3.1. Membrane characterization
3.1.1. Structural analysis by FTIR
The functional changes in the hydrophilized membrane were addressed using FTIR spectral studies. Fig.2a represents the FTIR spectra of GA cross-linked PVA-PES blend membrane, wherein the peaks at 1117.78 cm-1 and 1435 cm-1correspond to the stretching vibrations of ether (-O-) and sulfone (-SO2) groups of the PES. The peak at 1582.63 cm-1 denotes -C-H stretching vibrations of the aromatic benzene ring of PES. A broad-band observed between 3500 to 3200 cm−1 denotesinter and intramolecular hydrogen bonding (-OH) of PVA. The peaks in the range of 2789 and 3000 cm-1 refer to the C–H stretching vibration of PVA's alkyl groups. The formation of a new acetyl linkage is attributed to the peak at 1070 cm-1, representing the cross-linking between GA and the hydroxyl groups of PVA. The possible reaction mechanisms of the GA cross-linked PVA-PES blend membrane (HF-UF) is shown in Scheme 1.
3.1.2. XRD
The X-ray diffraction pattern of the HF-UF membrane is shown in Fig.2b, wherein the membrane appears to be semi-crystalline in nature. Amorphous regions were found to be within 0 to 37º on 2q scale, whereas sharper peaks of the membrane were found to be 38º and 48º on 2q, which represent crystalline nature. The GA cross-linker is found to increase the crystallinity in the PVA-PES blend membrane. Overall, the HF-UF membrane shows semi-crystalline nature.
3.1.3. SEM
The indigenously synthesized HF-UF membrane surface and cross-sectional morphologies are provided in Fig.2c and d. The membrane’s surface showed uniform distribution of PVA polymer throughout the PES matrix without any agglomeration, which is clearly evident at 5µm magnification in Fig.2c. The cross-sectional image in Fig.2d reveals that the PVA membrane effectively blended with the PES layer through the formation of ultra-pores having a sponge like structure. Moreover, the cross-linked blend was adequately interpenetrated into the PE non-woven fabric support through the formation of a finger-like structure (Fig. 2d), which can be seen at 200 µm magnifications. From the overall surface and cross-sectional morphologies, the PVA and PES membrane were successfully blended to form a hydrophilic porous layer on the PE support.
3.1.4. Determination of MWCO of the HF-UF membrane
PEG of 4000 and 6000 Da molecular weight (MW) in concentrations varying from 2000 to 20,000 ppm were prepared in de-ionized water and fed through the HF-UF membrane for the determination of MWCO of the indigenously synthesized membrane. The concentration values were correlated with RI data recorded for each sample to obtain standard graphs of 4000 and 6000 Da of linear relationship. The experiments were conducted at 3 bar pressure, and samples were collected to determine the RI values. Fig.3c represents the RI values of feed, permeate, and retentate samples, where the rejection efficiency was observed 91% for 6,000 Da and 85% for 4,000 Da of PEG. As the MW of PEG increases, it obviously results in an enhancement in rejection due to the smaller pore size and decreasing segmental gap between the polymeric chains during membrane formation. Hence, this experiment demonstrated that the synthesized membrane has an average MWCO of 5000 Da. Further, the obtained membrane’s porosity was 34.68%.
3.2. Experimental results for the treatment of RGW
3.2.1. Effect of feed temperature on flux and permeate characteristics
Laboratory experiments were carried out to understand membrane performance during UF of RGW at 30 ˚C and 80 ˚C feed temperatures. The feed and permeate characteristics at 30 ˚C and 80 ˚C are provided in Table 1. Permeate parameters such as TDS, turbidity, conductivity, COD, and pH at 30 ˚C were found to be 56,330 mgL-1, 1200 FAU, 84.3 mScm-1, 20,000 mgL-1, and 3.34, whereas the permeate values at 80 ˚C (hot condition) were found to be 62,330 mgL-1, 1,325 FAU, 34.3 mScm-1, 24,000 mgL-1, and 4.60, respectively. Further, the effect of operating time on permeate flux is provided in Fig.4a, where the average flux decreased from 8.33 to 4.81 Lm-2h-1 and from 7.14 to 4.63 Lm-2h-1 across the operating time of 0 to 30 min at 30 ˚Cand 80 ˚C feed temperatures, respectively. The declination of flux is due to the formation of scalants on the membrane surface, which accelerates membrane fouling at both the feed temperature conditions. Additionally, the membrane pore blockage is rapid within 30 min, due to direct filtration of the effluent through the spiral-wound HF-UF membrane. Processing of RGW by HF-UF membrane shows better performance at 30˚C rather than 80 ˚C of feed temperature. Moreover, it is preferable to conduct the experiments at room temperature for the reduction of load on membrane and effective treatment of DWW.
