Simultaneous removal of heavy metals and dye from wastewater: modelling and experimental study

Water pollution is one of the biggest threats to the environment. Water gets polluted due to the release of various pollutants from industrial activities. The presence of pollutants higher than the permissible limit (3, 0.1, 2, 5, 0.1, 0.01, 0.2, 3 mg/L for copper, chromium, cadmium, zinc, lead, mercury, arsenic and nickel, respectively) causes harmful effects on aquatic life. The excessive release of pollutants such as heavy metals, dyes and pigments, pesticides, pharmaceutical compounds, etc. released by textile, tannery, electroplating and pharmaceutical industries leads to water pollution. They can be toxic or carcinogenic in nature and can cause severe problems for humans (cancer, abdominal pain, nausea and liver damage) and aquatic ecosystems (Lin et al. 2017; Aigbe et al. 2018). Among various pollutants, dye and heavy metals are the most common and harmful/toxic to the environment.

Heavy metal pollution is one of the global threat because of the industrial revolution. However, in dyes metal complex dyes are prominent (Banat et al. 1996). The presence of dye in the water is highly visible and affects water transparency and gas solubility. Heavy metals and dyes come out simultaneously from the textile, printing and dye industries (Ren et al. 2008; Fu & Wang 2011). The presence of dyes and heavy metals together in the effluent stream necessitates the treatment for their simultaneous removal. There are various conventional techniques such as chemical precipitation, flotation, ion exchange, coagulation and flocculation, membrane filtration, reverse osmosis, and adsorption (Fu & Wang 2011). Among all these techniques, the adsorption process is the most efficient and widely used technique for the removal of heavy metals from wastewater because of its low cost, availability and eco-friendly nature.

Tovar-Gomez et al. (2012) has removed heavy metals such as nickel, cadmium, zinc and acid blue 25 dye simultaneously using Ca(PO₃)₂-modified carbon adsorbent. It was reported that presence of acid blue 25 dye has increased removal of heavy metals in binary mixture of dyes and heavy metals. Similarly, Aguayo-Villarreal et al. (2013) has removed cadmium, zinc and acid blue 25 simultaneously using activated carbons and it was observed that acid blue 25 has a synergistic effect on cadmium and zinc removal and in its presence cadmium and zinc removal has increased 60 times as compared to the results obtained in monometal system. Further, in simultaneous removal of dyes (methyl orange, methylene blue) and heavy metals (copper, cadmium, nickel) using fly ash, dyes were competing with heavy metals. The removal order was dye > copper > cadmium > nickel. However, various dyes such as basic violet 3, acid blue 25, basic blue 9 and heavy metals such as nickel, lead, cadmium and zinc were removed simultaneously using (CL) clinoptilolite and (erionite) ER. In this system, there is competition between heavy metals and dyes for the same adsorption sites and it reduces the adsorption capacity of both dye and heavy metals (Hernández-Montoya et al. 2013). Deng et al. (2013) has removed cadmium and dyes (methylene blue and orange G) simultaneously. Methylene blue was affecting cadmium antagonistically but cadmium does not effect on methylene blue dye in Cd(II)–methylene blue binary system. However, the removal of orange G increases with cadmium concentration but the presence of orange G does not effect the removal of cadmium. According to Taştan et al. (2010) when Cr(VI) and dye are present together then chromium will decrease the removal of Cr and dye as well because of the antagonistic effect. But for Cu(II) and dye, copper will not affect dye removal but Cu(II) removal rate will be increased from 29.06% to 37.91% by the presence of dye because the dye is acting antagonistically.

Although a good literature work is available for dye and single metal removal, the literature related to multimetal and dye is still not available. Therefore, in this paper, with the help of multimetal dye system, the effect of dye on metal removal capacity using synthesised adsorbent was analysed.

