Enhancing Morphology and Separation Performance of Polyamide 6,6 Membranes By Minimal Incorporation of Silver Decorated Graphene Oxide Nanoparticles

Nanomaterials can be incorporated in the synthesis of membrane to obtain mixed-matrix membrane with marked improvement in properties and performance. However, stability and dispersion of the nanomaterials in the membrane matrix, as well as the need to use high ratio of nanomaterials for obvious improvement of membrane properties, remain a major hurdle for commercialization. Hence, this study aims to investigate the improvement of polyamide 6,6 membrane properties with the incorporation of silver nanoparticles decorated on graphene oxide (Ag-GO) nanoplates and at the same time focus is given to the issues above. Graphene oxide nanoplates were synthesized using the modified Hummers’ method and decorated with silver before embedded into the polyamide 6,6 matrix. Physicochemical characterizations were conducted on both nanoplates and the mixed-matrix Ag-GO polyamide 6,6 membrane. The issues of Ag agglomeration and leaching were not observed, which could be attributed to the decoration of Ag on GO that helped to disperse the nanomaterials and provided a better anchor point for the attachment of Ag nanoparticles. The synthesized membrane showed marked improvement regarding flux (135% increment) and antifouling (40% lower irreversible fouling), which could be ascribed to the more negative charge of membrane surface (−14 ± 6 to −31 ± 3.8 mV) and hydrophilicity (46% enhancement) of the membranes. With minimal embedment of Ag nanoparticles, the membrane showed superior antibacterial property where the E. coli bacteria could not form a single colony on the membrane surface. Overall, the decoration of Ag on GO nanoplates could be a promising approach to resolve the agglomeration and leaching issues as well as reduce the amount of precious Ag in the synthesis of Ag-GO polyamide 6,6 membrane.

United States). In addition, the transmission electron microscopy (TEM) was utilized to observe the Ag-GO nanocomposite (Philips CM200, model JEOLJEM 2100, Netherlands). The surface charge of nanomaterial was assessed by zeta potential measurement using Malvern Zeta Sizer Nano ZS (Malvern Instruments, UK) by applying field strength of 25 V/cm.

Membrane Fabrication.
A wet phase inversion method has been chosen for the fabrication of polyamide-Ag-GO blend membrane. The casting solution was formulated by dispersing the Ag-GO nanoplates in formic acid. The polymer solution was prepared with different weight ratios of polymer (polyamide 6,6) to solvent (formic acid), as summarized in Table 1. The nanoplates were ultra-sonicated for 30 minutes before mixing with the polymer solution. The mixture was heated (using silicon oil bath) at a constant temperature of 60 °C with the slow stirring rate of 300 rpm for 5 hr to prevent the formation of any air bubbles. The solution mixture was then ultra-sonicated for 30 minutes to ensure a better dispersion of the GO nanoplates in the mixture. During the fabrication process, a small amount of the casting solution was poured onto a clean glass plate and cast using Filmographe Doctor Blade 360099003 (Braive Instrument, Germany). The thickness of the membranes was set at 0.2 mm using the casting knife. After 15 s of exposure to the air, the glass plate with the casted solution was quickly immersed in a water bath and left for 24 hours to ensure complete phase inversion.
Membrane Characterization. The hydrophilicity of the membrane surface was represented by contact angle that was measured using Kruss GmbH FNY12MKE Easy Drop, Germany under ambient conditions. Fourier-transform infrared spectroscopy (FTIR) analysis was conducted using a Nicolet 6700 Thermo Scientific-FITR spectrometer (United States) to identify the functional groups present on the membrane. Membrane surface zeta potential was measured using the Malvern Surface Zeta Potential Cell, Malvern Instruments, UK. The zeta potential of membrane surface was measured in 0.1 mM NaCl at pH 7 using 300-350 nm latex particles as the tracer particles (DTS1235 Malvern UK). The overall membrane porosity (ε) was calculated using the gravimetric method 29 , as defined in Equation (1): where ω 1 is the wet membrane weight (g), ω 2 is the dried membrane weight (g), A is the membrane surface area (m 2 ), l is the membrane thickness (m), and d w is the water density (998 kg/m 3 ). Using the porosity data and Guerout-Elford-Ferry equation 35,36 , the relative pore sizes of the fabricated membranes were calculated using Equation (2) Membrane Performance Evaluation. Membrane Flux. Sterlitech TM HP4750 dead end stirred cell was used to evaluate the flux of the fabricated polyamide membrane. In order to remove the residual chemicals within the produced membrane and to obtain a steadier solution flux, compaction was carried out for 20 min at a pressure of 5 bars. After the compaction step, membrane pure water flux was conducted for 120 min at a pressure of 4 bars. The permeate flux was calculated using Equation (3): where J is the measured permeate flux (L/m 2 h), ΔV is the permeate cumulative volume (L), A is the effective membrane surface area (m 2 ), and ΔT is the filtration duration (h).

