Introduction

Heavy metal contamination in the environment due to rapid industrialization has become a global concern. In recent times, copper (Cu(II)) ranks third as the most commonly utilized metal worldwide with an estimated annual discharge of 939,000 tons (Abdolali et al. 2017; Blagojev et al. 2019). Anthropogenic sources of Cu(II) in wastewaters include the pulp and paper industry, tanneries, refineries, electroplating, mining, and metal cleaning (Blagojev et al. 2019; Calero et al. 2018). Cu(II) is known as a vital trace element that performs several metabolic functions in the human body (de Freitas et al. 2018). Proteins in the presence of Cu(II) help maintain the body’s chemistry in terms of oxygen and electron transfer (Munagapati et al. 2010). Moreover, it also serves as a micronutrient for animals and plants (Rani et al. 2018). However, excessive ingestion of Cu(II) would result to severe health effects including cirrhosis, disorder of the central nervous system, kidney damage, stomach cramps, cancer, and vomiting (Georgopoulos et al. 2001; Li et al. 2017). In addition, high concentration of Cu(II) in water bodies proves to be highly toxic to aquatic organisms especially to fishes (de Freitas et al. 2017). The World Health Organization provided a maximum Cu(II) content of 2 mg/L in drinking water while the United States Environmental Protection Agency has set the maximum permissible level of Cu(II) to be 1 mg/L in wastewaters (Al-Saydeh et al. 2017; Babel and Kurniawan 2003).

Among the treatment technologies, adsorption is widely known due to its simple operation, versatility, low-cost, high efficiency even at low concentration of contaminants, ease of handling, possibility of metal recovery, and minimal generation of sludge (Bailey et al. 1999; Yun et al. 2001). Moreover, the wide application of adsorption is attributed to its practicability to be employed as a point-of-use treatment method (Arya and Philip 2016). Various adsorbent materials, inorganic and organic in nature, have been investigated in removing Cu(II) from synthetic and actual wastewaters. Alternative adsorbents derived from agricultural wastes or industry by-products including rice husks, orange peel, fly ash, clays, sand, and chitosan have been investigated in heavy metal removal. Reuse of industry by-products and waste materials implies reduced secondary pollution where the adsorption system is considered to be sustainable and inexpensive (Xiong et al. 2019).

Chitosan-based composites are a group of environment-friendly materials where powdered substances including sand, activated carbon, and clay are prepared in combination with chitosan. Chitosan-based composites have been utilized in extended applications due to its improved physico-chemical qualities such as increased mechanical and chemical strength, high density, enlarged pore diameter, and enhanced surface area (Wang et al. 2009; Wu et al. 2002). Literature review shows the extensive application of chitosan-based adsorbents such as chitosan/cotton fibers (Zhang et al. 2008), chitosan/cellulose (Balea et al. 2019; Sanchez-Salvador et al. 2018; Sun et al. 2009), chitosan-coated perlite (Kalayani et al. 2005), chitosan/clinoptilolite (Dinu and Dragan 2010), N-carboxylated carboxymethyl-chitosan (Escoda et al. 2013) and chitosan/silica hybrid microspheres (Zhao et al. 2013) in the removal of Cu(II) from aqueous solution. Chitosan, which is a natural cationic polymer derived from chitin, has several excellent properties including biodegradability, excellent chemical reactivity, non-toxicity, hydrophilicity, bactericidal property, and biocompatibility (Lewandowska et al. 2014; Vijaya et al. 2008). Natural clay minerals are typically composed of bentonite, chlorite, kaolinite, laponite, sepiolite, vermiculite, and hectorite. Clay minerals are hydrated phyllosilicates with various quantities of aluminum, silicon, magnesium, iron, calcium, sodium, and titanium (Kotal and Bhowmick 2015; Uddin 2017). Widespread application of clay as adsorbent materials is attributed to numerous features such as ease of surface modification, excellent cation exchange capacity, availability, and cost-effectiveness (Padilla-Ortega et al. 2013; Rekik et al. 2017). Bentonite is mainly composed of montmorillonite with a negatively charged, dioctahedral structure that has a 2:1 mineral to montmorillonite layer (Guggenheim et al. 2006; Şahin et al. 2015). Kaolinite has a 1:1 layered structure with a polar basal surface that is comprised of a tetrahedral [Si2O52−] sheet and octahedral [Al(OH)4]2+ sheet (Leonel et al. 2014). Sand is available in large quantities in numerous regions worldwide (Jiang et al. 2018). It is considered as an alternative adsorbent due to its reusability, high mechanical strength, abundance, and rapid attainment of adsorption equilibrium (Cestari et al. 2006). In literature, chitosan-coated bentonite (CCB), chitosan-coated kaolinite (CCK), and chitosan-coated sand (CCS) have been extensively applied in the removal of Cu(II) from aqueous solution under batch conditions (Chen et al. 2015; Dalida et al. 2011; Futalan et al. 2011; Wan et al. 2004). There are several limitations correlated to batch systems including reduced capacity in the handling of wastewater, short and intermittent duration of treatment period, and unsuitability of application in an actual wastewater treatment plant (Ai and Aris 2014; Podder and Majumder 2016). Only a handful of studies have been reported on the continuous adsorption of Cu(II) using chitosan-based composites. In the fixed-bed study of Futalan et al. (2011), the maximum adsorption capacity of Cu(II) at breakthrough was determined to be 14.92 mg/g using CCB. The scale-up and design of an industrial dynamic reactor would require essential data obtained from continuous fixed-bed systems. Moreover, a column reactor is described as a more systematic, cost-effective process (Dorado et al. 2014).

