Improved Surface Functional and Photocatalytic Properties of Hybrid ZnO-MoS2-Deposited Membrane for Photocatalysis-Assisted Dye Filtration

The synergistic mechanism of photocatalytic-assisted dye degradation has been demonstrated using a hybrid ZnO-MoS2-deposited photocatalytic membrane (PCM). Few layers of MoS2 sheets were produced using the facile and efficient surfactant-assisted liquid-phase exfoliation method. In this process, hydrophilic moieties of an anionic surfactant were adsorbed on the surface of MoS2, which aided exfoliation and promoted a stable dispersion due to the higher negative zeta potential of the exfoliated MoS2 sheets. Further, the decoration of ZnO on the exfoliated MoS2 sheets offered a bandgap energy reduction to about 2.77 eV, thus achieving an 87.12% degradation of methylene blue (MB) dye within 15 min of near UV-A irradiation (365 nm), as compared with pristine ZnO achieving only 56.89%. The photocatalysis-enhanced membrane filtration studies on the ZnO-MoS2 PCM showed a complete removal of MB dye (~99.95%). The UV-assisted dye degradation on the ZnO-MoS2 PCM offered a reduced membrane resistance, with the permeate flux gradually improving with the increase in the UV-irradiation time. The regeneration of the active ZnO-MoS2 layer also proved to be quite efficient with no compromise in the dye removal efficiency.


Introduction
The membrane filtration process is one of the most versatile and effective means of water/wastewater treatment technologies, owing to its reduced energy consumption and operating costs. The sustainable design of membrane processes is recently gaining more attention, especially with a focus to overcome the fundamental limitations on reject disposal associated with any membrane-based treatment [1]. Indeed, there are several evidences on integrating different unit operations or chemical processes with membrane separation, yet reject disposal remains challenging in terms of the economic and energy aspects [2,3]. Photocatalysis is one of the promising processes to be integrated with membrane separation for its advantage of the incomplete decomposition of organic pollutants under UV/visible light irradiation. Novel process designs in photocatalytic membrane reactors are continuously emerging to realise the maximum benefit from both photocatalysis and membrane filtration processes [4]. However, there are associated limitations with slurry/submerged or continuous photocatalytic membrane reactors in terms of photocatalyst regeneration, poor photocatalytic activity, excessive membrane resistance [5] and fouling [6,7]. Photocatalytic membranes (PCM) enable to overcome the above-mentioned limitations wherein the active photocatalytic layer is immobilised on the surface of the membrane rather than the bulk incorporation [8].
As an attempt to utilise the maximum photocatalytic efficiency, the membrane surface can be inherently tailored with desired multi-functionalities. By identifying a suitable material constituting photocatalytic, anti-fouling and selective separation properties, a hybrid PCM with a well-controlled structure and high performance could be developed. The recent advent of high-performance two-dimensional (2D) materials has been a significant impact on developing robust photocatalytic membrane materials [9,10], thanks to the advancements in material chemistry. Several research studies highlighted the photocatalytic activity of novel transition metal chalcogenides (TMDs) and 2D nanomaterials-based semiconductors [11][12][13]. However, studies related to developing membranes with photocatalytically active TMD-based semiconductor nanomaterials are limited.
The present work is central to achieving the desired surface functionality of the polymeric membranes by incorporating an active TMD layer with a photocatalytic property to perform a UV-assisted membrane filtration process. Special interest also lies in introducing heterojunctions between photocatalytic semiconductors and exfoliated 2D materials to realise a photocatalytic-enhanced membrane treatment. Emerging 2D TMDs, like MoS 2 , typically prepared based on liquid exfoliation [14], or particularly using solvents [15,16] and surfactants [17], hold greater potential in terms of nano-structural stability and permeation capability higher than that of its counterparts like graphene oxide, and has also been evidenced to offer membrane integrity towards pressure-driven water filtration applications [18,19]. As the present work necessitates the incorporation of exfoliated layers on MoS 2 , a facile and environmentally friendly surfactant-assisted exfoliation method [20] has been adopted to prevent any damage of the membrane substrate against solvent exposure and at the same time improve the stability of exfoliated MoS 2 sheets. MoS 2 can be effectively integrated as an active layer on the surface of the membrane by taking advantage of the robust van der Waals structure with a stable/narrow interlayer of MoS 2 in an aqueous environment [21]. Robust surface modification strategies to beneficially utilize the multi-functional properties of MoS 2 without altering the desired molecular channels within its interlayer spacing would hence bring breakthrough advances in the field of photocatalytic membranes.
Hybrid photocatalytic membranes constituting a selective and photocatalytic surface layer tend to offer synergistic benefits as well as mitigate the trade-off existing between the productivity and rejection efficiency of the membrane. Improved filtration properties and membrane life will be rendered while reducing the typical limitations of membrane filtration concerning reject disposal and fouling. The present study deals with developing a photocatalytically active polyvinylidene difluoride (PVDF) polymeric nanocomposite membrane with few layers of 2D hybrid ZnO-MoS 2 deposited on its surface. Methylene blue (MB) dye was chosen as a model pollutant to investigate the efficiency of the photocatalytic-assisted membrane filtration performance of a pristine ZnO-MoS 2 powder and hybrid ZnO-MoS 2 PCM, respectively. Chemically exfoliated MoS 2 (E-MoS 2 ), having a shorter conduction band gap energy, activates the electron transfer rate of the ZnO semiconductor by suppressing the faster recombination of its electron-hole pairs [1,22] which enable enhanced membrane properties such as surface adsorption, hydrophilic properties and membrane filtration performance [23,24].

