• No results found

Study of Effects of Binder Concentration on Properties of the Fly ash- based Tubular Ceramic Membrane

3.3 Results and discussions

3.3.1 Rheological behaviour of Na-CMC solutions

Fig. 3.2 (a) Stress-strain curve and (b) Viscosity-strain curve of Na-CMC solutions with different concentrations

The study of rheological behaviour of aqueous solutions of different concentrations of Na- CMC solution revealed that the solution viscosity increases drastically with an increase in the Na-CMC concentration. The detailed analysis reported in Fig. 3.2 and Table 3.1 has shown that increasing binder solution concentration from 1% to 4% resulted in a decrease in the values of

flow behaviour index (Reddy, 2015). This signifies that the increased concentration makes the solutions divert from Newtonian behaviour, thus making their mixing more difficult (Kao et al., 2015; Dickey, 2015). It has also been observed that the zero-shear viscosity showed a steep rise when the solution concentration changed from 3.5% to 4%. Hence, Na-CMC concentrations above 3.5% were not considered for membrane fabrication to avoid the formation of agglomerates owing to improper mixing. Though the zero-shear viscosity of 1 wt.% Na-CMC solution was quite low, it was not considered for membrane fabrication, as the solution was unable to hold the clay particles together properly, leading to the formation of deformed shaped membranes (Fig. 3.3).

Fig. 3.3 Deformed shaped membrane with 1 wt.% Na-CMC solution

Table 3.1 Rheological data of different concentrations of aqueous Na-CMC solution at 25 ˚C Na-CMC concentration

(wt.%)

Zero shear viscosity (Pa.s) Flow behaviour index (n)

1 1.21 0.782

2 9.13 0.677

3 28.62 0.613

3.5 39.72 0.613

4 193.70 0.498

The rheological observations mentioned above are in good agreement with the previous works reported in earlier literature (Reddy, 2015).

3.3.2 Morphology study of membranes

Once the membranes with different concentrations of binder solution were fabricated, the morphology of the membranes needs to be checked in order to confirm their usability in separation processes further. Fig. 3.4 corresponds to the inner as well as outer surface images of membranes M2, M3 and M3.5, respectively. The darker portions in the images represent pores, which are marked by red arrows, while the lighter areas represent the ceramic particles.

The membrane surfaces are smooth and are devoid of any defects such as pinholes or cracks.

However, it has been observed that the membrane fabricated with higher binder concentration shows non-uniformity in its pore size. The results illustrated in Fig. 3.5 also reveal the same, showing a narrow pore size distribution of M2 membrane compared to the M3 and M3.5 membranes. The pore sizes of M2 membrane are within the range of 0 - 1.5 µm, while the pores are scattered over a wide range of 0 - 5.0 µm and 0 - 3.5 µm in case of M3 and M3.5 membranes, respectively. The possible reason behind this non-uniformity may be the increased viscosity of the solutions, which causes stronger particle agglomeration in the membrane matrix. During sintering at 1100 oC, these agglomerates compact and lead to a wider pore size distribution (Das, 1999).

However, this problem of agglomeration is not observed in the case of M2 membrane due to comparatively lower solution viscosity. It needs to mention that FESEM images, using the same procedure as described in Chapter 2, were also being used to evaluate the average pore size of the membranes. The calculated average pore diameters of M2, M3, M3.5 membranes are 0.81, 1.96 and 1.16 µm, respectively (Bouazizi et al., 2016; Jana et al., 2010). This

observation is sufficient to prove the fact that comparatively smaller sized pores are present in M2 membrane as compared to the other two membranes.

Fig. 3.4 FESEM images of inner and outer surfaces of the membranes (M2-M3.5) (Red arrows in the picture denote the pores in the membrane while the brown boxes correspond to

the wider pores resulted due to agglomeration caused by increased binder content)

Fig. 3.5 Pore size distribution of M2, M3 and M3.5 membranes

3.3.3 Porosity

The average porosity values of M2, M3 and M3.5 membranes fabricated using various concentrations of binder solution are 40.17±1.04, 41.65±2.20 and 41.24±0.79%, respectively.

