Agricultural expansion, industrial development, population growth, and climate change have led to increased water withdrawals and water shortages in many regions of the world. The need to solve the water crisis is becoming so urgent that freshwater has taken the role of a strategic resource. Because oceans constitute nearly 97% of the world’s water, desalination has garnered significant attention as an effective method to extract freshwater from seawater, wastewater, or brackish water.

Commercial desalination technologies can be divided into two main categories: thermal distillation (multi-effect distillation (MED) and first multi-stage flash (MSF) distillation) and membrane separation (reverse osmosis (RO) and nanofiltration (NF)). The former involves phase change processes, whilst the latter involves semipermeable membrane systems [1]. RO technology is used in >90% of the current desalination plants and contributes to 69% of annual desalinated water production [2]. However, RO can recover max 50% due to the osmotic phenomena.

Recently, among the different emerging technologies, membrane distillation (MD) has received considerable attention both from an industrial and academic perspective. MD is a thermally driven separation process that uses a microporous hydrophobic membrane to separate dissolved molecules from a liquid stream [3]. Its advantages include a lower operating pressure compared to RO process which allows for the use of low-grade thermal energy and renewable energy for its operation. In addition, many efforts to further reduce MD energy consumption have been reported in recent years, including photothermal membrane distillation [4] and hybrid membrane distillation systems [5].

One of the most interesting and promising extension of MD concept is represented by the membrane crystallization (MCr) process which provides the possibility of recovering simultaneously pure water and high-quality salt products using the logic of zero liquid discharge (ZLD) [6].

Nevertheless, despite their great potential, MD and MCr have some drawbacks such as flux lowering due to poor hydrodynamics and inefficient module design, and distillate contamination due to membrane wetting; the latter is one of the main obstacles to the wider application of MD. Therefore, developing new advanced membrane materials with good performance and low cost has become essential to intensify the MD process. Ideal MD membranes are expected to exhibit features such as high liquid entry pressure (LEP), high permeability, low fouling rate, low thermal conductivity, excellent chemical and thermal stability and excellent mechanical strength.

Polymeric membranes have long been considered for MD applications due to their good processability, intrinsic hydrophobicity, and low cost.

However, they present disadvantages such as limited chemical stability, fouling, and short lifetime. In contrast, inorganic membranes (zeolite, silica, zeolite, carbon, and hybrid inorganic–organic one) exhibit higher chemical, thermal, and mechanical stabilities that make them potential candidates for MD applications. In particular, zeolites ^[7][8][9][7,8,9] and carbon nanotubes ^[10][11][10,11] were suitable for water desalination due to their subnanometer pores/channels. However, the high cost- and time-consuming process of preparing perfectly ordered subnanometer arrays of zeolites and carbon nanotubes hinder their commercialization [12].

Since the discovery of single-layer graphene, two-dimensional (2D) materials have garnered notable interest in many fields of science and technology including MD processes. In the last decade, research on the role of 2D materials in membrane science has grown as they have a high potential of providing new solutions to significant problems. Two-dimensional materials are suitable for the fabrication of high-performance membranes in terms of specificity, selectivity, high flux, chemical and mechanical resistance [13]. Members of the family of 2D materials include graphene-based materials, transition metal dichalcogenides (TMDs), 2D zeolites, metal organic frameworks (MOFs) and carbides (MXenes) ^{[7][14][15][16][17]}[7,14,15,16,17].