Surfactant Assisted Exfoliation of Bulk MoS 2
Bulk MoS 2 (50 mg) was added to a solution of deionised water (50 mL) and SDBS (25 mg). The solution was magnetically stirred while being heated to 45 • C for 20 min. The mixture was then ultra-sonicated for 3 h using a solid probe sonicator (Sonics, Newtown, CT, USA) with a 750W power output set at a 20 kHz frequency and maximum amplitude of 25%. In order to control the temperature, sonication was performed in a jacketed beaker with water circulation and the sonicator was set at a pulse rate of 7 s on and 2 s off [20,25]. The resultant dispersion containing the exfoliated (E-MoS 2 ) and bulk unexfoliated (B-MoS 2 ) was separated by two-step centrifugation. Initially, the dispersion was centrifuged at 4000 rpm for 1 h, and an upper green liquid part of the dispersion containing the exfoliated layers of MoS 2 was collected while leaving the larger flakes' sedimented residue for disposal. The collected supernatant was then centrifuged again for 30 min at 4000 rpm to avoid further bulk residues sedimented at the bottom of centrifuge tubes. Finally, the as-obtained E-MoS 2 dispersion was used for the membrane modification and further characterisation, as shown in Figure 1.

Surfactant Assisted Exfoliation of Bulk MoS2
Bulk MoS2 (50 mg) was added to a solution of deionised water (50 mL) and SDBS (25 mg). The solution was magnetically stirred while being heated to 45 C for 20 min. The mixture was then ultra-sonicated for 3 h using a solid probe sonicator (Sonics, Newtown, USA) with a 750W power output set at a 20 kHz frequency and maximum amplitude of 25%. In order to control the temperature, sonication was performed in a jacketed beaker with water circulation and the sonicator was set at a pulse rate of 7 s on and 2 s off [20,25]. The resultant dispersion containing the exfoliated (E-MoS2) and bulk unexfoliated (B-MoS2) was separated by two-step centrifugation. Initially, the dispersion was centrifuged at 4000 rpm for 1 h, and an upper green liquid part of the dispersion containing the exfoliated layers of MoS2 was collected while leaving the larger flakes' sedimented residue for disposal. The collected supernatant was then centrifuged again for 30 min at 4000 rpm to avoid further bulk residues sedimented at the bottom of centrifuge tubes. Finally, the as-obtained E-MoS2 dispersion was used for the membrane modification and further characterisation, as shown in Figure 1.

Preparation of Hybrid ZnO-MoS2 Nanocomposite
Twenty mg of ZnO nanopowder was taken per 10 ml of the E-MoS21 dispersion (1 mg/Ml concentration) and stirred for 20 min. Sonochemical-assisted doping of E-MoS2 was carried out by ultra-sonicating the entire solution for 30 min. A greenish white turbid dispersion was obtained, which was stirred further for 2 h [26]. The resultant dispersion was dried in a vacuum oven at 70 C for 12 h, followed by heating for 2 h at 200 °C to collect the dried ZnO-MoS2 nanocomposite [27,28]. The as-obtained hybrid ZnO-MoS2 ( Figure 1) was characterised and employed in pristine form to study its photocatalytic activity.

Characterisation of E-MoS2 Dispersion and MoS2 Doped ZnO
The stability of E-MoS2, ZnO and the hybrid ZnO-MoS2 nanocomposite was compared by measuring its zeta potential using a zetasizer nano ZS (Malvern Panalytical Ltd.,Worcestershire, UK). ZnO and the as-prepared dried samples of the hybrid ZnO-MoS2 were prepared using deionised water (pH 7.6) at a concentration of 3 mg/mL, and the dispersion of E-MoS2 was centrifuged and later re-dispersed in deionised water at a similar concentration. All measurements were performed in triplicates to determine the average zeta potential. Raman spectra of E-MoS2 and the hybrid ZnO-MoS2 photocatalyst were collected by means of a non-resonant 532 nm cw laser. The full-spectrum was recorded while placing the samples on a standard XY motorised stage and a coupled microscope objective (20× with 3 μm laser spot size) helped in scanning and collecting the back-scattered light using a grating spectrometer. Optical analysis was performed using UV-vis spectroscopy (Perkin-Elmer lambda 1050 spectrophotometer, Massachusetts, USA) to measure the band gap energy of the prepared photocatalyst by means of the well-known Tauc relation, Equation (1).

Preparation of Hybrid ZnO-MoS 2 Nanocomposite
Twenty mg of ZnO nanopowder was taken per 10 ml of the E-MoS 21 dispersion (1 mg/mL concentration) and stirred for 20 min. Sonochemical-assisted doping of E-MoS 2 was carried out by ultra-sonicating the entire solution for 30 min. A greenish white turbid dispersion was obtained, which was stirred further for 2 h [26]. The resultant dispersion was dried in a vacuum oven at 70 • C for 12 h, followed by heating for 2 h at 200 • C to collect the dried ZnO-MoS 2 nanocomposite [27,28]. The as-obtained hybrid ZnO-MoS 2 ( Figure 1) was characterised and employed in pristine form to study its photocatalytic activity.

Characterisation of E-MoS 2 Dispersion and MoS 2 Doped ZnO
The stability of E-MoS 2 , ZnO and the hybrid ZnO-MoS 2 nanocomposite was compared by measuring its zeta potential using a zetasizer nano ZS (Malvern Panalytical Ltd.,Worcestershire, UK). ZnO and the as-prepared dried samples of the hybrid ZnO-MoS 2 were prepared using deionised water (pH 7.6) at a concentration of 3 mg/mL, and the dispersion of E-MoS 2 was centrifuged and later re-dispersed in deionised water at a similar concentration. All measurements were performed in triplicates to determine the average zeta potential. Raman spectra of E-MoS 2 and the hybrid ZnO-MoS 2 photocatalyst were collected by means of a non-resonant 532 nm cw laser. The full-spectrum was recorded while placing the samples on a standard XY motorised stage and a coupled microscope objective (20× with 3 µm laser spot size) helped in scanning and collecting the back-scattered light using a grating spectrometer. Optical analysis was performed using UV-vis spectroscopy (Perkin-Elmer lambda 1050 spectrophotometer, Waltham, MA, USA) to measure the band gap energy of the prepared photocatalyst by means of the well-known Tauc relation, Equation (1).
where h refers to the Plank constant, ν is the incident of light frequency, A is the absorbance, E g is the bandgap and n equals to 1 2 , which refers to the direct band gap energy of the E-MoS 2 dispersion.

