Introduction

Chromium is a chemical species rarely found in natural water, and its presence can be an indicative of environmental pollution caused by electroplating, pigment production, metal plating, dyeing, wood-preserving industries, and so forth (Ravikumar et al. 2015; Deng et al. 2018). Although chromium exists in several oxidation states, the most common species are the chromium (VI) [Cr (VI)] and chromium (III) [Cr (III)] (Wan et al. 2018). Cr (VI) is mutagenic and carcinogenic to humans and highly toxic to animals and plants (Chen et al. 2018; Choudhary and Paul 2018; Pang et al. 2018). The Cr (VI) concentration in drinking water used for human consumption may not be higher than 0.050 mg L−1, and the maximum permitted concentration in water and wastewater for the final disposal in the environment is 0.10 mg L−1 (USEPA 2012).

Several methods have been developed and applied for the removal of heavy metals contained in aqueous solutions, including chemical precipitation, ionic exchange, filtration membrane, oxidation-reduction process, dialysis, electrodialysis, and electrolytic extraction (Aslani et al. 2018; Li et al. 2018). However, some of these methods are infeasible for the removal of certain chromium species due to their recalcitrant properties. Moreover, most of the developed methods have high cost, formation of secondary pollutants, high energy consumption, long processing times, poor adsorbent regeneration, and low chromium removal percentage (Li et al. 2018; Wan et al. 2018). Therefore, it is necessary to develop alternative methods for the removal of Cr (VI) ions contained in water and industrial wastewater.

The methods for the removal of heavy metals contained in aqueous solutions based on adsorption are simple, inexpensive, effective, and easy to operate and access (Vieira et al. 2018). In general, the adsorbents used for the adsorption and removal of heavy metals contained in water and wastewater need to be inexpensive, biodegradable, and non-toxic (Zhu et al. 2017). In this case, the most common adsorbents include zeolites, clays, silicates, biomass, polymers, hydrogels, composites, and so forth (Pereira et al. 2017; Vieira et al. 2018). Natural polysaccharide-based hydrogels are adsorbents formed by hydrophobic/hydrophilic three-dimensional polymer networks capable of absorbing large amounts of water. Hence, these adsorbents can be efficiently employed for the adsorption of dyes, proteins, agricultural nutrients, and heavy metals (Vieira et al. 2018).

The aim of this work was to study the adsorption and removal of Cr (VI) ions contained in aqueous solutions using a chitosan-based hydrogel as alternative adsorbent. These studies were performed at different contact times, pH, and initial metal concentrations. The Cr (VI) adsorption mechanism in the hydrogel was evaluated using non-linear Langmuir, Freundlich, Redlich-Peterson, and Sips isotherms, whereas the Cr (VI) adsorption kinetic was evaluated using non-linear pseudo-first, pseudo-second, and pseudo-nth order kinetic models.

Materials and methods

Reagents

Chitosan from shrimp (CS—deacetylation degree: 92.0 wt-%; 1.0 × 106 Da molar mass and 425 μm particle size), ammonium persulfate (APS), N,N′-methylenebisacrylamide (MBA), and acrylic acid (AA) were purchased from Aldrich®. Hydrochloridric acid and acetic acid (AAc) were purchased from Merck®. Sodium hydroxide (NaOH) and potassium dichromate (Cr2K2O7) were purchased from Dinamica®. All reagents were used as acquired and aqueous solutions were prepared in ultrapure water (Megapurity MEGA-UP).

Hydrogel synthesis

The chitosan-based hydrogel was synthesized by placing 30.0 mL of 1.0 wt-% CS solution, 3.40 mL of AA, 0.150 g of MBA, and 0.5215 mmol of APS (initiator) in a three-neck glass flask equipped with a mechanical stirrer, reflux condenser, and nitrogen line. The hydrogel synthesis was performed at 343.15 K for 3 h under inert atmosphere for complete crosslinking. The formed hydrogel was transferred to a 200.0-mL glass beaker containing 100.0 mL of 2.0 mol L−1 sodium hydroxide solution and left for 15 min to neutralize. Finally, the hydrogel was washed with ultrapure water and dried in an air circulation oven (Ethiktechnology 400/5TS) at 333.15 K for 72 h prior to characterization and application for the adsorption and removal of Cr (VI) ions contained in aqueous solutions.

