High-performance robust 2D graphitic carbon nitride nanosheets embedded membranes for desalination through direct contact membrane distillation

doi:10.21203/rs.3.rs-2497957/v1

Download PDF

Research Article

High-performance robust 2D graphitic carbon nitride nanosheets embedded membranes for desalination through direct contact membrane distillation

https://doi.org/10.21203/rs.3.rs-2497957/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Two-dimensional (2D) graphitic carbon nitride (g-C₃N₄)-based membranes have received widespread attention in the last ten years due to their fantastic separation performance. In the present investigation, for the first time, we tried to assess the desalination performance of membranes incorporated with g-C₃N₄ nanosheets through the membrane distillation (MD) process. for this purpose, different concentrations (0-0.06 wt.%) of synthesized nanosheets were embedded as fillers within the hydrophobic polyvinylidene fluoride (PVDF) matrix by using simple phase inversion method. Various techniques such as FTIR, XRD, SEM, AFM, water contact angle and LEP, tensile strength measurements as well as direct contact membrane distillation (DCMD) experiments were utilized to study the effect of the g-C₃N₄ contents on the elemental properties, crystalline structure, morphology, topography, surface hydrophobicity, mechanical resistance, and desalination performance of the membranes. The surface roughness parameters suggest that the membrane surface became even rougher by integrating g-C₃N₄ nanosheets into the membrane matrix. Upon loading with a small amount of 0.03 wt.% g-C₃N₄ in the polymer network, the water contact angle and LEP augmented from circa 80.5º and 6 bar for the pure PVDF membrane to 95.2º and 8 bar, respectively. Adding the g-C₃N₄ nanosheets also improved the mechanical properties of the PVDF membrane; so that, both the tensile strength and the elongation at break increased to 21.32% and 36.46% in comparison to the pristine PVDF membrane, respectively. The prepared mixed matrix membranes favorably improved MD performance; so that, 0.03 wt.% g-C₃N₄ embedded membrane exhibited a flux of 27.63 kg/m²h with an enhancement of 70% relative to its pure counterpart, despite having a trace amount of the nanosheets. Continuous testing for up to 24 h filtration of a 3.5 wt.% NaCl solution still showed a stable flux and almost complete salt rejection (i.e., 100%) for the mentioned mixed matrix membrane. This work holds promise for developing next-generation MD membranes with superior desalination performance in terms of water vapor permeability and salt rejection.

Graphitic carbon nitride nanosheets

Polyvinylidene fluoride

Membrane distillation

Desalination performance

Water scarcity is an urgent issue that has afflicted people throughout the world. Based on the United Nations' prediction, almost 6 billion people will encounter this scarcity by the middle of the century [1, 2]. Civilization expansion and world's population growth, agricultural development, industrialization, and global warming are the significant factors that have irrefutably escalated the freshwater shortage and have augmented the demand for it in the last two decades [1, 3, 4]. To overcome this challenge, not only management of water consumption is required, but also more water resources is essential. In this direction, unconventional water resources like seawater, industrial wastewater, and municipal brine have been attracting the attention of scientists and experts in this field. However, the ubiquity of seawater has made its desalination a convenient option for addressing the water supply challenge [1, 5].

The desalination process is broadly categorized as thermal and membrane-based technologies [2]. Over the years, the most conventional desalination technologies were reverse osmosis (RO), multi-effect distillation (MED), and multi-stage flash (MSF) distillation [4]. At the moment, membrane-based processes have surpassed thermal processes due to the raising awareness on energy saving topic [1, 2]. In addition, considerable amounts of electrical energy are required to conduct the large-scale separation with RO and commonly rely on non-renewable energy sources. This in itself can increase maintenance and investment costs. Consequently, alternative technologies which have the capability of performing efficiently but in a more stable way are presently being surveyed [6]. On the other hand, the merging of thermal distillation and membrane separation has presented a new technique which has the characteristics of both techniques, called as membrane distillation (MD), and seems to be a suitable option for desalination [3, 6]. Meanwhile, the thermal energy required for MD plants can be provided by utilizing available waste heat from industrial processes or low-grade energies such as solar, geothermal, and other renewable energy sources [4, 6–8].

MD was first patented by Bodell in 1963. Four years later, the first MD paper was published by Findley [9]. This technology has been put on the path of progress in the early 1980s with the appearance of novel membranes and modules with upgraded specifications [6]. Despite the slow start, both theoretical and experimental research on MD, particularly on desalination, have intensified significantly during the last few decades [1]. MD is a promising thermally-driven membrane filtration process at which only the vapor molecules from an aqueous feed solution pass through a hydrophobic microporous membrane and then condensate at the other side of the membrane, resulting in a highly pure distillate and a concentrated feed solution with non-volatile solutes [4, 7]. The mass transfer driving force in the MD process is the transmembrane vapor pressure difference, which arising from the transmembrane temperature difference or by the reduction of vapor pressure on the permeate side using vacuum or dry gas according to the used configuration [7, 10]. In the MD literature, it is easy to see that among the various configurations, the direct contact membrane distillation (DCMD) mode is the most reviewed [9, 11, 12]. This may be because its ease of handling since the condensation step is actualized inside the membrane module [6, 9].

MD has many inimitable benefits compared with the prevalent separation processes that the most important of which are: (1) desalting feed solutions of high salinity; (2) operating at much lower hydrostatic pressure from that utilized in pressure-driven membrane processes; (3) operating at much lower temperature than the boiling point of the solvent in comparison to those encountered in conventional distillation; (4) achieving a 100% theoretical rejection of solutions containing non-volatile solutes; and (5) low propensity for fouling due to the relatively larger membrane pore size [1, 6, 7, 9, 10]. Nevertheless, MD suffers from the absence of specially-designed economical membranes, and this is one of the main reasons why it has not yet been fully developed from the industrial implementation point of view despite the great potential for industrial applications [9, 13–15]. Although commercial hydrophobic microfiltration (MF) membranes can be employed for MD applications, the wetting phenomenon and sub-optimal flux performance are two obstacles they still face [13, 14]. Thus, to overcome these drawbacks, developing high-performance robust membranes special for the MD process with high level of hydrophobicity and porosity, appropriate pore size, narrow pore size distribution, and favorable chemical stability is of great importance [7, 16–18].

The membrane materials in the MD process must be hydrophobic to avoid pore wettability. The most common MD membrane materials are polyvinylidene fluoride (PVDF), polytetrafluoroethylene (PTFE), and polypropylene (PP), owing to their instinctive hydrophobicity [19]. Among them, PVDF, due to the dissolvability in common organic solvents, low thermal conductivity, high thermal stability, and good chemical and mechanical resistance, has demonstrated widespread practical applications [20]. Moreover, the two techniques comprising non-solvent induced phase separation (NIPS) and thermally induced phase separation (TIPS) are prevalently used to make MD membranes [21]. Although studies of the performance of membranes prepared with PVDF as a base polymer through phase inversion technique have confirmed their acceptable compatibility with MD systems, further improvement should be made to enhance the efficiency of MD. Over the past years, many novel techniques have been proposed for this purpose, including the usage of novel hydrophobic (co)polymers, electrospinning, and surface modification procedures [22].

The advent of nanotechnology has brought about new chances to promote water desalination technologies, which has attracted much attention recently [23–25]. Accordingly, MD, as one of the efficient technologies in the field of desalination, has not been deprived of its advantages. Up to now, there have been significant efforts in the development of polymeric nanocomposite membranes by using various nanomaterials in the form of nanoparticles, nanosheets and nanotubes as well as nanofibers toward MD having prominent features relative to their pristine counterparts and particular interest has been devoted to the nanocarbons [1, 8]. So that, the use of carbon-based nanomaterials such as carbon nanotube (CNT) [26–31], graphene [32, 33], or graphene-based ones [16–18, 23, 34–37] into the polymeric membrane matrix is promising as innovative MD membranes with optimal performance from the viewpoint of water vapor permeability and salt rejection [1, 2, 25, 38]. For instance, Roy et al. [27] were able to achieve higher fluxes in DCMD through immobilizing functionalized CNTs on porous polypropylene supports. In 2015, Silva and his coworkers [31] blended PVDF with multi-walled carbon nanotubes (MWCNTs) through the phase inversion strategy for DCMD of salty water. The fabricated MWCNT/PVDF membrane with optimum content of 0.2 wt.% MWCNTs represented complete salt removal and a greater permeate flux of 34.2 kg/m²h than that acquired with a commercial PVDF membrane (28.08 kg/m²h). Woo et al. [33] incorporated various concentrations of graphene into the electrospun poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP) membrane to increase the hydrophobicity and water flux in air gap membrane distillation (AGMD). Nowadays, some hydrophobically functionalized GO nanosheets have been successfully prepared and introduced within the hydrophobic membrane matrix by phase inversion for MD applications, such as 3-(aminopropyl)triethoxysilane (APTS)-GO [37], octadecylamine (ODA)-GO [23], n-butylamine (NBA)-GO [34], and reduced GO [16, 36]. These nano-structured materials can ensure the promotion of new generation desalination membranes with a remarkable performance for developing technology of MD [23].

In the same regard but from a different viewpoint, membranes based on 2D nanomaterials have exhibited exceptional properties in membrane separation, which has overcome the restrictions encountered by existing polymer membranes [39–41]. Graphene and its derivatives have paved the way for separation with 2D nanomaterials-based membranes [42–45]. Other 2D structures, like the MXenes, have been the object of new investigations in the context of membrane separation [46]. Recently, a new type of carbonaceous 2D nanomaterial called graphitic carbon nitride (g-C₃N₄) has revolutionized the construction and development of state-of-the-art water treatment membranes with highly desirable performance [25, 39, 47].

g-C₃N₄ is a 2D metal-free semiconductor polymeric nanomaterial that has excited much interest in the last few years due to its convenient, inexpensive, and green synthesis procedures with readily available precursors, non-toxic nature, large specific surface area, outstanding thermal, chemical, and mechanical stability, visible-light-driven photocatalytic activity, and (photo)electrochemical properties [25, 38, 39, 48–52]. The history of the genesis of g-C₃N₄ can be traced back to 1834, when Berzelius and Liebig first reported it as one of the oldest synthetic polymers and named “melon” [53], however, its potential application has not been wholly identified until Wang and his colleagues [54] synthesized g-C₃N₄ as a photocatalytic agent for water splitting to produce hydrogen under light irradiation in 2009 [39, 49, 50].

