Thermogravimetric analysis (TGA)
Thermogravimetric analysis (TGA) is considered one of the most powerful tools generally used to assess the thermal stability of polymers and their blends (Mahalle and Sangawar 2013). The stability of the membrane to withstand over wide range of temperatures is an important aspect to be considered at an industrial level as the feed may be subject to filtration at higher temperatures too (Mandal et al 2008). Hence in the current study, the prepared nanofiltration membrane was subjected to thermal analysis in order to know its stability and adaptability in operating conditions and temperatures.
TGA thermogram of pure chitosan showed two major weight losses (figure 1), one before 200 °C which was most probably due to evaporation of residual water present in chitosan, and the second stage around 270°C was due to the detachment of entangled chitosan chains (Lutfor Rahman et al 2022; Cheng-Ho Chen et al 2007) and the decomposition process of N-acetylated compound is overlapped by the N-deacetylated unit, thereby increasing the widening process seen at temperatures up to 400 °C (Cardenas and Patricia Miranda 2004). Whereas for CS/PVA/MMT nanofiltration membrane, showed three main subsequent weight loss stages (Figure 2). The first stage around 150 °C was due to the desorption of water molecules adsorbed in the membrane with a derivative weight percentage of 1.527 %/min. The second stage corresponds between 150° to 382.38° due to the partial thermal decomposition of the polymeric backbone of chitosan, poly(vinyl) alcohol, and montmorillonite clay with a derivative weight percentage of 15.301%/min and the third stage is around 446.15° corresponds to degradation of higher stability crystalline parts of chitosan and poly(vinyl) alcohol backbone with a derivative weight percentage of 6.714 %/min (Igberase et al 2019). At the end of the experiment, nearly 26.75% of the sample remained as residue showing the higher thermal stability of the membrane.
On adding poly(vinyl) alcohol and MMT clay to chitosan will eventually increase thermal stability without affecting its physical characteristics (Velu et al 2017). The increment in thermal stability of the membrane, as compared to pure chitosan implies that the individual homopolymers undergo better adhesion i.e., in-between nano clay, hydrophilic polymer, and chitosan resulting in homogeneous dispersion of nano clay within organic-inorganic matrix layers. A similar trend is reported by Velu and his research group (2017) in which the addition of aluminosilicate to polyethersulfone membrane will increase thermal stability without affecting the membrane's physical characteristics.
Differential Scanning Calorimetry (DSC)
Differential Scanning Calorimetry (DSC) is a commonly employed tool for assessing the thermophysical properties of polymers including melting temperature, crystallization temperature, and glass transition temperature (Tg). The glass transition temperature is an important parameter that reveals the compatibility of components in the polymeric membranes.
Figure 3 shows the DSC thermogram of chitosan with a single glass transition temperature at 203°C (Sakurai et al. 2000). From the figure 3, it is evident that the chitosan shows a broad endothermic event indicating crystallization temperature (Tc) at 89.04°C that was linked to the evaporation of volatile water contained in the sample and an exothermic peak (Tm) was observed at 312.26°C corresponding to the decomposition of amine units in the polymeric backbone of chitosan which includes saccharide rings dehydration, depolymerization and decomposition of deacetylated and acetylated chitosan units and this was in close agreement with previous studies (Flores-Hernandez et al 2014; Shang et al 2011; El-Hefian et al 2010).
Figure 4 shows the DSC thermogram of the CS/PVA/MMT (1:10:1) membrane. It exhibits broad endothermic and exothermic peaks at 112.5°C, and 320.12°C respectively corresponding to the evaporation of water linked to the hydrophilic segments of the polymers, this factor contributes to the robustness of the water-polymer interaction. As compared to pure chitosan, the endo and exo events in the membrane were shifted from 89.04°C to 112.5°C and 312.26°C to 320.12°C which was evident for the physical and molecular alterations resulting from the interaction of molecular chains between individual homopolymers of chitosan, poly(vinyl) alcohol and montmorillonite clay during phase inversion process (Table 1) (Swetha and Neetha 2018).