Modification of PVDF Membrane with Hybrid ZnO-MoS 2 Photocatalysts
For the surface modification, 20 mL of the prepared dispersion of the hybrid ZnO-MoS 2 (3 mg/mL concentration) was deposited on a PVDF membrane with the help of a pre-assembled filtration set-up (Sartorius Filtration apparatus, Microsart®, Göttingen, Germany) operated by a peristaltic pump. Deposition of the synthesised ZnO-MoS 2 dispersion was confirmed by visualising the clear permeate through the membrane. Finally, the ZnO-MoS 2 PCM was rinsed with deionised water, dried at 85 • C and immersed in deionised water for further use.

Hydrophilicity Characteristics
The contact angles of the virgin PVDF and modified ZnO-MoS 2 PCM were measured using an assembled goniometer set-up (First Ten Angstroms, FTÅ12, Cambridge, UK) to study its hydrophilicity characteristics. The static sessile drop method was adopted wherein a water droplet of 10 µl was placed on the membrane surface using a Gilmount syringe to measure the equilibrium water contact angle. The average of five readings of the equilibrium water contact angles was then calculated and reported as the mean contact angle of the prepared membranes.

Morphological Characteristics
The morphologies of the hybrid ZnO-MoS 2 photocatalyst and modified membranes were examined using a field emission scanning electron microscope (Zeiss Ultra plus, Carl Zeiss, Oberkochen, Germany). The membrane was initially pre-treated with gold (Au) sputtering to impart the electrical conductivity to get top surface and back-scattered micrographs of its cross-section morphology. The ZnO-MoS 2 photocatalyst was inherently conductive, and sample imaging was performed using a high-resolution lens at an accelerating voltage of 20 kV without gold sputtering.

Structural Characteristics
The surface structural characteristics of the ZnO-MoS 2 PCM were studied by performing X-ray photoelectron spectroscopy (XPS) (VG Scientific ECSA lab Mk II system) using Al Kα X-rays (1486.6 eV) under ultra-high vacuum conditions (<5 × 10 −10 mbar). Survey and elemental spectra were recorded while setting the analyser pass energy to 200 eV. An electron flood gun was employed for the charge compensation, and the binding energy scale of the adventitious carbon 1s core-level was referenced at 284.8 eV.

Hybrid ZnO Doped MoS 2 Photocatalysts for MB Dye Treatment
An amount of 100 ml of 20 ppm MB solution was stirred in two different beakers with 100 mg each of ZnO and the hybrid ZnO-MoS 2 photocatalyst. The solutions were initially left in the dark under stirring for 1 h to allow them to reach the absorption-desorption equilibrium. UV irradiation was then exposed while placing the solutions within an enclosed chamber. Sampling was done at every 15 min for hours, and collected samples were then centrifuged at 14,000 rpm for 10 min to remove any Membranes 2020, 10, 106 5 of 18 traces of the residual photocatalyst. The MB dye concentration was then determined by measuring the absorbance peak of the samples at 665 nm. The obtained data on the difference between the initial and final concentration were plotted against time to investigate the photocatalytic activity of ZnO and the hybrid ZnO-MoS 2 .

Photocatalysis Assisted Membrane Filtration Studies for MB Dye Treatment
A laboratory scale photocatalytic membrane reactor was set up, as shown in Figure 2. A UV lamp (Phillips-UVA, 9 W, RS Radionics, Dublin, Ireland) emitting a near-UV radiation of 365 nm wavelength was placed at about 3 cm above the membrane filtration set-up (Sartorius filtration apparatus, Microsart®, Göttingen, Germany) and kept together enclosed in a closed chamber as shown in Figure 2. The ZnO-MoS 2 PCM, having a diameter of 47 mm, was mounted in the filtration set-up connected to the peristaltic pump. The feed solution containing 20 ppm concentration of MB dye was pumped into the microfiltration set-up while irradiating the ZnO-MoS 2 PCM. Samples were collected every 30 min while recirculating the permeate back into the feed tank. All trials pertaining to the photocatalytic-assisted filtration performance evaluation and reusability of the ZnO-MoS 2 PCM are studied using the same experimental set-up.
Membranes 2020, 10, x FOR PEER REVIEW 5 of 18 between the initial and final concentration were plotted against time to investigate the photocatalytic activity of ZnO and the hybrid ZnO-MoS2.

Photocatalysis Assisted Membrane Filtration Studies for MB Dye Treatment
A laboratory scale photocatalytic membrane reactor was set up, as shown in Figure 2. A UV lamp (Phillips-UVA, 9 W, RS Radionics, Dublin, Ireland) emitting a near-UV radiation of 365 nm wavelength was placed at about 3 cm above the membrane filtration set-up (Sartorius filtration apparatus, Microsart® , Göttingen, Germany) and kept together enclosed in a closed chamber as shown in Figure 2. The ZnO-MoS2 PCM, having a diameter of 47 mm, was mounted in the filtration set-up connected to the peristaltic pump. The feed solution containing 20 ppm concentration of MB dye was pumped into the microfiltration set-up while irradiating the ZnO-MoS2 PCM. Samples were collected every 30 min while recirculating the permeate back into the feed tank. All trials pertaining to the photocatalytic-assisted filtration performance evaluation and reusability of the ZnO-MoS2 PCM are studied using the same experimental set-up. Initially, deionised water was passed through the virgin PVDF and ZnO-MoS2 PCM to compare the steady-state pure water flux. MB dye filtration studies were performed on the above-mentioned filtration set-up by employing (1) the virgin PVDF membrane, (2) the surface-modified ZnO-MoS2 PCM with filtration under dark (UV off) and (3) a UV-assisted filtration on the ZnO-MoS2 PCM. The permeate was collected to determine the flux and dye concentration to compare the efficiency and rejection ability of the virgin PVDF and modified ZnO-MoS2 membranes. The flux through the membranes was calculated using the following Equation (2): where J represents the permeate flux (Lm −2 h −1 ), V is the volume of the filtered water (m 3 ), A is the active membrane area (m 2 ) and Δt is the working time (h).
The dye rejection efficiency was quantified based on the concentration of the initial feed (Ci) and obtained permeate concentration (Cp) using the below Equation (3): Initially, deionised water was passed through the virgin PVDF and ZnO-MoS 2 PCM to compare the steady-state pure water flux. MB dye filtration studies were performed on the above-mentioned filtration set-up by employing (1) the virgin PVDF membrane, (2) the surface-modified ZnO-MoS 2 PCM with filtration under dark (UV off) and (3) a UV-assisted filtration on the ZnO-MoS 2 PCM. The permeate was collected to determine the flux and dye concentration to compare the efficiency and rejection ability of the virgin PVDF and modified ZnO-MoS 2 membranes. The flux through the membranes was calculated using the following Equation (2): where J represents the permeate flux (Lm −2 h −1 ), V is the volume of the filtered water (m 3 ), A is the active membrane area (m 2 ) and ∆t is the working time (h).
Membranes 2020, 10, 106 6 of 18 The dye rejection efficiency was quantified based on the concentration of the initial feed (C i ) and obtained permeate concentration (C p ) using the below Equation (3):