Membrane Fouling Analysis and Rejection Calculation.
After the water flux tests, a permeation experiment was also carried out using the Sterlitech stirred cell with 3 g/L BSA solution at 4 bars. A reservoir filled with 1 L of BSA solution at 3 g/L was connected to the stirred cell to continuously refill the stirred cell throughout the filtration process ( Fig. 1). A constant feed solution concentration of 3 g/L has been assumed in this study by averaging the bulk solution concentration. A constant filtration duration of 150 min was set in this study, and a mean concentration of permeate solution was then measured. Variation in solute concentration from the bulk solution to membrane surface during the filtration process was not directly measured in this study because it will involve complicated analyses and sophisticated instruments that can perform in situ measurement. Accumulation of foulants on the membrane surface, however, can be dynamically represented by continuous solution flux measurement. The flux variation profiles in this study have been represented by normalized flux by taking the pure water flux at the same operating condition as the basis of comparison. BSA was selected for this fouling test because it was one of the most common foulants used in membrane performance testing 37,38 . The normalized flux (F n ) was measured for 150-minute filtration period based on the ratio of flux for BSA solution (j BSA ) to pure water flux (j w ). Flux recovery ratio (FRR) was calculated based on the ratio of recovered ultrapure water flux after BSA solution (j w2 ) to pure water flux (j w ). Afterwards, the membranes were rinsed for 15 min using ultrapure water. Next, the membrane permeation flux with ultrapure water was conducted again. The total fouling ratio (R t ), reversible fouling ratio (R r ) and irreversible fouling ratio (R ir ) of the membranes were measured using Equations (4)(5)(6)(7)(8) to analyse the overall fouling pattern: where j BSA is the flux for BSA solution, j w is pure water flux, and j w2 is the recovered ultrapure water flux after BSA filtration.
In addition to BSA (3 g/L), Congo red (0.1 g/L) was also chosen as the foulant for rejection test due to its presence in various industrial effluents such as textiles, printing, dyeing, paper, and plastic 39 .
Membrane rejection capability for Congo red and BSA was calculated using Equation (9): where C p and C f are the concentrations of permeate and feed solutions, respectively.

Membrane Stability Test.
After each filtration test for 5 hours, the filtrate water was collected and sent to ICP-MS (Elan 9000 PerkinElmer, USA) analysis to detect the presence of Ag. The presence of Ag in the filtrate will reflect the stability of Ag in the membrane matrix.  Zeta Potential Measurement. Surface Zeta potential was measured using the Malvern Surface Zeta Potential Cell. The zeta potential of PES membrane surface was measured in 0.1 mM NaCl at pH 7 using 300-350 nm latex particles as the tracer particles (DTS1235 Malvern UK).
Membrane Antibacterial Analysis. Membrane antimicrobial analysis was conducted using Escherichia coli (E. coli) as a model bio-foulant 40 . E. coli were first cultured in a nutrient broth. After the growth of stock solution achieved 109 × 10 7 CFU/mL (OD600 1-1.5), the E. coli were serially weakened to 20 × 10 7 CFU/mL from the stock using nutrient broth medium 41 . After this stage, the fabricated membranes were soaked in the diluted E. coli solution for 10 min. They were then placed on the nutrient agar plates and incubated at 35 °C overnight. The colony forming units (CFUs) on the membrane surface were then observed using FESEM.

Results and Discussion
Characterization of Ag-GO nanoplates. Figure 2 demonstrates the FTIR spectroscopy of the synthesized GO and Ag-GO. The sharp peak at 1725 cm −1 and broad peaks around 3363.1 cm −1 and 1302 cm −1 can be seen in both GO and Ag-GO, which correspond to the stretching vibration and deformation vibration of O-H groups 42 .
