Cellulose acetate based Complexation-NF membranes for the removal of Pb(II) from waste water

This study investigates the removal of Pb(II) using polymer matrix membranes, cellulose acetate/vinyl triethoxysilane modified graphene oxide and gum Arabic (GuA) membranes. These complexation-NF membranes were successfully synthesized via dissolution casting method for better transport phenomenon. The varied concentrations of GuA were induced in the polymer matrix membrane. The prepared membranes M-GuA2–M-GuA10 were characterized by Fourier transform infrared spectroscopy, scanning electron microscopy, transmission electron microscopy, atomic force microscope and bio-fouling studies. Thermal stability of the membranes was determined by thermogravimetric analysis under nitrogen atmosphere. Dead end nanofiltration was carried out to study the perm- selectivity of all the membranes under varied pressure and concentration of Pb(NO3)2. The complexation-NF membrane performances were significantly improved after the addition of GuA in the polymer matrix membrane system. M-GuA8 membrane showed optimum result of permeation flux 8.6 l m−2 h−1. Rejection of Pb(II) ions was observed to be around 97.6% at pH 9 for all the membranes due to electrostatic interaction between CA and Gum Arabic. Moreover, with the passage of time, the rate of adsorption was also increased up to 15.7 mg g−1 until steady state was attained. Gum Arabic modified CA membranes can open up new possibilities in enhancing the permeability, hydrophilicity and anti-fouling properties.

Formation of graphene oxide (GO) by Hummer's method. Graphite powder (2 g) and NaNO 3 (2 g) were dissolved in 50 ml of H 2 SO 4 . Beaker was placed in an ice bath at continuous stirring for 2 h, maintaining a temperature at 5 °C. After that KMNO 4 (12 g) was added in solution dropwise as an oxidizing agent, maintaining a temperature at 15 °C for 4 h. The solution was stirred continuously for 48 h until a brownish suspension was formed. Distilled water (100 ml) was added slowly into the beaker; reaction mixture was heated up to 98 °C for 1 h. Then 200 ml of distilled water was added again for further dilution while stirring was kept on. H 2 O 2 (10 ml) was added dropwise in a solution, till appearance of yellow color which showed reaction termination. For purification purpose the mixture was rinsed out with distilled water (10×). The final product was vacuum dried at 25 °C, resulting in the formation of fine powdered GO. Casting of membranes. The five sample mixtures of CA/VTES-GO (1-5 wt%) were casted in glass petri dishes carefully maintaining same thickness of each membrane. Then petri dishes were placed at 50 °C in vacuum oven for 24 h. These membranes were removed from petri dishes using doctor's blade. Five final synthesized membranes were analyzed for lead rejection to bring forth the optimum filler CA/VTES-GO (3 wt%) was further named as M-GuA0, established optimal membrane for Pb(II) rejection (52%), then proceeded for further treatment.
Fabrication of complexation NF network by incorporation of GuA. Gum Arabic (GuA) was added in different concentrations into M-GuA0 sample as shown in Table 1. Primarily six different solutions were stirred continuously for 2 h maintaining a temperature at 60 °C. These homogeneous solutions were casted as mentioned in "Casting of membranes" section.  Contact angle. Contact angle measurements were carried out using a Goniometer (DIGI DROP, KSV Instruments). The sessile drop method was used to measure the contact angle of de ionized water on the dehydrated surface of the synthesized membranes. After the discharge of distilled water, on the membrane surface the image was captured and contact angle was measured. The given data were the mean of five contact angle values for each membrane sample.
Bio-fouling properties. Bio-fouling test was prepared by using Escherichia coli according to JIS L 1902-2002 approach. Conical flasks having solution, were placed in autoclave at 121 °C for 15 min at 15 psi. 100 ml of DH5 alpha E. coli strain was injected into the flasks. Then membranes of different compositions M-GuA0 to M-GuA10 were added into it, flasks were put into incubator at 37 °C for 18 h. Optical density (OD) was measured by using a spectrometer at 600 nm.