Studies conducted on the filtration performance of membranes based on g-C₃N₄ nanomaterial via MF, ultrafiltration (UF), nanofiltration (NF), reverse and forward osmosis (FO), gas separation, pervaporation, and ion exchange membranes reveal a high selectivity towards a wide range of contaminants, high permeability, and membrane fouling resistance by taking advantage of its photocatalytic properties during long-term filtration operation [38, 39, 47, 55].

g-C₃N₄ based hybrid membranes are also superior candidates for high-efficiency water desalination applications by NF [56], RO [57–60], and FO [61, 62]. Because of the laminar structure and the existence of nanopores (diameter of 1.48 nm), g-C₃N₄ nanosheets can provide channels for water transport and act as a molecular-sieving in desalination [50]. Seyyed Shahabi et al. [57] have fabricated the modified thin film nanocomposite reverse osmosis (TFN-RO) membranes by incorporating the synthesized g-C₃N₄ nanosheets into the polyamide (PA) active layer. The modified membrane showed improved pure water flux, NaCl rejection, and antifouling performance. To further enhance the permeance, Seyyed Shahabi et al. [58] have fabricated functionalized g-C₃N₄ nanosheets with –OH, –COOH, and –SO₃H groups and then introduced them into the PA layer of TFN-RO membranes. The resultant RO membrane modified with COOH-g-C₃N₄ nanosheets achieved a maximum pure water flux of 91.8 L/m²h with a 54% increase than that of the bare TFN-RO membrane without a decrease in salt rejection. In addition, dissociation of the functional groups in the water produced more negative charges on the surface of the membrane, which was fruitful in enhancing the membrane resistance against fouling. Gao et al. [60] have prepared acidified g-C₃N₄ and then embedded it into the PA selective layer. Therefore, the dispersion feature of acidified g-C₃N₄ was improved, which caused to alleviate the agglomeration of nanosheets and offered extra pathways for water transport. In conclusion, the optimized TFN membrane exhibited a 79.3% increase in permeate flux relative to the original membrane. Moreover, the introduction of acidified g-C₃N₄ generated a more hydrophilic and negatively charged membrane surface, and thus resulted in an improved antifouling performance. Poor anti-chlorine character is one of the main challenges for RO desalination membranes. Chlorine can break amide bonds and hence, reduce water flux and salt rejection rate. To improve the chlorine resistance of membranes in RO, Xu and his coworkers [59] have doped PA hybrid membranes with oxidized g-C₃N₄ as an antioxidant. The results showed that the salt rejection of the bare membrane decreased by 27% after soaking for 40 h in sodium hypochlorite solution. In contrast, the salt rejection of PA-oxidized g-C₃N₄ membranes only declined by 15%. This can be attributed to the increase in N$-$H bonds and oxygen-containing functional groups on the modified RO membranes.

With all these interpretations, membrane based on g-C₃N₄ is still absent in other typical membrane-based separation processes, including membrane distillation and electrodialysis; thus, there is yet room for exploration in these fields in academic sectors [47]. Therefore, in this study, we tried to unveil g-C₃N₄ as a 2D nanofiller with outstanding characteristics to improve the performance of the MD process. To the best of the authors' knowledge, this is the first time that g-C₃N₄ has been used to fabricate MD membranes. For this purpose, the g-C₃N₄ nanosheets have been synthesized, characterized, and then incorporated into the PVDF scaffold with various concentrations. The resultant mixed matrix membranes have been ultimately employed to desalinate aqueous solutions containing 35 g/L of sodium chloride via DCMD configuration.

2.1. Materials

PVDF powder (Kynar® 761 from Arkema) as the base polymer of the membrane matrix and N,N-dimethylformamide (DMF, ≥ 99.8%, Merck) as the solvent were used to prepare MD membranes. In order to synthesize the nanosheets, 1,3,5-triazine-2,4,6-triamine (melamine, 99%) was supplied from Khorasan Petrochemical Complex (KPC, Iran) and used as received without further purification. The performance of the prepared membranes was tested by sodium chloride (NaCl, > 99%, Dr.Mojallali) solution. Distilled water was applied all over this research.

2.2. Synthesis of g-C₃N₄ nanosheets

Thermal decomposition is a very common technique for the synthesis of graphite-like C₃N₄, in which the nitrogen-rich precursors such as melamine, urea, cyanamide or dicyandiamide, thiourea, and so on polymerized when heated up to 600 ºC [25, 39, 47–49, 51, 63]. In this research, the g-C₃N₄ nanosheets were prepared by a two-step thermal decomposition of melamine [57, 64]. In a typical method, a certain amount of melamine was poured into a closed crucible and sequentially heated in a homemade furnace to the target temperatures (i.e., 380 and 600 ºC) for one and two hours, respectively. The as-prepared bulk g-C₃N₄ was cooled to ambient temperature and finely ground into powder by a mortar. Eventually, the pale-yellow powder of g-C₃N₄ was gathered and denoted as gCN nanosheets.

2.3. Preparation of gCN embedded membranes

A series of asymmetric porous gCN embedded membranes were fabricated via the NIPS technique. The loading degree of gCN nanosheets according to the total weight of the solution was in the range of 0.015 to 0.06 wt.% (i.e., 0.015, 0.03, 0.045, and 0.06 wt.%). Applying the carbon-based nanomaterials in low concentrations for the preparation of the membranes without structural defects and nanoparticles agglomeration is a topic that has been emphasized by many literatures [57, 65–67]. First, precise amounts of the nanosheets were dispersed in DMF solvent using a probe sonicator (Hielscher UP400S, Germany) for 16 min to ensure uniform distribution of gCN, followed by the addition of PVDF powder (16 wt.%) batch-wise. Whereas, the PVDF powder was dried in an oven at 50 ºC for 24 h to remove its moisture content prior to use for the dope preparation. The dope solutions were stirred for 24 h at 45 ºC to form the homogeneous colloidal solutions. Afterward, the resultant polymeric solutions were kept standing in an oven at 30 ºC for 24 h to eliminate the trapped air bubbles. The as-prepared solutions were then cast on non-woven fabric polyester by a casting knife, set at a gate height of 200 µm. After being subjected to air for 5 s, the nascent films were immersed smoothly into a precipitation medium containing distilled water as the non-solvent at 25 ºC for 30 min. The flat-sheet membranes were then transferred into the fresh distilled water bath and stored for at least one day to ensure leaching out the leftover organic solvent. In the end, the membranes were dried between two sheets of filter paper for 24 h at room temperature. As a reference, the pure PVDF membrane was prepared without any additional gCN.

2.4. Characterization

The functional groups present on the surface of the prepared gCN nanosheets and flat-sheet membranes were appraised by attenuated total reflectance-Fourier transform infrared (ATR-FTIR, Thermo Nicolet Avatar, America) spectroscopy. The IR spectra were recorded over the wavenumber range 500–4000 cm^− 1 with a signal resolution of 1 cm^− 1. In order to identify the crystalline structure of the resultant gCN sheets and membranes, X-ray diffraction (XRD, Explorer, DNR, Italy) patterns were logged at 2θ range of 5–60° (with Cu Kα radiation, λ = 1.541874 Å). The top surface and cross-section morphologies of the fabricated gCN nanosheets and membranes were examined using a scanning electron microscope (SEM, Leo 1450VP, Zeiss) at different magnifications. Prior to SEM analysis, all specimens were coated with a thin layer of gold using a sputter coater (SC7620, Polaron) and by applying a current intensity of 5 mA for 180 s. The images were also captured at an accelerating voltage of 20 kV. Meanwhile, energy dispersive spectroscopy (EDS, Oxford) as an analytical method was applied to specify the elemental composition of the synthesized nanosheets. The thickness of the support and the resultant membranes was measured by a digital outside micrometer and then verified with SEM cross-section analysis. The transmission electron microscopy (TEM) images of gCN nanosheets were taken by a Leo 912AB (Zeiss) microscope. The desired sample was dispersed in water using an ultrasonic bath to gain a TEM image. Furthermore, the electron beam accelerating voltage equal to 120 kV was applied within an hour. Atomic force microscopy (AFM, Easyscan2 flex, Switzerland) was used to study the surface topography and roughness of the produced pristine and modified PVDF membranes. The AFM apparatus was Nanosurf scanning probe-optical microscope and the membrane surface images were obtained in a scan area of 10 µm $\times$ 10 µm. The surface roughness parameters of the membranes are introduced in the form of the mean roughness (S_a), the root mean square of the Z data (S_q), and the mean difference between the five highest peaks and lowest valleys (S_z), which were computed using DME SPM analysis software by tapping mode procedure and were applied to inspect the membranes morphology. In addition, the pore size distribution (PSD) of the resultant membranes was extracted from the elevation profile of the AFM images applying the Nanosurf report and the average pore size of them was specified from surface PSD graphs. The mechanical properties of the as-prepared membranes were estimated by a tensile test machine (SANTAM CO., Iran) at a crosshead speed of 10 mm/min for rectangular samples with an initial length of 7 cm and width of 1 cm. Three specimens were tested for each membrane sample and the average amount was presented. To assess the membranes surface hydrophobicity, water contact angle analysis was accomplished with an optical tensiometer (ADICO, Iran) using the sessile drop method. The volume of water droplets as the testing liquid used for the static contact angles measurement was 4 µl. To keep at minimum the experimental error, five random spots were considered for measuring the contact angle of each sample and finally, a valid value was reported.