Morphology of ZnO-MoS 2 Photocatalyst
The surface morphology of the ZnO-MoS 2 photocatalyst characterised by SEM is shown in Figure 3, wherein exfoliated few layers of MoS 2 sheets have been decorated with ZnO on the surface. Upon exfoliation, it could be observed that the lateral size of the MoS 2 sheets has decreased due to the sonication-induced scission. The size reduction of the lateral edges plays a significant role in rendering beneficial properties to the exfoliated 2D layers of MoS 2 , which in the present application have turned the ZnO-MoS 2 sheets catalytically more active than the bulk. The reduction in the lateral size of ZnO-MoS 2 is beneficial for photocatalytic applications as edge sites rather than basal planes of MoS 2 are known to be catalytically active [29].

Morphology of ZnO-MoS2 Photocatalyst
The surface morphology of the ZnO-MoS2 photocatalyst characterised by SEM is shown in Figure 3, wherein exfoliated few layers of MoS2 sheets have been decorated with ZnO on the surface. Upon exfoliation, it could be observed that the lateral size of the MoS2 sheets has decreased due to the sonication-induced scission. The size reduction of the lateral edges plays a significant role in rendering beneficial properties to the exfoliated 2D layers of MoS2, which in the present application have turned the ZnO-MoS2 sheets catalytically more active than the bulk. The reduction in the lateral size of ZnO-MoS2 is beneficial for photocatalytic applications as edge sites rather than basal planes of MoS2 are known to be catalytically active [29]. It is significant to visualise the dense surface porous structure of the MoS2 sheets, which would certainly offer improved permeability, owing to the higher cumulative porosity of the ZnO-MoS2-deposited membranes [30]. Other superior functionalities, including the catalytic activity and wettability, are expected to enhance by taking advantage of the surface pores [31].

Raman Spectral Analysis of MoS2 and ZnO-MoS2 Photocatalyst
Raman spectra for the bulk and exfoliated MoS2 dispersion are shown in Figure 4 and the displayed two MoS2 peaks are the in-plane E 1 2g and the out-of-plane A1g. Peak positions at 383 cm −1 for E 1 2g and 407 cm −1 for A1g confirm the presence of MoS2 [32,33]. The presence of a blue shift on the E 1 2g peak implies that a degree of exfoliation has occurred. The absence of the A1g red shift, which is typically more responsive to reducing the layer number, could be a result due to SDBS, which It is significant to visualise the dense surface porous structure of the MoS 2 sheets, which would certainly offer improved permeability, owing to the higher cumulative porosity of the ZnO-MoS 2 -deposited membranes [30]. Other superior functionalities, including the catalytic activity and wettability, are expected to enhance by taking advantage of the surface pores [31].

Raman Spectral Analysis of MoS 2 and ZnO-MoS 2 Photocatalyst
Raman spectra for the bulk and exfoliated MoS 2 dispersion are shown in Figure 4 and the displayed two MoS 2 peaks are the in-plane E 1 2g and the out-of-plane A 1g . Peak positions at 383 cm −1  [32,33]. The presence of a blue shift on the E 1 2g peak implies that a degree of exfoliation has occurred. The absence of the A 1g red shift, which is typically more responsive to reducing the layer number, could be a result due to SDBS, which adhered to the basal planes of the MoS 2 as it stabilises the dispersion through its ionic charge. When comparing the bulk and exfoliated MoS 2 , there is a blue shift from 381 to 383 cm −1 on the in-plane E 1 2g peak which is attributed to the reduction in the number of layers [34].
Membranes 2020, 10, x FOR PEER REVIEW 7 of 18 adhered to the basal planes of the MoS2 as it stabilises the dispersion through its ionic charge. When comparing the bulk and exfoliated MoS2, there is a blue shift from 381 to 383 cm −1 on the in-plane E 1 2g peak which is attributed to the reduction in the number of layers [34]. The Raman spectra of the hybrid ZnO-MoS2 ( Figure 4) show the characteristic peak of the A1g mode specific to the exfoliated sheets of MoS2 which have turned remarkably sharper, attributed to the improved crystallinity due to the formation of ZnO and MoS2 heterojunctions. Moreover, the presence of a weaker E2 (H) vibrational band exhibited at wavenumber 440 cm −1 represents the incorporation of a hexagonal wurtzite crystal structure of ZnO in the hybrid ZnO-MoS2 composite [27,35]. The reduction in the Raman shift occurring at the vibrational modes of the E2g and A1g peaks has also confirmed the 2H-hexagonal crystal plane of MoS2 with a reduced number of layers [36]. However, a more plausible rationale behind the reduced shift is attributed to the extent of exfoliation, which has been successful to only a few layers rather than a monolayer exfoliation. The Raman peak positions of the few layers of MoS2 are known to converge on its bulk profile when the layer number  4, as reported [37], thus suggesting the thickness of the obtained E-MoS2 sample has possibly around four layers.

