The particle size of the synthesised CZnO nano-adsorbent was found to be 88.35±2.89 d.nm. Sharp peaks obtained in the particles were of similar size and uniformly distributed throughout the mass (Fig. 2). The physicochemical characteristics of the activated sand and CSZOCS filter bed are shown in Table 1. The BET surface area of coated sand (6.20 m2/g) was nearly twice that of activated sand (2.87 m2/g). The developed filter bed was more acid resistant and less alkali resistant. Total zinc loaded on activated sand was found to be 40.47 mg/g.
Characterization of CSZOCS filter bed
The surface of the CSZOCS filter bed was found to be significantly rougher and had larger cracks compared to the activated sand (Fig. 3a and b). Rough surfaces and cracks are responsible for the higher adsorption capacity of the adsorbents (Rasheed and Meera 2016). Elemental compositions of the CSZOCS filter bed were studied by using SEM-EDS. To acquire precision measurement, map acquisitions such as beam energy of 18 kV, probe current of 250 pA, count rate of 67 cps and total counts of 10,720 for 160 s were used (Yordsri et al. 2018). Spectra of EDS (Fig. 3c) confirmed that 45.54% of zinc was loaded onto the activated sand. The remaining portion (54.46%) was considered as uncoated zinc, which is removed during the washing process. The small amounts of silica (28.51%) and carbon (25.96%) were considered impurities.
For employing the EDS analysis, two sets of elemental maps were generated (line scan and line map) by selecting the Zn element. The electron image size of 561 µm was selected for different coating dosages of CSZOCS (Fig. 4a). To obtain a line scan of the electron image (Fig. 4b), OXFORD Inka software was used. Net intensity maps appeared based on the f-ratio of the Zn element and different dosages at each pixel (Teng and Gauvin 2020). Considering recoup of the X-ray intensity, the point of interest maps selected for the latest word processing file reformation may be responsible for many iterative results. Fig. 4c shows the point-of-interest intensity maps of Zn-Lα recoup from the preferred dataset with any component segregation. Fig. 4d shows the intensity maps of particle distribution at the point of interest and full component of all the data sets. As for inorganic phases of the Zn-Lα element, it was ascertained that collected X-ray counts are not sufficient for analysis. Hence, the phase map of the point of interest was retried. These reformed intensity maps showed slightly more clarity than a point of interest map.
Interpretable dark particle profiles were also observed (as around bright), which might be due to interference of the other phase signals of carbon and silica on the surface of the CSZOCS filter bed. The python script for Zn element phases in the INCA software was slightly modified as to elemental f-ratio values in the phase map. The concentration of the Zn coatings on the sand surface corresponds to the f-ratio values. The chemical compositions of elements are generally determined by f-ratio values (Brożek-Mucha 2014). The area fraction of different coating dosages was calculated based on the intensity of the phase maps acquired (Miler and Mirtič 2013). The crucial phases were recognised in the major phase maps and the scattering point of line scan. The line scan was illustrated by overlaying the point of interest line scan with the particle distribution phase map image (Fig. 4e). Phase peaks in the line scan and line map depicted the deposition of the Zn on the surface of the CSZOCS, indicating the full spread of the Zn element. Sharp peaks in line scan denote a lower number of Zn particles deposited on the surface of the CSZOCS (Rokosz et al. 2016).
In the XRD patterns of ZnO, chitosan, and CZnO nano-particles, the typical strong peaks of ZnO at a 2-theta angle of 19.20° and 26.30o became weak in the XRD pattern of CZnO nanoadsorbent
(Fig. 5a). A one-week peak of ZnO was found at 48.44°, and it appeared in the XRD pattern of CZnO. Typical peaks of chitosan appeared at 31.80°, 34.48°, 36.32°, 47.50º, 48.40°, 56.60°, 62.76°, 68.02°, and 68.98°. These peaks become strong in the XRD pattern of CZnO nano-particles. From the XRD pattern of CZnO, the peaks at 31.80°, 34.50°, 36.32°, 47.53°, 48.44°, 56.64°, 62.88°, 68.00° and 69.10° correspond to (100), (002), (101), (102), (102), (110), (112), and (201) planes. These planes are indexed to the wurtzite structure with a hexagonal phase of CZnO, and also matched with JCPDS card no. 36-1451. Li et al. (2010), AbdElhady (2012), and Bashal et al. 2022) reported similar findings. The average crystallite size was calculated using Scherrer’s formula: D=0.9λ/β cosθ. Where "λ" wavelength of X-ray, β is FWHM (full width at half maximum) in radians, and θ is the diffraction angles (Preethi et al. 2020). The average crystallite size of the synthesized CZnO nano-adsorbent was calculated to be 49 nm.