Table 1: DSC thermogram details pure chitosan and CS/PVA/MMT (1:10:1) nanofiltration membranes
Samples
|
Tc (°C)
|
Tg (°C)
|
Tm (°C)
|
Pure Chitosan
|
89.04
|
203
|
312.26
|
CS/PVA/MMT membrane
|
112.5
|
215
|
320.12
|
Mechanical Properties
The mechanical properties of CS/PVA/MMT (1:10:1) nanofiltration membrane is calculated using stress-strain curves and it is compared with pure PVA membrane (figure 5). From stress–strain curve of membranes the following parameters were determined: proportionality limit, tensile strength (TS), and elongation at break (%), and the results were tabulated in Table 2.
Table 2: Mechanical parameters of CS/PVA/MMT (1:10:1) nanofiltration membrane
Sample
|
Proportionality limit (MPa)
|
Tensile Strength (MPa)
|
Elongation at break (%)
|
Pure Chitosan
|
2.356
|
2.368
|
31.68
|
CS/PVA/MMT membrane
|
22.16
|
22.32
|
41.4
|
The mechanical properties of nanofiltration membranes with different combinations are given in Table 2 and Figure 5. As evident from Table 2, the tensile strength is lower for pure PVA membrane with 2.368 MPa which is increased 10-fold times with the addition of chitosan and montmorillonite clay. The tensile strength and elongation at break for CS/PVA/MMT (1:10:1) nanofiltration membrane is 22.32 MPa and 41.4% respectively, and it is significantly increased as compared to the tensile strength of pure PVA membrane with 2.368 MPa and 31.68% respectively. The tensile strength of pure chitosan is reported as 13.73 MPa by Lutfor Rahman et al (2022) which is lesser than the nanofiltration membrane prepared in our present study (22.32 MPa). Hence it clearly indicates the tensile strength is significantly enhanced with the addition of montmorillonite clay.
The higher tensile strength of the membrane as compared to pure PVA membrane indicates the nano clay and chitosan embedded poly(vinyl) alcohol membrane requires more force to break the sample. The dispersion of nano clay within the polymeric matrix depends on van der Waals and electrostatic interactions, which in turn contribute to the formation of strong binding between individual homopolymers resulting in self-assembled structures in organic-inorganic matrix without agglomeration (Yogarathinam et al 2018).
Therefore, based on the aforementioned findings from thermal and mechanical investigations, it can be deduced that the fabricated membrane exhibits remarkable stability and superior properties to withstand efficiently a wide range of temperatures and pressure during the filtration process.
Scanning Electron Microscopy (SEM)
The top surface and cross-sectional views of the CS/PVA/MMT (1:10:1) nanofiltration membrane are investigated by recording SEM micrographs. From Figure 6a, it is evident that the membrane surface is rough, asymmetric without aggregation of clay particles, and has more pores throughout the surface as it is created during phase inversion reaction. The cross-section morphology reveals that the membrane possesses the desired asymmetric structure with thin skin top layer and bottom finger-like sublayer formations (Figure 6b). This observation indicates the exchange rate of solvent and non-solvent in a coagulation bath during membrane fabrication and is significantly increased during phase inversion resulting in the formation of micro voids.
It is well supported by the mechanical properties discussed above as there is no agglomeration of polymers with the weight ratios of chitosan, nano clay, and poly(vinyl) alcohol used to prepare this novel organic-inorganic membrane and this combination is found to be apt with well-distributed pores and best suitable for the filtration process.
Proposed mechanism of casted chitosan/poly(vinyl) alcohol/ montmorillonite clay membrane
During casting the individual homopolymers such as poly(vinyl) alcohol, chitosan and montmorillonite clay will joined together via hydrogen bonding with hydroxylated edge-edge interaction of silicate layers to form an organic-inorganic assembled matrix. The possible interaction of active sites between homopolymers during membrane formation is depicted in scheme 1.
Membrane filtration Performance
Copper Removal by altering parameters
The fabricated membrane undergoes copper rejection testing by using a synthetic aqueous copper solution as the feed, with the permeate being collected at scheduled time intervals. The membrane effectiveness is assessed by altering various parameters such as pH, metal ion concentration, applied pressure, and thickness of the membrane.
Effect of pH
pH of aqueous solution is an important parameter in deciding the performance of the membrane as it possesses a strong influence on the solubility of metal ions and the charge on the membrane surface (Angelin Vinodhini and Sudha 2017). The impact of pH on the Cu (II) removal process using CS/PVA/MMT nanofiltration membrane was studied and the result is shown in Figure 7. In order to evaluate the performance of the membrane over a wide range of pH the experiment was conducted in all three possible mediums i.e., acidic, alkaline, and neutral environment. The pH of the aqueous copper sulphate solution was adjusted by adding hydrochloric acid (HCl) and sodium hydroxide (NaOH).