Band Gap Energy and Stability Analysis of ZnO-MoS2 Photocatalyst
The zeta potential of E-MoS2 and the ZnO-MoS2 hybrid was found to be −38 mV and −33 mV, respectively, thus confirming its better stability in the SDBS exfoliation medium. Indeed, a higher negative value of the ZnO-MoS2 hybrid relative to ZnO (−3.8mV) confirms the anchoring of MoS2. Further, the negative zeta potential value is attributed to the negatively charged SDBS, which rendered the ionic charge stability to the hybrid ZnO-MoS2 dispersion through its electrostatic repulsions between the sheets [20]. Another factor is the tendency of the surfactant to absorb at the interface of the ZnO and MoS2 sheets wherein the aromatic ring of SDBS attaches only to the basal plane and leaves the catalytically active edge site of MoS2 free [38], thereby retaining the stability and improving the photocatalytic activity of ZnO-MoS2.
The optical band gap was estimated using the Tauc plot, by extrapolating a line from the linear region on the plot of (αhν) 2 vs. (hν), as indicated in Figure 5. The band gap of E-MoS2 was estimated The Raman spectra of the hybrid ZnO-MoS 2 (Figure 4) show the characteristic peak of the A 1g mode specific to the exfoliated sheets of MoS 2 which have turned remarkably sharper, attributed to the improved crystallinity due to the formation of ZnO and MoS 2 heterojunctions. Moreover, the presence of a weaker E 2 (H) vibrational band exhibited at wavenumber 440 cm −1 represents the incorporation of a hexagonal wurtzite crystal structure of ZnO in the hybrid ZnO-MoS 2 composite [27,35]. The reduction in the Raman shift occurring at the vibrational modes of the E 2g and A 1g peaks has also confirmed the 2H-hexagonal crystal plane of MoS 2 with a reduced number of layers [36]. However, a more plausible rationale behind the reduced shift is attributed to the extent of exfoliation, which has been successful to only a few layers rather than a monolayer exfoliation. The Raman peak positions of the few layers of MoS 2 are known to converge on its bulk profile when the layer number ≥4, as reported [37], thus suggesting the thickness of the obtained E-MoS 2 sample has possibly around four layers.

Band Gap Energy and Stability Analysis of ZnO-MoS 2 Photocatalyst
The zeta potential of E-MoS 2 and the ZnO-MoS 2 hybrid was found to be −38 mV and −33 mV, respectively, thus confirming its better stability in the SDBS exfoliation medium. Indeed, a higher negative value of the ZnO-MoS 2 hybrid relative to ZnO (−3.8mV) confirms the anchoring of MoS 2 . Further, the negative zeta potential value is attributed to the negatively charged SDBS, which rendered the ionic charge stability to the hybrid ZnO-MoS 2 dispersion through its electrostatic repulsions between the sheets [20]. Another factor is the tendency of the surfactant to absorb at the interface of the ZnO and MoS 2 sheets wherein the aromatic ring of SDBS attaches only to the basal plane and leaves the catalytically active edge site of MoS 2 free [38], thereby retaining the stability and improving the photocatalytic activity of ZnO-MoS 2 .
The optical band gap was estimated using the Tauc plot, by extrapolating a line from the linear region on the plot of (αhν) 2 vs. (hν), as indicated in Figure 5. The band gap of E-MoS 2 was estimated to be~1.55 eV; this value is in conjunction with the Raman data implying the exfoliation with most MoS 2 sheets is achieved when lying between two and six layers. As is obtained from the energy band gap analysis shown in Figure 5b, it is confirmed that the wide band gap of ZnO~3.2 eV [39] has been shortened to~2.77 eV upon doping E-MoS 2 , which typically holds a narrow direct band gap energy of 1.9 eV [40,41]. The addition of the narrow indirect gap p-type E-MoS 2 to ZnO has decreased its band gap through the creation of a p-n heterojunction [42]. The resultant energy band structure positively influenced the charge transfer of the photoinduced electrons in the conduction band of MoS 2 to ZnO, thus leaving behind photogenerated electron-hole pairs to produce hydroxyl and superoxide free radicals [43]. The rapid generation of those free radicals eventually helped in the MB dye degradation at a rate relatively faster than employing pristine ZnO.
Membranes 2020, 10, x FOR PEER REVIEW 8 of 18 to be ~1.55 eV; this value is in conjunction with the Raman data implying the exfoliation with most MoS2 sheets is achieved when lying between two and six layers. As is obtained from the energy band gap analysis shown in Figure 5b, it is confirmed that the wide band gap of ZnO ~3.2 eV [39] has been shortened to ~2.77 eV upon doping E-MoS2, which typically holds a narrow direct band gap energy of 1.9 eV [40,41]. The addition of the narrow indirect gap p-type E-MoS2 to ZnO has decreased its band gap through the creation of a p-n heterojunction [42]. The resultant energy band structure positively influenced the charge transfer of the photoinduced electrons in the conduction band of MoS2 to ZnO, thus leaving behind photogenerated electron-hole pairs to produce hydroxyl and superoxide free radicals [43]. The rapid generation of those free radicals eventually helped in the MB dye degradation at a rate relatively faster than employing pristine ZnO.

XPS Structural Analysis
The chemical structural analysis of the modified PVDF surface of the ZnO-MoS2 PCM, as identified through XPS measurements, is shown in Figure 6. The survey scan of PVDF/MoS2-ZnO indicates the presence of Zn 2p, O 1s, Mo 3d, S 2p and C 1s peaks, which are compared against the virgin PVDF showing a prominent peak for F 1s. The deconvoluted S 2p binding energy peak at 160 represents the presence of sulphide and clearly demarcates it from the peak at 168 which corresponds to a sulfate molecule from the SDBS dispersion medium. The presence of the S 2p peak at the binding energy of 160 eV is indexed to the S 2− ions in the Mo-S bonding of disulfide in MoS2. Moreover, peaks at the binding energies of 1020 eV and 1045 eV are characteristic of 2p 3/2 and 2p1/2, respectively, of Zn 2p, together corresponding to ZnO [44]. It is also important to mention that there is no doublet peak corresponding to Mo-3d, confirming no existence of Mo-O bond on the surface of the photocatalytic membrane [45].