Results and discussion
Fourier transform infrared spectroscopy. FTIR spectrum of VTES-GO, M-GuA0 (control) and modified membranes from M-GuA2-MGuA10 are given in Fig. 1. FTIR was employed to confirm the interaction between CA, GO and GuA. The observed band at 3431-3482 cm −1 was associated with the -OH stretching vibrations due to inter-molecular intra-molecular hydrogen bonds increased as concentration of Gum Arabic increased shown in (Scheme 2) 40,53 . The band observed from 2358 to 2400 cm −1 were specific for C-C bending at low vibration 54 . The GO sample showed a strong characteristic peak at 3431-3482 cm −1 for -OH stretching and 1640 cm −1 for aromatic C=C 55  Thermogravimetric analysis (TGA). Thermal degradation of control and modified membranes were studied in form of percentage weight loss given in Fig. 2. Three main steps were involved in thermal degradation analysis of polymer matrix membranes. First stage occurred between 30 and 250 °C, this degradation was due to removal of moisture contents and volatile matter from M-GuA0-M-GuA10. Second step involved onset temperature (T onset ) started from 250 to 400 °C showing 70 wt% loss . The reason was the degradation polymer backbone and by deacetylation of CA 62,63 . Third step involved the offset temperature (T offset ) started from 400 to 1160 °C, revealed carbonization of degraded products to ash because no change in mass occurred 64,65 . Table 2 revealed that degradation temperature of M-GuA8 was high as compare to control M-GuA0. The reason behind this was the homogenous dispersion of Gum Arabic in M-GuA8 as compared to control. M-GuA10 showed less degradation temperature, due to accumulation of Gum Arabic in the membrane which results in breakdown of main polymer chains at low temperature as compared to M-GuA8 27     Atomic force microscope (AFM). Figure 5 shows the Surface topography and roughness of the membranes. The root mean square (RMS) value of polymer matrix membranes increased as concentration of Gum Arabic was increased from M-GuA0-MGuA8. M-GuA0 showed 153 nm less than all other samples as shown in Table 1. The reason behind is the absence of GuA in it, which causes less valleys and ridges compared to other membrane. M-GuA2-MGuA6 the surface roughness values continuously increased from 210 to 310 nm, indicated that addition of GuA increased the surface roughness. Increment in valleys and ridges means increment in roughness parameter and adsorption 69,70 . RMS value of M-GuA8 was 361 nm which was optimum one showed better adsorption as compare to control membrane. The addition of Gum Arabic increased the hydrophilicity and surface adsorption in the membranes 70 . Sudden decrease in RMS value 220 nm of M-GuA10 due to greater accumulation of filler in vicinity of membrane, which will ultimately cause decrease in flux and adsorption rate 71 . Permeability raised up to a certain limit and then reduced when concentration of filler exceeded, because compactness occur in structure of membrane 67 .    Figure 6c shows the effect of change in pH (1-10) on Pb(II) % rejection. Sudden reduction was observed in case of pH 10. Whereas, at pH 1 Pb(II) % rejection was 56.7%. Optimum rejection (97.6%) was observed for M-GuA8 at pH 9. It was cleared from results that low (pH 1-6) was not obligatory for better rejection 74 . Furthermore, charge present on membrane surface is effected by pH and in turn influence the metal ion 76 . VTES-GO behave as absorbent in membrane, the adsorption capacity of metallic species was enhanced with pH. It was also observed that from pH (1-9) the absorption capacity of membrane reduced in acidic medium compared to basic medium. The reason was, functional groups -COOH and -OH present on VTES-GO, CA and Gum Arabic respectively, deprotonated in acidic medium 6 . Another fact was lower pH also leads to neutralization of functional group due to which absorption power of cation Pb(II) decreased on membranes. Furthermore, struggle between H 3 O + and metal ions leads to low adsorption 6,77 . In basic medium (pH 7-9), maximum Pb(II) rejection occurred due to conversion of Pb(NO 3 ) 2 to metal hydroxide. Electrostatic interaction is the reason when functional groups changes from -COOH and -OH to (-COOand -O -), which results high Pb(II) rejection. Moreover, at pH 10 rejection sudden decrease due to greater precipitation of metal hydroxides in solution except absorption 77 .   Fig. 6d. It was well-defined from figure that adsorption rate at zero time is 2.6 mg g −1 . Initially the value increased to 5.2 mg g −1 whereas, after 11 h values started continuously increasing from 10.4 to 15.7 mg g −11 till steady state was attained. The reason acclaimed the adsorption related to the hydrated radii of Pb(II). Hydrated radii of Pb(II) is 94 pm, smaller hydrated radii indicates that absorption of Pb(II) greater on the surface of membranes 78 . Greater absorption was observed in first 11 h due to larger pores in membrane structure, which leads to fast transport of Pb(II) toward membrane surface 77 . VTES-GO behave as absorptive nanomaterial in membrane structure 6 .

NF membrane performance. Effect of concentration of Pb
Contact angle. Contact angle of pure M-GuA0 (control) and modified membranes M-GuA2-M-GuA10 were analyzed to evaluate the surface hydrophilicity of the membrane are showed in Fig. 7, pure membrane (MGuA0) showed a larger value (68°) though modified membranes showed low values. The increase in Gum Arabic concentration illustrated decreased in contact angle from M-GuA2-MGuA 8. The membrane with lower contact angle would have higher hydrophilicity minimum value showed by M-GuA (56°) as the Gum Arabic decrease the contact angle, therefore increase the hydrophilicity 79,80 . In case of MGuA10 the contact angle value (57°) again increases because membrane become denser as discussed in "Scanning electron microscope (SEM)" section 66,67 . The contact angle of the averaged value for each membrane was presented in Fig. 7.

Conclusion
CA based membranes modified with VTES-GO/GuA as a filler enhanced the properties of permeation flux and Pb(II) ion rejection. The effect of pH, contact time, permeation flux and pressure on Pb(II) ion rejection has been studied using NF-membranes. It showed that optimum rejection was 97.6% at pH 9, which also revealed that low pH from 1 to 6 was not obligatory for salt rejection. The permeation flux due to varied pressure and concentration was 8.6 l m −2 h −1 . Moreover, the functional groups were confirmed by FTIR, surface morphology observed by SEM and topographical images by AFM were justified the even distribution of GuA in M-GuA8. TGA results was revealed that stability depend on uniform dispersion of GuA in membrane. Bio-fouling results further confirmed that greater concentration of GuA enhanced bacterial rejection properties. The described results are due to unique properties of nanomaterial-based membrane and their convergence with current treatment technique present great opportunities to revolutionize water and waste water treatment, and improve the performance understanding of NF-membrane and can be important parameter for the fabrication of membrane on commercial scale.