Study of a fixed-bed column in the adsorption of an azo dye from an aqueous medium using a chitosan

Abstract

A continuous adsorption study in a fixed-bed column was carried out using a chitosan–glutaraldehyde biosorbent for the removal of the textile dye Direct Blue 71 from an aqueous solution. The biosorbent was prepared from shrimp shells and characterized by scanning electron microscopy, X-ray diffraction, and nuclear magnetic resonance spectroscopy. The effects of chitosan–glutaraldehyde bed height (3–12 cm), inlet Direct Blue 71 concentration (15–50 mg l⁻¹), and feed flow rate (1–3 ml min⁻¹) on the column performance were analyzed. The highest bed capacity of 343.59 mg Direct Blue 71 per gram of chitosan–glutaraldehyde adsorbent was obtained using 1 ml min⁻¹ flow rate, 50 mg l⁻¹ inlet Direct Blue 71 concentration, and 3 cm bed height. The breakthrough curve was analyzed using the Adams–Bohart, Thomas, and bed depth service time mathematical models. The behaviors of the breakthrough curves were defined by the Thomas model at different conditions. The bed depth service time model showed good agreement with the experimental data, and the high values of correlation coefficients (R²≥ 0.9646) obtained indicate the validity of the bed depth service time model for the present column system.

Keywords

Adsorption breakthrough curves chitosan–glutaraldehyde Direct Blue 71 fixed-bed column modeling

Introduction

Water is a key factor for economic and social development and at the same time fulfills the basic function of maintaining the integrity of the natural environment. In addition to problems related to the amount of available water, problems associated with water quality are also presented. Pollution of water sources is one of the main problems faced by users of water resources and also threatens the preservation of natural ecosystems. Many industries, such as textile, paper, plastics, and food, generate a considerable amount of polluted waste water. Among the main pollutants are heavy metals and dyes. Industrial activities, such as textile production, emit various types of synthetic dyes in the water discharged to the environment, causing environmental problems worldwide (UN-Water, 2008). The composition of textile wastewater varies depending on the particular textile industry. However, most dyes are not biodegradable and tend to suppress photosynthetic activities in aquatic habitats by preventing the penetration of sunlight. Due to the complex structures and synthetic origins of dyes, textile effluents are very difficult to treat using conventional processes because these processes are costly and cannot effectively be used to treat the wide range of dyes present in wastewater (Ahmad and Hameed, 2010; Demarchi et al., 2013; Han et al., 2009). Several conventional technologies have been developed for removal of pollution from waste waters, such as precipitation, evaporation, filtration, ion exchange, membrane processes, and coagulation-flocculation. All of these methods have many limitations and involve complicated procedures that are economically infeasible. One of the most efficient methods for the removal of dye pollutants is adsorption, which has been more frequently used than other methods due to the simplicity in design, application, and regeneration of adsorbents. Besides, it does not produce sludge, as in chemical precipitation or coagulation-flocculation processes (Gupta and Suhas, 2009; Liu et al., 2012; Vieira et al., 2014). Activated carbon has a good capacity of removal of pollutants. However, its main disadvantages are the high price of treatment and difficult regeneration. Thus, potential low-cost adsorbents are in demand, such as wood (Sartape et al., 2013), lignite and peat (Ho and McKay, 1998), clay (Errais et al., 2011), agricultural wastes (Bulut et al., 2007), biomass (Chu and Chen, 2002), and chitin or chitosan (Guibal et al., 2003; Singh et al., 2009).