The two bands at 1066 cm −1 and 1200 cm −1 , originally from the C=O stretching vibrations of alkoxy groups were seen in the spectrum too. These data supported the existence of carboxylic acid groups on the GO surface. The FTIR spectrum of Ag-GO pointed out all the oxygenated functional groups of graphene oxide. However, an absolute decrease in the absorption intensity of the functional group bands has been observed for the Ag-GO sample. This decrease could be due to the shadowing effect of the silver nanoparticles decorated on the GO nanoplates 43 , as supported by XRD analysis. The XRD pattern for the Ag-GO (Fig. 3) showed peaks at about 38, 44, 64 and 77°, whereby these peaks can be assigned to the Miller indices of (111), (200), (220) and (311) crystallographic planes of face-centred cubic (fcc) of Ag nanoparticles, respectively 44 . Indeed, TEM micrograph in Fig. 4 shows that the silver nanoparticles (with an average size of 4 nm) were uniformly distributed across the GO sheet. Such observation indicates that agglomeration phenomenon did not occur in this case. Hence, the decoration of Ag nanoparticles on GO nanoplates has successfully resolved the issue of Ag agg Membrane characterization. Figure 5 displays the functional groups present in polyamide 6,6 and Ag-GO composite membranes. According to the FTIR spectra, the vibrational frequencies of peaks for both polyamide 6,6 and Ag-GO/polyamide 6,6 samples are determined and listed in Table S1 (Supplementary Document). The spectra obtained through FTIR analysis confirmed the standard footprints of polyamide 6,6 1 . The frequency bands of polyamide 6,6 and Ag-GO/polyamide 6,6 show some slight shifting (between 3-9 cm −1 ) after mixing with Ag-GO. The band shift exhibited that the Ag-GO has been successfully embedded into polyamide 6,6 matrixes, in which all peak intensities were reduced 2 .
The cross-sectional structure of the membranes (Fig. 6) presents a sponge-like structure for all the fabricated membranes. It can be observed that the membranes embedded with Ag-GO are more porous compared to the pristine polyamide 6,6 membrane. The addition of Ag-GO induced a high porosity structure. This development is favourable to the membrane permeability as it was generally reported that membrane with higher porosity recorded higher water flux 29 . In order to determine the distribution of silver nanoparticles in the matrix of the membranes, the EDX analysis and mapping mode has been employed. Figure 6(e) illustrates the EDX image of the membrane with 0.8 wt% of Ag-GO (NY4). The existence of the silver nanoparticles can be proven by the peaks  Fig. 7(f) showed a perfect distribution of the silver nanoparticles in the matrix of polyamide membrane. This clearly shows that Ag-GO nanocomposite was nicely embedded into the polyamide 6,6 polymer matrix without agglomeration issue. The overall porosities and the pore radius of the membranes can be summarized through the observation in Fig. 7(a). The fabricated membranes exhibited porosities in the range of 60 to 80%, which are greatly desirable for the polyamide 6,6 membranes. The modified membranes showed higher porosities compared to the pure polyamide 6,6 membranes, supporting the physical observation from FESEM in Fig. 6, which could be ascribed to the addition of Ag-GO nanoplates to the membrane matrices 29,36,45 . The pore radius of the fabricated membrane as presented in Fig. 7(b) revealed that the initial addition of Ag-GO increases the pore radius while the addition of 1 wt% of Ag-GO reduces the size of pore radius. This phenomenon can be related to the viscosity of    casting solution and mass transfer rate between the phases as reported by previous studies 46 . Coincidently, similar pore radii also can be obtained using Hagen-Poiseuille equation by assuming the cylindrical length as membrane thickness (0.2 mm). Ag-GO nanoplates have been reported to be one of the most hydrophilic derivatives of graphene-family 30,31,43,47,48 . The addition of hydrophilic nanoplates to the casting solution can decrease the thermodynamic stability of the casting solution (weakening the van der Waals interaction between the polymer and solvent molecules). This enabled the solvent molecules to diffuse rapidly from the polymer matrix to the coagulation bath and thus resulted in the formation of larger pore size and porosity structure of the polymer membrane 28 . As illustrated in Fig. 8, the addition of Ag-GO to the polyamide membrane has a direct impact on the contact angle reduction of the fabricated membranes. The pristine polyamide 6,6 membrane recorded the highest contact angle of 73.4 ± 6.12°. The addition of 0.2 to 1 wt% of Ag-GO reduced the water contact angle to 54.2 ± 6.03°, 43.6 ± 0.65°, 36.1 ± 5.14° and 39.4 ± 7.76°, respectively. The hydrophilicity of the membrane can be indicated by its surface wettability. The presence of epoxy, carboxyl, and the hydroxyl functional groups of the GO and the synergic effect presence of Ag-GO in the membrane structure can lead to the increment of interface energy in the polyamide membranes 28 . Consequently, the adhesive forces between the water molecules and the membrane surface will be higher than the cohesive forces exist between the molecules of the water. Thus the water molecules will be attracted to the surface of the membrane rather than to each other, resulting in lower contact angle (better hydrophilicity) 49 .
