A review of seawater desalination with membrane distillation: material development and energy requirements Open Access

Fresh water constitutes one of the most important resources in domestic as well as industrial activities. Globally, the concurrent growth in population and industrial development has led to a drastic increase in the demand for fresh water (Alkhudhiri et al. 2012). Unfortunately, there is a limit to the availability of water on earth. A recent statistic has revealed that 97.5% of the available water on earth is saline water. This means only about 2.5% of the water on earth is suitable for direct uses (Qasim et al. 2019). Furthermore, out of the 2.5% which constitute fresh water, 1.71% of it is in the form of ice and glaciers (Gopi et al. 2019). Yet, the demand for clean water is increasing due to the rapid increase in urbanization, and in economic and industrial development. Water treatment companies are therefore turning to seawater as the major source from which freshwater is produced.

Seawater can be treated using many physical, biological, and chemical methods. On a general note, the choice of any of these methods depends on many factors which include cost of treatment, nature of chemicals, space for installation, and possible generation of secondary pollution (Fakhru'l-Razi et al. 2009). For these reasons, membrane technology is being adopted for certain stages of water treatment processes. The commonly used types of membrane technology in seawater treatment are microfiltration, ultrafiltration, nanofiltration, and reverse osmosis (RO). All these membrane processes can be used either independently or combined as hybrid systems. The choice of any of these membranes will depend on the type of impurities that need to be removed from the wastewater. Some of the advantages of RO include its capability to remove nearly all contaminant ions and most dissolved non-ions, relative insensitivity to flow and total dissolved solids (TDS) level, suitability for small systems with a high degree of seasonal fluctuation in water demand, ability to operate immediately without any minimum break-in period, low effluent concentration, ability to remove bacteria and particles, and its operational simplicity and automation which allows for less operator attention (Minnesota Rural Waterworks Association 2009). Some of the major limitations of RO include high capital and operating costs, high level of pre-treatment requirement in some cases, and that membrane materials are prone to fouling and scaling.

One of the recommended solutions to the aforementioned challenges is the use of innovative technologies such as membrane distillation (MD). This membrane technique can substantially reduce the energy consumption of water treatment, reduce waste formulation, and make the treatment more cost-effective, safer, and sustainable (Stankiewicz & Moulijn 2000). MD is a class of membrane technology that operates at relatively low temperatures. This method has been investigated worldwide because it is well known to be a suitable technology for actualizing process intensification. It is therefore important to have reference material on this membrane separation technique which will serve as a source of data that will allow investors and decision-makers to be able to make a perfect economic decision on which method to adopt to provide a sustainable water treatment system. The objective of this review was to look at the current status of MD technology concerning membrane materials development and new methods of modifying the materials, new areas of applications of MD, and research work done so far on improving the energy requirements of the technique.

MD is based on the use of a hydrophobic microporous membrane that allows only the conveyance of water vapor and volatiles through micropores and prevents aqueous feed as a liquid to pass through it. This is achieved by the partial pressure difference established through the membrane (Criscuoli 2021).

Moreover, MD has turned out to be a significant and auspicious water treatment technology, expressly in water desalination. Its plentiful advantages depend on the fact that it can be carried out under low-pressure conditions compared to traditional technologies such as RO which requires a high feed pressure to overcome the osmotic pressure. Moreover, it runs at low temperature with a feed solution under its boiling point and pressure near to atmospheric pressure (Mendez et al. 2018).

Additionally, MD could be used for a large range of salt concentration, making it appropriate for initial water treatment along with final treatment without incidence of osmotic pressure on the driving force (Khayet & Matsuura 2011b). It necessitates a light pre-treatment of the feed water (Pangarkar et al. 2014). Also, it possesses a high salt rejection that yields recovered water of higher quality and purity. The reality that this process is a thermally driven operation makes implementation easier by using available, abundant, or less expensive energy such as solar energy, geothermal energy, and waste heat (Joo & Kwak 2016; Lokare et al. 2017).

MD is a compatible technology for separation processes in which water is the main constituent of the feed solution. In MD, at least one side of a microporous hydrophobic membrane is in direct interaction with an aqueous solution. The temperature gradient between the two sides of the membrane produces a partial pressure difference and that leads to mass transfer through membrane pores. Throughout the MD process, only vapor molecules are able to pass through the membrane while liquid molecules cannot pass through the membrane wall due to the membrane hydrophobicity. Evaporation of volatile compounds created by the partial pressure difference leads to the vapor particles being permitted across the pores and being condensed on the permeate side of the membrane. Numerous MD configurations are applied to control the driving force on both sides of the membrane (Jafari et al. 2018).

This review would not be complete without touching on the advantages and disadvantages of different desalination technologies as described by various researchers. Table 1 shows a comparison between different types of membrane-based desalination technologies (Feria-Díaz et al. 2021).

In 1963, the first patent on MD was filed by Bodell (1963) entitled ‘Silicone rubber vapor diffusion in saline water desalination’. A membrane material made from a viscoelastic polymer, i.e. silicone rubber, opened this new field of MD. It was not until 1967 that Findley (1967) investigated the suitability of other materials such as cellophane, nylon, gum wood, aluminium foil among others as membrane materials for MD. Following experimental observation, Findley established the qualities of an idealized membrane suitable for MD. These properties are:

• (i)

  Low thermal conductivity

• (ii)

  Moderate thickness

• (iii)

  A hydrophobic membrane (non-wettable)

• (iv)

  Negligible permeability to liquids.

The properties highlighted in Findley's paper later became the yardstick for fabricating membranes for MD application. In the same year, Weyl (1967) patented the idea of spiral wound membrane modules with membrane sheets fabricated from PTFE. The membrane utilized was very thin (thickness around 1/8 inch (3.175 mm)) and hydrophobic, thus, satisfying some of the desirable membrane qualities established by Findley. Other hydrophobic polymers such as polyethylene, polyvinylchloride, and polypropylene were also suggested (Weyl 1967). In 1972, Rodgers (1972) patented the idea of a plate and frame membrane module, including membranes possessing uniform pore distribution to the desirable properties of an ideal membrane, as listed above. Fluorocarbons such as polyvinylidene fluoride (PVDF) and PTFE were particularly preferred due to their high non-wettablity. Until 1979, the major challenge faced in using hydrophobic membrane for MD application was water-logging, a phenomenon whereby salty water fills up the pores of the membrane thereby preventing permeation of water vapor through the membrane. Water-logging is known to reduce demineralized water flux as the MD process progresses. On February 14, 1979, Cheng filed a patent on the use of a composite membrane to prevent water-logging. Both hydrophilic and hydrophobic membranes were used for an experimental investigation, whereby the hydrophobic membrane is sandwiched between two hydrophilic sides (Cheng & Wiersma 1983). Fluorocarbons such as PTFE and PVDF were individually used as a substrate on which a hydrophobic layer of polysulfone (PS) and polyallylamine (PAA) were formed. Research interest grew further in the 1990s, especially with respect to harnessing solar energy for MD heat requirement. Numerous research papers were published during this period which led to the development of Europern Commission projects such as SMADES and MEMDIS (Thomas et al. 2017). In 2009, another breakthough in material development with regard to superhydrophobic materials was implemented (Ma et al. 2009). It should be noted that the first experimental demonstration of superhydrophobic material was in 1996 with fractal surfaces made of alkyl ketene dimers (Onda et al. 1996). More recently, research directions in MD has focussed on thermoplasmonic (light-to-heat) nanomaterials (Santoro et al. 2022), MD crystallization (Yadav et al. 2022), membrane pore size tailoring for membrane scaling prevention (Tan et al. 2022), and in-situ photocatalytic semi-Volatile Organic Compound (sVOC) removal (Ning et al. 2022).