Chitosan is derived from a natural polysaccharide known as chitin, which can be obtained from crustaceans or fungi. Furthermore, chitosan is versatile because it can be manufactured into films, membranes, fibers, sponges, gels, beads, and nanoparticles or supported on inert materials. It is relatively cheap and exhibits a high dye adsorption capacity (Crini and Badot, 2008; Gupta and Suhas, 2009). Chitosan has been widely used as an adsorbent for dyes, transition metals, and organic species because the amino (–NH₂) and hydroxyl (–OH) groups on the chitosan chains can serve as coordination and reaction sites. However, this biopolymer is readily soluble in dilute acidic solutions below a pH of 5.5 ± 0.5, which tends to be a disadvantage when chitosan is used as an adsorbent in removing dyes in acidic effluents. One method to overcome this disadvantage for practical applications is to chemically modify the polymer by a cross-linking method. Chitosan can be cross-linked with glutaraldehyde, epichlorohydrin, ethylene glycol, diglycidyl ether, and sodium tripolyphosphate (Crini and Badot, 2008). After cross-linking, the material maintains its properties and original characteristics, particularly its adsorption capacity. The processes of dye adsorption with chitosan and its derivatives are especially interesting because in some cases dyes can be selectively adsorbed, concentrated, and recycled (Guibal et al., 2003). Ahmad and Hameed (2010) and Crini and Badot (2008) mention that there are numerous batch studies performed on the removal of dyes using pure chitosan or its derivatives, and these are usually conducted to evaluate the optimum adsorption conditions of the adsorbent materials. Batch experiments are usually done to measure the effectiveness of adsorption for removing specific adsorbates as well as to determine the maximum adsorption capacity. The data obtained from batch studies are, in most cases, limited to the laboratory scale and thus cannot be applied in industrial systems. Fixed-bed column study is necessary to obtain basic data for the design of continuous flow sorption. Hence, there is a need to perform column studies to provide data for direct applications in industrial systems (Ahmad and Hameed, 2010; Futalan et al., 2011; Han et al., 2008). Several research studies with batch method have been conducted for the adsorption of this dye, where are involved biosorbents such as sunflower stalks (Shi et al., 1999), cross-linked chitosan (Guibal et al., 2003), palm ash (Ahmad et al., 2010), and wheat shells (Bulut et al., 2007). However, little is known about adsorption test, where the chitosan-based gels are used as an adsorbent for Direct Blue 71 (DB71) dye removal, and even less in adsorption columns. Therefore, the aim of this study is to investigate the adsorption capacity of cross-linked chitosan–glutaraldehyde (QGlu) beads in a fixed-bed column with regards to the textile dye DB71. The effect of bed height, concentration, and flow rate on the column performance is analyzed. The breakthrough curves are analyzed using the Adams–Bohart, Thomas, and bed depth service time (BDST) models.

Materials and methods

Chemicals (materials)

Chitosan was prepared from shrimp shells (degree of deacetylation: 81.68% and average molecular weight: 178 kDa). The production of noncommercial chitosan was described in detail in a previous work (Sánchez-Duarte et al., 2012). Glutaraldehyde (25%, Mw = 100.12) was supplied by Sigma Aldrich (USA). The dye, DB71 (CI: 34140), was supplied by Sigma Aldrich and used without purification (see Figure 1). All other chemicals used were of HPLC or analytical grade. Distilled water was used to prepare all solutions.

Figure 1.

Direct Blue 71 (tris-azo dye).

Preparation of chitosan hydrogels

The preparation of cross-linked QGlu was described in detail in a previous work (Sánchez-Duarte et al., 2013). Briefly, 2.5 g of raw chitosan was dissolved in 2% acetic acid solution until complete dissolution and was allowed to set for 10–12 h. After 5 h, the chitosan solution was added from a pipette, drop by drop, into a 0.5 M NaOH solution to form hydrogels, i.e. chitosan beads. Finally, the wet chitosan beads were suspended in a 0.025 M glutaraldehyde solution (beads/solution ratio 1:10) at room temperature for 16 h. The QGlu beads were filtered and washed with distilled water and ethanol solution. The cross-linked QGlu beads were then stored in distilled water until further usage. All the QGlu beads were characterized in terms of diameter, moisture, and solubility in 5% (v/v/) acetic acid solution, whose results were reported in previous research (Sánchez-Duarte et al., 2013). The QGlu beads did not dissolve in acetic acid solution. The diameter (mm) and the moisture (%) were 2.81 ± 0.17 and 96.5 ± 0.04, respectively (means ± standard deviations, n = 50 for diameter, n = 3 for moisture).