XPS Structural Analysis
The chemical structural analysis of the modified PVDF surface of the ZnO-MoS 2 PCM, as identified through XPS measurements, is shown in Figure 6. The survey scan of PVDF/MoS 2 -ZnO indicates the presence of Zn 2p, O 1s, Mo 3d, S 2p and C 1s peaks, which are compared against the virgin PVDF showing a prominent peak for F 1s. The deconvoluted S 2p binding energy peak at 160 represents the presence of sulphide and clearly demarcates it from the peak at 168 which corresponds to a sulfate molecule from the SDBS dispersion medium. The presence of the S 2p peak at the binding energy of 160 eV is indexed to the S 2− ions in the Mo-S bonding of disulfide in MoS 2 . Moreover, peaks at the binding energies of 1020 eV and 1045 eV are characteristic of 2p 3/2 and 2p1/2, respectively, of Zn 2p, together corresponding to ZnO [44]. It is also important to mention that there is no doublet peak corresponding to Mo-3d, confirming no existence of Mo-O bond on the surface of the photocatalytic membrane [45].

Influence of ZnO-MoS2 on Morphological Characteristics
The SEM image, as shown in Figure 7a, compares the virgin PVDF membrane, with its surface modified, with the ZnO-MoS2 photocatalyst shown in Figure 7b. The random restacking of the MoS2 layered structure enables the size reduction of the larger pores, which helps to shorten the channel through which the water solute traverses. The average

Influence of ZnO-MoS 2 on Morphological Characteristics
The SEM image, as shown in Figure 7a, compares the virgin PVDF membrane, with its surface modified, with the ZnO-MoS 2 photocatalyst shown in Figure 7b.

Influence of ZnO-MoS2 on Morphological Characteristics
The SEM image, as shown in Figure 7a, compares the virgin PVDF membrane, with its surface modified, with the ZnO-MoS2 photocatalyst shown in Figure 7b. The random restacking of the MoS2 layered structure enables the size reduction of the larger pores, which helps to shorten the channel through which the water solute traverses. The average The random restacking of the MoS 2 layered structure enables the size reduction of the larger pores, which helps to shorten the channel through which the water solute traverses. The average height of the active ZnO-MoS 2 was identified to be nearly~6 microns based on the cross-section SEM imaging on the prepared ZnO-MoS 2 -modified PVDF membrane, which enabled us to infer a uniform thickness of the photocatalytic layer on the membrane surface. Hence, the orientation of a few layers of MoS 2 has altered the pore morphology of the PVDF membrane for achieving improved filtration properties.

Influence of ZnO-MoS 2 on Hydrophilicity and Porosity Characteristics
The contact angle was recorded to investigate the wettability of the surface-modified PVDF membrane against its unmodified counterpart. The average contact angle of the modified membrane is 41.04 • , which is lower than the virgin PVDF having the highest average contact angle of 72.24 • due to its inherent hydrophobic nature. Images shown in Figure 8 further illustrate the higher hydrophilic membrane surface of the ZnO-MoS 2 PCM. The significant reduction in the contact angle of the modified membrane is attributed to the hydrophilic nature of SDBS and ZnO, which has also contributed free hydroxyl groups on the surface of the ZnO-MoS 2 PCM.
Membranes 2020, 10, x FOR PEER REVIEW 10 of 18 height of the active ZnO-MoS2 was identified to be nearly ~6 microns based on the cross-section SEM imaging on the prepared ZnO-MoS2-modified PVDF membrane, which enabled us to infer a uniform thickness of the photocatalytic layer on the membrane surface. Hence, the orientation of a few layers of MoS2 has altered the pore morphology of the PVDF membrane for achieving improved filtration properties.

Influence of ZnO-MoS2 on Hydrophilicity and Porosity Characteristics
The contact angle was recorded to investigate the wettability of the surface-modified PVDF membrane against its unmodified counterpart. The average contact angle of the modified membrane is 41.04, which is lower than the virgin PVDF having the highest average contact angle of 72.24° due to its inherent hydrophobic nature. Images shown in Figure 8 further illustrate the higher hydrophilic membrane surface of the ZnO-MoS2 PCM. The significant reduction in the contact angle of the modified membrane is attributed to the hydrophilic nature of SDBS and ZnO, which has also contributed free hydroxyl groups on the surface of the ZnO-MoS2 PCM. The mean flow pore size was also calculated to be 0.20 µ m and 0.33 µ m for the virgin and the modified PVDF membranes, respectively ( Figure 9). In contrast to the virgin PVDF, the distribution of the cumulative porosity was across a wider pore diameter for the ZnO-MoS2 PCM. The improved pore volume could subsequently influence the permeation of the ZnO-MoS2-based PCM for filtration applications. The wet flow characteristics of the modified PVDF membrane achieving a higher porosity is due to the surface hydration of the MoS2 sheets, confirming the intercalation of -OH molecules. The inherent free volume of ZnO-MoS2 and intricate pores of the hybrid catalyst on the PVDF surface have offered enhanced porosity, as witnessed by the improved hydrophilic properties of the modified membrane. Moreover, the layered random orientation of the MoS2 nanosheets has also resulted in mitigating the larger pores/channels. The mean flow pore size was also calculated to be 0.20 µm and 0.33 µm for the virgin and the modified PVDF membranes, respectively ( Figure 9). In contrast to the virgin PVDF, the distribution of the cumulative porosity was across a wider pore diameter for the ZnO-MoS 2 PCM. The improved pore volume could subsequently influence the permeation of the ZnO-MoS 2 -based PCM for filtration applications. The wet flow characteristics of the modified PVDF membrane achieving a higher porosity is due to the surface hydration of the MoS 2 sheets, confirming the intercalation of -OH molecules. The inherent free volume of ZnO-MoS 2 and intricate pores of the hybrid catalyst on the PVDF surface have offered enhanced porosity, as witnessed by the improved hydrophilic properties of the modified membrane. Moreover, the layered random orientation of the MoS 2 nanosheets has also resulted in mitigating the larger pores/channels. Membranes 2020, 10, x FOR PEER REVIEW 11 of 18 Figure 9. Pore diameter vs. cumulative porosity of the virgin and the modified PVDF membranes.