In this work, modified wheat bran was used as an adsorbent for copper, chromium and cadmium removal from wastewater in batch and continuous modes. In batch mode, the Box–Behnken design was used in response surface methodology for carrying out experiments. Further, for simultaneous removal of heavy metals and dyes in continuous mode, pellets were prepared using modified wheat bran alongwith clay and chitosan as a binder. A dynamic mathematical model was developed for continuous fixed bed adsorption column to compare the breakthrough curve with experimental results. The effect of various parameters such as initial metal concentration of copper, chromium, cadmium and adsorbent dose was investigated. In continuous column study experiments were carried out for mixed metal system at various parameters bed height (0.30 m), flow rate (10 mL/min) and initial metal concentration (300 mg/L). Then breakthrough curves obtained from experimental data and model data were compared.

Bentonite clay and chitosan (Merck, Germany) were purchased from a local market. Wheat bran was collected from the local bakery shop and flour mill at Jaipur. Hydrochloric acid (HCl, Rankem) of analytical reagent grade was purchased from a local market. Potassium dichromate (K₂Cr₂O₇), cupric sulphate pentahydrate (CuSO₄.5H₂O) and cadmium chloride monohydrate (CdCl₂.H₂O) were purchased from Sigma Aldrich, New Delhi, India (Table 1).

In simultaneous metal study, copper, chromium, cadmium and acid black 60 dye containing equal volume of each solution in 250 mL flask was prepared. The concentration was ranged between 15 mg/L to 100 mg/L for each metal and 15 mg/L to 200 mg/L for acid black 60 dye. The multicomponent solution was prepared using salt of cupric sulphate pentahydrate, potassium dichromate and cadmium chloride monohydrate. The adsorbent dose was varied from 0.5 g to 5 g and then flasks were shaken at 180 rpm at 35 °C in incubator for 4 h. The pH of the solution was ranged between 2 and 10. Adsorbent was separated from the solution and absorbance was measured using atomic absorption spectrophotometer. The effect of several parameters such as pH, temperature, adsorbent dose, initial concentration of copper, initial concentration of chromium and initial concentration of cadmium were investigated using Box–Behnken design.

R² values tell that how well a model fits to the data points. The range of R² value is from 0 to 1 and 1 indicates the ideal value. R is predicted response (removal efficiency of heavy metal) where X_i = 1, 2, 3 are independent factors. While c_o, c_i (i = 1,2,3), c_ii (i = 1,2,3), c_ij (i = 1,2,3; j = 1,2,3) are model coefficients.

Elemental analysis of the raw wheat bran and chemically MWB was done using EDS to determine its chemical composition as shown in Figure 1(a) and 1(b), respectively. It was observed that there is the presence of carbon and oxygen elements in cellulose, having a weight percent of 54.65% and 65.85%, respectively. But after the acid treatment of wheat bran, the carbon wt % increases to 65.85% and oxygen wt% decreases to 31.82% because of loss of water molecules. Treatment with concentrated hydrochloric acid has caused dehydration and charring of the wheat bran, leading to increase in the carbon content and development of active sites for metal adsorption. This improves the extraction behaviour of wheat bran (Krishnani et al. 2004).

Further, Figure 2(b) shows FTIR of adsorbent pellets. In FTIR analysis of adsorbent pellets, peak at 3,621 cm⁻¹ is due to the asymmetric stretching of Al–OH–Al, 3,379 cm⁻¹ peak is due to vibrations of water molecules, 1,631 cm⁻¹ peak shows N–H bending vibrations in N–H, peak at 1,469 cm⁻¹ shows bending vibrations of amine group (–NH₂) present in chitosan before adsorption but after adsorption, this peak diminishes, peak at 1,009 cm⁻¹ is due to –CO stretching vibration in –COH, peak at 793 cm⁻¹ shows stretching vibration of Al-O-Si present in clay. Further, 933 cm⁻¹, 869 cm⁻¹, 691 cm⁻¹, 524 cm⁻¹, and 425 cm⁻¹ peaks show Al–Al–OH deformation, Al-Mg-OH deformation, coupled Al-O and Si-O, Al-O-Si deformation and Si-O-Si bending vibrations. Peaks at 2,922 cm⁻¹, 2,296 cm⁻¹ and 2,074 cm⁻¹ is due to the –C–H stretching in cellulose present in wheat bran. In Figure 2(c), there are various peaks of XRD pattern because of cellulose, clay and chitosan. Peaks at 2θ = 20°, 22° and 25.32°, 26.83°, 31.90°, 35.11°, 45.68°, 62.12°, 68.26° are due to chitosan, silica and bentonite clay. But after adsorption, these peaks shift to the lower intensity.