A high-intensity diffraction peak of activated sand was found at a 2-theta angle of 26.82º
(Fig. 5b). This result was confirmed by Eunice et al. (2013), who reported a 2-theta angle of 25º for river sand. The XRD pattern of CZnO found the presence of sharp peaks corresponding to the zinc blend crystal structure. The XRD profiles of activated sand and the CSZOCS filter bed depict the presence of sharp diffraction peaks. The presence of other weak XRD peaks in the CSZOCS can be attributed to the nature of CZnO loading on the sand surface. The XRD patterns of activated sand and CSZOCS were matched exactly at a 2-theta angle of 26.82°, 26.66°, 26.98°, 26.96° and 26.66°. This might be due to the presence of CZnO on the surface of the CSZOCS filter bed (Xia and Tang 2003). The intensity of the peak at 26.96º slightly decreased with the increase in CZnO dosage. The CSZOCS filter bed is crystalline, and the X-rays scattered with varying intensities due to interference effects of CZnO coating dosages (0.5, 1.0, 1.5, and 2.0 M).
The FT-IR spectra of CZnO showed bands at 3317 and 2870.07 cm-1 due to the stretching vibration mode of NH2 and OH groups (Fig. 6a). The peak at 1658 cm-1 was typical of carboxylic acid O-H stretching vibration, while the bands at 1377 and 1026 cm-1 were due to the amide-I group (N-H bend). Similar transition peaks of chitosan at 3317 and 2870.07 cm-1 are attributed to the C-H deformation. The special broad peaks at 1647 and 1419 cm-1 are attributed to the vibration mode of the alkoxy C-O (Bashal et al. 2022). In comparison with CZnO, the broader and stronger peak of chitosan shifted considerably to a similar wave number at 1026 cm−1, which indicated a strong attachment of ZnO to the amide groups of chitosan molecules. The weak transition peaks of ZnO detected at 2360.87, 1527.62 cm-1 are due to asymmetric stretching of Zn and OH of ZnO (AbdElhady 2012). New broad transmittance bands in the range of 1658–1026 cm−1 were found in the FT-IR spectra of CZnO nano-adsorbent, which were ascribed to the vibration of C-H and C-O groups. The reason for the above phenomena was the formation of a hydrogen bond between ZnO and chitosan. This result indicated that, the CZnO composite was prepared successfully without damaging the crystal structure of the ZnO core (Li et al. 2010).
To establish probable interactions between the sand and CZnO, FT-IR spectroscopy measurements were carried out for the CSZOCS filter bed. The FT-IR spectra of the CSZOCS filter bed showed a band at 2360.87 cm-1 due to the unbend quavering mode of the CO2 group (Fig. 6b). The hairy beard peak was found at 667.37 cm-1 due to Zn-O deformation. In comparison with different coating dosages of CZnO, the activated sand showed a silicon carbide (–Si–C–) stretching group at 3722 cm-1. The percentage transmittance of IR light declined slightly as the coating dosages increased from 0.5 to 2.0 M. These results confirmed the CZnO loadings on the activated sand surface.
A Raman spectroscopy analysis was used to determine the chemical composition present in the synthesised CZnO based on different Raman shifts (Fig. 7a). Three different Raman shifts were detected for synthesised CZnO nano-adsorbent at 43.77 cm-1, 1050.88 cm-1 and 2894.49 cm-1. Similarly, chitosan exhibited three shifts at 447.27 cm-1, 1106.80 cm-1, and 2878.10 cm-1, respectively. ZnO has shown a Zn peak at 433.92 cm-1. This suppressed Zn peak might be due to the photo-bleaching effect of X-rays on ZnO. In addition, a Raman spectroscopic experiment with variable coating dosage was carried out in this study. Each peak on the Raman spectrum corresponds to the specific vibration of the chemical bond present on the surface of the CSZOCS filter bed (Fig. 7b). A frequency shift was observed in each spectrum for the respective bonds. Based on the peak position and maximum vibration frequency, chemical species were identified. The Raman spectrum of activated sand and CSZOCS showed the Raman shift ranging from 500 to 550 cm-1 and 450 to 500 cm-1, respectively. CZnO coating dosage increased from 0.5 to 2.0 M, there was no change observed in the Raman shift due to the CZnO coating dosages, which appeared to affect the onset of Chitosan and ZnO dissolution (Honesty and Gewirth 2012).