MD can be classified according to the technique used to create the transmembrane vapor pressure on the permeate side. Based on this classification, there are basically four membrane configurations (see Table 2): (i) Direct Contact Membrane Distillation (DCMD) (ii) Vacuum Membrane Distillation (VMD) (iii) Sweep Gas Membrane Distillation (SGMD) and (iv) Air Gap Membrane Distillation (AGMD). The area of application, and the advantages and disadvantages of each of these configurations has been summarized in Table 2.

Table 2

Various membrane distillation configurations

MD configurations .	Application .	Advantages .	Disadvantages .	Diagram .
Direct Contact Membrane Distillation (DCMD)	– Seawater desalination – Crystallization – Treatment of dye effluents – Arsenic removal from aqueous solution.	– High permeate flux – Well thought-out at commercial scale.	– High conductive heat loss
Vacuum Membrane Distillation (VMD)	– Seawater desalination – Treatment of alcoholic solution – Recovery of aroma compounds – Treatment of textiles wastewaters.	– High permeate flux – Considered at commercial scale.	– High membrane risk pore wetting – Process difficulty.
Sweep Gas Membrane Distillation (SGMD)	– Brackish water desalination – Wastewater treatment – Azeotropic mixtures separation – VOC removal.	– Reduction of the barrier to the mass transport over forced flow	– High temperature polarization risk – Process complexity.
Air Gap Membrane Distillation (AGMD)	– Seawater desalination – Fruit juices concentration – Azeotropic mixtures separation – VOC removal.	– Low conductive heat loss – Process simplicity – Low temperature polarization risk.	– Lower flux than SCMD and VMD

MD configurations .	Application .	Advantages .	Disadvantages .	Diagram .
Direct Contact Membrane Distillation (DCMD)	– Seawater desalination – Crystallization – Treatment of dye effluents – Arsenic removal from aqueous solution.	– High permeate flux – Well thought-out at commercial scale.	– High conductive heat loss
Vacuum Membrane Distillation (VMD)	– Seawater desalination – Treatment of alcoholic solution – Recovery of aroma compounds – Treatment of textiles wastewaters.	– High permeate flux – Considered at commercial scale.	– High membrane risk pore wetting – Process difficulty.
Sweep Gas Membrane Distillation (SGMD)	– Brackish water desalination – Wastewater treatment – Azeotropic mixtures separation – VOC removal.	– Reduction of the barrier to the mass transport over forced flow	– High temperature polarization risk – Process complexity.
Air Gap Membrane Distillation (AGMD)	– Seawater desalination – Fruit juices concentration – Azeotropic mixtures separation – VOC removal.	– Low conductive heat loss – Process simplicity – Low temperature polarization risk.	– Lower flux than SCMD and VMD

Reprinted with permission from (Alkhudhiri & Hilal 2018). Copyright 2018 Elsevier.

View Large

Direct Contact Membrane Distillation (DCMD): The top diagram in Table 2 shows the schematic representation of DCMD. In DCMD, low temperature freshwater flows at the permeate side while hot salty feedwater flows at the other side of the membrane. This arrangement creates a water vapor pressure difference, which serves as the driving force, between the two sides of the membrane (Khayet & Matsuura 2011c). Considering the high temperature of the feedwater, the fugacity of water in the feed side will generally be higher than that in the low temperature permeate side. Even though temperature is a key parameter in MD, DCMD suffers from high heat loss by conduction through the membrane due to the high heat capacity of the sweeping fluid. This serves as one of the major drawbacks of using this technique as compared to others (Lawson & Lloyd 1996).

Vacuum Membrane Distillation (VMD): VMD utilizes a vacuum in the permeate side to create the required driving force for water permeation through the membrane wall. Thus, the permeate side of the membrane is essentially empty as incoming water through the membrane wall is sucked by the vacuum into a condenser where the water vapor is liquefied. The second diagram in Table 1 depicts the operation mechanism of VMD. Obviously, the pressure created by the vacuum pump should be lower than the vapor pressure of water in the feed side for continuous water permeation (Bandini et al. 1992).

Sweeping Gas Membrane Distillation (SGMD): This MD configuration is very similar to VMD. While a vacuum is used to create the driving force in VMD, dry inert gas is continuously passed through the permeate side in the case of SGMD (Table 2, third diagram). As in VMD, a condenser is usually required to condense the swept water vapor. Usually, the sweeping gas section is operated in a recirculation mode by first condensing the water before passing the dry gas through an external heat exchanger (Abdelrasoul 2020).

Air Gap Membrane Distillation (AGMD): AGMD is a membrane configuration type whereby permeated water vapor is directly condensed within the membrane module by stagnant cold air. As shown in Table 2 (fourth diagram), cold air is stored just between the membrane and the condensing wall. To maintain the low temperature of the air a cooling solution is usually passed just next to the air chamber at a regulated flowrate. By creating a low temperature region on the permeate side, the vapor pressure of water is significantly reduced and hence water vapor permeation can continuously take place. The condensed water is finally collected in a condensation compartment under the influence of gravity (Abdelrasoul 2020).

The plate and frame membrane module is the most commonly used membrane configuration in the laboratory for membrane testing due to the ease of fabrication, cleaning, replacement and maintenance of the membrane sheet. As shown in Figure 3(a), this membrane module takes a rectangular shape with the feed solution and sweeping fluid flowing in tiny channels. These channels which could either be rectangular or cylindrical are separated by the membrane sheets. It is to be noted that both flat sheet and capillary membranes can be used to fabricate plate and frame modules. Compared to hollow fiber membrane modules, plate and frame modules have a relatively low packing density which typically ranges between 100 and 400 m²/m³ (Khayet & Matsuura 2011a). Another limitation of plate and frame technology is the need of using a support especially at high flowrates.

The hollow fiber and tubular membrane modules take a form similar to a shell-and-tube heat exchanger as shown in Figure 3(b) and 3(c). The main difference between a hollow fiber and tubular membrane module is in the inner diameter which ranges between 50–100 μm for the former and 1–2.5 cm for the latter. The tiny holes in a hollow fiber membrane are the main reason for the high packing density they possess, which can be up to 3,000 m²/m³. Compared to plate and frame, hollow fiber and tubular membrane modules do not require support. However, maintenance of both hollow fiber and tubular membranes is extremely difficult if not impossible. Meanwhile, the pressure drop across the module length ranges between 0.2–0.5 kg/cm² for hollow fiber and 2–3 kg/cm² for tubular membrane modules.