In this work, design of experiments was used for optimization of operating conditions and removal efficiency, and parameters such as initial concentration, pH, temperature and adsorbent dose was considered (Table 2). 3-level, 4-factor Box–Behnken design (BBD) was used to determine effect of these parameters on removal of heavy metals. Box–Behnken design is appropriate to use because it gives few combinations of variables to determine complex response function (Muthukumar et al. 2003). In our work, total 62 experiments were carried out for each metal ion (Table 3).

Table 4 shows ANOVA for response surface quadratic model. The model F value is 2.97, 3.47, 3.46, 3 for copper, cadmium, chromium and dye. This value measures how well factors describe the variation in the data about its mean (Behbahani et al. 2021; Cheng et al. 2021; Pal et al. 2021). Probability value (P) is a measurement of effects in a model that should be less than 0.05 for being significant, and in our work probability values are less than 0.05 for parameters A₃, A₂, A₄, A₂D, A₃², A₄², C², B² for copper; A₃, A₁, A₄, B, A₁D, DB, A₁², A₄², C² for cadmium; A₃, A₁, A₄, DB, A₃², D² for chromium and A₃, A₂, A₁, A₄, B, A₃C, A₂B, A₄B, CB, D₂ for dye. It means that only these parameters are significant for copper, cadmium, chromium and dye removal. Adequate precision ratio is 7.158 for copper removal, 6.807 for cadmium removal, 7.860 for chromium removal, and 6.201 for dye removal which is much higher than the minimum desirable amount of 4 and it indicates that there is the presence of adequate signal to noise ratio for this model (Srinivasan & Viraraghavan 2010; Ahmadi & Harouni 2014; Afshin et al. 2021).

Chromium removal increases 80% to 100% with increase in copper concentration from 15 to 100 mg/L; it shows that increase of copper concentration has synergistic effect on chromium removal (Figure 10(b)). Cadmium removal decreases 100% to 50% with increase in copper concentration from 15 to 100 mg/L; it shows that increase of copper concentration has antagonistic effect on cadmium removal (Figure 9(a)).

Copper removal increases 20% to 60% with increase in cadmium concentration from 15 to 100 mg/L, it shows that increase of cadmium concentration has synergistic effect on copper removal (Figure 8(a)). Chromium removal decreases 100% to 90% with increase in cadmium concentration from 15 to 100 mg/L; it shows that increase of cadmium concentration has antagonistic effect on chromium removal (Figure 7(a)).

Copper removal increases from 20% to 40% with increase in chromium concentration from 15 to 100 mg/L, it shows that chromium concentration has synergistic effect on copper removal (Figure 8(a)). Cadmium removal decreases from 100% to 20% with increase in chromium concentration from 15 to 100 mg/L; it shows that chromium concentration has antagonistic effect on cadmium removal (Figure 9(a)).

In simultaneous removal of heavy metals and dye, dye is also added with the heavy metals. It is clear from Figures 8(b) and 9(b) that, when dye is added in ternary mixture of copper, cadmium and chromium, removal of copper, cadmium and chromium decreases. It shows that dye is acting as an interfering species in removal of all three heavy metals.

Thermodynamic study was done for individual metals and it suggest that the adsorption process is endothermic for copper and chromium, and exothermic for cadmium (Renu et al. 2018).