Batch adsorption study for %RE of BOD using CSZOCS
The effect of CZnO coating dosage and contact time on the %RE of BOD (at pH 7 and an initial BOD concentration of 100 mg/L) is shown in Fig. 8. The %RE of BOD increased from 40.14±0.30 to 95.22±0.71% at contact time varied from 20 to 120 min. This might be due to the initial number of active sites available for the binding of organic pollutants (Tsaneva et al. 2017). After reaching the plateaus, the equilibrium concentration was achieved at 95.22 mg/L of BOD adsorbed by the CSZOCS filter bed. The experiment was continued for up to 3 h of contact time to obtain equilibrium concentration at the solid/liquid interface. There was no change in BOD removal observed when the time was prolonged more than 3 h. As a result, the initial stage of adsorption was very quick, but the inner diffusion was considerably reduced (Nassar et al. 2014). Furthermore, the huge number of unoccupied active sites identified after a given period suggested that solute absorption was hampered by repulsive interactions between solute molecules on the CSZOCS surface as well as in the bulk phases (Alhooshani 2019).
The quantity of adsorbent is considered an important component since it can determine the adsorption efficiency for a given starting adsorbate concentration. The CZnO coating dosage increased from 0.5M to 1.5M, and the %RE of BOD increased from 75.22±0.56 to 95.22±0.75 mg/L at 120 min of contact time. This increase in trend could be attributed to the large vacant sites available at higher adsorbent dosages and for a constant concentration of CZnO (Nassar et al. 2014). The CZnO coating dosage increased further from 1.5 to 2.0 M, and the %RE of BOD slightly decreased from 95.22±x0.75 to 88.26±0.98%. It may be presumed that the availability of active sites on the adsorbent is decreased at higher doses (> 1.5M) (Thirugnanasambandham and Sivakumar 2016).
The effect of pH on the %RE of BOD (at CZnO coating dosage of 1.5 M, contact time of 120 min and initial BOD concentration of 100 mg/L) is depicted in Fig. 9. Synthetic BOD solution pH increased from 2 to 6, leading to an increased %RE of the BOD from 57.24±0.26 to 90.03±0.39%. Lower pH values were responsible for the production of H+ ions, resulting in organic anion diffusive resistance (Oladipo et al. 2017). From the figure, it is also noticed that a pH of 6 was established to be the finest condition for the maximum %RE of BOD (90.03±0.39%). Hence, the optimised pH value of 6 was used to carry out the additional batch adsorption experiments. Further increased the pH from 6 to 12, and the BOD%RE decreased from 90.030.39 to 84.270.36%. This might be due to the fact that at pH of 6, the positive charge on the chitosan surface decreased and it became insoluble. Hence, this negatively affected the treatment process and decreased the %RE of BOD (Thirugnanasambandham et al. 2014b).
The effect of initial BOD concentration on %RE of BOD (at CZnO coating dosage of 1.5 M, contact time of 120 min. and pH 6) is illustrated in Fig. 10. The initial BOD concentration increased from 50 to 300 mg/L, and the %RE of BOD decreased from 94.44±0.41 to 64.89±0.28%. This might be due to more organic substances being adsorbed on the surface of the CSZOCS filter bed, and the distribution coefficient was decreased. Thus, the limiting number of absorption sites available for absorption at a higher initial concentration of the BOD (Moradi Dehaghi et al. 2014). From the figure, it is also noticed that, %RE was more predominant at initial BOD concentrations (50 to 100 mg/L), and then it was slightly decreased from 100 to 250 mg/L. It is also observed that, there is no change in the %RE of BOD when the concentration is increased from 250 to 300 mg/L. In the initial stages, the adsorption of the organic ions on the CSZOCS surface was increased rapidly. With the abundant availability of active binding sites on the adsorbent and with gradual occupancy of these sites, the adsorption becomes less efficient in the later stages (Gupta et al. 2001; Tu et al. 2012). The equilibrium status was achieved after 250 mg/L of initial BOD concentration.