Fourier-transform infrared spectroscopy

Fourier-transform infrared spectroscopy (FT-IR) was conducted with the Bomem Easy MB-100 Nichelson spectrometer. The hydrogel sample was prepared using 1.0 wt-% KBr pellets and FTIR spectra obtained with 20 runs per minute at resolution of 4.0 cm−1.

Adsorption experiments

The Cr (VI) adsorption experiments were performed using 1000 mg L1 stock metal solutions at pH 4.5, 5.5, and 6.5. Dried hydrogel masses (100.0 mg) were immersed in 120.0-mL glass Erlenmeyer flasks containing 50.0 mL of Cr (VI) ion aqueous solutions at specific initial concentrations. The Cr (VI) concentrations were determined via flame atomic absorption spectrometry (FAAS—Analytic Jena AG, Jena, Germany, contra 700, air-acetylene flame, wavelength for chromium of 357.8687 nm). The removal percentage (R, %) was calculated using Eq. (1):

$$ \mathrm{R},\%=\frac{{\mathbf{C}}_{\mathbf{0}}-{\mathbf{C}}_{\mathbf{e}}}{{\mathbf{C}}_{\mathbf{0}}}100 $$
(1)

in which C0 and Ce are the initial and equilibrium concentrations (mg L−1), respectively.

The Cr (VI) adsorption capacities (qe, mg g−1) of the hydrogel were calculated using Eq. (2):

$$ {q}_e=\frac{{\mathrm{C}}_0-{\mathrm{C}}_{\mathrm{e}}}{m}\ V $$
(2)

in which m (g) is the dried hydrogel mass and V (mL) is the aqueous solution volume.

Adsorption isotherms

The Langmuir isotherm model is employed to explain the monolayer adsorption mechanism occurring at a defined number of active sites on the hydrogel network (Langmuir 1916). On the contrary, the Freundlich isotherm model is employed to explain the multilayer adsorption mechanism occurring on heterogeneous surfaces, with intermolecular interactions among adsorbed ions on neighbor active sites (Freundlich 1906). The three-parameter Redlich-Peterson and Sips isotherm models are employed to study homogeneous and heterogeneous adsorption systems (Foo and Hameed 2010). The Redlich-Peterson isotherm model describes the adsorption occurring with monolayer formation and multi-site interactions simultaneously (Redlich and Peterson 1959), whereas the Sips isotherm model describes the adsorption occurring only on heterogeneous systems (Sips 1948). The non-linear Langmuir, Freundlich, Redlich-Peterson, and Sips isotherm models are represented by Eqs. (3), (4), (5), and (6), respectively:

$$ {q}_e=\frac{{\mathrm{q}}_{max}{K}_L{C}_e}{1+{K}_L{C}_e\ } $$
(3)

in which qe (mg g−1) is the adsorption capacity of the hydrogel, qmax (mg g−1) is the maximum adsorption capacity, KL (L mg−1) is the Langmuir adsorption constant, and Ce (mg L−1) is the metal concentration at equilibrium:

$$ {q}_e={K}_F{C}_e^{\frac{1}{n}} $$
(4)

in which KF (mg g−1) is the Freundlich constant related with the adsorption capacity and 1/n is the Freundlich constant related with the surface heterogeneity of the specific adsorbate.

$$ {q}_e=\frac{K_R\ {C}_e}{1+{a}_R\ {C}_e^{\beta }} $$
(5)

in which aR (mg−1) and KR (L g−1) are the Redlich-Peterson constants, and β is an exponent having values between 0 and 1:

$$ {q}_e=\frac{q_{ms}{K}_s\ {C_e}^{\beta_S}}{1+{K}_S{C}_e^{\beta_S}} $$
(6)

in which qms (mg g−1) is the maximum Sips adsorption capacity, Ks (L mg−1) is the Sips constant at equilibrium, and βS is the Sips exponent describing the homogeneity and heterogeneity of an adsorption mechanism.