A dynamic adsorption model was developed including mass transfer resistance and dispersion phenomena. Operating system is in isothermal conditions, Langmuir isotherm was used for characterising the adsorption process, external-film mass transfer coefficient term was used and adsorbent particles are homogeneous in density and also in size.

The Equations (8) and (9) form a coupled system of differential equations and solved by finite difference technique using MATLAB software after discretizing the equations.

In simultaneous removal, heavy metals (copper, cadmium and chromium) and dye (acid black 60) were mixed together and passed through the continuous column. Simultaneous removal of heavy metals and dye was performed for flow rate of 10 mL/min, initial metal concentration of 300 mg/L and bed height of 0.30 m (Table 7).

Figure 15(b) shows the breakthrough curve for copper, cadmium and chromium in mixed system at initial metal concentration of 300 mg/L, flow rate of 5 mL/min and bed height of 0.15 m. It was observed that the breakthrough time has been decreased from 2,037 s to 922 s for copper, 825 to 512 s for cadmium, 1,514 to 565 s for chromium and 4,653 s to 2,870 s for dye. However, the exhaustion time reaches from 22,500 to 20,000 s for copper, 14,000 to 11,981 s for cadmium, 17,500 to 16,406 s for chromium and 20,000 to 18,393 s for dye. In mixed metal study, breakthrough and exhaustion time has been decreased for copper, cadmium, chromium and dye due to the antagonistic effect.

Figure 15(c) shows the breakthrough curve for copper, cadmium and chromium in mixed system at initial metal concentration of 100 mg/L, flow rate of 5 mL/min and bed height of 0.30 m. It was observed that due to antagonistic effect, breakthrough time has been decreased from 2,783 s to 1,925 s for copper, 1,436 s to 1,034 s for cadmium, 966 to 542 s for chromium and 4,826 s to 3,712 s for dye. However, the exhaustion time reaches from 12,000 to 10,184 s for copper, 7,000 to 6,749 s for cadmium, 14,656 to 12,406 s for chromium and 20,000 to 17,000 for dye.

The adsorption mechanism of heavy metals and dyes on adsorbent pellet has been examined by comparing XRD and FTIR of adsorbent pellet before and after adsorption. After the adsorption of heavy metals on adsorbent pellet, all XRD peaks shift towards the low diffraction (Figure 2(c)). It is clear that peaks at 2θ = 25.32°, 26.83°, 35.11° and 62.12° shift to 25.17°, 26.81°, 35.03° and 62.11° due to expansion in d-spacing caused by intercalation of the metals and dyes in the interlayer of adsorbent pellets. Similarly, in Figure 2(b), FTIR of adsorbent pellets show heavy metals and dyes adsorption on adsorbent pellets because after adsorption 1,469 cm⁻¹ peak of amine group (–NH₂) diminishes due to the adsorbent loading with heavy metals and dyes.

The results in the present study show that modified wheat bran is an efficient adsorbent for copper, chromium, and cadmium removal. The best local maximum value found was pH 7.59, initial metal concentration 93.50 mg/L for chromium, 15 mg/L for cadmium and 48.79 mg/L for copper, temperature 33.23 °C and adsorbent dose 2.90 g. Removal efficiency obtained at these conditions is 99.99% for cadmium, 87.275% for copper, and 79.70% for chromium and the desirability for these heavy metals is 0.862.

A dynamic mathematical model was developed for continuous fixed bed adsorption column to compare the breakthrough curve with experimental results. The high value of correlation coefficient R², low value of chi-square and mean absolute percent error (MAPE) for all the parameters indicate that the present model can predict the column behaviour with good accuracy and there is a good agreement in experimental data and model-predicted data obtained in MATLAB software. In simultaneous removal of dye and heavy metals, dye acted as an interferring species in heavy metal removal.

The authors thank to Department of Chemical Engineering for the financial support and materials research centre, MNIT, Jaipur for carrying out characterization analysis.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.