a)
	Alumina NPs
PP	PVDF	36.4	50	233.9 ± 7.3	60	140.5	3.5wt% NaCl	VMD/ (Li et al. 2020a)
	GO

Triple-layer nanofiber membrane

Each layer in the triple-layer nanofiber membrane has a specific function.

A novel triple-layer membrane (Prince et al. 2013) was prepared by electrospinning PVDF solution on casted microporous PVDF membrane onto a PET substrate for AGMD application. The results showed that the formation of a thin hydrophobic nanofiber layer on the dual-layer membrane resulted in augmentation of the flux, salt rejection efficiency, and long-time performance due to enlargement of the LEP and hydrophobicity of the resultant membrane than the single-layer and dual-layer nanofiber membrane.

A triple-layer nanofiber-based membrane (Prince et al. 2014) was manufactured by wet casting and electrospinning method. (1) The middle layer was manufactured by the conventional phase inversion method while PVDF was dissolved in DMAc/ethanol solvent mixture. Then PVP (poly(vinyl pyrrolidone)) as pore former was added to prepare the casting polymer dope. Then, a small amount (0–3wt%) of surface modifying macromolecule (SMM (hydrophobic)) was added to the above mixture to alter the membrane surface. (2) PVDF-SMM(hydrophilic) composite nanofibers coated the bottom side of the intermediate layer via electrospinning. (3) The top side of the middle layer was coated with PVDF nanofibers via electrospinning.

A novel triple-layer nanofiber membrane possessing a diameter gradient (Ebrahimi et al. 2018) was fabricated: outer layers with bead-free and thinner diameters manufactured with the minimum concentration of PVDF (20wt%) and middle layer with thicker diameters produced with higher polymer concentrations (21.5–26wt%) via electrospinning for DCMD application. The results showed that all prepared membranes displayed about two times the permeate flux than a single-layer membrane while maintaining salt rejection above 99.9% due to lower heat loss and mass transfer resistance of the proposed membrane attributed to the increased pore size of the middle layer. The results showed that a permeate flux of the resultant membranes (27.8–31.5 kg/m2·h) was higher than 15.4 kg/m2·h for single-layer membranes under 42 g/L NaCl of feed solution.

A triple-layer nanocomposite membrane, which is composed of polyethersulfone (PES)/carbon nanotubes (CNTs) as a core and PVDF-HFP/CNTs as the bottom and top surfaces of the membrane (Elmarghany et al. 2020) was fabricated via electrospinning method, and the DCMD test was conducted.

A novel triple-layer membrane comprised of an electrospun PVDF-PTFE hydrophobic layer, a PET middle support layer, and an electrospun chitosan–polyethylene oxide (CS-PEO) hydrophilic layer (Li et al. 2020c) was developed to improve wetting resistance and water flux because a hydrophobic membrane via electrospinning only is difficult to guarantee a high flux. The experimental results revealed that the fabricated triple-layer composite membrane with an average flux of 19 kg/m2 h and rejection rate of 99.92% displayed better performance than the dual-layer membrane without a hydrophilic layer possessing an average flux of 15 kg/ m2 h and rejection rate of 99.88%. The PVDF-PTFE hydrophobic layer did not allow the water molecules to pass through the membranes. Then, CS-PEO hydrophilic layer will attract the water vapor from the PET supporting layer in time to enhance MD flux and delay the pore wetting.

However, the amphiphilic organic contaminants and low surface tension surfactants exist everywhere in hypersaline industrial effluents, like shale gas effluent or electroplating effluent, make superhydrophobic membranes susceptible to membrane wetting by significantly reducing the LEPw (Rezaei et al. 2018; Zheng et al. 2018). Therefore, superhydrophobic membranes lose rejection characters. Therefore, to extend the MD application for challenging wastewater’s desalination, a robust omniphobic membrane has been developed and widely investigated as a promising MD membrane that can resist wetting to water and low surface tension liquid.