Performance Evaluation of ZnO-MoS2 on MB Dye Degradation
The photocatalytic activity of the raw ZnO-MoS2 powder was evaluated by the photodegradation of MB dye (10 ppm) under a long-wave 365 nm UV-A light. Based on the correlation between the absorbance recorded for the MB concentration and wavelength (~660 nm), as shown in Figure 10a, the residual MB concentration as a function of illumination time was determined and plotted, as shown in Figure 10b, to study the photocatalytic activity of pristine ZnO and ZnO-MoS2 for two cycles. Under a long-wave UV light, the photocatalytic efficiency of the ZnO-MoS2 composite is higher compared with pure ZnO with ZnO-MoS2 achieving 97.21% degradation and ZnO achieving 89% after 180 min, and the hybrid ZnO-MoS2 displayed a better initial efficiency of 87.12% compared with ZnO showing only 56.89% after 15 min, resulting from its improved adsorption of MB on to its surface. The efficient contact between the interface of the heterojunction of ZnO and MoS2 enables the charge separation of the electron-hole pairs owing to the diffusion of electrons in the ZnO nanoparticles to MoS2 at the interface [23]. The improved charge carrier density gradient and subsequent generation of an internal electrostatic field together influence the charge separation of the photogenerated electron−hole pairs in the hybrid ZnO-MoS2 composite, thus leading to an enhanced photocatalytic performance in the MB dye degradation. The photocatalytic degradation of MB by ZnO-MoS2 has been reported to follow pseudo first-order kinetics [46], governed by the rate equation where k is the rate constant (min −1 ), and t is the illumination time in minutes. Figure 10c shows the plot of

Performance Evaluation of ZnO-MoS 2 on MB Dye Degradation
The photocatalytic activity of the raw ZnO-MoS 2 powder was evaluated by the photodegradation of MB dye (10 ppm) under a long-wave 365 nm UV-A light. Based on the correlation between the absorbance recorded for the MB concentration and wavelength (~660 nm), as shown in Figure 10a, the residual MB concentration as a function of illumination time was determined and plotted, as shown in Figure 10b, to study the photocatalytic activity of pristine ZnO and ZnO-MoS 2 for two cycles. Under a long-wave UV light, the photocatalytic efficiency of the ZnO-MoS 2 composite is higher compared with pure ZnO with ZnO-MoS 2 achieving 97.21% degradation and ZnO achieving 89% after 180 min, and the hybrid ZnO-MoS 2 displayed a better initial efficiency of 87.12% compared with ZnO showing only 56.89% after 15 min, resulting from its improved adsorption of MB on to its surface. The efficient contact between the interface of the heterojunction of ZnO and MoS 2 enables the charge separation of the electron-hole pairs owing to the diffusion of electrons in the ZnO nanoparticles to MoS 2 at the interface [23]. The improved charge carrier density gradient and subsequent generation of an internal electrostatic field together influence the charge separation of the photogenerated electron−hole pairs in the hybrid ZnO-MoS 2 composite, thus leading to an enhanced photocatalytic performance in the MB dye degradation. The photocatalytic degradation of MB by ZnO-MoS 2 has been reported to follow pseudo first-order kinetics [46], governed by the rate equation − ln C C0 = kt where k is the rate constant (min −1 ), and t is the illumination time in minutes. Figure 10c shows the plot of − ln C C0 vs.

Flux Performance
In complementary to the improved hydrophilicity and porosity, the pure water permeability of the ZnO-MoS2 PCM has increased 25% relative to the virgin PVDF as shown in Figure 11. There has been a steady flux decline, which is evident as a result of the membrane compaction under a hydraulic pressure. It is also apparent that the interlayer spacing of the stacked MoS2 in its completely hydrated state has exhibited a substantial improvement of the water flux and transport nature. Similarly, the flux values from the MB dye treatment, as shown in Figure 12, indicate that the MB dye permeation from the ZnO-MoS2 PCM is ~70.86% higher than the virgin PVDF membrane, which is attributed to the smooth MoS2 nanochannels causing a lower hydraulic resistance [19]. After exposure to UV illumination at 60 min, the MB flux of the ZnO-MoS2 PCM has exhibited a considerable increase of about 20% at the end of the 2 h of UV exposure. This prominent effect is In complementary to the improved hydrophilicity and porosity, the pure water permeability of the ZnO-MoS 2 PCM has increased 25% relative to the virgin PVDF as shown in Figure 11. There has been a steady flux decline, which is evident as a result of the membrane compaction under a hydraulic pressure. It is also apparent that the interlayer spacing of the stacked MoS 2 in its completely hydrated state has exhibited a substantial improvement of the water flux and transport nature. Similarly, the flux values from the MB dye treatment, as shown in Figure 12, indicate that the MB dye permeation from the ZnO-MoS 2 PCM is~70.86% higher than the virgin PVDF membrane, which is attributed to the smooth MoS 2 nanochannels causing a lower hydraulic resistance [19]. After exposure to UV illumination at 60 min, the MB flux of the ZnO-MoS 2 PCM has exhibited a considerable increase of about 20% at the end of the 2 h of UV exposure. This prominent effect is certainly attributed to the photocatalytic activity of the hybrid ZnO-MoS 2 layer that eventually reduced the fouling layer resistance caused by the MB dye solutes [47]. The surface-deposited layer has offered to overcome the membrane resistance and enabled to improve the productivity while rejecting the MB feed solution. The effect of the surface properties of the ZnO-MoS 2 PCM in overcoming the trade-off between permeability and rejection [48] is evident from the MB dye rejection performance discussed in further sections.
certainly attributed to the photocatalytic activity of the hybrid ZnO-MoS2 layer that eventually reduced the fouling layer resistance caused by the MB dye solutes [47]. The surface-deposited layer has offered to overcome the membrane resistance and enabled to improve the productivity while rejecting the MB feed solution. The effect of the surface properties of the ZnO-MoS2 PCM in overcoming the trade-off between permeability and rejection [48] is evident from the MB dye rejection performance discussed in further sections. Membranes 2020, 10, x FOR PEER REVIEW 13 of 18 certainly attributed to the photocatalytic activity of the hybrid ZnO-MoS2 layer that eventually reduced the fouling layer resistance caused by the MB dye solutes [47]. The surface-deposited layer has offered to overcome the membrane resistance and enabled to improve the productivity while rejecting the MB feed solution. The effect of the surface properties of the ZnO-MoS2 PCM in overcoming the trade-off between permeability and rejection [48] is evident from the MB dye rejection performance discussed in further sections.   In order to realise the synergistic effect of photocatalysis and membrane filtration performance, the as-prepared ZnO-MoS 2 PCM was subjected to only pressure-assisted filtration under dark. After 1 h, the MB dye rejection rate of the virgin PVDF and ZnO-MoS 2 PCM was found to be about 8.5% and 27%, respectively, thus markedly indicating no significant influence of a photocatalytic layer under dark. When exposed to UV illumination, the MB concentration reduced to about 0.12 mg/mL in 30 min with a maximum degradation of about 99.95% in 2 h (Figure 13).
In order to realise the synergistic effect of photocatalysis and membrane filtration performance, the as-prepared ZnO-MoS2 PCM was subjected to only pressure-assisted filtration under dark. After 1 h, the MB dye rejection rate of the virgin PVDF and ZnO-MoS2 PCM was found to be about 8.5% and 27%, respectively, thus markedly indicating no significant influence of a photocatalytic layer under dark. When exposed to UV illumination, the MB concentration reduced to about 0.12 mg/mL in 30 min with a maximum degradation of about 99.95% in 2 h (Figure 13). This clearly shows a significant improvement in the MB rejection performance of the ZnO-MoS2 PCM and evidences the pronounced synergism of photocatalysis and filtration. The catalytically active edges of MoS2 induced the photocatalytic production of radicals which bind to the pollutants, thus resulting in a complete mineralisation [29]. The enhanced photocatalytic performance of PVDF/MoS2-ZnO is attributed to the catalytically active site of MoS2 wherein stacking few layers of the hybrid ZnO-doped MoS2 sheets [21] on the PVDF membrane substrate has offered advantageous effects on improving the photocatalytic-assisted dye filtration when compared with a randomly packed or reaggregated ZnO-MoS2 dispersion suspended in an MB dye solution. Further, the influence of surface-selective pore characteristics has also been evident from the extent of the rejection rate of about 45.5% achieved by the ZnO-MoS2 PCM membrane under no UV irradiation, as against the virgin PVDF membrane exhibiting only a 25% rejection.