In the present work, a comparative evaluation of the adsorption capacity of CCB, CCK, and CCS in the removal of Cu(II) under a fixed-bed system was investigated. The influence of solution pH and initial concentration on the time to reach breakthrough and adsorption capacity at breakthrough was examined. Experimental data was fitted with the Thomas model. The regeneration capacity of the adsorbents was studied by carrying out one cycle of adsorption-desorption using HCl and NaOH as eluents.

Materials and method

Chemicals and reagents

Copper sulfate (Cu2SO4), hydrochloric acid (HCl, fuming 37%), and sodium hydroxide (NaOH) were acquired from Merck (Germany) while chitosan (low-molecular weight, ≥ 75% deacetylation) was purchased from Sigma-Aldrich (USA). Quartz sand, bentonite, and kaolinite were procured from New Anxing Company (Taiwan).

Synthesis of CCB, CCK, and CCS

The synthesis of the chitosan-based composites followed the method of Wan et al. (2010). Firstly, chitosan (5 g) was dissolved in 300 mL (5% v/v) HCl via continuous stirring for 2 h. Then, 100 g material (bentonite, kaolinite, or sand) was added into the mixture and stirred for 3 h. Drop by drop, 1 N NaOH was slowly added into the solution until pH 7.0 was attained. Adsorbent beads were washed with deionized water, filtered, and dried in an oven (Channel DV425) at 105 °C for 24 h. Adsorbent with particle size range of 0.5 to 0.7 mm was utilized in the study.

Characterization of adsorbents

The surface properties of CCB, CCK, and CCS were determined using the Brunauer-Emmett-Teller (BET) method (Micrometrics, ASAP 2020). Samples were degassed at a high temperature of 230 °C that would remove the water molecules present on the adsorbent surface. Thermogravimetric analysis (TGA) was performed using TGA 2950 to examine the changes in adsorbent weight as a function of temperature. The thermodynamic stability was conducted by heating about 5 to 10 mg adsorbent up to 800 °C at a rate of 10 °C min−1. The identification of chemical groups in the chitosan-based adsorbents before and after adsorption was carried out using Fourier transform infrared (FT-IR) spectroscopy. The spectra were recorded with Jasco FT-IR 410 using a 2-cm−1 resolution within wavelength from 4000 to 400 cm−1.