Apart from the improvement of hydrophilicity, Fig. 9(a) shows the tensile strength test of the fabricated membranes where a significant mechanical strength enhancement was observed. The peak load for the pure polyamide membrane was 1159.5 g, and the deformation peak was at 7.9 mm. On the other hand, the peak load for the NY 5 membrane (1% of Ag-GO) was at 1733 g, and the deformation peak was at 13.6 mm, with an increment of 39.69% in mechanical strength. This phenomenon could be attributed to the attractive forces between the Ag-GO nanoplates structure and the polyamide 6,6 membrane, resulting in the formation of sea-island structure between polyamide 6,6 and Ag-GO 50 . Homogeneous dispersion of Ag-GO in the matrix of membrane resulted in efficient and uniform load transfer throughout the membrane matrix and thus increasing the mechanical properties of the modified membranes. In a nutshell, the enhanced mechanical properties were due to the uniform dispersion of Ag-GO nanoplates and effective stress transfer 50,51 .  Figure 9(b) illustrates the surface zeta potentials of the fabricated membranes. The addition of the Ag-GO to the structure of the membrane propels the charge towards more negative range, changing from −14 ± 6 to −31 ± 3.8 (mV) for pure membrane and membrane with 1 wt% of Ag-GO, respectively. This phenomenon could be attributed to the addition of negatively charged Ag-GO (−47.4 mV at pH 7) to the membrane matrix 52 , and could be favourable for biofouling prevention as the modification of the polymer surface charge has been proven to be an effective mean of biofilm prevention 53,54 . Membrane Performance Evaluation. Overall, the incorporation of Ag-GO nanoplates has improved the membrane properties, and the issue of nanoparticles agglomeration was resolved by the decoration of Ag onto GO nanoplates. Figure 10 shows that the membrane pure water flux is linked directly with hydrophilicity and porosity structure of the fabricated membranes (Figs 7 and 8). The incorporation of 0.8 wt% Ag-GO in the matrix of the membrane (NY4) resulted in the highest water flux, corresponding with the lowest contact angle and largest pore radius. Addition of nanoplates in the membrane matrices in this study increased the membrane porosity, which could be explained with the presence of holes in nanoplates and formation of nanochannels on membrane surfaces as observed in the SEM image on Fig. 6. This could be one of the reasons for higher membrane fluxes as observed in this study 55 . On the other hand, the membranes incorporated with 1 wt% (NY5) of the nanoplates recorded a lower flux. The addition of 1 wt% might increase the viscosity of the casting solution and thus slow down the mass transfer between the solvent and non-solvent phase. The slow mass transfer between the solvent and non-solvent phase causes the formation of smaller pores size (Fig. 7b) as explained in the previous section. This phenomenon was in accordance with studies reported by previous authors 18,19,23,56,57 .

Membrane Stability Tests.
The results from ICP-MS of the filtrate water show no trace of silver in the permeate (Table S2 in the Supplementary Document), proving that the silver did not leach out of the membranes after 5 hours of filtration. As reported by other authors, pure silver nanoparticles have the tendency to leach out from the membrane. However, this phenomenon did not occur in the current study, showing that GO provided numerous anchor points for the attachment of Ag nanoparticles 24,58 .  Fouling and Rejection analysis. The flux performance for BSA filtration process is presented in Fig. 11(a).