Batch adsorption study for %RE of COD using CSZOCS
The effect of CZnO coating dosage and contact time on the %RE of COD (at pH 2 and an initial COD concentration of 200 mg/L) is shown in Table 2. From the table, it is noticed that the %RE of COD increased from 47.47±0.35 to 84.55±0.63% with initial contact time varied from 20 to 120 min. and CZnO nano-adsorbent coating dosages varied from 0.5 to 1.5 M. The %RE of the COD was enhanced rapidly due to the ample binding sites available on the circumference of the CSZOCS. At the end of the adsorption process, the capacity of the adsorbent will be very low. The equilibrium concentration of the COD was achieved after 120 min. at 1.5 M (Öztaş et al. 2008; Hodaifa et al. 2013). From the results, it was noticed that the 1.5 M CZnO coating dosage was found to be an optimised dosage. Hence, a 1.5 M CZnO coated sand filter bed was used for further adsorption experiments. The CZnO coating dosage increased further from 1.5 to 2.0 M, and the %RE of COD slightly decreased from 84.55±0.63 to 82.86±0.62%. This might be due to the fact that above 1.5 M coating dosage, the aggregation of particles takes place, as a result of a lower %RE of COD (Padmavathy et al. 2016). Thirugnanasambandham et al. (2014) reported that 95% RE of COD was the best at a 16 mg/L dosage of chitosan nano-particles. Thirugnanasambandham and Sivakumar (2016) reported 97% COD reduction efficiency achieved at a CZnO composite dose of 1.5 g/L with a contact time of 45 min. and an amount of dilution of 12%. These results were slightly higher compared to the present investigation.
The effect of pH on the %RE of COD (at 1.5 M CZnO coating dosage, 120 min. contact time and an initial COD concentration of 200 mg/L) is presented in Table 3. The pH will affect the surface charge of CZnO and the stabilisation of the suspension. The %RE of COD was found to be highest (85.44±0.57%) at a pH of 6. The optimum pH for the effective adsorption of COD is 6, and above this pH, the %RE of COD was decreased. This might be due to the weak interaction between the amino groups of chitosan and the oppositely charged ions present in the effluent (Devi et al. 2008). Hence, this optimised pH value of 6 was used for further experiments. Previous studies indicate that the solubility of chitosan decreases with an increase in pH above 6 (Takahashi et al. 2005).
The effect of initial COD concentration on %RE of COD (at 1.5 M CZnO coating dosage, 120 min. contact time and pH 6) is shown in Table 4. The %RE of COD was found to be increased with an increase in the initial COD concentration and after reaching saturation level. The %RE decreased from 85.08±0.64 to 59.56±0.44%, when the initial COD concentration increased from 50 to 300 mg/L. From the table, it could be observed that the %RE of COD is high at a lower initial substrate concentration, and the rate of adsorption of the organic ions increased progressively due to the increased driving force (Oladipo et al. 2017). However, at 200 mg/L of initial COD concentration, the CZnO coated sand reached its saturation point (Oladipo and Gazi 2015).
Realistic treatment of MPIW
MPIW was realistically examined through a batch adsorption study under predefined optimal conditions. Realistic application of the developed CZOCS filter bed will show the feasibility of organic pollutants' reduction in MPIW. The most prominent physico-chemical characteristics such as total dissolved solids (TDS), total suspended solids (TSS), turbidity, electrical conductivity (EC), pH, BOD, COD, sulphate, phosphate, ammoniacal nitrogen, nitrate-nitrogen, and chloride were analysed. In addition, co-existing pollutants such as aluminium (Al), copper (Cu), iron (Fe), magnesium (Mg), manganese (Mn), nickel (Ni), strontium (Sr), and zinc (Zn) were also determined, and %RE values were shown in Table 5. Among the analysed physical parameters of MPIW, TSS, TDS, and turbidity were decreased by 73.38%, 88.36%, and 98.97%, respectively. In addition, the EC of treated samples exhibited a lesser reduction efficiency (33.59%), and the pH of treated water increased substantially from 5.28 to 8.67. The organic pollutants of MPIW resulted in a significant reduction of BOD (90.68%), COD (97.98%), sulphate (98.98%), phosphate (88.15%), ammoniacal nitrogen (63.01%), nitrate-nitrogen (98.08%), and chloride (88.45%), respectively. Similarly, co-existing contaminants were eliminated to be 85.40%, 71.18%, 57.46%, 52.70%, 62.79%, and 76.59% for Al, Fe, Mg, Mn, Ni, and Zn, respectively. However, copper and strontium are limited in concentration, and no ions were detected in the samples after treatment. The obtained findings demonstrated that, the unique CZOCS is a potential adsorbent for commercial MPIW treatment.