Reusability of ZnO-MoS2 PCM
In order to evaluate the reusability of the ZnO-MoS2 PCM, a UV-assisted membrane filtration was performed on the same membrane back washed with water for 10 min and fed with the fresh MB dye solution. The performance, as measured by the permeate flux during the second run, has shown a decrease of about 30% compared with the initial run. The reduction in MB flux rate to about 118 Lm −2 h −1 has been observed during the second run as against 134 Lm −2 h −1 obtained for the initial filtration run. However, the rejection rate of the reused ZnO-MoS2 PCM has only reduced marginally (~1%), thus demonstrating the stability and restoration of the photocatalytic efficiency of the ZnO-MoS2 layer (Figure 14), reaching nearly a 99.45% removal of MB dye. The reusability of the prepared ZnO-MoS2 PCM is much faster, unlike using the hybrid ZnO-MoS2 in powder form, where additional steps like centrifugation and drying are required to recover the photocatalyst [49]. This This clearly shows a significant improvement in the MB rejection performance of the ZnO-MoS 2 PCM and evidences the pronounced synergism of photocatalysis and filtration. The catalytically active edges of MoS 2 induced the photocatalytic production of radicals which bind to the pollutants, thus resulting in a complete mineralisation [29]. The enhanced photocatalytic performance of PVDF/MoS 2 -ZnO is attributed to the catalytically active site of MoS 2 wherein stacking few layers of the hybrid ZnO-doped MoS 2 sheets [21] on the PVDF membrane substrate has offered advantageous effects on improving the photocatalytic-assisted dye filtration when compared with a randomly packed or reaggregated ZnO-MoS 2 dispersion suspended in an MB dye solution. Further, the influence of surface-selective pore characteristics has also been evident from the extent of the rejection rate of about 45.5% achieved by the ZnO-MoS 2 PCM membrane under no UV irradiation, as against the virgin PVDF membrane exhibiting only a 25% rejection.

Reusability of ZnO-MoS 2 PCM
In order to evaluate the reusability of the ZnO-MoS 2 PCM, a UV-assisted membrane filtration was performed on the same membrane back washed with water for 10 min and fed with the fresh MB dye solution. The performance, as measured by the permeate flux during the second run, has shown a decrease of about 30% compared with the initial run. The reduction in MB flux rate to about 118 Lm −2 h −1 has been observed during the second run as against 134 Lm −2 h −1 obtained for the initial filtration run. However, the rejection rate of the reused ZnO-MoS 2 PCM has only reduced marginally (~1%), thus demonstrating the stability and restoration of the photocatalytic efficiency of the ZnO-MoS 2 layer (Figure 14), reaching nearly a 99.45% removal of MB dye. The reusability of the prepared ZnO-MoS 2 PCM is much faster, unlike using the hybrid ZnO-MoS 2 in powder form, where additional steps like centrifugation and drying are required to recover the photocatalyst [49]. This confirms that the reusability of the ZnO-MoS 2 PCM is quite facile and efficient upon a comparison with the hybrid ZnO-MoS 2 photocatalyst in powder form.

Conclusions
This photocatalytic membrane fabrication by depositing the exfoliated dispersion of the hybrid ZnO-MoS 2 photocatalyst implicates the proof of the concept of utilising the highly stable organic polymeric substrate with a simultaneous catalytic activity for an overall improvement of a water/ wastewater treatment. The influence of catalyst immobilisation on the PVDF membrane substrate has offered superior surface functional properties. Photocatalytic degradation studies have also proved that the heterogeneously structured ZnO-MoS 2 nanocomposite is photocatalytically more efficient than pristine ZnO. Robust advances in membrane surface modification strategies enable to utilise the multi-functional properties of MoS 2 without altering its desired molecular channels to eventually result in breakthrough advances in the field of membrane separation processes. The current work also brings scope for research studies on precisely tuning the physico-chemical properties of surface-modified photocatalytic membranes to design scalable and continuous photocatalytic membrane reactors.