Omniphobic membrane

MD, as a hybrid thermal/membrane separation process, allows only water vapors to pass through a hydrophobic membrane by a vapor pressure gradient (Wang and Chung 2015), and salt is rejected regardless of salt concentration. The available MD membranes show excellent hydrophobicity and can be successfully employed to demineralize relatively clean waters like seawater and saline water (Hou et al. 2010). A wetting resistance of the substrate pore in the hydrophobic membrane is ensured at the MD process’s beginning step. The chemical and physical properties at the interface between the membrane and aqueous solutions are modified as the feed chemicals (for example, surfactants and other organic materials) and the membrane materials interact, resulting in fouling/wetting.

Usually, the superhydrophobic surface exhibits a high water contact angle of above 150° and a water SA of below 10°. However, those values are measured by using high purity water in the absence of any surfactants or contaminants with low surface tension. Although surfactants are fusible in water, they usually exist as oil-in-water emulsions in the effluent from oil/gas and chemical industries, so they are an essential factor to influence membrane wetting (Chen et al. 2017).

Numerous MD membranes, i.e., superhydrophobic (Celia et al. 2013), composite hydrophilic–hydrophobic (Edwie et al. 2012; Peng et al. 2005), and omniphobic (possessing hydrophobic and oleophobic) membranes (Wang et al. 2016) have been specifically designed and developed to prevent surfactant and other low surface-tension materials wetting. Superhydrophobic membranes essentially improve the anti-wetting properties, but unavoidable electrostatic interaction makes them easily wet by surfactants, resulting in the reduction of the MD performance for the more prolonged operation (Chen et al. 2017). Then, in the composite hydrophilic–hydrophobic membranes, the hydrophilic side contacts the hot feed, and the hydrophobic side contacts the cold permeate. They can prevent the wetting problem and enhance permeation flux. The biggest challenge is to possess the possibility of producing wetting defects owing to the possible coating of the inner pores within the hydrophobic membranes (Khayet et al. 2018). An active layer in the composite hydrophilic–hydrophobic membranes may be delaminated from the support layer because of the difference in the interfacial tension (Tijing et al. 2014b). Omniphobic membranes prone to repel water (hydrophobic) and low-surface-tension liquids like oil (oleophobic) have demonstrated a stable performance for the MD process under the surfactant’s existence in an aqueous solution for a long time (Boo et al. 2016). The omniphobicity that repels water and low surface tension liquids such as oil is created via producing surfaces possessing low surface tension and multilevel re-entrant structures that together promote the existence of the metastable Cassie–Baxter state for interfacial contact between liquid, solid, and vapor phases (Anish Tuteja et al. 2008; Butt et al. 2013; Tuteja et al. 2008). The superhydrophobic surface could be realized by a cooperation of roughness (nano/macroscale architectures) and a low surface energy chemistry (Li et al. 2008). However, an omniphobic surface needs a more accurate micro-/nano-scaled hybrid architecture design than a superhydrophobic surface (Schlaich et al. 2016). Although the roughness of a surface possessing a re-entrant morphology is not high enough, it will result in omniphobicity if the material has low surface energy (Anish Tuteja et al. 2007).

An omniphobic surface possessing a re-entrant texture that can support strongly metastable composite solid–liquid-air interfaces with both polar and non-polar liquids maintaining low surface tensions was developed (Tuteja et al. 2008) (Fig. 14). The experimental results revealed that in case of a liquid drop placed on the resultant surfaces keeping re-entrant texture (ψ < 90°), the interacting traction upon the liquid–vapor interface is directed upward to support the metastable Cassie state for the composite solid–liquid-air interface, even though θ < 90°.

Fig. 14

a) and b) Schematic portrayals illustrating possible liquid–vapor interfaces on two different surfaces possessing the same solid surface energy and the same equilibrium contact angle (θ), but different geometric angles (ψ) (Tuteja et al. 2008)

Many omniphobic membranes have been fabricated via deposition of nanoparticles to produce re-entrant micro/nano-structures on an electrospun substrate membrane surface, followed by surface fluorination to lower surface tension energy (Dizge et al. 2019; Hou et al. 2019). An omniphobic SiNP (silica nanoparticle) modified electrospun cellulose nanofiber membranes have been fabricated through three steps for the DCMD process: fabrication of cellulose nanofiber membranes by electrospinning; achievement of SiNPs modified nanofiber membranes by a sol–gel method; silanization using fluorinated alkylsilanes onto SiNPs modified nanofiber membranes to obtain the membrane with low surface energy via chemical vapor deposition (CVD). Heptadecafluoro-1,1,2,2-tetrahydrodecyl trichlorosilane (FDTS) was used for silanization (Dizge et al. 2019).