From this figure, it can be concluded that the fouling tendency (reflected by the steepness of flux decline) of the membrane decreases with the increase of Ag-GO ratio in membrane matrix. However, NY4 is an exception. This might be due to the fact that the foulant particles can easily penetrate and clog the larger membrane pores appear on NY4 membrane 59 . The flux recovery ratios (FRR) of membranes are depicted in Fig. 11(b). In general, a high FRR% indicates a better fouling resistance for the membrane. It is obvious that the FRR% of the modified membranes is higher than the pure membrane (~25% to ~60% respectively). This indicates that the blending of Ag-GO increased the organic fouling resistance of the membranes. The highest FRR% was related to NY5 membrane  1%. It is well established that hydrophilic surface can adsorb water molecules and form a water layer on the membrane surface, which retards the absorption of protein and other fouling agents to foul the membrane 19 . The trend that was observed for the flux recovery ratio is matched by the contact angle reduction of the membranes. However, the NY4 membrane is an exception to this trend due to the larger pore radius, possibly caused by the protein that might be able to penetrate into the pores and cause the fouling. It could be seen that the effect of the pore size overwhelmed the improvement in terms of surface charge ( Fig. 9) and hydrophilicity. Furthermore, Fig. 11(c) shows the overall fouling tendency of the membranes. The total fouling of the mixed matrix membranes was considerably reduced from 92% for NY1 to 49.72% for NY5. Reversible and irreversible fouling of the membranes is also demonstrated in Fig. 11(c). These results showed that the modified membranes displayed remarkable antifouling properties in comparison to the pristine polyamide 6,6 membrane. The irreversible fouling of the polyamide 6,6 membrane was reduced from 76% to 37% (NY5). The improvement could be attributed to the more negatively charged and hydrophilicity of the NY5 membrane that helped to repel the foulant from persistently adsorbed to the membrane surface 1 . This exhibits that reversible fouling was the dominant phenomenon in total fouling for the modified membranes and hydraulic cleaning would easily reverse the fouling effect. The phenomenon is attributed solely to the fact that a low percentage of the nanoplates were uniformly and perfectly distributed across the membrane matrix providing enhanced surface properties such as charge and hydrophilicity that resulted in lower overall fouling. Thus, this proves that the addition of small amount of Ag-GO to the membrane helped to alter the membrane surface properties and enhanced the overall properties of the fabricated membrane.
According to Fig. 12, the rejection percentage of Congo red is relatively high (89% to 92%) despite the fact that the pore size and water flux of the modified membranes increased with the addition of Ag-GO. This phenomenon was possibly attributed to the hydrophilicity enhancement of the modified membranes. The establishment of highly hydrophilic membrane structure enhances the affinity of membrane surface to water molecules rather than to Congo red, resulting in lower hydraulic resistance. On top of that, the negatively charged surface of the membrane might repel the negatively charged Congo red too 60 . Hence, the change in pore size and water flux does not affect the rejection dramatically.
On the other hand, similar findings were observed for BSA rejection tests. All the modified membranes rejected more than 90% of BSA. As shown in Fig. 12, the addition of 0.8 wt% and 1 wt% of Ag-GO would result in even better BSA rejection, approaching 95% and 97%, respectively. The rejection capabilities of the membranes were not compromised, even though the modified membranes have higher flux and bigger pore radius. This could be attributed to the fact that the modified membranes have higher hydrophilic surface and negative charge that could minimize the attachment of BSA on the membrane surface 35,61,62 .
Looking at both sets of rejection data, it can be concluded that the rejection capability of polyamide 6,6 membranes is thought to occur in few different mechanisms. Firstly, sieving (steric) rejection that depends on the pore size. However, the addition of the graphene oxide increased the pore size of the membrane and porosity simultaneously. The increase in pore size and porosity contributed to higher membrane fluxes. Surprisingly, the rejection of both pollutants were not compromised much from larger pore size and higher flux. This indicated the role played by other factors that overwhelmed the negative effect of larger pore size on the rejection. The interactions between Ag-GO and polyamide made the membrane more hydrophilic (as can be seen by the decrease in contact angle), which would reduce the adsorption of organic pollutants on the surface of the membrane. Thus, it reduced the tendency of foulants to permeate through the membrane. The establishment of highly hydrophilic membrane structure could increase the affinity of the membrane to water rather than the pollutant, resulting in lower hydraulic resistance. Besides, the enhanced rejection properties could be correlated with the zeta potential of the membrane. The negatively charged membranes would cause Donnan repulsion to the