According to experimental data, the greatest %RE of BOD was 94.44% for synthetic effluent and 90.68% for real effluent. MPIW has a significant number of coexisting pollutants; these may fight for adsorption sites, reducing the effectiveness of nano-adsorbents. The competitive adsorption ability of metals varies from one ion to the next and is influenced by various parameters, including molecular weights, ionic energies, hydrated ionic radii, and hydrating activity (Eldeeb et al. 2021). This suggests that the developed CZnO surface is very receptive to a wide range of MPIW pollutants, implying that multi-adsorption is also possible.
Regeneration of filter bed
Recovery of adsorbed BOD and COD from the surface of the CSZOCS filter bed was tested through four cycles of adsorption and desorption experiments (Fig. 11). In the first cycle of the regeneration study, the percent desorption efficiency was shown to be 97.50% and 93.86% for both synthetic BOD and COD. During the second to fourth cycle, the desorption ratio was drastically reduced to 62.35% and 60.40%. Thus, the reduction could be attributed to the detachment of coated CZnO nano-adsorbent from the sand surface during the regeneration process (Dinesha et al. 2021b). The results may guide the regenerated CSZOCS filter bed to find a higher potential ability to adsorb BOD and COD in collective regeneration cycles for organic pollutant recovery. After four cycles, the desorption efficiency of CSZOCS nano-adsorbent coated sand was reduced to 63.35% and 52.79% for BOD (synthetic and real effluent) and 60.40% and 56.41%, indicating that CSZOCS possesses excellent regeneration potential.
Comparison with other studies
Comparison of maximum %RE of BOD and COD using nano-adsorbents with different agro-processing industrial effluents such as milk, bagasse, egg, tannery, olive mill, and coffee processing industries was reported in this study (Table 6). Thirugnanasambandham and Sivakumar (2016) worked on CZO nano beds for application in wastewater treatment in the milk processing industry. The results of this experiment suggested that, 89 % of turbidity and 97% of COD were reduced at process parameters of 4.52 pH, 1.5 g/L CZO dose, 45 min contact time, and 12% dilution. Thirugnanasambandham et al. (2014a) studied the effect of nano-chitosan on bagasse processing industry wastewater treatment. The study was conducted with the four independent variables, and the experimental process was optimised using the response surface methodology. The maximum %RE of BOD (84.90%) and COD (93%) were achieved at optimal conditions of pH 6, chitosan dose of 1.8 g/L, and settling time of 60 min. At the same time, Thirugnanasambandham et al. (2014b) worked on chitosan as an effective adsorbent for egg processing industry wastewater treatment. Under the optimal process variables such as pH 4, chitosan dose of 1.1 g/L and settling time of 40 min, the maximum reductions of BOD, COD, and turbidity were 83%, 88%, and 94%. Various kinetic and isotherm models were used to predict the data of COD by using the experimental values. A green synthesised MgO nano-adsorbent was synthesised by Oladipo et al. (2017), and MgO was tested for tannery wastewater. The batch adsorption experimental results were shown, 95.35% COD reduction, 94.49% BOD reduction, and 83.33% turbidity reduction at 120 min contact time. Nassar et al. (2014) studied the use of commercially available ᵞ-Fe2O3 nano-particles on decolourization, dephenolization, and COD removal from olive mill wastewater. Devi et al. (2008) achieved 99.18% BOD adsorption from avocado peel carbon and 98.20% COD reduction by using commercial activated carbon.
Above, literature findings showed that, the maximum %RE of BOD and COD was achieved in the real and synthetic effluent samples. This might be due to the prior investigations directly using nano-particles as an adsorbent. As a result, it will almost certainly yield a higher %RE in batch adsorption experiments. The direct use of nano-particles as an adsorbent is not environmentally friendly, cost-effective, and leads to nano-particle leaching, resulting in nano-particle residual contamination in water and soil (Pathak et al. 2016). In order to address these issues, in the present investigation, CSZOCS is used as an effective adsorbent for MPIW treatment.