The last type of membrane module is the spiral wound module which is schematically represented in Figure 3(d). As shown in this figure, flat sheet membranes are rolled around a perforated collection tube located at the center of the module. As the feed water flows tangentially in the feed channel, water permeates through the membrane to the permeate channel which is designed to allow the permeate water to whirl around until it reaches the perforated collection tube. Packing density for this kind of module ranges between 300 and 1,000 m²/m³ typically creating pressure drop in the range 3–6 kg/cm².

Among the difficulties encountered in the development of MD is the selection of the membrane materials. MD requires specific properties of membrane materials to handle the working principle and performance. The membranes utilized in the MD process ought to be hydrophobic, porous, non-dampened, with low thermal conductivity at high temperature, and show fit thermal stability (Zare & Kargari 2018).

Classification of membrane materials is based on several features. Membrane materials can be synthetic or natural. The natural membrane materials are often found in plants and animals, such as polymers including polysaccharides and rubbers (Drioli & Giorno 2009). Membranes can be classified into organic, like polymers, inorganic like ceramic, metals, zeolites, and carbon (Chen et al. 2015), or mixed-matrix that contain both organic and inorganic materials (Adewole & Sultan 2019). Popular materials are described below.

As pointed out by an early researcher (Rodgers 1972), fluorocarbon membranes such as PTFE and PVDF are preferred for MD application due to their non-wettability. Other commonly used membrane materials for MD are PP and PE. However, PE is not as thermally stable as the other three membrane materials. Polydimethylsiloxane (PDMS) is another polymer membrane material, but it is rarely used in MD. Table 3 summarizes some of the important properties of the polymeric membranes used in MD desalination.

Table 3

Characteristics of commercial polymer materials (Wang & Chung 2015)

Polymer materials .	Chemical structure .	Surface energy (×10−3 N m−1) .	Thermal conductivity (W m−1 K−1) .	Thermal stability .	Chemical stability .	Fabrication methods .	References .
PTFE		9–20	0.25	Good	Good	Sintering Melt-extrusion	Khayet (2011); Camacho et al. (2013); Zhang & Gray (2011); Lalia et al. (2013a)
PP		30	0.17	Moderate	Good	Melt-extrusion TIPS
PE		28–33	0.40	Poor	Good	Melt-extrusion TIPS
PVDF		30.3	0.19	Moderate	Good	NIPS TIPS Electro-spinning

Polymer materials .	Chemical structure .	Surface energy (×10−3 N m−1) .	Thermal conductivity (W m−1 K−1) .	Thermal stability .	Chemical stability .	Fabrication methods .	References .
PTFE		9–20	0.25	Good	Good	Sintering Melt-extrusion	Khayet (2011); Camacho et al. (2013); Zhang & Gray (2011); Lalia et al. (2013a)
PP		30	0.17	Moderate	Good	Melt-extrusion TIPS
PE		28–33	0.40	Poor	Good	Melt-extrusion TIPS
PVDF		30.3	0.19	Moderate	Good	NIPS TIPS Electro-spinning

View Large

Arguably, PTFE and PVDF are the most employed polymeric membranes for MD. To this end, Adnan et al. (2012) investigated the influence that PTFE microstructure tuning will have on water flux. A series of commercial PTFE membranes were utilized for the study where the effects of porosity, tortuosity, pore size and thickness were investigated. In this study, porosity ranged between 62 and 82% while pore size was between 0.2 and 0.45 μm. The membrane thickness was between 30 and 160 μm and tortuosity was calculated from the inverse of porosity. Meanwhile, feed temperature was varied between 30 and 70 °C. The authors found porosity to be the most critical factor among the factors studied. Indeed, high porosity enhances water diffusion due to the accessibility of more pore channels. Consequently, water flux was found to increase. Although reduction of membrane thickness also enhances water flux, there is a threshold beyond which membrane thickness reduction becomes detrimental to water flux. The reason for this occurrence is the blockage of the membrane pores by water-logging (Cheng & Wiersma 1983). Meanwhile, the importance of uniformity of pore sizes was demonstrated on some commercial PTFE membranes (Shirazi et al. 2014). Two commercial membranes – one from Chang-Qi (M1) and the other from Millipore (M7) – were used for MD application. The authors reported that the manufacturers' specifications for M1 were: pore size (0.2 μm), thickness (230 μm) and porosity (80%); M7 has the same pore size, a thickness of 175 μm and porosity of 70%. Even though these specifications suggest that M1 should have a higher flux than M7 due to the higher porosity and smaller thickness, the reverse was found to be true. The main reason for this observation was because of the higher uniformity of the pore size in M7 as compared to M1. It will be recalled that Rodgers already emphasized the importance of pore size distribution in water permeation through membranes in MD (Rodgers 1972).

Compositing hydrophobic membranes with hydrophilic layers is one of the techniques present-day researchers are using to improve membrane performance in terms of flux improvement. Li et al. (2020) used an electrospinning technique to prepare composite membranes having a PVDF/PTFE hydrophobic layer on the feed side and a chitosan-polyethylene oxide (CS-PEO) hydrophilic layer on the permeate side. Sandwiched in between these layers is a PET support layer. The triple layer membrane (TLM) was compared with a dual layer membrane (DLM) fabricated without the hydrophilic layer. The results of their findings showed that there is a marginal increase in flux from 15 L/m²/h in the case of the TLM to 19 L/m²/h in DLM. However, the salt rejection was c. 99.9% in both cases. It is to be noted that the experiment performed by Li goes against the recommendation in the patent filed by Cheng (Cheng & Wiersma 1983) with respect to placing the hydrophilic layer next to saline water. According to Cheng, membrane wetting will generally reduce the flux obtained in an experiment similar to that of Li; in fact, Li reported a decline in the flux of both TLM and DLM as time progressed (Li et al. 2020). However, the authors failed to investigate the idea of preventing membrane wetting by making the hydrophilic side of the membrane face the saline water.

Another hydrophobic polymer commonly used for MD application is PVDF. Among the popular polymers in MD, PVDF is the only one that can be processed into membranes with asymmetric morphology using the non-solvent induced phase separation (NIPS) technique (Wu et al. 2017). The influence of incorporating nanoparticles (to form mixed matrix membranes) such as multi-wall CNT, SiO₂, hydrophobic CaCO₃, SiO₂, CuO, rGO etc on the morphological structure of PVDF has been extensively studied (Hou et al. 2012; Baghbanzadeh et al. 2015; Efome et al. 2015; Tijing et al. 2016; Wu et al. 2017; Abdel-Karim et al. 2019). For instance, Hou et al. (2012) studied the influence of hydrophobic CaCO₃ nanoparticle addition in PVDF membrane fabrication. The membrane was fabricated using the NIPS technique and membrane performance such as thermal loss and water flux were assessed. The obtained membrane showed low thermal loss and improved water flux. This improved performance is a direct consequence of narrower pore size distribution, larger pore sizes, and increased surface roughness and contact angle exhibited by the mixed matrix membranes. The weight fraction of CaCO₃ in the PVDF membrane was varied from 0 to 50 wt% and it was found that 20 wt% gave the best water flux and thermal efficiency.