Characterization techniques of chitosan hydrogels

Scanning electron microscopy (SEM)

The surface morphology of chitosan hydrogels was examined by SEM using an EVO LS 15 (Carl Zeiss SMT, Rudolf-Eber-Straße, Oberkochen, Germany). Prior to observation, the samples were lyophilized. For the magnification, the scanning electron microscope was operated at ×255 magnification and at 25 kV accelerated voltage in order to obtain high resolution micrographs.

X-ray diffraction

X-ray diffractograms of chitosan and QGlu beads (previously lyophilized samples) were measured on a Philips type powder diffractometer fitted with a Philips “PW1710” control unit, Vertical Philips “PW1820/00” goniometer, and FR590 Enraf-Nonius generator. The instrument was equipped with a graphite diffracted beam monochromator and a copper radiation source (λ(Kα1) = 1.5406 Å), operating at 40 kV and 30 mA. The X-ray powder diffraction pattern was collected by measuring the scintillation response to CuKα radiation versus the 2Θ value over a 2Θ range of 5–60, with a step size of 0.02° and a counting time of 2 s per step.

NMR spectroscopy

¹H nuclear magnetic resonance (¹H NMR) spectrum was recorded with a NMR spectrometer (Varian Inova 750, USA) at 750 MHz. Deuterated hydrochloride was used as the solvent for dissolving the chitosan sample.

Column setup

The experiments were conducted using a glass column with an internal diameter of 2.54 cm and a length of 37 cm. A fixed amount of glass beads was placed at the bottom of the column to serve as the support material for QGlu. Glass beads were placed on top of the adsorbent bed to prevent the bed from being pulled with the outflow. All of the experiments were conducted at room temperature, and the direction of flow was from bottom to top. The pH of the inlet DB71 solution was set to 4. The pH of the dye solution was adjusted with HCl (0.1–5 mol l⁻¹). Effluent samples were collected at the top of the column at different time intervals. The residual concentration of DB71 dye in the outlet was analyzed using a GENESYS 10 S UV–Vis Spectrophotometer (Thermo Scientific, Madison, Wisconsin, USA) at 587 nm.

Column experiments

The effects of the bed height of the adsorbent, initial influent concentration, and flow rate on the column performance were analyzed.

Effect of the bed height

The effect of varying the bed height (3, 6, and 12 cm) on the column parameters was studied. A flow rate of 1 ml min⁻¹ and an initial influent concentration of 15 mg l⁻¹ were kept constant.

Effect of initial influent concentration

Initial influent DB 71 concentrations of 15, 30, and 50 mg l⁻¹ were examined. A flow rate of 1 ml min⁻¹ and a bed height of 3 cm were kept constant.

Effect of flow rate

Flow rates of 1, 2, and 3 ml min⁻¹ were used to analyze the effect on the column performance. An initial influent concentration of 15 mg l⁻¹ and a bed height of 3 cm were kept constant.

Modeling of the breakthrough curves

The performance of a column can be evaluated based on the shape of the breakthrough curve obtained from the plot of C_i/C_o versus time t, where C_i and C_o are the effluent DB71 dye and inlet concentrations in milligram per liter, respectively. This is very important for determining the operation and the dynamic response of an adsorption column (Futalan et al., 2011; Han et al., 2009). The total weight of DB71 adsorbed by QGlu in the column, i.e. q_total (mg), for a given feed concentration and flow rate, is equal to the area under the plot of the adsorbed DB71 concentration, C_ad (C_o−C_t) (mg l⁻¹) (C_o and C_t are the initial and final concentration, respectively) versus t (min), and is expressed by equation (1)

q_{total} = \frac{Q}{1000} \int_{t = 0}^{t_{total}} C_{ad} d t

(1)

where Q is the volumetric flow rate (ml min⁻¹), t_total is the total flow time (min).

The experimental uptake capacity, q_e(exp) (mg g⁻¹), is calculated by equation (2), where X is the total dry weight of QGlu in the column (g)

q_{e (exp)} = \frac{q_{total}}{X}

(2)

The total amount of DB71 (W_total) added to the column (mg) is calculated by equation (3)

W_{total} = \frac{C_{o} {Qt}_{total}}{1000}

(3)

The total DB71 removal (% removal ) is calculated by equation (4)

% removal = \frac{q_{total}}{W_{total}} * 100

(4)

The breakthrough curves were analyzed with the mathematical models of Adams–Bohart, Thomas, and BDST.