As can be evident from Figure 7, the highest copper ion rejection was observed at a pH of 5 i.e. in acidic medium as compared to neutral and basic pH conditions. This is due to the fact that pH has a direct influence on the membrane surface and in acidic conditions the membrane surface will get protonated to a greater extent with excess presence of H3O+ ions. The amino group of chitosan within the membrane was converted to NH3+, resulting in an increase in the number of charged groups along the chain. This likely led to highly positively charged entrances to the membrane’s pores, potentially obstructing copper ions from entering through the pores, thus contributing to the high rejection rates.
On further increment of pH to neutral and alkaline medium will show lower rejection rate as compared to acidic medium. Copper exists as Cu2+ in acidic pH and precipitates as hydroxides at higher pH and exists as Cu(OH)2, Cu(OH)3- and Cu(OH)42- (Fan et al 2020; Ahmadi et al 2017). Here increasing the pH to a higher level will cause the copper ions to precipitate as insoluble hydroxides and it will block the active sites in the membrane surface hence this will affect the efficiency rate of copper rejection. For the CS/PVA/MMT membrane, the percentage removal of copper initially at 15 min at pH of 5, 7, and 9 were found to be 79.78%, 60.80%, and 59.68% respectively.
This result was in agreement with the literature papers discussed here. Abu-Saied and his co-workers (2017) fabricated sulfated chitosan/poly(vinyl) alcohol adsorbent membrane to remove Cu2+, and Ni2+ ions from an aqueous solution and reported maximum rejection of copper was observed at a pH of 5.5-6 similar to our observations. Bessbousse et al (2008) have also reported the maximum rejection of Cu (II) ions at pH 5 filtered with poly(vinyl) alcohol/ poly(ethyleneimine). Under these acidic conditions, the metal ions do not precipitate as insoluble hydroxides, and the protonation of PEI is reduced sufficiently to enable the formation of metal complexes. Similar to the observation in our present work, the maximum removal rate of Cu (II) was observed at pH 5, and further increasing the pH to neutral and basic conditions will result in the precipitation of copper ions as copper hydroxides which hinders the active sites on the membrane surface and hence may reduces the removal rate.
The pH was kept constant at 5 throughout the experiment while testing remaining parameters such as initial metal ion concentration, thickness, and applied pressure.
Effect of metal ion concentration
The impact of metal ion concentration on the membrane’s copper rejection rate was investigated at room temperature and the result is depicted in Figure 8. The experiment was performed at three different concentrations of 50 ppm, 100 ppm, and 200 ppm and the filtration behavior of the membrane was discussed below with suitable references.
As shown in Figure 8, the copper rejection rate decreases with increment in concentrations of metal ions. The maximum rejection rate of copper was observed at 50 ppm with a percentage of 78.24%. At lower initial metal ion concentrations i.e. 50 ppm, there are ample adsorption sites available for trapping Cu (II) ions from the aqueous solution. Consequently, this results in an elevated percentage of removal. However, at higher concentrations i.e., at 100 ppm and 200 ppm the number of heavy metal ions is relatively higher compared to the availability of adsorption sites (Thakur and Parmar 2013) and hence it shows a decreased percentage removal of Cu (II) ions with the removal rate of 71.96% and 69.51% for membrane at concentration of 100 ppm and 200 ppm respectively. This sorption behavior suggests that surface saturation relies on the initial metal ion concentration, while functional adsorption remains unaffected by the initial metal ion concentration.
Similar results were reported by Mokhter and their research group (2018) they have treated Cu (II), Ni (II), and Zn (II) synthetic metal ion solutions with modified polyether sulfone membranes surfaces by assembling multilayer films of polyelectrolytes. Initially, maximum removal of heavy metal ions was observed for feed concentration 50 mg L−1 with the rejection values of 91%, 93%, and 98% for Cu(II), Ni(II), and Zn(II) respectively and this rate was further decreased to around 62-67% for all the three metal ions with the feed concentration of 1200 mg L−1. This behavior was logically explained by the fact that an increase in feed concentration leads to a decrease in retention rate. This occurs because the chelating functional groups on the polymeric membrane surface may not be able to chelate all the metal cations present in the solution. As evident from Figure 8, a similar trend is observed in our present investigation.