A comparative cost analysis was performed on the use of PP and PVDF for MD saline water treatment (Macedonio et al. 2014). Water production cost was found to be lower when PVDF was used as compared to a PP membrane. The lower cost obtained with a PVDF membrane was because of the higher pore size which translated to higher water flux. Apart from the increase in water flux obtained when membrane pore size is increased, membranes with larger pore sizes also have lower heat loss. In a separate study, Gryta investigated a number of commercial capillary PP membranes for water desalination (Gryta 2018). The membranes employed in the study were fabricated using the thermal induced phase separation (TIPS) technique. The performance of these commercial membranes was continuously monitored over several hours. It was found that both surface porosity and pore size reduction can be used to control membrane wetting. Consequently, the author recommended that the pore size should be in the range 1–3 μm.

In a very rare study, the implementation of a TIPS-made PE membrane for MD desalination was investigated (Zuo et al. 2016). Membrane performance indicators such as energy efficiency and water flux were investigated for a series of commercial PE membranes. It was found that PE membranes having large pores had lower heat loss and higher water flux than those with smaller pores. Most importantly, the authors reported a superior water flux from PE membranes as compared to flatsheet/hollow fiber membranes fabricated from PVDF, PTFE and PP. Although the pore size of the PE membranes utilized in the study was lower than those of PVDF, PTFE and PP, the authors rationalized this result by pointing out the importance of pore interconnectivity in membrane fabrication for MD application. However, the authors did not report the energy efficiency of any of the PVDF, PTFE and PP membranes.

Inorganic materials mostly own superior chemical, mechanical and thermal stability compared to polymeric materials. Yet, employment of inorganic membrane materials has been limited, although growing interest can be observed today. Ceramic, glass, and metallic materials are the three types of inorganic materials that are used frequently for membrane distillation. Ceramic membranes are made of materials formed by a grouping of a metal such as aluminium, titanium, or zirconium with a non-metal in form of an oxide, carbide, and nitride. They are taken from such materials from the core class of inorganic membranes with aluminium oxide and zirconium oxide as the most significant representatives. They are fabricated by sol-gel processes, while metal membranes are made of materials formed by the sintering of metal powders such as tungsten and molybdenum. They have had limited attention until now. Glass membranes are made of glass materials such as silicon oxide. They are prepared by techniques involving leaching on demixed glasses and by sol-gel processes (Mulder & Mulder 1996).

Mixed-matrix membranes have been used as an alternate membrane technology. They are made of organic (polymers) and inorganic materials. The selection of polymeric and inorganic materials is fundamentally significant in the improvement of the mixed-matrix membranes since the properties of materials can impact their morphology and separation performance. A mixed matrix membrane that contains highly selective polymers can produce a better separation performance (Aroon et al. 2010). Overall, a mixed matrix membrane can offer the physicochemical stability of a ceramic material and the membrane creating ease of polymeric materials as well as high thermal stability (Qadir et al. 2017).

Table 4 shows the comparison between different types of membranes utilized for water purification (Qadir et al. 2017).

Two of the most important characteristics of MD membrane are hydrophobicity and porosity. MD membranes with a balance between these two characteristics are commonly fabricated using different methods. These methods (which include phase inversion, stretching, interfacial polymerization, track-etching, and electrospinning) are used to make polymeric, inorganic, and mixed-matrix MD membranes (Lalia et al. 2013b). Furthermore, the membranes use in MD can be in several forms such as single-layer membranes (hydrophobic), composite dual-layer membranes (hydrophobic/hydrophilic), or composite triple-layer membranes (hydrophobic/hydrophilic/hydrophobic). The pore size of MD membranes are between 100 Ǻ to 1 μm (Abdelrasoul 2020).

Phase inversion is based on the solidification of an analogous polymeric solution. The manufacturing of MD membranes by phase inversion can be divided into the following phases: first, forming a casting solution by dissolving polymer pellets in a solvent, which is then cast on a plate. Second, casting the semi-liquid film on the plate and immersing it into a utensil for precipitation. Finally, creating a polymeric film with asymmetric or symmetric structures (García-Fernández et al. 2015). Moreover, this method can be utilized to make both asymmetric and symmetric porous membranes using one of three procedures: Non-solvent Induced Phase Separation (NIPS), Thermally Induced Phase Separation (TIPS), and Vapor-induced Phase Separation (VIPS). NIPS and VIPS are the most commonly used for fabricating hydrophobic membranes (Yeszhanov et al. 2021).

Stretching is a solvent-free technique in which membranes are made by squeezing out a polymer at a temperature near its melting point to create micropores. Fabrication by the stretching method is cheaper than other methods. The MD membrane manufacturing procedure involves extending a polymer with partial crystallinity to the axis of crystallite orientation. Then, a film is produced by extruding the polymer at below melting point temperature. So, by this technique, an MD membrane with a high porosity of 90% and a uniform porous structure can be obtained (Yeszhanov et al. 2021).

Interfacial polymerization (PI) is the most significant technique for the commercial manufacturing of thin-film composite (TFC) membranes. The procedures to fabricate MD membranes by using the PI method involves soaking a microporous polysulfone (PSU) support in an aqueous solution of the polymeric amine, then dipping the amine saturated membrane into a solution of a di-isocyanate in hexane, and finally cross-linking the membrane by heat-treatment at 110 °C (Lalia et al. 2013b).

The track-etching method is used to set the pore geometry of a membrane and is commonly used to manufacture membranes with the lowest pore size. The procedure to create MD membranes using this technique is: first, subject the film to high energy radiation which is perpendicularly applied to the film; then, immerse the film in an acid or alkaline bath and etch the polymeric materials away along tracks to create uniform cylinder-shaped pores with thin pore size distribution (Mulder & Mulder 1996).

The electrospinning technique was developed to produce nano-fiber MD membranes. It is an active technique used to manufacture nano-fiber MD membranes with high porosity and roughness. It is made up of a high voltage electric source which is used to form an electrically charged jet of polymer solution leaving a syringe with a metallic needle, then collected on a roller. It creates a polymer membrane with a high surface area to pore volume ratio (Yeszhanov et al. 2021).

Liquid entry pressure (LEP), membrane thickness, porosity, thermal conductivity, and water contact angle are the main characteristics of MD membranes (Kebria & Rahimpour 2020).

Membrane thickness is the most important operative characteristic of MD membranes. The relationship between membrane thickness and permeate flux of a membrane is an inverse relationship because permeate flux is enhanced when the membrane becomes thinner due to the reduction of mass resistance, while increase in the membrane thickness leads to reduced heat loss. So, there is a tradeoff between an advantage, which is minor heat loss, and a disadvantage, which is a minor permeate flux, for a thicker membrane (Kebria & Rahimpour 2020).

Liquid contact angle analysis is used to measure a membrane surface's propensity to be wetted by a liquid. The major component of the feed solution in MD is water. The calculation of water contact angle is made to determine the extent that surface tends to droplets of water. The angle between the surface of the membrane and the water droplets is calculated in this method. The contact angle of water can be calculated through numerous randomly selected membrane surfaces locations to minimize the error of calculation. The impact of mean pore and roughness of the surface should be considered to specify the exact water contact angle (Kebria & Rahimpour 2020).