Next, various omniphobic membranes have been manufactured by electrospinning of hydrophobic polymer solution as substrate membrane, followed by surface fluorination to lower surface tension energy via solution submerging without any assistance of roughening treatments (i.e., the auxiliary nanoparticles) and thermal treatment (An et al. 2018; Deng et al. 2018; Qing et al. 2020). A novel amphiphobic PVDF-HFP nanofibrous membrane (An et al. 2018) was prepared via electrospinning and followed by the surface coating by 1H,1H,2H,2H-perfluorodecyl-triethoxysilane (FAS) through a facile immersion-coating process and thermal treatment. The results showed that the resultant FAS-coated PVDF-HFP nanofibrous membranes displayed prominent amphiphobicity against both water and oil even under rigorous conditions without much reduction of permeate flux and salt rejection of the consequent membrane by FAS coating in the DCMD process. A new nanoparticle-free omniphobic PVDF nanofibrous membrane has been fabricated through three steps for the DCMD process: the fabrication of PVDF nanofibrous membrane via electrospinning; the creation of multiscale hierarchical nanofin structures on electrospun PVDF nanofibers by solvent-thermal-induced roughening method; polydopamine-anchored surface fluorination to low the surface energy of the nanofibrous membrane (Qing et al. 2020). 1H,1H,2H,2H-perfluorooctyltrichlorosilane (PFTS) was used for fluorination treatment. The experimental results demonstrated that the resultant membrane displayed a contact angle of 173.2° and 153.8° for water and mineral oil.

Then, some omniphobic membranes have been manufactured via simple one-step electrospinning. A novel omniphobic F-POSS/PVDF-HFP membrane (Lu et al. 2018) fabricated from PVDF-HFP and fluorinated-decyl polyhedral oligomeric silsesquioxane (F-POSS) colloidal suspension solution via one-step electrospinning. The experimental results revealed that the resultant omniphobic F-POSS/PVDF-HFP membrane had uniform fiber structures and very high F-POSS concentration on the surface. Furthermore, this membrane displayed excellent omniphobicity with a contact angle of 128.2° toward several low surface tension liquids such as ethanol and surfactant (sodium dodecyl sulfate, SDS).

And then, various omniphobic membranes have been manufactured via surface modification methods to produce re-entrant micro/nano-structures on an electrospun substrate membrane surface, followed by surface fluorosilanization and fluorination to lower surface tension energy (Jongho Lee et al. 2016; Koh and Lee 2021; Lee et al. 2016b; Li et al. 2019). A new omniphobic PVDF-HFP nanofibrous membrane with pine-needle-like hierarchical TiO2-nanorods has been fabricated by using electrospinning, hydrothermal technique with polydopamine, and a fluorination treatment for the DCMD process (Li et al. 2019). Trichloro (1H,1H,2H,2H-tridecafluoro-n-octyl) silane (FOTS) was used for fluorination. The experimental results demonstrated that the TiO2 nanorods form uniform pine-needle-like hierarchical nanostructure on PVDF-HFP nanofibers. Then, the resultant membrane exhibited high contact angles of 168° and 153° towards the water and mineral oil, respectively.

A new omniphobic polyamide 6 (PA6)-based nanofibrous membrane has been fabricated by chemical surface modification methods for the MD process (Koh and Lee 2021). PA6 nanofibrous membrane was fabricated via electrospinning, and electrospun PA6 nanofibrous membrane was treated with trichloro(1H,1H,2H,2H-perfluorooctyl) silane (F-POSS) to endow a hydrophobic property. The fluorine-treated PA6 nanofiber membrane was altered with the PVDF-g-F-POSS solution by the grafting reaction between PVDF and F-POSS, and the polymer chains were flocculated to create Si–F nanoparticles (NPs). The resultant membrane displayed the omniphobicity against low surface tension liquids thanks to the formation of Si–F NPs on the membrane surface.