Fixed-bed experiments

Figure 1 illustrates the schematic setup for the fixed-bed system. Experiments were carried out using a borosilicate glass column reactor (working height = 30 cm; inner diameter = 3 cm) where about 10 g of adsorbent was packed. The influent was fed at the top of the column under a flow rate of 5 mL/min using a peristaltic pump (Sigma-Aldrich, Z678430). Samples were collected at the bottom of the reactor under pre-determined time intervals. The effect of solution pH (3.0 to 5.0) and initial concentration (200 and 1000 ppm) on the shape of the breakthrough curve and adsorption capacity at breakthrough was determined. Regeneration studies were performed by washing the spent adsorbent bed using 0.1 M HCl and 0.1 M NaOH as eluents. Residual Cu(II) concentration was analyzed using an inductively coupled plasma optical emission spectrometer (ICP-OES, Perkin-Elmer 2100DV).

Fig. 1
figure 1

Schematic diagram of the fixed-bed experimental setup

Full size image

Theoretical background

Several column parameters are derived from the breakthrough curve including time to reach breakthrough, total volume of effluent treated, breakthrough capacity, removal efficiency, and length of mass transfer zone. Time of breakthrough (tb) is described as the time when the effluent concentration reaches 10% of the influent concentration while time of exhaustion (te) refers to the time when effluent concentration reaches 90% of the influent concentration. The total volume of effluent treated (Veff) can be computed using Eq. 1 (Aksu and Gönen 2004):

$$ {V}_{\mathrm{eff}}=Q{t}_{\mathrm{total}} $$
(1)

where Q refers to the flow rate (mL/min) and ttotal refers to the time (min). The total Cu(II) adsorbed (qtotal, mg) in the column and total quantity of Cu(II) sent to the reactor (mtotal, mg) were estimated using Eqs. 2 and 3:

$$ {q}_{\mathrm{total}}=\frac{Q}{1000}{\int}_{t=0}^{t={t}_{\mathrm{total}}}\left({C}_0-{C}_t\right) dt $$
(2)
$$ {m}_{\mathrm{total}}=\frac{C_0Q{t}_{\mathrm{total}}}{1000} $$
(3)

where C0 and Ct refer to the initial concentration and concentration at any time t (mg/L), respectively. The removal efficiency is obtained using Eq. 4:

$$ \mathrm{Total}\_\mathrm{removal}\left(\%\right)=\frac{q_{\mathrm{total}}}{m_{\mathrm{total}}}\times 100 $$
(4)

The length of the mass transfer zone (Zm, cm) was calculated by Eq. 5 (Hernandezeudave et al. 2016):

$$ {Z}_{\mathrm{m}}=\frac{\left({t}_{\mathrm{e}}-{t}_{\mathrm{b}}\right)}{t_{\mathrm{e}}}\times h $$
(5)

where h is the bed height of the adsorbent (cm).

Results and Discussion

Characterization analysis

Textual parameters of bentonite, chitosan, CCB, CCK, CCS, kaolinite, and sand are provided in Table 1. Among the three adsorbents, the largest surface area was exhibited by CCB (8.29 m2/g) followed by CCK (5.13 m2/g) and CCS (0.4944 m2/g). Meanwhile, the average pore diameter was highest for CCS (737.70 nm) followed by CCB (85.14 nm) and CCK (72.40 nm). It was observed that reduced surface area and average pore diameter of bentonite, kaolinite, and sand occurred upon modification with chitosan, which was attributed to pore blocking.

Table 1 Textual parameters of bentonite, chitosan, kaolinite, sand, and chitosan-based composites
Full size table

The thermogravimetric (TG) curves of bentonite, chitosan, CCB, CCK, kaolinite, sand, and CCS are presented in Fig. 2. For bentonite, there are two peaks observed that are attributed to weight loss. The first peak in the temperature range 40 to 70 °C with an 8% weight loss was attributed to the evaporation of water molecules present in the interparticle and interlayer structure. The second weight loss (7%) in the range 600 to 700 °C was due to the degradation of hydroxide groups (Attar et al. 2018). For the TGA profile of kaolinite, minimal weight loss of < 15 up to 650 °C corresponds to the dehydroxylation where breakdown of the crystallographic structure occurs (Panda et al. 2010). For sand, a weight loss of < 1.0% up to 500 °C was related to evaporation of moisture. The TGA profile of chitosan presented two weight losses. The weight loss of 12% that occurred in the range 40 to 120 °C indicates the loss of physically adsorbed water. About 80% weight loss in the temperature range 240 to 570 °C is related to the degradation of hydrogen bonds in the polymer chains (Rajamani and Maliyekkal 2018). At 570 °C, chitosan was burnt out completely. Based on the thermograms of CCB, CCK, and CCS, about 3.50%, 3.79%, and 4.05% of chitosan have coated sand, bentonite, and kaolinite, respectively.