Adsorption kinetics

Kinetic studies are employed to evaluate the adsorption dynamic with base on the rate constants (Norouzi et al. 2018). The steps of an adsorption process are governed by (i) transport of solute from the solution bulk to a thin film formed on the adsorbent surface, (ii) transport of solute from this thin film to the adsorbent surface, and (iii) transport of solute from the adsorbent surface to the interior of the porous structure (Sharma and Bhattacharyya 2005). The Cr (VI) adsorption kinetics in the chitosan-based hydrogel were studied using non-linear pseudo-first, pseudo-second, and pseudo-nth order kinetic models. The non-linear pseudo-first and pseudo-second order kinetic models are represented by Eqs. (7) and (8), respectively:

$$ {q}_t={q}_e\left(1-{e}^{- kt}\right) $$
(7)
$$ {q}_t=\frac{\mathrm{k}{q}_e^2\mathrm{t}\ }{1+\mathrm{k}{q}_et} $$
(8)

The non-linear pseudo-nth order kinetic model is represented by Eq. (9):

$$ {q}_t={q}_e-{\left[\left(n-1\right) kt+{q_e}^{\left(1-n\right)}\right]}^{\frac{1}{1-n}} $$
(9)

in which k (min−1) is the adsorption rate, t (mim) is the adsorption time, qt (mg g−1) is the adsorption capacity at adsorption time t, and n is the reaction order with regard to the effective concentration.

Error analysis

The experimental data obtained from the Cr (VI) adsorption isotherms and kinetics in the chitosan-based hydrogel were evaluated using chi-square test statistic (χ2) (Tang et al. 2018) which is represented by Eq. (10):

$$ {\chi}^2=\sum \frac{{\left({q}_{experimental}-{q}_{e, theoretical}\right)}^2\ }{q_{e, theoretical}} $$
(10)

The coefficient of determination (R2) based on the chi-square test statistic is represented by Eq. (11):

$$ {R}^2=\frac{\sum {\left({q}_{e, theoretical}-\overline{q_e}\right)}^2\ }{\sum {\left({q}_{e, theoretical}-\overline{q_e}\right)}^2+\sum {\left({q}_{e, theoretical}-{q}_{experimental}\right)}^2} $$
(11)

in which qe,theoretical and qexperimental are the theoretical adsorption capacities (mg g−1) predicted from each model and experimental adsorption capacities (mg g−1) calculated from the experimental adsorption results, respectively. \( \overline{q_e} \) (mg g−1) is the average of qexperimental.

Results and discussion

Fourier-transform infrared spectroscopy

Figure 1a shows a Fourier-transform infrared spectrum for the chitosan-based hydrogel. The absorption bands at 3400 and 3200 cm−1 are attributed to stretching vibrations of O–H and N–H groups, respectively, from CS molecules. The absorption band at 3400 cm−1 can also be attributed to stretching vibrations of O–H groups from AA molecules. The absorption band at 2925 cm−1 is related to the vibrations of aliphatic compounds containing C–H groups. Carbonyl groups (C=O) are frequently observed at absorption bands from 1820 to 1680 cm−1 (He et al. 2015; Falamarzpour et al. 2017). Hence, the absorption band at 1750 cm−1 indicates the presence of C=O groups from polyacrylic acid (PAA), MBA, and radical CS. The absorption bands between 1500 and 1600 cm−1 are associated with primary and secondary amide groups due to occurrence of crosslinking reactions involving high deacetylation-degree chitosan (Li and Tang 2016; Falamarzpour et al. 2017; Mohamed and El-Ghany 2017). The absorption band at 1577 cm−1 is related to the stretching deformation of N–H groups from CS and MBA molecules. Finally, the absorption band at 1463 cm−1 is related to the stretching of C–O–C groups.