Liquid entry pressure of water (LEPw) is a parameter representing the resistance of the membrane against wetting. In other words, LEPw is interpreted as the least amount of hydrostatic pressure at which the liquid feed enters inside the pores of the hydrophobic membrane. Based on the guidelines described by Smolders and Franken [68], LEPw measurement was took place through a homemade dead-end filtration setup. The samples of pure and mixed matrix membranes with an effective area of 38.5 cm² were housed in the steel cell that included two feed and permeate chambers. The feed chamber was filled with 200 ml of distilled water and linked to a cylinder containing nitrogen gas. At first, a pressure equal to 0.5 bar was applied on the feed side. Then, the pressure was elevated stepwise with intervals of 0.5 bar after a ten-minute stabilization time. As soon as the first drop of water appeared on the permeate side, the corresponding pressure amount was noted down as the LEPw.

In order to designate the porosity of the membranes, the gravimetric procedure was applied [69]. Square pieces of membranes with specific sizes were dried in an oven at 60 ºC for 24 h and weighed precisely. The membrane samples were then immersed in ethanol for 24 h and weighed again after wiping by a blotting paper. The porosity (ε) of the respective membranes was calculated using Eq. (1) [69]:

$$\epsilon \left(\%\right)=\frac{\left({m}_{2}-{m}_{1}\right){\rho }_{1}}{{\rho }_{1}{m}_{2}+\left({\rho }_{2}-{\rho }_{1}\right){m}_{1}}\times 100 \left(1\right)$$

where m₁ and m₂ are the dry and wet membrane weights, respectively. ρ₁ and ρ₂ refer to the density of PVDF (1.78 g/cm³) and ethanol (0.789 g/cm³), respectively. The membranes porosity was measured three times in different spots of the same sample, and the average amount has been reported.

2.5. DCMD experiments

The desalination performance of the prepared pristine and gCN embedded PVDF membranes was evaluated through a homemade DCMD test setup. Figure 1 demonstrates a schematic representation of the setup used in this investigation. The membrane module made of Plexiglas with an effective area of 36.92 cm² has composed of two chambers, feed (top side) and distillate (bottom side). The hot synthetic feed (i.e., 35 g/L NaCl aqueous solution) is passed on the top side of the module and the coolant (i.e., distilled water with electrical conductivity$<$ 3 µS/cm) from the bottom side in counter-current mode. The circulation of feed and coolant into the membrane module was fulfilled using two diaphragm pumps (JET-50, HEADON, Taiwan). The saline feed was heated via a temperature-controllable oil bath to reach a certain temperature of 68 ºC. In comparison, the resultant permeate was maintained at a constant temperature of 23 ºC during the operation by passing through a chiller circulator. During the experiments, the volumetric flow rates of feed and permeate were respectively adjusted at 400 and 200 ml/min using two flowmeters. A bypass valve was mounted to regulate the feed pressure on the membrane surface. A pressure gauge was also used to measure the surface pressure of the membranes. The system was further equipped with a digital balance (AND, d = 0.1 g, Japan) to measure the weight increment of coolant water because of the permeate collected from the tested MD membranes. During the MD test, the electrical permeate conductivity has also been recorded by using a portable online conductivity meter (AZ 8306, China). The whole experiments were performed under the conditions mentioned above and each experiment took a long time about 300 min. The water vapor permeation flux and salt rejection quantities were calculated according to the equations below:

$$Flux=\frac{M}{A\times t} \left(2\right)$$

$$R \left(\%\right)=\left(1-\frac{{C}_{P}}{{C}_{F}}\right)\times 100 \left(3\right)$$

where flux (Kg/m²h) is stated as the value of permeated water M (Kg) per unit effective area of the membrane A (m²) per unit time of operation t (h). In addition, R, C_F, and C_P correspond with the salt rejection and the feed and permeate concentrations, respectively.

"The location of Fig. 1"

3.1. Characterization of gCN nanosheets

To corroborate the formation of 2D gCN nanosheets, the functional groups, crystalline structure, and morphology of the synthesized nanosheets were conducted using FTIR, XRD, SEM, and TEM analysis, respectively.

The chemical nature of different functional groups present on the surface of the gCN sample was explored through FTIR analysis (Fig. 2a). The spectrum of the as-synthesized gCN illustrates several characteristic infrared absorption bands that are in coordination with those of the reported spectra by many other publications [56, 57, 59–61, 67, 70–78]. The sharp absorption peak, known as a specific characteristic band of gCN, was recorded at 812 cm^− 1, which represents the ring-sextant out-of-plane bending vibration of heptazine rings. In addition, the absorption band at 890 cm^− 1 is related to the deformation mode of N$-$H in amino groups [56, 57, 59–61, 72, 73, 76]. The absorption bands in the 1200–1700 cm^− 1 region correspond to the typical stretching patterns of aromatic heterocyclic C$-$N and C$=$N in gCN structure [56, 57, 59, 67, 70–72, 76–78]. With a more detailed review on the peaks of the mentioned region, it can be found that the peaks at 1237 and 1316 cm⁻¹ are allocated to the stretching vibration of interconnected trigonal units of C$-$N($-$C)$-$C (full condensation) or bridging C$-$NH$-$C (partial condensation) [60, 70, 75]. Moreover, the peaks at 1407, 1462, 1575, and 1641 cm^− 1 are ascribed to the typical stretching vibration modes of heptazine-derived repeating units [70, 76]. On the other hand, as narrated in some references [57, 74], the aromatic C$-$N stretching bond of tri-s-triazine is identified by the peaks at the wavenumbers of 1237, 1316, 1407, and 1462 cm^− 1. The existence of the C$=$N heterocycle stretching vibration bond is also proved by the peaks appearing in the vicinity of 1540, 1575, and 1641 cm^− 1 wavenumbers. Finally, the broad absorption bands in the range of 3000–3350 cm⁻¹ are derived from the stretching vibration modes of the terminal N$-$H band, which is at the deficiency sites of the aromatic ring encompassing the residual $-$NH or $-$NH₂ groups [70, 73, 74, 76, 78]. It is worth mentioning that such primary and secondary amines arise from the incomplete condensation of melamine precursor [52, 56, 72]. Furthermore, a group of researchers [57, 60, 67, 72] has pointed out the presence of the O$-$H stretching vibration modes in this region due to the physically adsorbed water molecules.

Depending on the synthesis method and the precursor type, diverse allotropes of carbon nitride with different properties can be obtained. Due to its unique physicochemical properties, gCN is known as the most energetically stable allotrope in ambient conditions [38, 48–50]. Generally, the configuration of gCN has been constituted of carbon and nitrogen atoms, which are covalently linked by sp² hybrid orbitals [48, 50]. The periodically connected tri-s-triazine units have also formed the whole structural framework of gCN. Furthermore, the layered material has been contained of a lattice structure analogous to graphite, which are coupled with weak van der Waals interactions. Within the layers, nitrogen atoms alternate instead of carbon atoms in a hexagonal arrangement [39, 49, 50]. The structural formula of gCN displays many intrinsic triangular nanopores and defective pore structures (see the inset in Fig. 2a).

In order to inquire about the crystalline structure of the as-synthesized gCN nanosheets, XRD analysis was performed and the resultant pattern is represented in Fig. 2b. The diffractogram displays two typical diffraction peaks at around 13.08º and 27.68º of 2θ, indicating the successful synthesis of the gCN by thermal polymerization from melamine precursor. In detail, the weak reflection peak located at 2θ = 13.08º is attributed to the (100) in-planar consecutive motif of tri-s-triazine polymeric building units. Meanwhile, the strong peak with high intensity at 2θ = 27.68º corresponding to the (002) crystallographic planes of graphitic materials, which originates from the presence of the interlayer periodic stacking of the conjugated aromatic graphite-like framework in gCN nanosheets. These values are in accordance with JCPDS No. 87-1526 [56, 57, 72–74, 77, 79, 80].

The morphological properties of the synthesized nanosheets together with EDS analysis are presented in Fig. 2. The images of SEM and TEM of gCN are demonstrated in Fig. 2c and d, respectively. As can be observed from the SEM images, a sheet-like lamellar structure with irregular wrinkles and ripples has been the result of the nanomaterial synthesis. The TEM images, however, can provide detailed descriptions concerning the morphology of the nanosheets. They exhibit a transparent graphene-like 2D morphology composed of poly-crystallized texture with a size of about several hundred nanometers. Figure 2e shows the EDS spectrum of the gCN nanosheets and confirms the presence of carbon and nitrogen elements without any impurities.

"The location of Fig. 2"

3.2. Membrane characterization

Figure 3a shows the digital photo of different dope solutions prepared in this study for membrane fabrication after standing for 24 h at 30 ºC. Compared with the PVDF clear solution (vial number 1), adding gCN nanofillers caused pale-yellow coloring opaque solutions (vials number 2–5), indicating that an even mixing as well as sustainable solutions were being achieved. Moreover, from Fig. 3a, it can be evidently observed that the color of the colloidal solutions is gradually getting darker as the amount of gCN is increased.

ATR-FTIR and XRD were used to characterize the chemical composition and crystalline phase structure of the as-prepared membranes, respectively. The typical spectra of the pristine PVDF membrane and membranes with various contents of gCN nanosheets are depicted in Fig. 3(b and c).

The FTIR spectrum of PVDF makes it possible to discern between the different crystalline forms. Besides, the same film of PVDF can contain one or more crystalline phase structures depending on the preparation conditions. In the FTIR spectrum of pure PVDF (Fig. 3b), the peak at 878 cm^− 1 is assigned to the amorphous phase. From literatures on the spectral features of the crystalline phases in PVDF, the peaks at 762, 795, and 976 cm^− 1 represent typical bands of α phase, while the absorption bands at 840 and 1277 cm^− 1 are imputed to the all-trans conformations of β phase vibration [23, 70, 76, 81]. Likewise, the characteristic peaks at 1403, (1072 and 1179), and 878 cm^− 1 are related to C$-$H stretching vibrations of $-$CH₂ groups, C$-$F stretching vibrations of $-$CF₂ groups, and C$-$C stretching vibration bond, respectively [23, 69]. Additionally, two vibrational bands and reflection peaks at the wavenumbers of 1235 and 1277 cm^− 1 can be ascribed to the stretching frequencies of C$-$F and C$-$F₂ [18]. The weak absorption bonds at 2981 and 3019 cm^− 1 could also correspond to the sp³ C$-$H symmetric and asymmetric stretch of methylene groups of PVDF [69, 70].