Due to its excellent oleophobic property, robust mechanical property, low water–adhesion property, and good durability, the omniphobic membrane has fulfilled unique potential in the MD process. The omniphobic membrane’s manufacturing process contains hierarchical re-entrant structures and surface fluorination like fluorosilanization and fluorination to lower the resultant membrane’s surface energy with the hierarchical re-entrant structures (Table 4).

Table 4

The characters of representative omniphobic membrane and the performance in the MD process

Polymer	Post-treatment agent	Flux, kg/m2·h	Feed side temperature (°C)	Operation time (h)	WCA (°)	Feed solution	MD type/reference
Additives
Cellulose acetate	1H, 1H, 2H, 2H-perfluorodecyl-triethoxysilane (PDTS)	13.6	53	120	155.6 ± 3.9 (for water); 95.3 ± 2.5 (for decane)	3.5wt% NaCl; SDS	DCMD (Hou et al. 2019)
Silica nanoparticles
PVDF	1H, 1H, 2H, 2H-perfluorododecyl-trichlorosilane (CF3(CF2)9CH2CH2SiCl3, FTCS)	36.9	60	24	> 150	3.5wt% NaCl; SDS	DCMD (Deng et al. 2018)
PVDF-HFP			60	8	154.5 ± 2.6 (for water); 149.5 ± 4.7 (for 3 mM SDS)	5.85wt% NaCl; 3 mM SDS	DCMD (Lu et al. 2018)
F-POSS
PVDF-HFP	1H, 1H, 2H, 2H-perfluorodecyl-triethoxysilane, (FAS)	19	65	2	138 ± 2; 127 ± 3 (oil contact angles)	3.5wt% NaCl; 0.1 mM SDS;	DCMD (An et al. 2018)

Remarks of electrospinning technology and MD process

Current status and challenges of electrospinning technology

Many fusible or soluble polymers, including proteins, polysaccharides, and nucleic acids, have been used for electrospinning by controlling process parameters and environmental conditions and selecting the appropriate solvents. Their physicochemical properties and their merit and demerit of them have been studied overall.

Based on recognizing the solvent’s importance during electrospinning, numerous solvents are distinguished from different characteristics such as boiling point, density, volatility, solubility, dipole moment, vapor pressure, and dielectric constant. Their effects on electrospun nanofiber’s characteristics (for example, electrospinnability, morphology, thermal characters, crystal architecture, and chemical structure) have been discovered in detail.

In applying electrospinning technology to practical applications, there are many challenges to solve. Poor mechanical properties and low productivity are the main challenges limiting their application in practical industrial applications of ENMs. Then, another challenge is the tailored alignment of fabricated fibers, especially fabricating electrospun fibers to calibrate morphology to ensure electron transport to be maximum possibly. To obtain uniform nanofibres with a diameter below 50 nm is also a significant challenge. Continuous development of electrospinning equipment and adding chemical additives such as organosoluble salts (for example, tetrabutyl ammonium bromide (TBAB), lithium chloride (LiCl), and ionic liquid) are more practical to obtain uniform nanofibers.

Current status and challenges of the MD process

Numerous experimental and modeling studies for the MD process application such as DCMD, AGMD, VMD, and SGMD using the electrospun membranes have been carried out, and many mass and heat transfer mechanisms for each MD process using different modes of membranes (multi-layer, composite and omniphobic membranes) have been studied in detail (Fig. 15).

Fig. 15

Schematic of heat and mass transfer portrayal in MD process of water transportation through single (a) (Zhao et al. 2020), composite (b) (Zhao et al. 2020), triple-layer (c) (Li et al. 2020c), and omniphobic membranes (d) (Hou et al. 2019)

q, q, q0, q mean the heat transferred through the boundaries of the feed side and permeate side, and through the hydrophobic layer and hydrophilic layer of the membrane; q corresponds to the overall heat transfer flux through the entire thickness of the membrane; J indicates vapor permeation flux: Tb T, T’, T, and T indicate the temperature of the bulk solution in feed side, the membrane in feed side, the interface of the hydrophobic/hydrophilic layers, the membrane in permeate side and bulk solution in permeate side; T, T, T, T, T, T correspond feed bulk temperature, the surface temperature in the hydrophobic layer, the surface temperature in the hydrophilic layer, the interface of the hydrophobic/middle layers, the interface of the middle/hydrophilic layers and permeate bulk temperature; P1 and P2 indicate pressure on feed and permeate side of the membrane.