1.1. The Advent of 2D Materials

Two-dimensional materials are crystalline solids with a high ratio between their lateral size (∼1–10,000 μm) and thickness (<1 nm) [18]. They consist of single- or few-layer atoms in which the in-plane interatomic interactions are much stronger than those along the staking direction [19]. Being ultrathin, 2D materials exhibit exceptional physical, electrical, chemical, and optical properties ^[20][21][22][20,21,22]. Two-dimensional materials have garnered great interest in seawater treatment applications because ion-sieving, either via in-plane pores or stacked nanosheet channels, can be achieved and both can be economically produced on a large scale. The hyper-thinness of 2D monolayers makes it possible to fabricate ultra-thin membranes with excellent mass transport and high transmembrane fluxes [23]. Furthermore, 2D materials can significantly improve the selectivity of solutes due to their structural and chemical homogeneity [24]. Currently, nanofiltration is the membrane process where 2D materials are most commonly used. In 2004, graphene was the first 2D material discovered, becoming the most representative of 2D materials ^[25][26][25,26]. The possibility of cheaper synthesis of large-area graphene sheets makes this material very interesting for large-scale applications [27]. Furthermore, graphene has shown great potential as filler for composite materials, giving it outstanding mechanical and thermal properties ^[28][29][28,29]. The early successes of graphene have inspired much research on the use of 2D materials for next-generation membranes [13]. As a result, a variety of new 2D materials have been discovered. Examples include 2D boron nitride (BN), 2D transition metal dicalcogenides (e.g., molybdenum disulfide (MoS₂), tungsten chalcogenides (WS₂)), monoelemental 2D materials (such as borophene or phosphorene) and carbides or transition metal carbonitrides (MXenes) ^{[4][30][31][32][33]}[4,30,31,32,33]. Some of them (MoS₂, BN and MXenes) were already extensively studied in membrane technologies for water purification. Two-dimensional transition metal dichalcogenides (TMDCs), in recent years, were mainly investigated as an alternative route to replace the metal oxide-based photocatalyst, thanks to their suitable bandgap (1.6–2.03 eV) for the collection of visible light. Bidimensional TMDCs are endowed with high mobility of charge carriers which provide large active surface for dye degradation ^{[34][35][36][37]}[34,35,36,37].

1.1.1. Graphene

Graphene is a single layer of carbon atoms arranged in a flat honeycomb lattice. Over the years, graphene and its derivatives have found application in many fields, from water treatment to electronic and optics, due to its peculiarity such as excellent electronic conductivity, hydrophobic and anti-wetting nature, anti-fouling properties, great mechanical strength, low density and selective absorption of water vapors [38]. The introduction of graphene in membrane science is a promising approach to support sustainable growth in industrial development and research [39]. Few-layers graphene membranes exhibit ideal characteristics for the membrane contactor process, such as narrow pore size, high performance in terms of permeability and selectivity ^[40][41][42][40,41,42]. Generally, graphene is produced via exfoliation of graphite sheets where the particles of uncontaminated or expanded graphite (obtained from the thermal expansion of graphite intercalation compounds) are first dispersed in a solvent and subsequently, through an applied driving force (ultrasound, electric field or cutting), the graphite exfoliates into high quality graphene sheets. Dispersion in a solvent is necessary to reduce the strength of the van der Waals attraction between the graphene layers. A drawback of this process is the necessity to remove the non-exfoliated graphite residue. On the contrary, the advantages of this process are very the high scalability and the low cost of liquid phase exfoliation, which make it suitable for the production of graphene in large quantities. The exfoliation technique is fundamental for the quality of graphene in terms of the number of layers, lateral dimension and thickness of the graphene flakes. Various techniques for graphene exfoliation are reported in the literature such as mechanical exfoliation ^[43][44][43,44], epitaxial graphene in silicon carbide (SiC) [45], direct growth of graphene on thin nickel [46], chemical vapor deposition (CVD) ^[47][48][47,48], etc. The most used method is liquid phase exfoliation (LPE), which (in a liquid environment) allows ultrasound or shear forces to break the van der Waals bond between the graphite layers, thereby causing the exfoliation of graphite [49]. In recent years, an innovative and advanced technique known as wet jet milling (MJW) was developed for high-quality graphene exfoliation. In particular, this technique allows to prepare grapheme films with well-controlled thickness and lateral size in an easy and scalable way [44]. When chemical oxidation occurs on the graphene, several oxygen-containing groups, such as carboxylic, hydroxyl, and epoxide functional groups are formed on the carbon surface. This is called graphene oxide (GO), which is usually prepared using the Hummers method [50]. This method consists of the addition of potassium permanganate to a solution of graphite, sodium nitrate, and sulfuric acid. GO is a nonstoichiometric compound that varies in compositions depending on the synthesis conditions (oxidizing agent, solvents, time, exfoliation, etc.). The exact composition and structure of chemically synthesized GO are still being discussed [51]. Li et al. [52] prepared graphene membranes to study the ripple of GO nanosheets and found that GO nanosheets could be successfully used in the nanofiltration process with a 67% rejection rate for direct yellow (i.e., a dye molecule that is often used as a model molecule in the characterization of separation membranes).