MD membrane characterization techniques are required in membrane investigation, development, and engineering to evaluate performance, correlating specification characteristics such as flux and salt rejection, etc. Though commonly used methods for membrane characterization can be found in much of the literature, some of these methods have been included here so as to make this review a more complete reference point for information about MD. Characterization methods can be classified into dynamic and static methods. Information on the morphology, structure, physical and chemical properties of a membrane can be provided by using static methods, while the dynamic methods are of essential importance when studying the performance of a membrane. Scanning Electron Microscopy (SEM), Atomic Force Microscopy (AFM), and the wet and dry flow method are the common techniques used to optimize MD membrane characterization (Tylkowski & Tsibranska 2015).

Scanning electron microscopy (SEM) is a grading technique used in the characterization of membranes. To reach the optimal performance of field emission scanning microscopes and to reduce artifacts caused during the process, a careful specimen preparation technique is required. The investigation of SEM is a valued tool utilized to improve membranes for use in industry. Separation characteristics and mechanical strength of porous membranes due to their morphology can be detected in a secondary electron image. It is feasible to restrict the concentrations of the elements on the surface inspected by using a backscattered electron imaging mode and elemental analysis. In this technique, the distribution of functional parts or materials may be identified, allowing the membrane to be reinforced against mechanical strength (Schossig-Tiedemann & Paul 2001).

Atomic force microscopy (AFM) is a modern characterization technique. It is used to characterize membranes to gain an excellent knowledge of their performance and to fabricate particular membranes for the MD process. Moreover, AFM is a recently improved high-resolution technique for analyzing the topography of the surface of different types of membranes. And it can obtain 3D images of the surface of membranes without specific sample preparation (Shirazi et al. 2013).

The wet and dry flow method is a combined bubble point and gas permeation test. This technique includes the displacement of a liquid from the wetted membrane by using pure gas. It is used to determine the maximum pore size, mean pore size, and pore size distribution of MD membranes (Khayet 2011).

Our research covered the great efforts being made to improve and develop the performance of direct contact MD for seawater desalination being done by testing operational parameters such as feed flow rate, feed temperature, feed concentration, permeate flow rate, and cold permeate temperature. Also, detection of fouling to find the impact on the performance of a DCMD system was investigated.

Furthermore, the influences of feed and cold permeate flow rates on the performance of DCMD flux have been investigated by other researchers. Increasing the feed flow rate from 40 to 120 L/h led to an increase in the system flux from 4.3 to 8.2 kg/m² h as well as increasing the permeate flow rate from 1.4 to 3.29 L/h; the consequence of feed and cold permeate flow rates on the performance of DCMD flux was detected (Zhu et al. 2015).

In a separate study by Ameen et al. (2020), the influence of feed flow rate on the performance of DCMD was also investigated. The results also demonstrated that the DMCD flux increased with an increased feed flow rate. To illustrate this, the researchers increased the feed flow rate from 0.3 to 1.07 L/min with feed temperature in the range of 45 to 65 °C and with a constant salt concentration equal to 35 g/L. The flux of the system increased from 46.63% to 67.1% at feed temperature of 55 to 65 °C.

Furthermore, another study investigated the influence of feed and cold permeate flow rate on DCMD productivity for desalination of seawater. It was observed that increasing feed flow rate is more effective than increasing cold permeate flow rate to increase the productivity of the system. For example, when the feed flow rate increased from 0.4 to 0.8 L/min with feed temperature of 80 °C and cold permeate flow rate of 0.4 L/min with temperature in the range of 5 to 25 °C, the permeate flux increased from 39.5 to 47.5 L/m²h. While increasing the cold permeate flow rate from 0.1 to 0.4 L/min with temperature in the range of 5 to 25 °C and feed flow rate of 0.8 L/min with the temperature of 80 °C, the permeate flux increased but the effect of increasing feed flow rate was greater (Shirazi et al. 2012).

Overall, feed and permeate flow rate are the parts of operating conditions that affect the performance of a DCMD system. They increase the performance of DCMD when the countercurrent flow is used with a high rate of feed and permeate flow rates in the desalinating process.

Feed temperature measures the temperature of the fluids used in a system and it instantly affects the productivity of a DCMD system. A high temperature of the feed is very significant for increasing system performance, especially for seawater desalination using DCMD, because bringing seawater closer to the point of evaporation leads to increased productivity and that allows vapor only to pass through the hydrophobic membrane sheet in a DCMD system. Cold permeate refers to the coldness of the permeate that is reused in the system to increase productivity. Cold permeate temperature is a very imperative factor that influences the performance of a DCMD system. The low temperature of permeate helps to increase the surface area of heat and mass transfer that leads to condensing the vapor produced in the fastest way, to get a high amount of system productivity.

To illustrate the effect of feed and cold permeate temperatures, Khalifa et al. (2017) tested the performance of direct contact membrane distillation. The experimenters explored the effect of feed temperature on system flux by increasing the feed temperature from 50 to 90 °C with a feed flow rate of 4.6 L/min with salinity equal to 2 g/L and a cold permeate flow rate of 3.65 L/min, with the temperature of the permeate decreasing in 5 degree stages from 25 to 5 °C, the flux increased from 302% to 560%. The experimenters also investigated the effect of cold permeate temperature on system flux by decreasing the temperature of permeate from 25 to 5 °C with a cold permeate flow rate of 3.65 L/min and with feed flow rate equal to 4.6 L/min with a temperature of 90 °C and 2 g/L salinity of feed, and the system flux increased from 80 to 100 kg/m²h. The experimenters detected that at the highest feed temperature and lowest permeate temperature, the system flux reached to a high level of DCMD system productivity.

Shirazi et al. (2012) experimented with seawater desalination by using different hydrophobic membrane sheets in a DCMD system. To evaluate performance, the experimenters considered the influence of numerous operating conditions involving feed temperature, etc., and increased the feed temperature from 50 to 80 °C with a feed flow rate equal to 0.8 L/min and cold permeate flow rate equal to 0.4 L/min with decreasing temperature in 5 degree stages from 25 to 5 °C. The result of the experiment showed that by increasing feed temperature to 80 °C, the system flux increased to 47.5 L/m²h. Following this result, it can be deduced that there is a direct relationship between hot feed temperature and permeate flux.

Al-Obaidani et al. (2008) investigated the effect of feed temperature on the flux of DCMD. DCMD flux increased with increased feed temperature. To clarify this, the experimenters tested numerous membrane sheets and detected the feed temperature influence on permeate flux for each membrane sheet. The experiments ran with a feed temperature increase from 25 to 70 °C with a mass feed flow rate of 0.055 kg/s and with feed concentration of 35 g/L, with a mass permeate flow rate equal to 0.026 kg/s with constant permeate temperature of 15 °C.

Alwatban et al. (2019) investigated the effect of feed temperature on the Reynolds number of the feed flow rate on the performance of DCMD, and increased the temperature of the feed from 50 to 80 °C with the Reynolds number of the feed equal to 1,500. The system flux increased from 18.5 to 59.05 kg/m²h.