As can be seen from the curves, all the gCN incorporated membranes exhibit similar IR spectra to the pure PVDF membrane without any new peaks appearing related to gCN. There could be several reasons for this issue, including: (1) The amount of gCN powder embedded in the polymer matrix is not enough to be detected. In other words, adding a trace amount of nanosheets does not change the surface chemical structure of the PVDF membrane [18, 71]; (2) The sheets are covered by or buried within the PVDF polymer matrix; and (3) The nanosheets are mixed with PVDF physically or by weak chemical bonding [32, 82].

It is worth noting that the interference and interfacial interaction between polymer and nanoparticles results in a slight diminution of bond intensity in IR spectra for fabricated mixed matrix membranes. This decline of transmittance intensity becomes bolder with the enhancement of nanofiller content as a consequence of the increased interfacial influence [82]. These results indicate a good compatibility between gCN nanosheets and PVDF molecules [18].

Figure 3c illustrates XRD patterns of the membranes embedded with 0.03 and 0.06 wt.% of gCN nanosheets compared to the pristine PVDF membrane. As is well known, PVDF is a semi-crystalline polymer, which has two main crystalline phases, namely nonpolar α and polar β forms [70, 83]. For pristine PVDF membrane, two characteristic diffraction peaks at angles (2θ) of 18.04º and 26.28º represent the crystal plane reflections of (020) and (021), respectively, which belong to the characteristic α-phase of PVDF. Moreover, there is a relatively small peak at 2θ = 20.4º which indicates the characteristic β-phase of PVDF [70, 76, 81]. After the addition of gCN nanosheets, the former packing of the backbone is maintained. In other words, it can be seen that the XRD patterns of the mixed matrix membranes have not changed significantly when compared to that of the pristine membrane. It should also be acknowledged that the peaks imputed to the reflections of α and β phases of the PVDF still appeared in nearly the same situations, but their intensities are lower when compared to those of the pristine PVDF membrane and decrease with increasing the concentration of nanofillers as represented in Fig. 3c [70]. In addition, the characteristic peaks of gCN are not visible in the resultant mixed matrix membranes, which is attributed to the meager amount of 2D nanofillers [67]. These observations were further confirmed by the IR spectra.

"The location of Fig. 3"

To investigate the impact of gCN on the structure of the resultant mixed matrix membranes, visual characterization under the SEM analysis was performed (Figs. 4 and 5). The surface morphology of the unmodified and the gCN modified PVDF membranes at two different magnifications are depicted by top-view SEM micrographs in Fig. 4(a-e1). All membranes exhibit a uniform and defect-free surface. The gCN nanosheets are not visible on the surface of the mixed matrix membranes owing to the low concentration as well as good dispersion of the nanosheets. Comparing the surface SEM images of the membranes indicates that adding gCN into the polymeric solution up to 0.03 wt.% leads to an increment in the number and the size of surface pores. However, it is worth noting that no tangible change was observed in the number of surface holes following the addition of 0.03 and 0.045 wt.% of gCN in the PVDF matrix. This phenomenon can be clarified by the fact as follows:

In general, thermodynamics and kinetics are two determining factors, which control the format of membranes prepared with phase inversion. The presence of polymeric nanosheets into the dope solution enhances its thermodynamic instability, which favor the progress of phase separation and increasing the rate of demixing [84].

Moreover, the number of surface pores diminishes obviously since the percentage of nanofillers raises to 0.06 wt.%. This is maybe because the relatively higher viscosity of the dope solution comprising 0.06 wt.% of gCN, which hinders the exchange rate of the solvent and non-solvent during the phase inversion process, resulting in lower membrane pore size and porosity. Generally, the addition of inorganic or carbon-based nanomaterials will increase the solution viscosity and thus hinder the solution demixing process [85].

"The location of Fig. 4"

As previously mentioned, the membranes in this study were formed via the phase inversion of the PVDF induced by the immersion precipitation (IP) method in the presence of water as the non-solvent. Phase inversion induced by IP can be considered as the key technique in the production of industrial asymmetric membranes owing to the simplicity of the process. Most of the available PVDF membranes are also fabricated via the same method due to their easy dissolution in prevalent organic solvents. Since PVDF is well-known as a semi-crystalline polymer, its behavior during the phase separation process is more complicated than amorphous polymers and is governed by two mechanisms (i.e., liquid–liquid demixing and crystallization). The coagulation bath is one of the main factors in determining the sequence of the mechanisms as mentioned above, which dictates the ultimate morphology of the PVDF membrane formed through the IP method. The presence of water (acts as the harsh non-solvent) as the coagulation medium during IP often accelerates the liquid–liquid demixing process and consequently, an asymmetric structure like ordinary amorphous membranes consisting of dense skin layer along with finger-like voids could be created [86].

Figure 5(a-e1) illustrates the cross-sectional morphology of the pristine and the gCN embedded membranes. It can easily be recognized by these SEM images that all membranes have an asymmetric structure composed of a thin compact top layer and a porous sublayer divided into two distinct regions (i.e., finger-like and sponge-like). These observed morphologies are consistent with the contents of water immersion induced structures mentioned above. As shown in Fig. 5, the incorporating of the gCN nanosheets changes the cross-section microstructure of the prepared PVDF/gCN membranes. Beneath the skin is a region composed of parallel columnar macro voids, which expand to the middle part of the membrane. Adding the relatively hydrophobic gCN nanosheets in the dope solution, turns the texture of this region into spherical-like nodules connected to each other with pores that are resided between these adjacent nodule aggregates. Moreover, the size of these spherical nodules appears to be growing over the entire membrane cross-section by increasing the content of the polymeric nanosheets. The lower half of the gCN embedded membranes cross-section also demonstrates a cellular morphology that became more porous and mixed with spherical nodules compared with pristine membrane. However, for the pure PVDF membrane, the sponge-like morphology instead of the cellular one was observed both at the bottom and among the finger-like macropores. As can be apparently seen, interconnected inner pores of the pure PVDF membrane joined to form larger voids in the 0.015 wt. % gCN embedded membrane. Moreover, by adding 0.015 wt.% of gCN nanosheets, the finger-like voids changed to pear-shaped macro voids with larger sizes and looser than those of the pure membrane. In addition, by incorporating the nanofillers, a slight decline in the skin layer thickness was detected. The finger-like structures seem to become slightly narrow, more regular, and with a rather uniform size after adding 0.03 wt.% gCN into the PVDF membranes. In the case of the 0.045 wt.% gCN embedded membrane, the finger-like structure is similar to the 0.03 wt.% counterpart but, to some extent, narrower. However, a further increase in the loading of gCN to 0.06 wt.%, increased the viscosity of the solution and, as already explained, resulted in a lag in the phase inversion kinetics (i.e., the rate of solvent-outflow and nonsolvent-inflow) during the NIPS process and the cellular structure was more pronounced than the finger-like one.

Besides, the upper layer of the finger-like zone in the resultant gCN embedded membranes was composed of small cavities. Also, the number of these cavities in the membrane incorporating 0.03 wt.% of gCN seems to be more than the other three mixed matrix membranes. It is expected that such regular and uniform finger-like voids combined with more pores underneath the top layer led to an increase in porosity and would promote the permeability of the MD membrane.

On the other hand, from higher magnification SEM images in Fig. 5(c1, d1, and e1), it can be seen that the gCN nanosheets are well embedded in the spherical nodules of the resultant mixed matrix membranes, as marked with yellow arrows in the corresponding images. Furthermore, it can be concluded that when using higher loadings of gCN (i.e., 0.045 wt.% and particularly 0.06 wt.%) in the polymer matrix, the aggregation of nanomaterials takes place.

"The location of Fig. 5"

Generally, membrane porosity is defined as the void space volume fraction of the membrane. The higher the porosity, the higher the permeate flux and consequently the lower the heat loss by conduction [10]. Anyway, there must be a counterbalance between the high porosity and the mechanical stability of the membranes because the porosity is restricted by the required mechanical strength. El-Bourawi et al. [6] have reported that the porosity for MD membranes typically stands within the range of 30–85% and indeed, the values obtained in this study lie within this range. The porosity amounts for the as-fabricated membranes are given in Table 1, which demonstrate an increase by adding the nanosheets into the polymer solution. The highest and the lowest porosity values of 84.27 $\pm$ 1.84% and 53.01 $\pm$ 3.45% were obtained for the PVDF membranes with 0.03 wt.% gCN loading and without nanofiller, respectively. It is likely that the higher nanosheets concentration increases the thermodynamic instability, accelerates the rate of precipitation in the water coagulation bath, and results in a membrane with higher porosity, as previously discussed in the section of SEM images analysis of the membranes. It is noteworthy that the membrane with 0.045 wt.% gCN has shown a porosity close but a little less (i.e., 80.88 $\pm$ 1.2%) to that of the membrane embedded with 0.03 wt.% gCN. However, the increment in gCN content to 0.06 wt.% produces a far more viscous dope solution, which prevents the solution demixing process due to the enhanced kinetic hindrance, again as explained earlier in the SEM section, resulting in lower membrane pore size and porosity (Table 1). These results are in keeping with the calculated flux values that are rendered in the coming sections.