3.2.2. Effect of operating time on flux and permeate characteristics using coagulation integrated HF-UF membrane processes
In this study, the RGW was pre-treated with various coagulants such as methanol, ethanol, and HCl, followed by membrane process at 30 ˚C feed temperature. Permeate characteristics of TDS, turbidity, conductivity, COD, and pH using various coagulation integrated HF-UF membrane systems is denoted as HCl + HF-UF, ethanol + HF-UF, and methanol + HF-UF as seen in Table 2. Initially, the feed color was found to be turbid milky white, which was completely removed from permeate after the treatment with various coagulation + HF-UF combinations. From Table 2, permeate quality for methanol + HF-UF combination was found to be 4,200 mgL-1 TDS, 70 FAU turbidity, 6.528 mScm-1 conductivity, 6,000 mgL-1 COD, and 7.23 pH which was better than the data recorded in case of other coagulants. The effects of operating time on permeate flux and % rejection for coagulation integrated HF-UF processes is provided in Fig.4b. The average permeate flux values for HCl + HF-UF, ethanol + HF-UF, and methanol coagulation + HF-UF processes (Fig.4b) were found to be 7.7, 8.2, and 13.62 Lm-2h-1 for an operating time of 30 min. Furthermore, from Fig.4b the % removal efficiencies of TDS, turbidity, conductivity, and COD were noted to be 87.4, 99.7, 87.4 and 28.6 for HCl + HF-UF, and 36.3, 93.77, 99.93, and 53.6 for ethanol + HF-UF and finally 93.77, 95.4, 93.9, 78.6 for methanol + HF-UF processes, respectively. Since the effluent is already acidic, the coagulation by HCl acid was not that effective in sedimentation of suspended solids. In contrast, ethanol has greater van der Waals force within the molecules and particulate matter may take a longer time to settle down. On the contrary, the coagulant methanol forms very large flocs within a short period of time because of its smaller size, that enables effective interaction with particulates, which are broken into smaller particles, that subsequently agglomerate into large sized flocs. Only 1% (v/v) methanol coagulant was mixed with RGW in volumetric ratio to facilitate economic feasibility. Methanol + HF-UF membrane process was found to be optimum for subsequent studies with AMBR at room temperature.
3.2.3. Effect on permeate flux and characteristic parameters with operating time by coagulation integrated AMBR
In this process, RGW was pre-treated with methanol and subjected to AMBR consisting of HF-UF membrane in continuous recirculation mode for 24 days. The effect of operating time on permeate flux is provided in Fig.5a, wherein the flux was found to decrease from 4.87 to 3.4 Lm-2h-1 across a time period of 0 to 30 min. Particles that are retained and form a cake-like structure on the membrane surface, cause an increase in the thickness, as the filtration progresses. However, the declination in the flux is moderate, as coagulation helps to reduce the load of suspended solids on the membrane surface. Moreover, from Fig.5b-5e, high efficiency can be observed in the TDS, turbidity, conductivity, and COD removal from the feed to permeate from 67,392 to 5,760 mg L-1, 1500 to 60 FAU, 105.3 to 9.5 mS cm-1, and 28,000 to 1500 mg L-1 over the operating time of 1 to 24 days, respectively. From graphical observations (Fig.5b-5e), the TDS, conductivity, and COD values in the bioreactor reduced gradually during 1 to 9 days of treatment. From the 9th day onwards, a drastic reduction was observed up to the 18th day with respect to TDS, conductivity, and COD. On the other hand, the turbidity reduced slowly up to 6 days of treatment after which it decreased steeply up to the 9th day and then slowly reduced until the 18thday. These observations can be explained on the basis of the varying extent of activity of microbial consortium on the RGW. After 24 days, there is no considerable reduction observed in all the parameters. The overall percentage removal efficiencies of the methanol coagulation + AMBR process with respect to TDS, turbidity, conductivity, and COD were noted to be as considerable as 91.0, 96.0, 91.0, and 94.6%, respectively. The integrated process exhibits a greater efficacy in degradation and separation of organic constituents that constitute COD. As per some literature reports, more than 90% of COD removal was achieved during the MBR processes, which substantiates results observed in this study [34, 35]. Moreover, after coagulation, RGW's conductivity and TDS reduced slowly in the first 9 days and then rapidly decreased while culture growth progressed in the bioreactor under continuous operation mode. When feed at 30 ˚C was treated with AMBR, the pH of the effluent increased from acidic value of 4.85 to 7.04 neutral medium (Fig.5f). Further, from Fig.5b-5f, it can be concluded that the AMBR becomes stable after 18 days of operation. Further, the amount of mixed liquor suspended solids (MLSS) obtained in the bioreactor was around 350 mgL-1 which was further subjected to an anaerobic digester for the production of the biogas as a by-product for cooking purpose. Between the 18th day and 24th day, negligible reduction in all the wastewater parameters was observed, due to no further degradation of the RGW in the bioreactor. Based on these observations, the sludge age can be considered 18 days to optimize the process. In this sense, methanol pre-treatment + AMBR, is the best-suited system for treating RGW. Additionally, from the obtained results, the produced water can be recycled for domestic applications such as gardening, washing, laundry, etc.