1.1.2. Transition Metal Dichalcogenides

Transition metal dichalcogenides are layered metal oxides classified as two-dimensional (2D) materials. Their intrinsic properties make them interesting materials for energy storage [13], catalysis [53], sensors [15] and biomedical applications [54]. Two-dimensional transition metal dichalcogenides include different materials such as molybdenum sulphide (MoS₂), tungsten disulfide (WS₂), MXene, boron nitride (BN), carbon nitride g-(C₃N₄), metal–organic framework (MOF) and covalent–organic framework (COF) [55]. Recently, single-layer or few-layer TMDs were used in the field of membranes due to their excellent molecular sieve [56] and high flux and antifouling proprieties in water treatment. They have strong covalent bonds in the plane and weak van der Waals bonds outside the plane [57]. TMD nanosheets, depending on the coordination environment of the transition metal and chalcogen, can be found in the two different polymorphic states. These two polymorphs are the stable 2H trigonal prism polymorph (semiconducting) and the 1T metastable octahedral polymorph (metallic) [58]. Lithium interception predominantly produces the 1T phase, but exfoliation into the liquid phase produces the 2H phase. The use of 2D materials in membranes has shown promising results in gas separation processes, especially for MXene membranes [59], demonstrating high separation performance. In the case of layered TMD, there are limited reports on the use of MoS₂ as an additive for the preparation of composite membrane for gas separation. For example, Coleman et al. [60] prepared MoS₂ membranes with different thicknesses after exfoliation with butyllithium, which showed an H₂/CO₂ separation factor of 4.4–3.4 and an H₂ permeability in the range of 2454–27,400 GPU [60]. Chari and Eswaramoorthy prepared MoS₂ nanosheets [61] in 1T phase using butyllithium and converted them to 2H phase by heating the membrane [62]. The 2H phase showed better perselectivity than the 1T phase (H₂/CO₂ selectivity of 6.2 and 8.2 with an H₂ permeability of 1740 and 1175 barrers, respectively).

1.2. Exfoliation of 2D Materials

Exfoliation consists of separating the individual layers more or less regularly to obtain materials with few layers. Nanofabrication methods can be classified into two groups: top-down and bottom-up approach ^[63][64][63,64] (Figure 1). The top-down method allows for the formation of 2D or nanometric materials starting from the raw material, reducing its size until the desired nanometric particle size is reached. Although top-down approaches are relatively simple to use, they are unsuitable for producing irregularly shaped and extremely small particles. The difficulty in obtaining the correct particle size and shape is the main drawback of this approach. The top-down method can be achieved through several techniques including tape exfoliation, metal exfoliation, intercalation and liquid-mediated exfoliation ^[65][66][65,66]. The exfoliation of layered bulk crystals into single- or few-layer sheets is generally achieved using chemical/physical driving forces to overcome the weak van der Waals interactions between the layers. Tape exfoliation of thin metal film is a very simple technique which mechanically strips loose crystals and produces single crystal flakes [44]. The single crystal flakes produced are often of high purity and clean lines, mainly suitable for the production of main devices and the characterization of new 2D systems. Regarding intercalation strategies, ionic species are interdispersed in the spaces between the layers. Intercalation creates expansion forces through the reaction and allows for the separation of layers in the liquid. Ultrasound, or liquid-mediated exfoliation, is used to produce a high local temperature and extreme pressure, which tends to break the bonds between the various layers [43]. Liquid phase preparations are considered more suitable for obtaining large production, although submicrometric scales with high defects and modified chemical or physical properties are often obtained. Monolayers prepared via fluid mediation are particularly useful for catalysis or other chemical applications [38]. The bottom-up approach is based on the direct construction of 2D nanomaterials with precursor atoms via the growth or self-assembly of atoms and molecules. The nanomaterials (NMs) produced using these techniques have a well-defined shape, size and chemical composition depending on the constituent elements. During deposition, nucleation at different local positions leads to continuous growth of individual domains, which subsequently coalesce into a continuous monolayer thin film constituting a polycrystalline film [67]. With tuned growth conditions, isolated single-crystalline flakes, down to a few hundred µm in the lateral dimension, can be obtained. Controlling the number of layers through targeted growth is particularly difficult due to the high barrier of nucleation [68]. Figure 2 illustrates the main exfoliation techniques. The illustrated techniques were utilized especially for graphene production. In some cases, several techniques were also utilized with other materials (such as carbon-based nanomaterials, silicate clays, TMDs, MOFs layered double hydroxides (LDHs), titanium dioxide (TiO₂), Tetramethyl orthosilicate (TMOs), BN, black phosphorous (BP), etc.) ^[69][70][71,72]. High production costs are still the major barrier hindering commercial adoption. For example, in the case of graphene, on average, the cost is USD 550 per square centimeter; however, it could also drop down to between USD 100 and USD 200 when demand is expected to increase ^[71][72][73,74].