Fig. 1
figure 1

a, b Fourier-transform infrared spectrum for the chitosan-based hydrogel (a). Photos of the dried hydrogel particles, dried hydrogel piece, and swollen hydrogel piece (b)

Full size image

Other characterizations

Scanning electron microscopy (SEM) images for the chitosan-based hydrogel were recently published by our research group (Varnier et al. 2018). Mechanical properties were confirmed by the values of the modulus of elasticity (0.30 kPa) and applied compression force (0.38 N). Photos of dried hydrogel particles, dried hydrogel piece, and swollen hydrogel piece are shown in Fig. 1b.

Effect of contact time and pH

Figure 2 shows the Cr (VI) adsorption capacities of the chitosan-based hydrogel versus time at pH 4.5, 5.5, and 6.5. The Cr (VI) adsorption capacities were 47.36, 43.90, and 40.79 mg metal per g dried hydrogel at pH 4.5, 5.5, and 6.5, respectively, after 1440 min. The high initial adsorption rates are results of the high amounts of available active sites in the hydrogel. These adsorption rates decreased with the increase of the contact time due to partial saturation of the active sites. It confirms that the adsorption kinetic depends on the adsorbate rate transported from the solution bulk to the adsorbent surface (Nascimento et al. 2014). The Cr (VI) adsorption capacity increased with the decrease of pH due to changes in the ionic network of the chitosan-based hydrogel (Paulino et al. 2011a) and formation of specific Cr (VI) species in aqueous solutions at different pH values (Wan et al. 2018). The COO− and NH2 groups contained in the hydrophilic three-dimensional network of the chitosan-based hydrogel are protonated at pH values lower than 6.0, forming COOH and NH3+. This decreases the intermolecular interaction forces between Cr (VI) ions and active sites. However, the high H+ ion concentration can be predominant to form significant amounts of either NH3+ or NH42+ and increase the electrostatic attraction forces between NH3+ or NH42+ and chromate (CrO4−2), dichromate (Cr2O7−2), hydrogen chromate (HCrO4−), and hydrogen dichromate (HCr2O7−) (Luz and Oliveira 2011; Ba et al. 2018). HCrO4− ions are considered the main ionic forms of chromium at pH from 2.0 to 6.4, whereas CrO4−2 ions are considered the main ionic forms of chromium at pH higher than 6.4. Hence, higher HCrO4− ion concentrations were adsorbed at lower pH values, increasing the Cr (VI) adsorption capacities of the hydrogel (Chai et al. 2018).

Fig. 2
figure 2

Cr (VI) adsorption capacities of the chitosan-based hydrogel versus time at pH 4.5, 5.5, and 6.5. Experimental conditions 100.0 mg L−1 initial metal concentration, 100.0 mg hydrogel mass and room temperature

Full size image

Effect of initial Cr (VI) concentration at different pH values

Figure 3 shows the Cr (VI) adsorption capacities of the chitosan-based hydrogel versus pH at initial metal concentrations from 100.0 to 400.0 mg L−1. The Cr (VI) adsorption capacities ranged from 47.36 to 75.80 mg metal per g dried hydrogel with the increase in the initial metal concentration from 100.0 to 400.0 mg L−1, respectively, at pH 4.5. The Cr (VI) adsorption capacities ranged from 43.90 to 60.85 mg metal per g dried hydrogel, respectively, at pH 5.5 and 40.79 to 23.63 mg metal per g dried hydrogel, respectively, at pH 6.5. The Cr (VI) adsorption capacities decreased with the increase of pH due to the decrease of the HCrO4− ion concentrations in aqueous solution and cationic group concentrations in the hydrophilic three-dimensional network of the chitosan-based hydrogel. Moreover, the increase of pH increased the electrostatic repulsion forces among CrO4−2, Cr2O7−2, HCrO4−, and HCr2O7− ions and carboxylate groups (COO−) in the hydrogel network (Paulino et al. 2011a), decreasing the adsorption capacities. The stronger intermolecular interaction forces that occurred between HCrO4− ions and cationic groups, at pH lower than 6.0, overcame the lower degrees of swelling of the chitosan-based hydrogel (Vieira et al. 2018), increasing the Cr (VI) adsorption capacities. The Cr (VI) removal percentage from an aqueous solution with pH 4.5 and 100.0 mg L−1 initial metal concentration was 94.72%. The Cr (VI) removal capacities and percentages increased with the increase of the initial metal concentrations at pH 4.5 and 5.5 up to saturation of the active sites in the hydrogel network. Otherwise, the Cr (VI) removal capacities and percentages decreased with the increase of the initial metal concentration at pH 6.5. This indicates that the CrO4−2 ion affinity by active sites in the chitosan-based hydrogel, at higher pH values, is lower than HCrO4− ion affinity, which predominates in more acidic solutions (Wan et al. 2018).