Figure 6 offers the two- and three-dimensional surface AFM images of the pristine and gCN embedded PVDF membranes at a scan size of 10 µm $\times$ 10 µm. Although all fabricated membranes presented ridge and valley topology, it is visually observed that their surface topography has been affected by the addition of nanosheets in the polymer scaffold. The surface roughness parameters of the membranes obtained from AFM images were measured in a scan area of 10 µm $\times$ 10 µm utilizing DME SPM software and the results are summarized in Table 1. It can be seen in Table 1 that the mixed matrix membranes possess a rougher appearance relative to the pristine PVDF membrane originating from the presence of the nanofillers. Here, the question may arise that the gCN nanosheets are not visible in the surface SEM images. However, it should be said in response that those same images prove the morphological changes of the surface and/or underneath the skin of the membranes due to the addition of the nanosheets. As we know, the rough surface is usually desirable for the MD process, whereas it enables higher free surface energy, and thus, more resistance against wetting [87].

According to Table 1, when the nanosheets are integrated into the membrane matrix, the S_a value is increased from 20.02 nm for the pure membrane to 26.51 nm for the membrane comprising 0.015 wt.% gCN. By adding the gCN content to 0.03 wt.%, the S_a value is considerably increased to 46.6 nm. By more adding gCN to 0.045 wt.% into the polymer network, the mean roughness is slightly decreased to 43.88 nm. And finally, the value of S_a reaches to 32.23 nm at the content of 0.06 wt.% gCN. These observations can be justified by the morphological alterations during the phase separation from two aspects of the number of surface pores and the size of subsurface spherical nodules, which were discussed earlier in the analysis of membranes surface and cross-sectional SEM images. Some authors have also pointed out that the entrapping of large quantities of the gCN nanomaterials (i.e., in the case of 0.06 wt.% gCN) into the PVDF matrix may lead to filling up the membrane surface pores, defects, or vacancies and accordingly, a smoother surface is obtained [50, 88, 89]. Anyhow, a large number of surface pores as well as smaller subsurface spherical nodules lead to the increased surface roughness by virtue of the increase in bumps and depressions and vice versa. It can be finally concluded from the data of Table 1 that the resultant surface roughness parameters might have been derived from a tradeoff between the number of surface pores and the subsurface spherical nodules.

"The location of Fig. 6"

Additionally, the size of the pores on the membrane surface is considered as an effective factor of surface roughness according to the reviewed literatures [89, 90]. Regarding the dependence of the roughness parameter on the Z value, when the surface of the membrane includes relatively larger pores, the vertical distance to be moved up and down by the piezoelectric scanner tip increases in comparison to shallow depressions, thus, the roughness parameter also increases [91, 92].

In order to further clarify this issue, the pore size distribution of the membranes was calculated from AFM images by using the Nanosurf Report (please see Fig. 7) and the mean surface pore sizes are presented in Table 1. The results reveal that the mean surface pore size of the PVDF membranes enhances after loading gCN nanosheets up to 0.03 wt.% and then follow a downward trend (Table 1). In other words, the pore size distribution shifts to the larger pore diameters by increasing the gCN concentration up to 0.03 wt.% in the dope solution and then it moves towards smaller pore diameters (Fig. 7). These evidences are in good agreement with the results of the surface roughness analysis.

"The location of Fig. 7"

The obtained contact angles are also in harmony with the resulting surface roughness parameters, so that the maximum contact angle has shown the maximum surface roughness [88]. The water contact angle (θ) on the surface of the membrane can quantify the wettability of the material and reflect the strength of the interaction on the interface [93]. As can be seen in Table 1, the water contact angle of the mixed matrix membranes is enhanced by increasing the content of the gCN nanosheets in the polymeric solution and then decreased by continuing this trend. Among the mixed matrix membranes, the highest contact angle belongs to the membrane incorporated with 0.03 wt.% of gCN (θ = 95.2 $\pm$ 1º). The pristine PVDF membrane has also rendered the lowest contact angle (θ = 80.5 $\pm$ 1º).

LEPw was measured to inquire about the effect of the gCN nanosheets on membrane resistance against wetting and the obtained values can be observed in Table 1. The PVDF membrane embedded with 0.03 wt.% gCN has demonstrated the highest LEPw equals 8 $\pm$ 0.5 bar, making it a better candidate for MD. LEPw is strongly influenced by surface hydrophobicity and mean surface pore size [94]. A membrane with higher hydrophobicity and smaller mean surface pore size will possess a higher LEP value according to the Cantor-Laplace equation [6]. Although the PVDF membrane modified with 0.03 wt.% gCN has a larger mean surface pore size than all other membranes, the higher hydrophobicity of the former compensates for its larger pore size and leads to a more water-resistant membrane. From the measured values of LEPw in Table 1, it can be proven that all membranes behave in an identical manner.

Table 1

Characteristics of different fabricated membranes.
Membrane	Thickness (µm)	Mean surface pore size (nm)	Porosity (%)	Contact angle, θ (º)	LEPw (bar)	Roughness parameters (nm)
Membrane	Thickness (µm)	Mean surface pore size (nm)	Porosity (%)	Contact angle, θ (º)	LEPw (bar)	S_a	S_q	S_z
Pure PVDF	26.9 $\pm$ 0.84	289.63	53.01 $\pm$ 3.45	80.5 $\pm$ 1	6 $\pm$ 0.5	20.02	25.19	198.01
gCN-PVDF 0.015	25 $\pm$ 1.5	327.03	67.85 $\pm$ 2.27	84 $\pm$ 2	6.5 $\pm$ 0.5	26.51	33.36	258.50
gCN-PVDF 0.03	29.3 $\pm$ 1.6	486.28	84.27 $\pm$ 1.84	95.2 $\pm$ 1	8 $\pm$ 0.5	46.60	62.13	472.35
gCN-PVDF 0.045	26.2 $\pm$ 0.92	403.15	80.88 $\pm$ 1.2	91.3 $\pm$ 2	7.5	43.88	55.29	351.55
gCN-PVDF 0.06	30.6 $\pm$ 1.1	356.5	71.73 $\pm$ 2.07	87.6 $\pm$ 2	7	32.23	40.25	333.73

"The location of Table 1"

MD is typically operated without hydraulic pressure and it is taken for granted that unlike pressure-driven membrane separation technologies, it has less membrane mechanical robustness requirements. Nevertheless, having sufficient mechanical resistance of MD membranes are still needed by virtue of withstanding the stress from compression in modules under the high-temperature long-term operation [30]. The effect of the gCN nanosheets on the mechanical behavior of the mixed matrix membranes against the pristine PVDF membrane was evaluated and the related data of stress-strain measurements are summarized in Table 2. As can be seen in Table 2, the ever-increasing tensile strength by increasing gCN content in the PVDF matrix occurs in all the mixed matrix membranes compared with the pure PVDF membrane. So that, tensile strength increases from 6.05 $\pm$ 0.21 MPa for the pure PVDF membrane to its maximum value of 7.34 $\pm$ 0.26 MPa for the membrane containing 0.06 wt.% gCN. In other words, the maximal increase by 21.32% is obtained for 0.06 wt.% gCN embedded membrane relative to the pure one. Thus, the increase in membranes' tensile strength is mainly interpreted by the presence of gCN nanosheets with high intrinsic mechanical stability within the polymer matrix [75, 95]. Also, the improvement of membranes' tensile strength may be due to the creation of morphological changes through introducing gCN nanosheets in the structure of the membranes. In line with the previous sentence, it should be said that the good reinforcement property of gCN is the fundamental reason to employ it as a filler material, as investigated by Shi and his coworkers [96]. They have developed sodium alginate (SA) nanocomposite films using 0.5-6 wt.% gCN and observed that with the addition of 6 wt.% gCN, the tensile strength of SA nanocomposite films was dramatically enhanced by 103%. At the same time, the Young’s modulus increased to an extremely high value of 3540 MPa. Regarding the latter, the cross-sectional SEM images had proved that when a high concentration of the nanosheets (with special emphasis on 0.06 wt.% gCN) was incorporated into the membrane scaffold, the sponge-like morphology became more apparent. Since membranes with a sponge-like structure possess superior mechanical attributes according to the literature reviews [97, 98], it is expected that the mechanical resistance of membranes with higher levels of gCN loading will increase.

Furthermore, the elongation at break increases with the addition of the nanofiller content into the polymeric matrix up to 0.03 wt.%. When the nanofiller content exceeded 0.03 wt.%, the elongation at break started reversing, probably due to the agglomeration of gCN nanosheets [95, 99]. The mixed matrix membrane with 0.03 wt.% gCN possesses the maximum level of elongation at break equal to 134.54 $\pm$ 4.2%. The elongation at break of the mixed matrix membranes is still higher than that of the pure PVDF membranes even at high gCN content, confirming that the incorporating of gCN nanosheets into the PVDF matrix can effectively enhance the mechanical properties of mixed matrix membranes [96].

Table 2

Mechanical strength of different fabricated membranes.
Membrane	Tensile strength (MPa)	Elongation at break (%)
Pure PVDF	6.05 $\pm$ 0.21	98.62 $\pm$ 2.5
gCN-PVDF 0.015	6.21 $\pm$ 0.23	116.44 $\pm$ 6.5
gCN-PVDF 0.03	6.88 $\pm$ 0.18	134.54 $\pm$ 4.2
gCN-PVDF 0.045	7.16 $\pm$ 0.11	129.08 $\pm$ 3.68
gCN-PVDF 0.06	7.34 $\pm$ 0.26	122.78 $\pm$ 5.01

"The location of Table 2"

3.3. Membrane desalination performance using DCMD

The desalination performance of the as-prepared membranes was evaluated through the DCMD configuration under the temperature of 68°C and with a feed concentration of 35000 ppm. The obtained values for water flux and salt rejection of the pristine PVDF and mixed matrix membranes over a duration of 5 h are displayed in Fig. 8. As shown, the pure PVDF membrane exhibits the lowest flux value of 16.25 $\pm$ 0.8 kg/m²h, while much higher fluxes are obtained for the mixed matrix membranes. By adding a small amount of nanofiller to 0.015 wt.%, the flux enhances to 20.64 $\pm$ 0.95 kg/m²h. The membrane incorporated with 0.03 wt.% gCN renders the highest flux of 27.63 $\pm$ 0.75 kg/m²h; however, at a loading of 0.045 and 0.06 wt.%, this drops to 25.03 $\pm$ 0.85 and 21.94 $\pm$ 0.7 kg/m²h, respectively. These findings conform truly to the alterations in pore size and porosity of the mentioned membranes (see Table 1). The treatment efficiency in the field of water and wastewater via gCN based hybrid membranes could be influenced by their physicochemical properties such as surface roughness, surface contact angle, pore size, and porosity, etc. [50]. While these morphological differences play a crucial role in the membranes' performance, other factors may also be influential. In order to figure out the flux improvement for membranes containing graphene-based nanofillers, various suggestions have been offered in the literatures [16, 17]. Based on the analogy with graphene-like nanomaterials structure, it can be assumed that gCN nanosheets are able to provide effectual sorption sites for water vapor but repel liquid water molecules. Moreover, the existence of the remaining terminal polar functional groups in the gCN structure (i.e., amino groups) can desirably alter the interaction between the membrane and water vapor molecules, thus enhancing the overall rate of vapor permeation [16]. Certainly, the combination of all these has been effective in the flux promotion of the membranes, but in some cases, one or more factors may prevail over the other(s). For example, in the cases of the 0.045 and 0.06 wt.% gCN embedded membranes, the morphological factors have overcome the gCN concentration.