Inspired by natural phenomenons such as lotus leaf, fresh rose petal, rice leaf, Salvinia molesta, and sharkskin, many membranes with excellent characteristics such as superhydrophobic, superhydrophilic, superomniphobic, and superoleophobic properties have been developed. They also have been applied in the MD process to demonstrate their outstanding performances.

Many studies about the module design and module optimization in MD configuration (Zhao et al. 2013) and the membrane’s structural design have been performed: a superhydrophobic single layer, multi-layer (composite hydrophilic–hydrophobic/hydrophobic–oleophobic underwater dual layer, superhydrophobic dual layer, triple-layer nanofiber membrane), and omniphobic nanofiber membrane fabricated in various structure (hollow, core–shell, lotus, pillar, acicular, multilevel re-entrant, pine-needle-like hierarchical structure and so on).

Figure 16 shows the structure of a single stage in the V-MEMD module. V-MEMD modules are composed of a steam raiser, multiple stages, and a condenser. The single stage consists of foil frames that function as a condenser and “distillate channels,” and membrane frames that serve as “vapor channels.” The space between foil frames and membrane frames plays a role as “Feed channels.” The vapor that flows into the parallel foil frames heats the feed and is then condensed. A vacuum is applied to a small channel at the bottom of the frame to remove non-condensable gases. The condensed vapor then enters a “distillate channel.” The heat energy of condensation transported through the foil frame is converted into evaporation energy to generate new vapor in the vapor channel. The vapor leaves the vapor channels and flows into the next stage. This process is repeatedly performed in the following stages (Fig 17).

Fig. 16

Schematic portrayal of streams in vacuum-multi-effect-membrane-distillate (V-MEMD) module (Zhao et al. 2013)

Fig. 17

The architecture of the representative triple-layer membrane (Prince et al. 2014)

In order to mitigate the membrane wetting, many kinds of research for feed pre-treatment operations have also been performed widely: removal of the wetting constituents of the feed solution before MD operation (Rezaei et al. 2018), filtering and boiling feed solutions containing protein (Gryta 2008), alleviating protein fouling by using ultrasonic waves (Hou et al. 2017), nanofiltration for less fusible complexes with divalent salts (Roy et al. 2017), aggregation and anaerobic/aerobic digestions (Noel Dow et al. 2017), and chemical adjusting of feed solution using antiscalant (Peng et al. 2015). Along with the development of new membranes with a new structure for MD applications, it is important to develop a pre-treatment technology to reduce the components as much as possible that cause membrane wetting according to the characteristics of industrial wastewater.

As mentioned earlier, the MD process has not yet been industrialized due to the lack of reasonably designed membranes that can satisfy the MD process. The design of a membrane that can satisfy the MD process faces two critical challenges: low permeate flux and membrane pore wetting. Fabricating membranes possessing high porosity, open-cell pore structures, and narrow pore size distributions, especially electrospun nanofiber membranes with hierarchical re-entrant structures, could be considered a more efficient method for higher flux and enhancing the wetting resistance of the prepared membrane. The omniphobic property depends greatly on the hierarchical re-entrant structures. The air trapped in hierarchical re-entrant structures performs a vital role in omniphobic membranes. However, because the captured air may get away from air pockets during long-time MD operations, it is imperative to develop the omniphobic membrane with new designs such as multi-layer and nanorods architecture to support the omniphobicity during long-time MD operations. On the other hand, membrane fouling is a crucial drawback of MD, causing a decline in permeate flux because of the accumulation of undesirable material on the membrane surface and pores. Therefore, pre-treatment of wastewater before applying MD should be performed to decrease fouling and wetting problems and to ameliorate energy consumption and efficiency in the treatment and recovery process.

The electrospinnable polymeric materials possessing inherent hydrophobicity are minimal. A superhydrophobic surface with hierarchical architecture would not be efficient in enhancing the VMD process’s membrane wetting because the membrane material is not inherently hydrophobic (Jiao et al. 2020). Therefore, it is crucial to develop novel electrospinnable membrane materials with intrinsic hydrophobicity. The water drops will be rolled/slid off as soon as they contact the electrospun membrane surface, possessing a “lotus effect” or the self-cleaning property during the MD process. It helps to prevent membrane wetting. Therefore, it is important to fabricate membranes with low SA as well as high WCA.