Ameen et al. (2020) were examined the influence of feed temperature on the performance of DCMD. The experimenters detected that when feed temperature increased, the DMD flux increased. To illustrate this, the experimenters increased feed temperature from 45 to 65 °C with a constant feed flow rate of 1.07 L/min and by using different amounts of feed concentration. The experimenters discovered that with high feed temperature with high feed flow rate and low concentration of feed, the flux of DCMD was at highest peak. Figures 5a and 5b represent the typical effect of feed temperature at different salt concentrations and cold permeate temperature at different feed temperatures on DCMD permeate flux, respectively.

Feed and permeate temperatures are the biggest operating condition factors affecting the performance of a DCMD system. The performance of DCMD increased when the lowest permeate temperature and highest feed temperature were used in the desalinating process.

The performance of DCMD (in terms of water flux) strongly depends on seawater salt concentration. DCMD performs excellently when the salt concentration is very low and vice versa. This is due to the salt particle rejection of hydrophobic membranes which are typically used for this application. To illustrate this, Khalifa et al. (2017) experimentally investigated DCMD performance using different values of feed concentration in the range 0.14 to 100 g/L. The experimenters distinguished that the DCMD flux increases when a feed has a low concentration compared with system flux decreasing when a feed has a high concentration. In general, the lowest feed concentration provides more system productivity than the highest feed concentration.

Moreover, the impact of salt concentration was studied concerning the pore size (0.20, 0.45 and 1 μm) and thickness of membrane sheets (100, 130, and 170 μm) used on a DCMD system and the experimenters observed that the concentration polarization has an antagonistic consequence over the membrane system as it decreases the system flux (Alwatban et al. 2019).

The influence of salt concentrations of feed on the performance of DCMD was deliberated concerning different feed flow rates. The experimenters investigated DCMD flux by using three values of feed flow rate with salt concentration starting from 0 to 200 g/L and with constant feed temperature at 60 °C. The experimenters detected that the DCMD flux decreased by 23.28% when feed concentration increased from 0 to 35 g/L. Also, the DCMD flux reduced by around 24.91% when salt concentration increased from 23 g/L and 32% when salt concentration increased from 100 to 200 g/L (Ameen et al. 2020). Figure 6 shows the effect of salt concentration at different feed flowrates on the performance of DCMD.

Membrane fouling is the deposition and buildup of undesirable materials such as salts, suspended solids, and algae on the surface of membrane sheets that blocks the membrane sheet pores. Furthermore, feeding solutions which are used as a feed stream in a DCMD system are the main cause of fouling, helped by other operating parameters and manufacturing materials of the system (Gryta 2008).

For illustration, Gryta (2008) analyzed an experimental study of fouling in the DCMD process by using several feed solutions such as oily wastewater (bilge water) and saline wastewater. The researcher observed that the effect of different types of fouling occurs in the DCMD process such as organic fouling and crystallization fouling. Organic fouling is caused in DCMD by oily and saline wastewater ingredients. Additionally, the bilge water used in the experiment was supplied from a UF process product which contained an amount of 7.2 mg/L of oil and 382 mg TOC/L of water. Moreover, the distillate of the DCMD process obtained exhibited conductivity of 1.8 μS/cm and turbidity of 2.1 mg/dm³. And during the experiment, the researcher increased the amount of oil in the feed stream continuously from 7.2 up to 27 mg/dm³ to observe the influence of oil concentration in the feed. Furthermore, a gradual decrease in maximum permeate flux was detected, indicating that the oil was adsorbed on the membrane surface. On the other hand, crystallization fouling was caused in DCMD by calcium carbonate (CaCO₃) when using tap water as a feed with feed properties such as conductivity of 610–625 μS/cm, TOC of 6.3 mg/dm³, and inorganic compounds of 28 mg/dm³, and feed temperature of 323 to 363 K. The researcher observed that the feed heating led to form numerous compounds on the membrane surface, and after 85 h of module operation a deposit isolated the membrane surface and led to reducing the performance of the membrane sheets. The researcher recommended filtrating the feed solutions to avoid the fouling that occurred in the DCMD system; using HCL solution did not dissolve all the deposits.

Moreover, Hung et al. (2016) investigated the influence of scaling on the performance of DCMD by using pre-filtrate seawater as feed. The experimenters measured the water recoveries by observing feed temperature and water circulation influence. As well as detecting that scaling occurs on the membrane when the process operated at high feed temperature, it increased the supersaturation of calcium sulfate and magnesium sulfate on the membrane sheet, and that led to more relentless membrane scaling. To exemplify that, the experimenters ran the experiment with a feed flow rate equal to 1 L/min with TDS equal to 37,000 mg/L and with feed temperature increased from 40 to 60 °C and permeate flow rate of 1 L/min with a constant temperature of 25 °C. However, the experimenters found that the water flux of the DCMD process reduced when the scaling formed on the surface of the membrane at high water recoveries, and they used two ways to mitigate the scaling: decreasing feed temperature and adding anti-sealants to the feed.

In all, the performance of DCMD is affected by several operation conditions such as feed and permeate flow rates, feed and permeate temperature, feed concentration, fouling, and scaling. Each condition has specific influences on a DCMD system, and each condition is related together to affect the system productivity. Table 5 illustrates the operating condition parameters of different experiments.

MD process performance is mostly affected by the qualities of the membrane. A good MD membrane should possess a certain level of porosity, be hydrophobic, and be stable under high-temperature conditions. These properties are fundamental to foster a highly proficient membrane for the practical operation of MD technology. Thus recently, different investigations have been published on the modification of MD membrane by incorporating nanomaterials (Ray et al. 2020), by grafting, coating, electrospinning, electroblotting and many other techniques. In this section, modern modification techniques for improving the performance of MD membrane materials are covered.

Grafting is one of the most encouraging strategies to adjust a membrane's surface through covalent bonding interaction between the joined chains and the membrane. Covalent addon of graft chains on the membrane surface evades their delamination and offers the long-term chemical stability of the grafted chain (Kato et al. 2003). Moreover, it can be accomplished by versatile techniques such as UV photo-irradiation, plasma, high energy irradiation, and controlled polymerization. It uses different types of monomers and can be classified into two categories: grafting with a single monomer and grafting with two or more mixtures of monomer (Liu et al. 2011).

As an example, Lee et al. (2021) modified a polyvinylidene fluoride-co-chlorotrifluoroethylene (PVDF-CTFE) three-layer membrane for hollow fiber and grafted the prepared membrane with pentafluorostyrene (PFS) and adding it to 10 wt% of PFTE. The researchers evaluated the performance of the modified membrane by using the VMD process. The results showed an increasing value of LEP to 117 and 154% as well as increased flux to 137% due to an increase in the size of the internal macro void. Correspondingly, this showed an enhancement in the wetting resistance and permeability.

The coating technique is considered one of the most promising methods to modify a membrane surface through coating a hydrophilic layer on hydrophobic membranes to increase performance. It can be applied by physical adsorption. It has several types such as plasma (Liu et al. 2011). To illustrate more, (Eykens et al. 2018) investigated the MD performance of SMMs-PS membranes coated using several approaches and evaluated them for seawater desalination. For 3 wt% NaCl, the modified membrane showed the highest flux compared with the original and commercial membranes such as PTFE.