Furthermore, the rejection of NaCl is > 99.9% for all tested membranes, while it is approximately 100% for the membrane comprising 0.03 wt.% gCN nanosheets. To have a better comparison between the permeate quality of the membranes, the permeate conductivity values after 5 h continuous desalting operation were arranged in tabular form, as can be observed in Table 3. The addition of gCN results in a lower permeate conductivity relative to the pure PVDF membrane. The comparison of the data in Tables 1 and 3 shows that the resultant permeate conductivities follows the contact angle values of the corresponding membranes. Accordingly, the lowest permeate conductivity, even lower than that for the coolant (i.e., 1.71 $\pm$ 0.3 µS/cm), is achieved for 0.03 wt.% gCN embedded membrane with the highest contact angle value. The presence of the relatively hydrophobic gCN nanosheets in the structure of the mixed matrix membranes may be considered as the main cause for the reduction of the permeate conductivity in addition to the measured contact angle values.

In association with the surface wettability of gCN, many scientific theories have been proposed. All of them agree on the same concept and that is the relative hydrophobicity of this 2D nanosheet. Some references have pointed out that gCN possesses a hydrophobic surface [64, 73, 100, 101], and some others have used the phrase “the relative hydrophobic surface” for gCN [102–104]. Liu et al. [93] have studied the mechanisms of water desalination across the nanoporous multilayer gCN membranes with different pore diameters using non-equilibrium molecular dynamics simulations. They have reported that the surface of gCN is somewhat hydrophobic with a contact angle of about 70º. Cao and coworkers [103] have investigated the mechanisms of humic acid impact on the photocatalytic antialgal activities of gCN and copper oxide (CuO) nanoparticles doped gCN to harmful algae. They have introduced gCN as a hydrophobic nanomaterial with a contact angle equal to 77.4 $\pm$ 0.9º. In addition, the performance of nanoparticulate gCN as an amphiphile has been evaluated in research by Xu et al. [105]. Amphiphiles refer to substances that own both hydrophilic and hydrophobic moieties. According to the authors, the amphiphilic feature of gCN is rooted in its intrinsic structural characteristics, i.e., the hydrophobic conjugated basal framework and hydrophilic edge groups and that’s why it is expected to be capable of dispersing hydrophobic solids in the aqueous phase.

Finally, to evaluate the long-time performance stability, the pristine and the optimum-performing membrane containing 0.03 wt.% gCN nanosheets were tested for 24 h. Results are demonstrated in Fig. 9 and offer a good stability of the 0.03 wt.% gCN embedded membrane under the test conditions, without diminution in salt rejection.

Table 3

Permeate conductivity of all tested membranes after five hours continuous operation. Feed conductivity was (50 $\pm$ 3) $\times$ 10³ µS/cm.
Membrane	Permeate conductivity (µS/cm)
Pure PVDF	39.6 $\pm$ 4
gCN-PVDF 0.015	26.5 $\pm$ 3
gCN-PVDF 0.03	1.71 $\pm$ 0.3
gCN-PVDF 0.045	8.12 $\pm$ 2
gCN-PVDF 0.06	22.83 $\pm$ 3

"The location of Fig. 8"

"The location of Fig. 9"

"The location of Table 3"

3.4. Performance comparison with other reported MD membranes

The desalting performance of reported PVDF membranes containing carbon-based nanofillers like CNTs, graphene, graphene oxide (GO), and reduced graphene oxide (rGO) is presented in Table 4. As seen in the previous section, the maximum permeate fluxes for each mixed matrix membrane in this study lie in the range of 20.64–27.63 kg/m²h, which are commonly higher than the optimum fluxes reported in some other literature given in Table 4 [16, 23, 28, 32, 33, 37]. Generally, it can be concluded from this literature surveys that adding the graphene-like nanomaterials ameliorates the separation performance of the original polymeric membranes. Besides, a comparison of these studies with the present study reveals the potential of the fabricated membranes embedded with small contents of gCN nanosheets. It is noteworthy that there are simultaneously a few flux values higher than those obtained in our work. For example, membranes prepared by Silva et al. [31], Roy et al. [27], Li et al. [18], and Dastbaz et al. [35], have higher water fluxes of about 34.2, 36.8, 36.4, and 34.1 kg/m²h, respectively. Our lower flux can be attributed to the different test specifications, such as differences in feed composition, flow rate, or temperature, module configuration (flat-sheet, hollow fiber, nanofibrous, etc.), concentration used for nanofillers as well as their modification with chemical agents.

Table 4

Comparison of the obtained results in the present study with other types of nanocarbon-based membranes used in MD process.
Membrane type	Nanofiller concentration (wt.%)	MD configuration	Module configuration	Feed solution			Permeate solution		Permeate flux (kg/m²h)	Salt rejection (%)	References
Membrane type	Nanofiller concentration (wt.%)	MD configuration	Module configuration	NaCl concentration (wt.%)	flow rate (L/min)	Temperature (ºC)	flow rate (L/min)	Temperature (ºC)	Permeate flux (kg/m²h)	Salt rejection (%)	References
PVDF/MWCNT	0.1-1	DCMD	Flat sheet	3.5	0.048	82	0.048	20	34.2	100	[31]
PVDF/CNT-COOH (CNIM)	10 mg	DCMD	Flat sheet	1	0.32	70	0.25	–	36.8	> 99.9	[27]
PVDF/CNT-COOH (CNIM)	10 mg	SGMD	Flat sheet	3.4	0.02	90	–	–	19.2	> 99	[28]
PVDF/graphene	0.1-2	AGMD	Flat sheet	Produced water	0.4	60	0.4	20	20.5	99.99	[33]
PVDF-HFP/graphene	1–10	AGMD	Nanofibrous	3.5	0.2	60	0.2	20	22.9	100	[32]
PVDF /APTS-GO	0.1-1	AGMD	Flat sheet	3.5	0.38	85	–	20	6.2	> 99.9	[37]
PVDF /FTES-GO	0.5-7	VMD	Nanofibrous	3.5	0.85	50	–	–	36.4	> 99.9	[18]
PVDF/GO-ODA	–	AGMD	Flat sheet	3.5	0.25	80	0.25	15	16.7	98.3	[23]
PVDF-HFP/GO-ODS	1–5	DCMD	Hollow fiber	Sea water	0.3	90	–	–	34.1	> 99	[35]
PVDF/rGO	0.1–0.7	AGMD	Flat sheet	3.5	0.385	80	–	20	~ 7	> 99.99	[16]
PVDF/gCN	0.015–0.06	DCMD	Flat sheet	3.5	0.4	68	0.2	23	27.63	~ 100	Present study
SGMD: sweep gas membrane distillation;
VMD: vacuum membrane distillation;
CNIM: carbon nanotube immobilized membrane;
FTES: 1H, 1H, 2H, 2H-perfluorooctyltriethoxysilane;
ODS: octadecyltrichlorosilane.

"The location of Table 4"

Considering that the expectations we had at the start of this research were fulfilled, now it is possible to claim that adding the gCN nanosheets, prepared using a simple method with the inexpensive and available precursor(s), into the PVDF matrix considerably enhances the MD desalination performance. This increment can be described by alterations in morphology, as is evident by the data analyses of SEM, AFM, pore size, and porosity as well as the interactions of nanosheets with water vapor molecules. The dispersion of only 0.03 wt.% gCN nanosheets into the polymeric solution could provide a membrane with optimum performance for the membrane distillation process. Such membrane could enhance the water flux by 70% relative to the pure PVDF membrane, reach the amount of 27.63 kg/m²h. More importantly, this could be achieved without increment in the produced permeate conductivity value. Moreover, this best-performing membrane has maintained its performance stability under a longer time experiment for 24 h. The current achievements can move the frontiers of knowledge forward and add to a progressing body of literature on combining 2D gCN-based nanofillers in hydrophobic polymeric matrices for MD applications. This is definitely the beginning of the way, and innovative and constructive ideas should be disseminated to fill the gap of utilizing gCN in membrane distillation technology in order to address the problems of water shortage in the world.