Effect of Membrane Thickness
Membrane thickness plays a crucial role in filtration technique as it will affect the percentage rejection of heavy metal ions (copper) from aqueous solution. The percentage of copper rejection from aqueous solution by varying thickness from 0.1 mm to 0.2 mm was conducted to evaluate its performance and the result is shown in Figure 9.
As evident from figure 9, the percentage of copper rejection increases with an increase in the thickness of the membrane from 0.1 mm to 0.2 mm, as the diffusion of any chemical species through a pore of finite length and thickness also determines the transport rate. The initial rejection rate was increased from 70.28% to 82.47% as the thickness of the membrane was doubled to 0.2 mm for the CS/PVA/MMT membrane (i.e. 15 min). There was a gradual decrease in the rejection rate as time passed, which may be due to saturation of active sites and at 90 min the rejection value was found to be 53.31% for 0.1 mm, 69.67% for 0.2 mm at 90 min. Practical challenges have led to the avoidance of using a 0.3 mm thickness. Furthermore, it has been observed that the 0.2 mm membrane thickness provides superior copper removal results.
Yasemin and their coworkers (2014) have also studied the effect of membrane thickness on the rejection efficiency of metal ion solutions of cobalt, nickel, cadmium, and copper using polymer inclusion membranes containing Alamine 336 and Tributyl phosphate (TBP). The thickness of the membrane was varied from 20 mm to 45 mm for optimization and from the experiment it was concluded that the maximum rejection of metal ion solutions was observed at 25 mm. If the thickness was further increased it caused a negative effect due to an increase in flux which results in faster escape of metal ions from the membrane surface. A similar observation is reported in our present study.
Effect of Pressure
The effect of operating pressure is one of the significant factors that influence the flow rate of solution during the filtration process as it possesses a direct influence on the adsorption phenomenon of metal ions from aqueous solution. It is obvious from reported results from the literature that the applied pressure has a direct influence on the performance of the membrane and it is considered as a critical aspect to be studied to evaluate its efficiency under given operating conditions. The experiment was conducted under two distinct pressure conditions namely 50 kPa and 100 kPa and the filtration behavior is explained in detail below.
The rejection percentage of copper with respect to pressure is shown in Figure 10. For the CS/PVA/MMT membrane, the rejection rate was increased with an increase in trans-membrane pressure i.e. from 50 to 100 kPa. The rejection of copper ions was increased from 66.95% to 78.24% and this may be due to the fact that at a lower pressure of 50 kPa, the interaction forces between membrane surface and metal ion copper were lower causing slow sorption through the membrane surface. Since there exists electrostatic repulsion between positive surfaces of the membrane due to protonated amino groups of chitosan and Cu2+ ions, this repulsive force will result in a lower rejection rate of copper within a given time of 15-90 min.
Corroborating to our present investigation, Cheng and his research group (2016) have also reported the effect of operating pressure on the removal rate of titan yellow dye. The pressure was varied from 0.01 to 0.4 bar showing a maximum rejection rate of 99.6% at 0.05 bar. Further increment in pressure however will decrease the rate of removal instead due to a rapid increase in flux of membrane which results in running off of titan yellow without being adsorbed on the ceramic membrane surface.
The obtained results showed the more desirable characteristic of nanofiltration membrane for effective copper removal from aqueous solution as tested with various parameters. Further, the results and effectiveness of CS/PVA/MMT nanofiltration membrane were compared with other similar commercially available and fabricated membranes reported in literature papers and were tabulated in Table 3.
Table 3: Comparison table of chitosan based membranes in rejecting metal ions from aqueous solutions
Membrane
|
Foulants
|
Optimized parameter
|
Removal efficiency
|
Ref.
|
Stable chitosan gel material
|
Copper removal
|
pH = 5
|
75.4%
|
(Yang et al. 2019)
|
Chitosan stacking membrane
|
Copper removal
|
Initial feed concentration = 50 ppm
|
|
(Zhang et al. 2019)
|
Chitosan/TGIC- nanofiltration membrane
|
Magnesium chloride,
Sodium sulphate, Magnesium suphate, and sodium chloride
|
Reduced water hardness
|
|
Binbin Yang et al. 2021
|
Chitosan/polyvinl alcohol/montmorillonite clay membrane
|
Chromium removal
|
Applied pressure = 100 kPa
|
84-88.34%
|
(Sangeetha et al. 2019)
|
Chitosan/polyvinl alcohol/montmorillonite clay membrane
|
Copper removal
|
pH = 5
Concentration = 50 ppm
Membrane Thickness
= 0.2 mm
Applied pressure
= 100 kPa
|
78-79.78%
|
Present study
|
The trapping of copper onto the membrane surface was evidenced by comparing the FT-IR spectrum of membranes taken before and after filtration and also from EDAX spectral details. The possible interaction of metal ion copper with membrane surface has been proposed and has been explained in detail below.