This technique is one of the most hopeful approaches to adjust the membrane surface through applied high voltage. As an example, Wanke et al. (2021) investigated a modified membrane with a hydrophilic surface with high pure water flux compared with a PVDF membrane using wastewater as a feed of the experiment and by electrospinning a layer of polyvinylpyrrolidone-co-polymethyl methacrylate (PVP-co-PMMA) over the modified polydopamine membrane.

Moreover, Li et al. (2021) have been fabricated a superhydrophobic nanofiber membrane with a high contact angle and sliding angle reach of 162.3° and 9.8° through fluorinating a zinc oxide (ZnO) and electrospun PVDF membrane (PDTS-ZnO-PVDF membrane). The researchers investigated that the modified membrane has a much better anti-wetting property and higher value of LPE (reaching 160 kPa) than a neat and ZnO blended PVDF membrane for desalinating pure NaCl solution and NaCl solution.

Electroblowing is the modern method of electrospinning aided with air blowing. The strategy includes the preparation of a polymer solution by melting the polymer in a solvent and feeding the polymer solution through a spinning nozzle applied with high voltage. Compressed air is passed through the lower end of the spinning nozzle, gathering the fibers in the form of a web on a suitably grounded collector to form a new membrane surface. Moreover, the presence of two applied forces interact to manufacture the nano-fibers from the polymeric fluid by electroblotting (Nayak et al. 2012).

Niknejad et al. (2021) evaluated the performance of modified membranes produced by the elecrtoblowing technique and coating them by utilizing a dope solution including 6 wt % polystyrene (PS) in dimethylformamide to compare it with the original membrane, which was a acrylonitrile-butadiene-styrene (ABS) polymer produced by using the electroblowing technique. The results show the performance of the modified membrane increased by 70% compared with the ABS membrane. In addition, it showed a stable permeate flux and salt rejection factor because of its high LEP value and sufficient hydrophobicity.

In terms of development and improvement of the properties of MD membranes, innovators have explored several ways to improve the different qualities of a membrane to increase performance and increase membrane life without any problems that influenced the permeability of the membrane using the addition of another chemical element or compound in the original MD membrane. So, Tibi et al. (2021) have embedded TiO2 modified with PVDF to improve the ammonia rejection and permeate flux in MD application by adding saline to a modified membrane (PVDF-TiO2-Saline). The innovators found out that the performance of the modified membrane was better than the original membrane and ammonia rejection reached more than 90% additionally the researchers expect that ammonia rejection could reach 100% with a modified membrane made of 31.3% silane, 2.50% TiO2, and 15.48% of polymer concentration.

Due to the development of membrane distillation technologies in recent years, the use of MD technology in many economic and commercial fields has also increased. As the interest in the use of MD grows, researchers are always looking into ways of improving MD technologies for use in several other areas. So, MD technologies could be used in the desalination and pure water production industry, as well as the chemical industry, food industry, textile industry, pharmaceutical and biomedical industry, and nuclear industry (Kiss & Kattan Readi 2018).

MD is the most used technology in the lab to test readiness and performance in desalination and wastewater treatment compared with other sectors. Niknejad et al. (2021) experimented with testing the performance of a modified polymethyl methacrylate (PMMA) membrane and compared it with commercial membranes such as PP and PTFE. The results showed better performance of the modified membrane than commercial membranes for desalinating 35 and 150 g/L saline water. The highest permeate flux reached 41.04 and 35.94 kg/m² using a fabricated membrane and due to the superhydrophobicity of 164.2° and high liquid entry pressure (LEP) of 227.3 kPa, while PTFE and PP produced a good permeat flux reaching 35 kg/m². Moreover, Bandar et al. (2021) found average flux of 13.10 kg/m²h and salt rejection of 98.96% during testing of a modified membrane by using well water as feed.

Furthermore, Foureaux et al. (2021) examined DCMD for water reclamation by using PTFE and PVDF membranes for eight months. The researchers found salt rejection was stable after a long-term experiment and it was maintained to 99.2% by both membranes. Also the permeate flux obtained was 6.82 L/m²h. DCMD was indicated as a promising type of MD for water reclamation in the future by the researchers.

MD innovations have been well utilized in the lab to test readiness and performance in the chemical industry and consequently also used in the field. To illustrate that, Lewandowicz et al. (2011) implemented MD for ethanol recovery. The experimenters distinguished that MD could be a straightforward method that leads to an increase in ethanol production by using Hsp70 and Hsp104 membranes in batch and continuous fermentation processes. The Hsp104 membrane reached a high flux of 28.20% of ethanol by using a continuous fermentation process.

Furthermore, Veleva et al. (2021) examined hydrophobic and oleophobic membranes by using DCMD to remove the main components of petrochemical pollution such as acetate, phenol, and propionate. The researchers found that an oleophobic membrane would have the highest value of petrochemical pollution rejection reaching 97% at a pH equal to 13.

Moreover, Yan et al. (2019) applied DCMD to treat anaerobic digestion effluent containing a big amount of ammonia and phosphates to increase toxin removal by using feed acidification for potential recovery. The experimenters obtained a superior rejection reaching 99% of ammonia, phosphates, sulfanilamides, and DOC in anaerobic digestion effluent. Also, feed acidification contributed to increasing pollutant removal from 66% to 99%. However, Al-Salmi et al. (2020) studied DCMD for producing water by treatment from an oil field by using a PP membrane. The results obtained indicate the impressive DCMD potential for treating hypersaline oilfield-produced water. The overall salt rejection reached higher than 99.9% while TOC was more than 93.3%.

In recent years, the food industry has been a part of MD application studies. Our research included the different areas in the food industry investigating the performance of MD technology and considering it as the best alternative modern technology in place of old technologies. As an example, Quist-Jensen et al. (2016) evaluated the performance of DCMD for concentrating clarified orange juice after treating it by an ultrafiltration process. The researchers found that DCMD concentrated clarified orange juice from 9.5 to 65 degrees Brix. Therefore, utilizing DCMD as a second stage integrated process a high quality of concentrated juices could be obtained. Moreover, Criscuoli & Drioli (2020) studied VMD for concentrating clarified date juice by using a PP membrane. The researchers found that the VMD concentrated the clarified juice from 18 to 70 degrees Brix, and so recommended implementing this technology as a secondary process for increasing juice concentration. Furthermore, Moejes et al. (2020) assessed AGMD for concentrating milk after treating the milk by using the RO process. The researchers found that the milk concentration increased from 18% to 50% by using AGMD after the RO unit, and that AGMD has the benefit of internal heat recovery that could be utilized as a second stage process for milk concentration.