Fig. 3
figure 3

Cr (VI) adsorption capacities of the chitosan-based hydrogel versus initial metal concentration at pH 4.5, 5.5, and 6.5. Experimental conditions contact time of 2880 min; initial metal concentration from 100.0 to 400.0 mg L−1; 100.0 mg adsorbent mass; 50.0 mL solution volume and room temperature

Full size image

Adsorption isotherms

Figure 4 shows the results of the non-linear Langmuir, Freundlich, Redlich-Peterson, and Sips isotherms for the adsorption of Cr (VI) ions to the chitosan-based hydrogel. Table 1 shows the isotherm parameters. The best fits for the adsorption of Cr (VI) ions were obtained using the non-linear Redlich-Peterson and Sips isotherm models, with higher R2 value and lower chi-square test statistic (χ2) for the non-linear Redlich-Peterson isotherm. It indicates that the adsorption of Cr (VI) ions is governed by formation of monolayer and multi-site interactions concomitantly, modifying the hydrogel structure (Vieira et al. 2018; Paulino et al. 2011b). As the exponent β of the non-linear Redlich-Peterson isotherm model was not equal to 1, the adsorption process is not exclusively governed by the Langmuir isotherm (Redlich and Peterson 1959). Hence, the maximum Cr (VI) adsorption capacity should be better represented by the non-linear Sips isotherm model. The maximum Cr (VI) adsorption capacities of the chitosan-based hydrogel were 73.14 and 93.03 mg metal per g dried hydrogel, according to the non-linear Langmuir and Sips isotherm models, respectively. The maximum Cr (VI) adsorption capacities of the chitosan-based hydrogel were compared to the adsorption capacities of other adsorbent materials studied at different pH values as shown in Table 2. As it can be seen, the chitosan-based hydrogel could be efficiently applied for the removal of Cr (VI) ions contained in aqueous solutions compared to some adsorbent materials.

Fig. 4
figure 4

Non-linear Langmuir, Freundlich, Redlich-Peterson and Sips isotherms for the adsorption of Cr (VI) ions to the chitosan-based hydrogel. Experimental conditions: contact time of 1440 min; initial metal concentration from 50.0 to 500.0 mg L−1; 100.0 mg adsorbent mass; 50.0 mL solution volume; pH 4.5 and room temperature

Full size image
Table 1 Parameters of the non-linear Langmuir, Freundlich, Redlich-Peterson, and Sips isotherms for the adsorption of Cr(VI) ions to the chitosan-based hydrogel
Full size table
Table 2 Maximum Cr(VI) adsorption capacities of the chitosan-based hydrogel compared to other adsorbent materials studied at different pH values
Full size table