It has been observed that there is no significant variation in the membrane porosities with increasing the concentration of binder. Compared to M2 membrane, the membranes (M3, M3.5) prepared with higher binder content (> 2 wt.% Na-CMC solution) demonstrate similar porosity values even after showing the presence of bigger pores, as evidenced from FESEM

analysis (Figs. 3.4 and 3.5). This may be attributed to the fact that the membrane M2 may possess larger numbers of smaller pores as compared to membranes M3 and M3.5, where lesser number of pores with bigger diameters may present, thus compensating for the overall membrane porosity. The pore density of M2, M3 and M3.5 membranes are found out to be 17.7×109, 12.2×109 and 14.5×109 pores/m2, respectively, which is in accordance with the above-mentioned justification. Similar trend of variation in membrane porosity with binder content has been observed in the works of Das (Das, 1999). The obtained porosity values in this work are consistent with the other ceramic membranes prepared by various authors. As evident from the literature, ceramic membranes with 40% porosity are considered to be highly suitable for separation applications (Benito et al., 2015; Vasanth et al., 2011; Kumar et al., 2015a).

3.3.4 Mechanical strength

The mechanical strength of the membranes is observed to decrease with an increase in binder concentration. The mechanical strength of M2, M3 and M3.5 membranes, as obtained from compression strength evaluation test, are 20.28±2.09, 14.98±1.44 and 12.60±2.00 MPa, respectively. As evident from FESEM images (Figs. 3.4 and 3.5), at higher binder concentrations, the formation of bigger sized pores in the membrane is noticed due to the burn out of binder agglomerates, which suppresses the driving force of densification. This results in a comparatively hollow structure inside the membrane as compared to the membranes fabricated using lower binder content. Consequently, a significant decline in the mechanical strength is noticed at higher binder concentrations. It is well documented in the literature that membranes having dense structures possess higher membrane strength than those having bigger pores as bigger pores are very much prone to defect formation on application of compressive force (Adam et al., 2020). To sum up, the high porosity fly ash-based ceramic

membranes prepared in this work display moderate mechanical strength compared with those of clay-based porous ceramic membranes reported by other authors (Zou et al., 2019c; Jedidi et al., 2009).

3.3.5 Chemical stability

Results of chemical stability tests indicated that only minimal weight loss of the membranes is observed in basic medium (2.77±0.67, 3.60±0.65 and 3.02±1.26% for M2, M3 and M3.5 membrane, respectively), thus signifying the potential applicability of the membranes in harsh basic condition. However, the weight loss of the membranes is on the higher side in acidic environments, restricting their use in low pH environments. The corresponding weight loss for M2, M3 and M3.5 membrane in acidic medium is 6.56±1.67, 10.83±1.16 and 12.23±1.67%, respectively. As mentioned in Chapter 2, CaO present in the sintered membranes is mainly responsible for the decrease in membrane weight after acid treatment. CaO reacts with hydrochloric acid, forming precipitates of calcium chloride, thus leading to loss of weight (Suresh et al., 2016).

However, it is seen that membranes fabricated with increased binder content demonstrate higher weight loss in acidic condition. As evident from FESEM images (Fig. 3.4), at higher binder concentrations, the formation of abnormally large pores takes place in the membrane and such large pores can be considered as the main cause of poor chemical resistance of the membranes. The large sized pores allow the acidic solution to enter into the pores more easily by offering lesser diffusional resistance and thus, contribute to higher weight loss (Saxena et al., 1974). The findings obtained in this work are in good agreement with those reported by Bose and Das (Bose and Das, 2014).

3.3.6 Water permeability and pore size evaluation

Fig. 3.6 Pure water flux as a function of operating time for membranes M2, M3 and M3.5

Fig. 3.7 Pure water flux at different applied pressures for membranes M2, M3 and M3.5 Results of pure water flux of the three membranes (M2, M3 and M3.5) performed using cross- flow filtration setup are depicted in Fig. 3.6. The constant water flux for the entire duration of experiment at a specific pressure signifies the saturation of the membrane pores with water molecules. For all the membranes, the water flux rises with an increase in the applied pressure (Fig. 3.7) and this effect is solely because of an enhancement in the driving force, which pushes more water through the membrane pores. As evident from Fig. 3.7, an increased pure water flux is observed at higher concentration of binder (Na-CMC). This can be ascribed to the presence of bigger pores in the membrane. Moreover, the pore size evaluated using water permeability data reveals that the membrane pores are in the similar size range for M3 and M3.5 membranes, average pore size being 0.190 µm for M3 membrane and 0.177 µm for M3.5 membrane. However, the pore sizes of these membranes are slightly higher than that of M2 membrane (0.133 µm), which may be contributed by the few bigger pores present in the former membranes as it has already been mentioned that the agglomeration effect is more pronounced in those two membranes (Das, 1999). This trend of pore size variation of membranes is quite similar to that calculated from FESEM images using ImageJ software. However, the size of

pores obtained from FESEM images is comparatively larger as this method considers all the surface pores, including the dead-end ones (Sinha and Purkait, 2013).