Fig. 2
figure 2

TGA curve of a CCS, b CCB, and c CCK

Full size image

The FT-IR spectra before and after adsorption of CCS, CCB, and CCK are illustrated in Fig. 3. Bentonite is characterized by the following peaks: 3664 cm−1 due to stretching of O–H, 1040 cm−1 due to stretching of Si–O and 917 cm−1 due to stretching of Al–O (Xing et al. 2012). The peaks of chitosan are the following: 3431 cm−1 for the stretching of O–H and N–H and 1642 cm−1 due to stretching of C=O and amide II band (Devi and Dutta 2017). Several peaks are identified that belong to kaolinite: 3674 cm−1 due to Al–OH stretching vibration of inner surface hydroxyl, 1047 cm−1 of Si–O–Si vibrations, 946 cm−1 for surface Al–Al–OH bonds and 480 cm−1 for the in-plane bending of Si–O–Si (Belver et al. 2002; Zuo et al. 2017). The following peaks are attributed to sand: 3412 cm−1 corresponds to O–H group stretching and 1082 cm−1asymmetric stretching of tetrahedral oxygen atoms in Si–O–Si (Primeau et al. 1997; Sharma et al. 2019). After adsorption, several peaks in CCB were observed to have shifted from 3431 to 3430 cm−1 and 1642 to 1638 cm−1. Meanwhile, changes in peaks in CCS were determined to be from 3474 to 3472 cm−1 and 1642 to 1640 cm−1 while the peak shifted from 1657 to 1655 cm−1 in CCK. Results suggest that amine (–NH2) and hydroxyl (O–H) groups of chitosan in CCS, CCB, and CCK are involved in the uptake of Cu(II) under fixed-bed conditions.

Fig. 3
figure 3

Infrared spectra of before and after Cu(II) adsorption of CCK, CCB, and CCS

Full size image

Effect of initial concentration and pH

Figure 4 illustrates the effect of varying initial concentration on the breakthrough performance of chitosan-based composites. As the initial concentration was increased from 200 (Fig. 4a–c) to 1000 (Fig. 4d–f) mg/L, the breakthrough curves were observed to become steeper. The longest duration for time of exhaustion was attained at 90, 130, and 300 min for CCS, CCK, and CCB, respectively (Fig. 4a). In Table 2, a shorter time of breakthrough and time of exhaustion were attained at higher initial concentration. Table 3 provides the column parameters derived from the breakthrough curve. Under pH 3.0, increasing the initial concentration from 200 (Fig. 4a) to 1000 (Fig. 4d) mg/L caused the time of breakthrough to decrease from 19 to 16 min for CCS, 82 to 35 min for CCB, and 70 to 30 min for CCK. A Cu(II) concentration of 1000 mg/L would imply a higher number of Cu(II) molecules in the system that would lead to an adsorbent bed that is easily saturated (Abdolali et al. 2017). High Zm values were attained at high Cu(II) concentration, which signifies that the active adsorption zone in the fixed-bed system would quickly exit the reactor. In addition, reduced Veff and lower removal efficiencies are attributed to low adsorbent dose present in the fixed-bed that is insufficient in removing high Cu(II) concentration at 1000 mg/L. The adsorption capacity at breakthrough (qb) and adsorption capacity at exhaustion (qe) were observed to increase at higher Cu(II) concentration. The main driving force for the adsorption in a fixed-bed system is the concentration gradient. The improved fixed-bed performance at 1000 mg/L is due to the difference in the Cu(II) concentration present in solution and on the adsorbent surface. This would result in higher values of the mass transfer coefficient and rapid transport of Cu(II) ions from solution onto the adsorbent surface (Abdolali et al. 2017; Aksu and Gönen 2004). Figure 4 illustrates the breakthrough curves under various pH of 3.0 (Fig. 4a, d), 4 (Fig. 4b, e), and 5.0 (Fig. 4c, f). As the pH was increased from 3.0 to 5.0, sharper breakthrough curves are formed where curves appear to have shifted from right to left. As the pH values become more acidic, reduced values of mass transfer zone were obtained that leads to longer breakthrough time and exhaustion time. In addition, a solution pH of 3.0 caused improvement in column performance such as an increase in treated volume of effluent, better removal efficiency, and improved adsorption capacity at breakthrough and exhaustion (Table 3). As shown in Fig. 4a, the maximum Cu(II) removal efficiency of 56.10% for CCS, 58.90% for CCB, and 68.60% for CCK at pH 3.0 and 200 mg/L was attained. The performance of the fixed-bed system can be arranged in the following order: pH 3.0 > pH 4.0 > pH 5.0. An acidic solution would contain excess Cl− ions where various Cu(II) complexes are formed that would utilize Cl− as a bridge. This implies that a binding site on the adsorbent surface would be able to retain more than one Cu(II) ion, which would result in better removal efficiency (Taboada et al. 2003). Similar results in previous studies were reported by Taboada et al. (2003) and Wan et al. (2010).