Adsorption kinetics

Figure 5 shows the experimental and predicted results of the non-linear pseudo-first, pseudo-second, and pseudo-nth order kinetic models for the adsorption of Cr (VI) ions to the chitosan-based hydrogel at pH 4.5 (a), 5.5 (b), and 6.5 (c). Table 3 shows the parameters of the non-linear pseudo-first, pseudo-second, and pseudo-nth order kinetic models. The best kinetic fit for the adsorption of Cr (VI) ions to the chitosan-based hydrogel at pH 4.5 was obtained using the non-linear pseudo-nth order kinetic model, with higher R2 value and lower chi-square test statistic (χ2). However, the adsorption kinetics changed with the increase of pH since the best kinetic fit at pH 6.5 was found with the non-linear pseudo-first order kinetic model when comparing R2 and χ2. This confirms that CrO4−2 ions predominate at less acidic solutions. As the reaction order (n) was approximately 2 at pH 4.5, both non-linear pseudo-nth and pseudo-second-order kinetic models can be used to fit the experimental adsorption data (Guechi and Hamdaoui 2013). The best kinetic fit for the adsorption of Cr (VI) ions to the chitosan-based hydrogel at pH 5.5 was found with the pseudo-second-order kinetic model, confirming the reaction order (n = 2.000). This suggests that the adsorptions of different chromium species to the chitosan-based hydrogel are governed by the active-binding site number (Norouzi et al. 2018). Overall, the adsorption of Cr (VI) ions to the chitosan-based hydrogel is more favorable at lower pH values.

Fig. 5
figure 5

Non-linear pseudo-first, pseudo-second, and pseudo-nth-order kinetic models for the adsorption of Cr (VI) ions to the chitosan-based hydrogel at pH 4.5 (a), 5.5 (b), and 6.5 (c). Experimental conditions: 100.0 mg L−1 initial metal concentration; 100.0 mg adsorbent mass; 50.0 mL solution volume and room temperature

Full size image
Table 3 Parameters of the non-linear pseudo-first, pseudo-second and pseudo-nth order kinetic models for the adsorption of Cr(VI) ions to the chitosan-based hydrogel at different pH values
Full size table

Figure 6 shows the experimental and predicted results of the non-linear pseudo-first, pseudo-second, and pseudo-nth-order kinetic models for the adsorption Cr (VI) ions to the chitosan-based hydrogel at initial metal concentrations of 100.0 (a), 200.0 (b), 300.0 (c), and 400.0 mg L−1 (d). Table 4 shows the parameters of the non-linear pseudo-first, pseudo-second, and pseudo-nth-order kinetic models. All coefficients of determination (R2) were higher, and chi-square test statistics (χ2) were lower when using the non-linear pseudo-nth-order kinetic model, independently of the initial Cr (VI) concentration. Hence, the best kinetic fits for the adsorption of Cr (VI) ions to the chitosan-based hydrogel at all initial concentrations were found with the non-linear pseudo-nth-order kinetic model. Moreover, the calculated and predicted qe values were similar to the qe values determined experimentally during the adsorption of Cr (VI) ions, confirming that the best fit was found with the non-linear pseudo-nth-order kinetic model. These results can be employed for confirming a favorable Cr (VI) adsorption process in hydrogels.

Fig. 6
figure 6

Non-linear pseudo-first, pseudo-second, and pseudo-nth-order kinetic models for the adsorption of Cr (VI) ions to the chitosan-based hydrogel at initial metal concentrations of 100.0 (a), 200.0 (b), 300.0 (c), and 400.0 mg L−1 (d). Experimental conditions: 100.0 mg adsorbent mass; 50.0 mL solution volume; pH 4.5 and room temperature

Full size image
Table 4 Parameters of the non-linear pseudo-first, pseudo-second, and pseudo-nth-order kinetic models for the adsorption of Cr(VI) ions to the chitosan-based hydrogel at different initial metal concentrations (C0)
Full size table

Conclusion

  • A chitosan-based hydrogel could be efficiently employed as alternative adsorbent for the adsorption and removal of Cr (VI) ions contained in aqueous solutions.

  • The best fit for the adsorption of Cr (VI) ions to the chitosan-based hydrogel was found with the non-linear Redlich–Peterson isotherm model.

  • The Cr (VI) adsorption mechanism was governed by formation of monolayer and multi-site interactions concomitantly.

  • The best kinetic fit for the adsorption of Cr (VI) ions to the chitosan-based hydrogel was found with the non-linear pseudo-nth-order kinetic model.

  • A chitosan-based hydrogel could be efficiently applied as cost-effective and eco-friendly alternative adsorbent for the water and wastewater treatment containing heavy metals.