Most fabricating superhydrophobic and omniphobic surfaces by reducing the surface energy depends on fluorinated materials that have potential environmental hazards. Therefore, research on developing eco-friendly materials such as non-fluorinated silicon compounds should be performed to fabricate these surfaces. Good selection and incorporation of suitable additives and appropriate post-treatment to improve the MD process’s membrane performance during electrospinning or post-electrospinning are essential to satisfy the MD application’s desalination requirement. Since the electrospun membrane has poor mechanical strength, it is used as a functional layer on other substrates. Therefore, studies to develop a competitive electrospinning substrate should be performed in the future.

Conclusion and the future prospects

This review performed comparisons for membrane design manufactured in various structures (hollow, core–shell, lotus, pillar, acicular, multilevel re-entrant, pine-needle-like hierarchical structure, and so on). From the fact that MD has not yet been introduced into the water industry due to a lack of suitable membranes, it is very important to design and manufacture suitable ENMs that can meet the needs of MD so as to fulfill the advantages of MD. Because the MD process is a thermally driven process, research on the development of heat transfer and mass transfer models that can comprehensively predict the properties of the manufactured ENMs based on various conditions (feed concentration, flow rate, temperature difference over the membrane, feed temperature pore-wetting, long period performance, membrane fouling, and so on) for various MD processes should be carried out.

Due to the nanofiber membrane’s intrinsic properties, they demonstrated their notable performances in seawater desalination and wastewater remediation.

Then, the membrane’s fabrication methods with excellent characteristics such as superhydrophobic, superomniphobic, and superoleophobic characteristics via electrospinning, especially the application and the structural design of these membranes in the MD process, were discussed. At present, composite dual-layer or multi-layer membranes, especially omniphobic membranes, are regarded as the only solution. Therefore, research should be performed extensively to develop a composite membrane using the omniphobic membrane as a hydrophobic layer. Nano-additives with hydrophobicity and hydrophilicity in composite membranes and omniphobic membranes endow better permeability, good salt rejection and excellent stability, and good antifouling performance by improving the hydrophobicity and hydrophilicity of the membrane, so research on the development of effective nano-additives is very important.

In this review, we also emphasize the importance of pre-treatment with the development of an optimally designed membrane. If the pre-treatment of chemical components and materials in the feed solution affecting the wetting of the membrane before applying the MD process is performed, the designed membrane can display its potential by reducing the wetting elements on the membrane. Studies on a suitable post-treatment method for the ENMs and a suitable pre-treatment for the feed solution of the MD should be performed in the future.

The development of ENMs with antiviral, antibacterial, antifouling properties, and efficient filtration properties from the continuously increasing water pollution and global coronavirus disease (COVID-19) pandemic has recently attracted great interest. By incorporating a functionalized filler with antibacterial and antiviral properties such as graphitic carbon nitride (g-C3N4), Ag nanoparticles, and tourmaline nanoparticles into the ENMs, the ENMs have high potential as an antibacterial and antiviral material for water purification, air filtration, and protective clothing while maintaining the nanoscale properties. Therefore, studies on preparing ENMs for producing distilled water from wastewater containing bacteria and viruses through the MD process should be conducted in the future.

The cost of manufacturing nanofibrous membrane through electrospinning technology on a laboratory scale is estimated at about $ 200/m2, which is affordable (Liu et al. 2022). Reducing the manufacturing cost of the ENMs is very important in expanding the application potential of the ENMs.

Nature provides excellent clues to thought and creative inspiration. Therefore, developing biomimetic approaches is very important to design and fabricate a more suitable membrane in the MD process.

The challenges mentioned above in fabricating nanofiber membranes are complicated; however, they will be solved. It is undeniable that with the rapid development of science and technology, ENMs manufactured with optimized design will display their enormous potential for addressing numerous emerging challenges to humankind, such as seawater desalination, metal removal from wastewater, and wastewater purification.