Fig. 4
figure 4

Effect of solution pH and initial concentration on the breakthrough curve of adsorption of Cu(II) using CCB, CCK, and CCS at Q = 5 min/L and bed height = 3 cm (a = pH 3.0, 200 ppm; b = pH 4.0, 200 ppm; c = pH 5.0, 200 ppm; d = pH 3.0, 1000 ppm; e = pH 4.0, 1000 ppm; f = pH 5.0, 1000 ppm)

Full size image
Table 2 Effect of initial concentration and solution pH on the time of breakthrough and time of exhaustion
Full size table
Table 3 Calculated values of the fixed-bed parameters under varying initial concentration and solution pH
Full size table

Among the adsorbents, CCS has attained the shortest time to reach breakthrough and exhaustion while CCB provided the longest time to reach breakthrough and exhaustion as well as the highest removal efficiency. The difference in the behavior of adsorbents in terms of column performance could be attributed to specific surface characteristics. The longer time to reach saturation by CCB followed by CCK and CCS could be attributed to CCB having the highest surface area. This indicates that there are more binding sites available for Cu(II) adsorption using CCB, which would result in better adsorption capacity at breakthrough and exhaustion as well as improved removal efficiency.

Thomas model

Fixed-bed kinetic models are utilized in the prediction of the breakthrough curve for the design of a commercial column reactor (Han et al. 2009). The Thomas model follows the assumption that the axial dispersion is insignificant in the fixed-bed system since the second-order reversible kinetics can adequately describe the rate constant (Thomas 1944). The linear form of the model is given as Eq. 6:

$$ \ln \left(\frac{C_0}{C_t}-1\right)=\frac{k_{\mathrm{Th}}{q}_{\mathrm{e}}m}{Q}-{k}_{\mathrm{th}}{C}_0t $$
(6)

where kTh is the Thomas rate constant (mL/mg min) and qe is the adsorption capacity obtained under equilibrium dynamic conditions (mg/g).

Table 4 presents the calculated Thomas model parameters. Based on high values of coefficient of determination (R2 ≥ 0.8503), the experimental data can be well-described using the Thomas model. This is confirmed by Fig. 5, wherein there is a good agreement between the experimental data and predicted breakthrough curves generated by the Thomas model. As the initial Cu(II) concentration was increased from 200 (Fig. 5a–c) to 1000 (Fig. 5d–f) mg/L, a corresponding increase in kTh was observed. At 1000 mg/L, a high concentration gradient exists due to a greater difference between the initial concentration of Cu(II) present on the CCB surface and solution. This would result in a better driving force resulting in improved mass transfer and faster uptake of Cu(II) ions by CCB, CCK and CCS (Aksu and Gönen 2004; Du et al. 2018). Under various pH of 3.0 (Fig. 5a, d), 4.0 (Fig. 5b, e), and 5.0 (Fig. 5c, f), values of kTh were observed to decrease as pH becomes more basic. An acidic solution would imply more than one Cu(II) ion would occupy an available adsorption site. Previous reports of Han et al. (2007) and Du et al. (2018) obtained similar results.