Funding

There is no funder for this research

Author Contribution

S.A: Investigation, Writing - Original Draft

E.S: Supervision, Review & Editing, Project administration

M.M: Supervision, Review & Editing

M.J: Investigation

All authors reviewed the manuscript

Conflicts of interests/Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Ravi J et al (2020) Polymeric membranes for desalination using membrane distillation: A review. Desalination 490:114530
Subramani A, Jacangelo JG (2015) Emerging desalination technologies for water treatment: A critical review. Water Res 75:164–187
Alsebaeai MK, Ahmad AL (2020) Membrane distillation: Progress in the improvement of dedicated membranes for enhanced hydrophobicity and desalination performance. J Ind Eng Chem 86:13–34
González D, Amigo J, Suárez F (2017) Membrane distillation: Perspectives for sustainable and improved desalination. Renew Sustain Energy Rev 80:238–259
Tibi F et al (2020) Fabrication of polymeric membranes for membrane distillation process and application for wastewater treatment: Critical review, vol 141. Process Safety and Environmental Protection, pp 190–201
El-Bourawi MS et al (2006) A framework for better understanding membrane distillation separation process. J Membr Sci 285(1):4–29
Rezaei M et al (2018) Wetting phenomena in membrane distillation: Mechanisms, reversal, and prevention. Water Res 139:329–352
Bhoj Y et al (2021) Recent advancements in practices related to desalination by means of nanotechnology. Chem Phys Impact 2:100025
Khayet M (2011) Membranes and theoretical modeling of membrane distillation: A review. Adv Colloid Interface Sci 164(1):56–88
Alkhudhiri A, Darwish N, Hilal N (2012) Membrane distillation: A comprehensive review. Desalination 287:2–18
Hsu ST, Cheng KT, Chiou JS (2002) Seawater desalination by direct contact membrane distillation. Desalination 143(3):279–287
Cath TY, Adams VD, Childress AE (2004) Experimental study of desalination using direct contact membrane distillation: a new approach to flux enhancement. J Membr Sci 228(1):5–16
Francis L et al (2014) PVDF hollow fiber and nanofiber membranes for fresh water reclamation using membrane distillation. J Mater Sci 49(5):2045–2053
Wang P, Chung T-S (2015) Recent advances in membrane distillation processes: Membrane development, configuration design and application exploring. J Membr Sci 474:39–56
Eykens L et al (2016) How To Optimize the Membrane Properties for Membrane Distillation: A Review. Ind Eng Chem Res 55(35):9333–9343
Abdel-Karim A et al (2019) PVDF membranes containing reduced graphene oxide: Effect of degree of reduction on membrane distillation performance. Desalination 452:196–207
Bhadra M, Roy S, Mitra S (2016) Desalination across a graphene oxide membrane via direct contact membrane distillation. Desalination 378:37–43
Li H et al (2020) Improved desalination properties of hydrophobic GO-incorporated PVDF electrospun nanofibrous composites for vacuum membrane distillation. Sep Purif Technol 230:115889
Zhang J et al (2010) Identification of material and physical features of membrane distillation membranes for high performance desalination. J Membr Sci 349(1):295–303
Kang G-d, Cao Y-m (2014) Application and modification of poly(vinylidene fluoride) (PVDF) membranes – A review. J Membr Sci 463:145–165
Basile A, Figoli A, Khayet M (2015) Pervaporation, Vapour Permeation and Membrane Distillation: Principles and Applications. Elsevier
Eykens L et al (2017) Membrane synthesis for membrane distillation: A review. Sep Purif Technol 182:36–51
Zahirifar J et al (2018) Fabrication of a novel octadecylamine functionalized graphene oxide/PVDF dual-layer flat sheet membrane for desalination via air gap membrane distillation. Desalination 428:227–239
Pendergast MM, Hoek EMV (2011) A review of water treatment membrane nanotechnologies.Energy Environ. Sci.,4p. 1946–1971
Cui Y et al (2021) Emerging graphitic carbon nitride-based membranes for water purification. Water Res 200:117207
Bhadra M, Roy S, Mitra S (2016) Flux enhancement in direct contact membrane distillation by implementing carbon nanotube immobilized PTFE membrane. Sep Purif Technol 161:136–143
Roy S, Bhadra M, Mitra S (2014) Enhanced desalination via functionalized carbon nanotube immobilized membrane in direct contact membrane distillation. Sep Purif Technol 136:58–65
Bhadra M, Roy S, Mitra S (2013) Enhanced desalination using carboxylated carbon nanotube immobilized membranes. Sep Purif Technol 120:373–377
Gethard K, Sae-Khow O, Mitra S (2011) Water Desalination Using Carbon-Nanotube-Enhanced Membrane Distillation. ACS Appl Mater Interfaces 3(2):110–114
Tijing LD et al (2016) Superhydrophobic nanofiber membrane containing carbon nanotubes for high-performance direct contact membrane distillation. J Membr Sci 502:158–170
Silva TLS et al (2015) Multi-walled carbon nanotube/PVDF blended membranes with sponge- and finger-like pores for direct contact membrane distillation. Desalination 357:233–245
Woo YC et al (2016) Water desalination using graphene-enhanced electrospun nanofiber membrane via air gap membrane distillation. J Membr Sci 520:99–110
Woo YC et al (2016) Graphene/PVDF flat-sheet membrane for the treatment of RO brine from coal seam gas produced water by air gap membrane distillation. J Membr Sci 513:74–84
Lu K-J, Zuo J, Chung T-S (2017) Novel PVDF membranes comprising n-butylamine functionalized graphene oxide for direct contact membrane distillation. J Membr Sci 539:34–42
Dastbaz A et al (2017) Preparation and characterization of novel modified PVDF-HFP/GO/ODS composite hollow fiber membrane for Caspian Sea water desalination. Desalination 424:62–73
Qiu H et al (2018) Pore channel surface modification for enhancing anti-fouling membrane distillation. Appl Surf Sci 443:217–226
Leaper S et al (2018) Flux-enhanced PVDF mixed matrix membranes incorporating APTS-functionalized graphene oxide for membrane distillation. J Membr Sci 554:309–323
Safikhani A et al (2021) Application of graphitic carbon nitrides in developing polymeric membranes: A review. Chem Eng Res Des 173:234–252
Akonkwa Mulungulungu G, Mao T, Han K (2022) Two-dimensional graphitic carbon nitride-based membranes for filtration process: Progresses and challenges. Chem Eng J 427:130955
Hyun T et al (2019) 2D-enabled membranes: materials and beyond. BMC Chem Eng 1(1):12
Liu G, Jin W, Xu N (2016) Two-Dimensional-Material Membranes: A New Family of High-Performance Separation Membranes.Angew Chem Int Ed Engl55p. 13384–13397
Liu G, Jin W, Xu N (2015) Graphene-based membranes. Chem Soc Rev 44(15):5016–5030
Homaeigohar S, Elbahri M (2017) Graphene membranes for water desalination. NPG Asia Materials 9(8):e427–e427
Pei J et al (2016) Fabrication of reduced graphene oxide membranes for highly efficient water desalination. RSC Adv 6(104):101948–101952
Yoon HW, Cho YH, Park HB (2016) Graphene-Based Membranes: Status and Prospects.Philos Trans A Math Phys, Eng Sci,374p. 20150024
Ding L et al (2017) Two-Dimensional Lamellar Membrane: Mxene Nanosheet Stacks.Angew Chem Int Ed Engl56p. 1825–1829
Wang Y et al (2020) Graphitic carbon nitride (g-C3N4)-based membranes for advanced separation. J Mater Chem A 8(37):19133–19155
Darkwah WK, Ao Y (2018) Mini Review on the Structure and Properties (Photocatalysis), and Preparation Techniques of Graphitic Carbon Nitride Nano-Based Particle, and Its Applications. Nanoscale Res Lett 13(1):388
Baig U, Faizan M, Sajid M (2021) Semiconducting graphitic carbon nitride integrated membranes for sustainable production of clean water: A review. Chemosphere 282:130898
Li X et al (2021) A review on graphitic carbon nitride (g-C3N4) based hybrid membranes for water and wastewater treatment. Sci Total Environ 792:148462
Zhao W, Yan Z, Qian L (2020) Graphitic Carbon Nitride: Preparation, Properties and Applications in Energy Storage. Eng Sci 10:24–34
Tan SC, Lee HK (2020) Graphitic carbon nitride as sorbent for the emulsification-enhanced disposable pipette extraction of eight organochlorine pesticides prior to GC-MS analysis. Microchim Acta 187(2):129
Aggarwal M et al (2021) Photocatalytic carbon dioxide reduction: Exploring the role of ultrathin 2D graphitic carbon nitride (g-C3N4). Chemical Engineering Journal, 425: p. 131402
Wang X et al (2009) A metal-free polymeric photocatalyst for hydrogen production from water under visible light. Nat Mater 8(1):76–80
Zheng T et al (2022) Two-dimensional graphitic carbon nitride for membrane separation. Chin J Chem Eng 42:297–311
Baig U et al (2021) Facile fabrication of graphitic carbon nitride nanosheets and its integrated polyamide hyper-cross-linked TFC nanofiltration membrane with intrinsic molecular porosity for salts and organic pollutant rejection from water. J Mater Res Technol 15:6319–6328
Seyyed Shahabi S, Azizi N, Vatanpour V (2019) Synthesis and characterization of novel g-C3N4 modified thin film nanocomposite reverse osmosis membranes to enhance desalination performance and fouling resistance. Sep Purif Technol 215:430–440
Seyyed Shahabi S et al (2020) Novel functionalized graphitic carbon nitride incorporated thin film nanocomposite membranes for high-performance reverse osmosis desalination. Sep Purif Technol 235:116134
Xu C et al (2019) Highly chlorine resistance polyamide reverse osmosis membranes with oxidized graphitic carbon nitride by ontology doping method. Sep Purif Technol 223:178–185
Gao X et al (2017) Highly permeable and antifouling reverse osmosis membranes with acidified graphitic carbon nitride nanosheets as nanofillers. J Mater Chem A 5(37):19875–19883
Rezaei-DashtArzhandi M et al (2018) Development of novel thin film nanocomposite forward osmosis membranes containing halloysite/graphitic carbon nitride nanoparticles towards enhanced desalination performance. Desalination 447:18–28
Wang Y et al (2015) Graphene oxide modified graphitic carbon nitride as a modifier for thin film composite forward osmosis membrane. J Membr Sci 475:281–289
Zhao Z, Sun Y, Dong F (2015) Graphitic carbon nitride based nanocomposites: a review. Nanoscale 7(1):15–37
Zarei M et al (2015) Graphitic carbon nitride embedded hydrogels for enhanced gel electrophoresis. Anal Chim Acta 887:245–252
Farahbakhsh J, Delnavaz M, Vatanpour V (2017) Investigation of raw and oxidized multiwalled carbon nanotubes in fabrication of reverse osmosis polyamide membranes for improvement in desalination and antifouling properties. Desalination 410:1–9
Vatanpour V, Zoqi N (2017) Surface modification of commercial seawater reverse osmosis membranes by grafting of hydrophilic monomer blended with carboxylated multiwalled carbon nanotubes. Appl Surf Sci 396:1478–1489
Vatanpour V et al (2021) Enhancing the permeability and antifouling properties of cellulose acetate ultrafiltration membrane by incorporation of ZnO@graphitic carbon nitride nanocomposite. Carbohydr Polym 256:117413
Smolders K, Franken ACM (1989) Terminology for Membrane Distillation. Desalination 72(3):249–262
Ardeshiri F et al (2019) PDADMAC/PAA semi-IPN hydrogel-coated PVDF membrane for robust anti-wetting in membrane distillation. J Ind Eng Chem 74:14–25
Zhu L et al (2015) Structural and friction characteristics of g-C3N4/PVDF composites. Appl Surf Sci 345:349–354
Chen Y et al (2018) Graphitic carbon nitride embedded in hot-melt adhesive polyester and hydrophilic cellulose blend fibers for the efficient elimination of antibiotics under solar irradiation. Appl Surf Sci 453:110–119
Nguyen TKA et al (2021) The effect of graphitic carbon nitride precursors on the photocatalytic dye degradation of water-dispersible graphitic carbon nitride photocatalysts. Appl Surf Sci 537:148027
Hu R et al (2015) Application of graphitic carbon nitride for the removal of Pb(II) and aniline from aqueous solutions. Chem Eng J 260:469–477
Lima MJ et al (2017) Graphitic carbon nitride modified by thermal, chemical and mechanical processes as metal-free photocatalyst for the selective synthesis of benzaldehyde from benzyl alcohol. J Catal 353:44–53
Wang J et al (2017) Graphitic carbon nitride nanosheets embedded in poly(vinyl alcohol) nanocomposite membranes for ethanol dehydration via pervaporation. Sep Purif Technol 188:24–37
Yang F et al (2020) Self-cleaning, antimicrobial, and antifouling membrane via integrating mesoporous graphitic carbon nitride into polyvinylidene fluoride. J Membr Sci 606:118146
Yue R, Saifur M, Rahaman (2022) Hydrophilic and underwater superoleophobic porous graphitic carbon nitride (g-C3N4) membranes with photo-Fenton self-cleaning ability for efficient oil/water separation. J Colloid Interface Sci 608:1960–1972
Li R et al (2019) Graphitic carbon nitride (g-C3N4) nanosheets functionalized composite membrane with self-cleaning and antibacterial performance. J Hazard Mater 365:606–614
Nadig AR et al (2021) Impact of graphitic carbon nitride nanosheets in mixed- matrix membranes for removal of heavy metals from water. J Water Process Eng 41:102026
Wu C et al (2019) Tape casting and phase inversion derived easily recyclable PVDF@g-C3N4/Bi2W0.98Nb0.02O6 porous film for high performance and long-term photocatalytic pollutants degradation. Appl Surf Sci 496:143731
Martins P, Lopes AC, Lanceros-Mendez S (2014) Electroactive phases of poly(vinylidene fluoride): Determination, processing and applications. Prog Polym Sci 39(4):683–706
Zhang J et al (2017) Sulfonated Ti3C2Tx to construct proton transfer pathways in polymer electrolyte membrane for enhanced conduction. Solid State Ionics 310:100–111
Munirasu S et al (2017) Intrinsically superhydrophobic PVDF membrane by phase inversion for membrane distillation. Desalination 417:77–86
Ashrafian S, Ataei SA, Jahanshahi M (2016) Novel composite membranes embedded with molecularly imprinted porous polymeric nanospheres for targeted phenol. Polym Adv Technol 27:789–804
Barnes HA (2003) A REVIEW OF THE RHEOLOGY OF FILLED VISCOELASTIC SYSTEMS.
Liu F et al (2011) Progress in the production and modification of PVDF membranes. J Membr Sci 375(1):1–27
Zhan N et al (2010) A novel multinozzle electrospinning process for preparing superhydrophobic PS films with controllable bead-on-string/microfiber morphology. J Colloid Interface Sci 345(2):491–495
Jahanshahi M, Rahimpour A, Peyravi M (2010) Developing thin film composite poly(piperazine-amide) and poly(vinyl-alcohol) nanofiltration membranes. Desalination 257(1):129–136
Yang Y et al (2007) The influence of nano-sized TiO2 fillers on the morphologies and properties of PSF UF membrane. J Membr Sci 288(1):231–238
Rahimpour A et al (2011) TiO2 entrapped nano-composite PVDF/SPES membranes: Preparation, characterization, antifouling and antibacterial properties. Desalination 278(1):343–353
Rajesh S et al (2013) Preparation and performance evaluation of poly (amide–imide) and TiO2 nanoparticles impregnated polysulfone nanofiltration membranes in the removal of humic substances. Colloids Surf A 418:92–104
Essalhi M, Khayet M (2012) Surface segregation of fluorinated modifying macromolecule for hydrophobic/hydrophilic membrane preparation and application in air gap and direct contact membrane distillation.Journal of Membrane Science,417–418: p. 163–173
Liu Y et al (2018) Water desalination across multilayer graphitic carbon nitride membrane: Insights from non-equilibrium molecular dynamics simulations. Carbon 140:131–138
Kebria MRS et al (2020) Hyper-branched dendritic structure modified PVDF electrospun membranes for air gap membrane distillation. Desalination 479:114307
Cao K et al (2015) Highly water-selective hybrid membrane by incorporating g-C3N4 nanosheets into polymer matrix. J Membr Sci 490:72–83
Shi Y et al (2014) Influence of g-C3N4 Nanosheets on Thermal Stability and Mechanical Properties of Biopolymer Electrolyte Nanocomposite Films: A Novel Investigation. ACS Appl Mater Interfaces 6(1):429–437
Li JF, Xu ZL (2008) Microporous polyethersulfone membranes prepared under the combined precipitation conditions with non-solvent additives. Polym Adv Technol 19:251–257
Sun D et al (2015) Preparation and characterization of PDMS-PVDF hydrophobic microporous membrane for membrane distillation. Desalination 370:63–71
Zhang L et al (2011) High strength graphene oxide/polyvinyl alcohol composite hydrogels. J Mater Chem 21(28):10399–10406
Shi Y et al (2016) Processable Dispersions of Graphitic Carbon Nitride Based Nanohybrids and Application in Polymer Nanocomposites. Ind Eng Chem Res 55(28):7646–7654
Li Z, Meng X, Zhang Z (2019) An Effective Approach to Improve the Photocatalytic Activity of Graphitic Carbon Nitride via Hydroxyl Surface Modification. Catalysts 9. DOI: 10.3390/catal9010017
Wang M et al (2015) A Taiji-principle-designed magnetic porous C-doped graphitic carbon nitride for environment-friendly solid phase extraction of pollutants from water samples. J Chromatogr A 1412:12–21
Cao X et al (2021) CuO nanoparticles doping recovered the photocatalytic antialgal activity of graphitic carbon nitride. J Hazard Mater 403:123621
Chen S et al (2020) Use of fibrous TiO2@graphitic carbon nitride nanocomposites in dispersive micro-solid phase extraction for arsenic species before inductively coupled plasma mass spectrometry determination. Microchem J 158:105211
Xu J, Antonietti M (2017) The Performance of Nanoparticulate Graphitic Carbon Nitride as an Amphiphile. J Am Chem Soc 139(17):6026–6029