3.2.4. Effect of membrane washing on pure water flux
The membrane performance depends on chemical cleaning, which affects the pure water flux. After RGW treatment, the HF-UF membrane was washed thoroughly with both organic and inorganic chemicals, following which the pure water flux was determined to help understand the membrane cleaning efficiency. Fig.6 reveals the decrease in the pure water flux from 136 to 87 Lm-2h-1 whereas, after 3 min, the flux increased from 27.27 to 80.35 Lm-2h-1 respectively, after the membrane cleaning. From the graphical observations, it can be concluded that after chemical washing, the pores on the membrane were opened again by removing scale forming foulants and particulate matter from the membrane surface.
3.3. Construction of the molecular dynamic (MD) simulation
A PES polymer chain for the MD simulation with repeating monomer units was constructed and simulated using the Accelrys commercial software with the condensed phase optimized molecular potential for atomistic simulation study force field [36]. The 3D structure of the PES, PVA, and DMF are shown in Fig.7a, b, c. The amorphous builder module of PVA is used to construct the membrane system to minimize the PES system. PES has been built with repeat units of the monomer of PVA in DMF solvent and minimized by the steepest-descent method. Several cycles of energy minimization (EM) in MD simulation were considered to allow the polymer chain to fold until the polymer structure had a density of 1.0003 gcm-3. Moreover, this value is closer to experimental frequency of PES, whose density was 1.37 gcm-3 under ambient conditions. After the first EM and MD simulation were performed at a constant temperature of 298 K in the NVT, wherein the quantity is denoted by N, volume by V, and Temperature by T, a statistical ensemble was made to speed up the process of folding of the polymer chain. The minimized PES/PVA blend membrane structure was determined at different hydration levels under ambient conditions and the final composition of the PES bulk polymer model is shown in Fig. 7d.
3.3.1. EM
Fig.7e shows the potential energy to increase during molecular minimization, due to the energy replacement of the hydroxyl group (-OH) in PVA. The amorphous systems were constructed with a periodic boundary condition, and the density was found to be 1.0003 gcm-3 after the minimization of each molecule. The minimization energy of the amorphous system in terms of potential energy is -227.549 kcal mol-1. Therefore, Fig.7e describes the information related to how the water molecules facilitate an increase in the total strength of the system.
3.3.2. Economic estimation of integrated AMBR
Usually, conventional processes require enormous space, high energy consumption, chemical usage, and labour cost compared to membrane processes [37], which are one-time investments to obtain good quality of water from effluents for reuse in various applications to reduce fresh water consumption. Low pressure membrane processes can be used for the production of reusable water in various applications. Table 3 represents a detailed description of the materials used in the present study to install an experimental setup. The economic estimate of a 1000 Lh-1 capacity AMBR pilot plant is provided in Table 4, including operating and maintenance costs. The capital investment was found to be 5501.84 USD (INR 4.20 Lacs), with a payback period of 2.72 years.
3.3.3. Practical implications
Based on the efficiency observed in removal of impurities present in the wastewater, coagulation pre-treatment + AMBR appear to possess several advantages over conventional processes. In the present study, an innovative approach of methanol coagulation + AMBR integration was explored to treat the highly gelatinous and sticky RGW to produce reusable water. In this integrated process, methanol coagulant effectively reduced the load of suspended and dissolved solids from the wastewater to enhance the efficiency of AMBR. The utilization of HF-UF membranes in bioreactors would effectively reduce the fouling propensity owing to greater hydrophilicity of the membrane. The overall process flow diagram of the RGW treatment by methanol coagulation integrated AMBR for reusable water production is provided in Scheme 2. Through this process, the reclaimed water can be used for domestic application such as washing, gardening and flushing, while in industries it could be used in cooling towers.