Table 4 Calculated parameters of the Thomas model for removal of Cu(II) using CCB, CCK, and CCS under various operating conditions
Full size table
Fig. 5
figure 5

Experimental data fitted with theoretical breakthrough curves derived from the Thomas model under varying solution pH and initial concentration (a = pH 3.0, 200 ppm; b = pH 4.0, 200 ppm; c = pH 5.0, 200 ppm; d = pH 3.0, 1000 ppm; e = pH 4.0, 1000 ppm; f = pH 5.0, 1000 ppm).

Full size image

In Table 5, the adsorption capacity at breakthrough of CCB, CCS, and CCK was compared with adsorbents utilized in previous researches. The highest adsorption capacity at breakthrough was attained by Kenaf fibers and carboxylated sugarcane bagasse, which is followed by CCB and polyethylenimine-modified wheat straw. Meanwhile, CCS displayed one of the lowest adsorption capacity at breakthrough.

Table 5 List of adsorbents utilized in the removal of Cu(II) using a fixed-bed system
Full size table

Regeneration study

Regeneration of spent adsorbents, which is considered as a sustainable method, was performed to determine reusability of the media, decreased treatment costs, and lower generation of solid wastes (Daneshvar et al. 2017). Figure 6 illustrates the desorption performance of Cu(II) using 0.1 M HCl and 0.1 M NaOH from spent CCB, CCK, and CCS. Results show that high desorption efficiency was obtained using HCl, which indicates that HCl is a better eluent when compared to NaOH. The highest desorption efficiency of 84.59% for pH 3.0, 82.38% for pH 4.0 and 82.13% for pH 5.0 was obtained for CCB using HCl as an eluent. A more acidic eluent implies a high amount of H+ ions are present in the solution that could replace the Cu(II) ions bound on the adsorbent surface (Li et al. 2009). Moreover, H+ ions may diminish the strength of the bond between Cu(II) and binding sites of the adsorbent that would lead to the release of Cu(II) to the solution (Vakili et al. 2019). The low desorption performance using NaOH can be attributed to the deprotonation of the adsorbent’s binding sites, which would make it more difficult for Cu(II) ions to detach from the adsorbent surface. The effective desorption performance of HCl implies that the regenerated adsorbents would still have the capacity to be reused in the removal of Cu(II) even after the first cycle of adsorption-desorption. This demonstrated the reusability of CCB in fixed-bed systems using 0.1 M HCl as the elution agent. Moreover, regeneration studies displayed the possible recovery of Cu(II) from wastewater that could be reused in several industries. Instead of disposal, past researches have utilized spent adsorbents in the production of cement mortars as insulators and aggregates (Petrella et al. 2016, 2018).

Fig. 6
figure 6

Desorption capacity of Cu(II) from CCB, CCK, and CCS using a HCl and b NaOH as eluents in a fixed-bed system under various solution pH

Full size image

Conclusion

In the present work, the application of CCB, CCK, and CCS in the adsorptive removal of Cu(II) from aqueous media using a fixed-bed system was investigated. Results demonstrated that lower initial Cu(II) concentration and pH 3.0 would cause lower values of Zm, longer time to reach breakthrough and exhaustion, higher effluent volume treated, and improved removal efficiency. The column data was fitted with the Thomas model where high coefficient of determination values were attained, which implies that the model is suitable in predicting the breakthrough curves. The preference for Cu(II) adsorption can be arranged in the following order: CCB > CCK > CCS. The higher affinity of CCB could be attributed to its high surface area available for Cu(II) adsorption. Regeneration studies showed that HCl is an effective elution agent for the reusability of CCB, CCK, and CCS. After one cycle of adsorption-desorption, regeneration efficiency of 84.59%, 82.51%, and 82.08% was attained for CCB, CCK, and CCS, respectively.