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Membrane	Tensile strength (MPa)	Elongation at break (%)
Pure PVDF	6.05 \(\pm\) 0.21	98.62 \(\pm\) 2.5
gCN-PVDF 0.015	6.21 \(\pm\) 0.23	116.44 \(\pm\) 6.5
gCN-PVDF 0.03	6.88 \(\pm\) 0.18	134.54 \(\pm\) 4.2
gCN-PVDF 0.045	7.16 \(\pm\) 0.11	129.08 \(\pm\) 3.68
gCN-PVDF 0.06	7.34 \(\pm\) 0.26	122.78 \(\pm\) 5.01

High-performance robust 2D graphitic carbon nitride nanosheets embedded membranes for desalination through direct contact membrane distillation

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1. Materials

2.2. Synthesis of g-C₃N₄ nanosheets

2.3. Preparation of gCN embedded membranes

2.4. Characterization

2.5. DCMD experiments

3. Results And Discussion

3.1. Characterization of gCN nanosheets

3.2. Membrane characterization

3.3. Membrane desalination performance using DCMD

3.4. Performance comparison with other reported MD membranes

4. Conclusions

Declarations

References

Additional Declarations

Status:

Version 1

Membrane	Permeate conductivity (µS/cm)
Pure PVDF	39.6 \(\pm\) 4
gCN-PVDF 0.015	26.5 \(\pm\) 3
gCN-PVDF 0.03	1.71 \(\pm\) 0.3
gCN-PVDF 0.045	8.12 \(\pm\) 2
gCN-PVDF 0.06	22.83 \(\pm\) 3

High-performance robust 2D graphitic carbon nitride nanosheets embedded membranes for desalination through direct contact membrane distillation

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1. Materials

2.2. Synthesis of g-C3N4 nanosheets

2.3. Preparation of gCN embedded membranes

2.4. Characterization

2.5. DCMD experiments

3. Results And Discussion

3.1. Characterization of gCN nanosheets

3.2. Membrane characterization

3.3. Membrane desalination performance using DCMD

3.4. Performance comparison with other reported MD membranes

4. Conclusions

Declarations

References

Additional Declarations

Status:

Version 1

2.2. Synthesis of g-C₃N₄ nanosheets