pollutants that prevent it from permeating through the membrane. The synergistic effect of surface hydrophilicity and charge repulsion resulted in the observed membrane performance, where the rejection capability was not compromised even though at higher flux. Antibacterial Performance. Figures 13(a,b) illustrates the FESEM images of the pristine polyamide 6,6 membrane after the antibacterial test. Theoretically, surfaces with higher hydrophobicity are more immune to bacterial growth, as the microbial adhesion to the surface will be more difficult 63,64 . However, it can be seen that the surface of the pristine polyamide 6,6 membrane is covered by bacterial colonies. E. coli grew into colonies and formed a layer that completely covered the pores of the pure polyamide 6,6 membrane 65 . In contrast, the addition of Ag-GO nanoplates to the membrane stopped the colonial growth of the bacteria. Figures 13(c) and 12(d) show the FESEM image of the membrane containing 0.8 wt% of the Ag-GO nanoplates. There was no colony formation or bacterial growth of the surface of the membrane, proving that the decoration of silver nanoparticles of graphene oxide is an effective way of biofilm formation control. As aforementioned, the membrane surface showed a more negative charge when embedded with Ag-GO. The synergetic effect of static antimicrobial of GO and oligodynamic antimicrobial of silver attributed to the production of a membrane that successfully restricted the growth of gram-negative E. coli despite the fact that E. coli shows resistive towards most antibiotics 66 . Benchmarking with Past Studies. The findings from this study were benchmarked against other similar studies on the issues of nanoparticles agglomeration, leaching and amount of nanoparticles used in the synthesis of membrane. It is not possible to draw a solid correlation between the amount of Ag used in the study and the performance improved by the incorporation of nanoparticles. However, the best performing membrane in this study only employed 0.2 wt% of pure silver, which is much lower than most of the amount reported in other literature 24,[67][68][69][70][71][72][73][74][75] . In this study, flux improvement of 133% and irreversible fouling reduction of 40% were observed while the rejection maintained at the same level.
There are several literatures reported on the leaching of Ag after filtration as reported in Table 2. It was discovered that all the studies with leaching issues did not decorate the Ag nanoparticles on any medium. The lack of strong anchor sites for Ag nanoparticles could be the reason that contributed to the leaching issue. Such a phenomenon was not observed in this study as the GO provided a better anchor point for the Ag nanoparticles.
The agglomeration of pure silver nanoparticles was pointed out by many authors 24,68,74 while some of the past studies did not reveal the distribution of nanoparticles in their membranes 27,67,71,72 . Various modification methods (as shown in Table 2) have been attempted to reduce the agglomeration with success 24,[69][70][71]74 . However, based on their reports, leaching still remained a problem due to the lack of proper anchor sites for Ag nanoparticles. In this study, no agglomeration of silver nanoparticles has been observed. Hence it could be concluded that the decoration of Ag on GO help alleviating the issue of agglomeration and leaching of the nanoparticles simultaneously.

Conclusion
This study has shown that the decoration of Ag onto GO nanoplates prevented the agglomeration of Ag especially when embedded into the membrane matrix. In addition, the GO provided the anchor sites for the Ag to attach and thus eliminated the leaching issue. The EDX mapping result presented that the silver nanoparticles distribution is even and there was no agglomeration observed. Overall, the incorporation of minimal amount of Ag-GO brought marked improvement on the membrane properties, such as flux, rejection, fouling, and hydrophilicity. Generally, Ag-GO enlarged the mean pore radius and enhanced the porosity of the polyamide membrane. Membranes embedded with Ag-GO possessed better hydrophilicity (46% enhancement) and stronger negative surface charge (−14 ± 6 to −31 ± 3.8 (mV)) as compared to pristine membrane. The combination of all these enhanced properties contributed to higher water flux (135% increment) and lower fouling propensity (40% enhancement) compared to pristine polyamide membrane. Despite the increment in flux and pore radius, the rejection of the membranes was not compromised and remained higher than 95% for both BSA and Congo red. This was due to the stronger electrostatic repulsion (more negative surface charge) and hydrophilicity of the membranes. The antimicrobial test results confirmed that the addition of Ag-GO prevented the bacterial growth on the surface of the membrane. The synergistic effect of GO static antibacterial and oligodynamic of silver nanoparticles decorated on GO restricted the growth of gram-negative E. coli. Overall, the incorporation of minimal amount of Ag-GO nanoplates managed to enhance the properties of the polyamide 6,6 membranes without the issues of Ag agglomeration and leaching. This shows the encouraging potential of Ag-GO as a nanofiller for the fabrication of polyamide 6,6 membranes with superior properties in water and wastewater treatment processes.