Proposed Mechanism of copper interaction on PVA/CS/MMT membrane during filtration
The schematic representation of copper removal during filtration is shown in Scheme 5. Based on the literature reports we have proposed a mechanism to explain the interaction between the heavy metal ion copper onto the membrane surface. The retention of metal ion copper with the membrane is due to three simultaneous actions:
- The electrostatic repulsion occurring between membranes and cation (Cu2+),
- The binding of cations by the polymers deposited onto the membrane surface (complex formation), and
- The steric hindrance caused by the polymer films partially filling of the membrane’s pores (Figure 11).
As shown in Scheme 2, the interaction between functional groups present in the membrane with copper ion was through the N atom of amino groups of chitosan and the O atom in the hydroxyl groups present in all the individual homopolymers. This observed interaction between copper and the polymeric membrane surface was additionally substantiated by the data obtained from FT-IR spectrum, and EDAX micrographs of the membrane both before and after copper removal.
FT-IR spectrum of CS/PVA/MMT membrane before and after Filtration
To access the changes occurred in the functional groups during the filtration of copper ions, the filtered membranes were studied using FT-IR. The membrane after filtration exhibited several peaks similar to the spectral peaks observed before filtration. However, few peaks with different intensities and shifts in wavelength were observed compared to untreated membranes. These results confirmed the role of functional groups along with metal ions during the filtration process. The information regarding the changes in peak values and their corresponding functional groups is shown in Figure 12 and explained in detail below:
For the CS/PVAMMT membrane (Figure 12) the distinct peaks at 1581.63 cm-1and 3448.72 cm-1 indicate the bending of -NH and –OH bonds, the -NH stretching peak shifted to 1591.27 cm-1 and 3423.64 cm-1 indicating the presence of the amide (II) functional group, and suggesting that hydroxyl groups may be involved in the copper removal process. Additionally, the line shape of the characteristic glycosidic bond peak at 1020.34 cm-1 changed to 1024.20 cm-1, possibly due to copper (II) cross-coordination with adjacent chitosan chains, affirming a potential interaction between copper and membrane surface (Gritsch et al., 2018). In general metal ions may chelate with the hydroxyl and amino groups within molecular chains through a combination of partially ionic and coordination bonds. This interaction can lead to a shift in the carboxylate groups to lower wavenumber (Li and Bai 2005).
The stretching frequencies of metal-carbon, metal-nitrogen, and metal-oxygen bonds in the spectra of the adsorbed membrane surface were observed in the range of approximately 200-450 cm−1 i.e in the fingerprint regions (Niemantsverdriet 2007). In this study, the evidence of metal ion attachment to the membrane surface became evident through the appearance of a sharp peak at 481.23 cm−1, attributed to MN/MO stretching vibrations after filtration. Therefore, based on these findings, it can be concluded that the reduction in this peak indicates the involvement of these functional groups in the chelation of metal ions. Furthermore, the presence of a copper band in the fingerprint region is evident in the presence of copper on the membrane surface after filtration.
Energy Dispersive Spectroscopy (EDAX)
From Figure 13, it was incurred that the presence of carbon and oxygen content in the membranes confirms the existence of polymers (PVA, chitosan, and polyethylene glycol) and the Si, Al, Fe, and K content confirms the inclusion of montmorillonite clay within the membrane. These EDAX images confirm the uniform dispersion of individual homopolymers in the hybrid organic-inorganic polymeric membrane surface.
For this novel nanofiltration membrane, peaks for copper ions have emerged in the region of 1.5-3 keV. As evident from Figure 13, it was concluded that the arrival of new peaks for copper affirms the interaction between heavy metal ions and the membrane surface during filtration, likely indicating chelation and adsorption.