The study of MD applications has not yet covered all processes in the textile industry and researchers are still looking to test the ability of MD to handle textile industry processes to the fullest or be a significant part of it. Therefore, Zhang et al. (2021) investigated a pretreatment method for treating a combined textile effluent by joining DCMD with fractionation and ozonation. The researchers found that water recovery reached 98% when treating fractionated effluent and 38% for treating ozonated effluent by DCMD. Moreover, de Sousa Silva et al. (2021) analyzed the influence of operational conditions on the performance of DCMD for synthetic effluents that contain reactive and disperse dyes. The investigators found that the performance of DCMD increased with feed temperature and flow rate and led DCMD to distillate 97.3% of reactive and 98.7% of disperse dyes with color rejection reaching more than 98% of dyes. Also, Fortunato et al. (2021) investigated the impact of the efficiency of DCMD for treating a synthetic textile dye solution compared with other research works. The experimenters found that performance of DCMD increased with feed temperature increase and fouling formed with increasing temperature of the feed. This finding is in agreement with a previous study Laqbaqbi et al. (2019).

MD technology is known as a modern technology in the separation process and is widely studied to use in different sectors in the industrial world. Researchers have tested the ability of MD technology in the pharmaceutical and biomedical industry sector. For illustration, Guo et al. (2020) studied membrane fouling in MD in pharmaceutical wastewater treatment and found that the solution for reducing membrane fouling throughout the MD process was to use DCMD to treat pharmaceutical wastewater. The researchers fabricated a new membrane to increase the hydrophobic properties and increase fouling resistance by coating PVDF with perfluorooctyltriethoxysline (PTFS)-TiO2. The researchers found no visible foulant layers in the membrane surface and the salt rejection reached 78% after DCMD operation.

Jeong et al. (2021) experimented with testing the feasibility of the MD process for potable water reuse by using MD as a barrier for dissolved pharmaceuticals and organic particles. The experimenters found that all the selected dissolved pharmaceutical particles were completely removed by MD and the flux of MD obtained was 20 L/m²h. Furthermore, Nellessen et al. (2021) examined MD for producing water for pharmaceutical use using VMD and AGMD, finding that both VMD and AGMD were able to produce distillates with a limiting value of conductivity of PW. The researchers found that a single MD process is stable for producing pharmaceutical-grade water.

The nuclear industry is one of the significant sectors in which researchers have started investigating the use of modern technology like MD to increase the quality of production and save energy. To illustrate this, Liu & Wang (2013) used DCMD for treating low-level radioactive wastewater. The researchers found that DCMD was able to eliminate Ca⁺, Sr⁺² and Co⁺² from wastewater and the DCMD flux decreased linearly by 60% due to increasing NaNO₃ concatenation, and found that DCMD is a promising technology in the separation of radioactive wastewater. Furthermore, Jia et al. (2021) designed a VMD pilot-scale desalination plant for decontaminating radioactive wastewater. The researchers studied the impact of operating conditions on VMD performance. The results obtained show that the VMD flux increased with an increase in feed temperature. The flux of VMD reached 6.82 kg/m²h when feed temperature increased to 90 °C. Moreover, Nie et al. (2021) used VMD for decontaminating radioactive wastewater that contained a low level of uranium from the UO₂ fuel element industry. The researchers found an average flux of VMD that reached 7.3 L/m² h, and considered that VMD showed excellent preservation of all major particles like uranium, COD and NH₄⁺-N.

Thus, researchers have utilized MD in several areas to see the performance suitability of an MD unit and include it in the different industrial areas.

Energy requirements are the major challenge faced in the desalination sector. As the world heads for sustainable energy resources rather than non-renewable resources due to cost and environmental reasons, renewable energy is considered as the alternative sustainable energy that could be used in place of non-renewable energy. Solar, wind, geothermal, hydro, and biomass are all examples of renewable energy (Harjanne & Korhonen 2019).

Moreover, renewable energy use and desalination are both critical needs for huge portions of the world facing energy and water challenges. So, membrane distillation is considered as the alternative promising thermal technology that could be used to desalinate water with renewable energy. Most researchers have focused research on solar energy in the last years due to its stability to heat feed to the level needed for MD. For illustration, Zarzoum et al. (2019) have experimented with operating DCMD by using solar energy to produce water. The researchers used solar energy to heat feed by using a heat exchanger in the desalination process with 12.29 μV/Wm² intensity of solar energy, and the DCMD productivity obtained was 15.2 L/h on a sunny day. The experimenters observed that DCMD is a very suitable and promising unit configuration for the Arab world and North Africa due to the arid area.

Furthermore, Huang et al. (2020) evaluated solar membrane distillation SMD for desalting water by using a photothermal membrane with thermal concentration. The experimenters found that the average SMD productivity reached 0.64 kg/m²h with a light intensity of 1 kW/m². The researchers observed that using solar energy, the waste heat of industrial and other renewable energies would reduce the cost and energy consumption of the MD process. Moreover, Said et al. (2020) investigated the performance of Nanophotonic Enhanced Solar Membrane Distillation (NESMD) for treating oil-produced water. Regardless of the method of water sample collection, an excellent salt rejection and productivity was observed in both cases. Specifically, water flux was 2.77 L/m²/h for samples collected using a pyrex bottle and 2.80 L/m²/h for samples collected using a recycled plastic container. Also, Ma et al. (2022) analyzed direct solar heating schemes for VMD by using a photothermal membrane (VMD-PM) and solar absorber plate (VMD-FPC). The experimenters found that the performance of VMD-PM was higher than VMD-FPC, reaching 7.14 L/m² and 6.86 L/m², respectively, with energy consumptions of 886.6 kWh/m³ and 926.8 kWh/m². A summary of recent articles on MD using solar energy can be found in Table 6.

Thus, researchers have analyzed the performance of MD by using renewable solar energy to reduce the cost and energy consumption for desalinating water and save the environment.

This review has given a general overview of different aspects of membrane distillation (MD) such as module types and configurations, membrane material classifications, membrane fabrication techniques, areas of application of MD, membrane characterization techniques and influence of MD operating parameters. Particular attention has been given to polymeric membranes and energy requirements due to their importance in MD. Literature findings have demonstrated the prospect of producing potable water from seawater desalination using MD technology. Despite these prospects, many issues still need to be resolved before its commercialization. Some of these issues are highlighted below:

• Most of the studies focus on membrane development and/or performance evaluation of commercially available membranes. Research into MD water treatment plant requirements and water production cost evaluation are very limited in the literature. More emphasis should be given to water production cost to allow investors to identify the hidden potentials of MD.

• The low thermal stability of polyethylene (PE) has discouraged most researchers from studying PE membranes for MD application. Nevertheless, PE has been demonstrated to have higher water flux than some commercial PTFE, PVDF and PP. More studies on compositing PE with other materials are required.

• Currently, there are no standardized procedures for evaluating the long-term applicability of membranes for MD; some researchers test the membrane for less than 100 hours while others test the membrane for above 1,000 hours. This testing time should be standardized to allow easy comparison of membranes.

• Membrane performance depends on membrane properties such as pore size, porosity, pore interconnectivity, tortuosity and on operating parameters such as feed temperature, feed concentration, feed flowrate, permeate flowrate and temperature. Efficient optimization of these factors requires advanced machine learning techniques such neural network, support vector machine, and so on. This area is yet to be fully exploited.

The authors gratefully acknowledge the support of the International Maritime College Oman (IMCO).

All relevant data are included in the paper or its Supplementary Information.

The authors declare no conflict of interest.