Adams–Bohart model

The Adams–Bohart model assumes that the adsorption rate is proportional to both the residual capacity of the adsorbent and the concentration of the adsorbing species. This model is used for the description of the initial part of the breakthrough curve (Han et al., 2008)

ln (\frac{C_{i}}{C_{o}}) = k_{AB} C_{o} t - k_{AB} N_{o} \frac{Z}{F}

(5)

where k_AB is the kinetic constant (l mg⁻¹min⁻¹), F is the linear velocity calculated by dividing the flow rate by the column section area (cm min⁻¹), Z is the bed depth of the column (cm), and N_o is the maximum adsorption capacity (mg l⁻¹). The values of k_AB and N_o can be obtained from a plot of ln(C_i/C_o) versus t.

The Thomas model

This model is based on the assumption that the process follows Langmuir kinetics with no axial dispersion in the column adsorption, since the rate driving force obeys the second-order reversible reaction kinetics (Atar et al., 2011; Futalan et al., 2011). The linearized form of the Thomas model is the following

ln (\frac{C_{o}}{C_{i}} - 1) = \frac{k_{Th} q_{o} w}{Q} - k_{Th} C_{o} t

(6)

where k_Th is the Thomas model constant (ml min⁻¹mg⁻¹), q_o is the equilibrium adsorption amount of dye on QGlu (mg g⁻¹), w is the mass of the adsorbent (g), and Q is the flow rate (ml min⁻¹).

BDST

The BDST is a simple semiempirical model in the fixed-bed analysis that enables most rapid prediction of adsorbent performance. The BDST model is based on the assumption that the rate of adsorption is controlled by the surface reaction between adsorbate and the unused capacity of the adsorbent (Han et al., 2009; Singha and Sarkar, 2015). This model can predict the relationship of bed depth (Z) of the fixed-bed column with service time (t). A linear relationship between bed depth (Z) and service time is given by equation (7)

t = \frac{N_{o}'}{C_{o} F} Z - \frac{1}{k_{a} C_{o}} ln (\frac{C_{o}}{C_{i}} - 1)

(7)

where

N_{o}'

is the adsorption capacity from BDST model (mg l⁻¹), and k_a is the rate constant in BDST model (l mg⁻¹min⁻¹).

A simplified form of the BDST model is

t = aZ - b

(8)

where a (min cm⁻¹) and b (min) are the first and second terms of equation (7).

The slope constant for a different flow rate can be directly calculated by equation (9)

a' = a \frac{F}{F^{'}} = a \frac{Q}{Q^{'}}

(9)

where a and F are the old slope and inlet linear velocity; a′ and F ′ are the new slope and inlet linear velocity, respectively.

For other inlet concentrations, the desired equation is given by new slope, and a new intercept is obtained as follows

a' = a \frac{C_{o}}{C_{o}'}

(10)

b' = b \frac{C_{o}}{C_{o}'} \frac{\ln (\frac{C_{o}'}{C_{i}'} - 1)}{ln (\frac{C_{o}}{C_{i}} - 1)}

(11)

where b′ and b are the new and old intercepts (min), respectively;

C_{o}'

and C_o are the new and old inlet concentrations (mg l⁻¹), respectively.

C_{i}'

and C_i are the effluent concentrations (mg l⁻¹) at

C_{o}'

and C_o, respectively.

Error analysis

All the adsorption experiments were performed in duplicate and the results are presented as means of the replicates along with standard deviation (represented as error bars in the breakthrough curves). The linear regression coefficients R² were determined using Microsoft Excel (Microsoft Inc., WA, USA) to test the adequacy and accuracy of the linearized forms of the model equations of the rupture curves.

Results and discussion

SEM

Figure 2(a) to (c) shows the morphologies of dry beads of chitosan, QGlu, and QGlu after dye adsorption, respectively. Figure 2(a) represents pure chitosan showing a smooth surface, which is opaque and homogeneous with some wrinkles and without pores. The dry beads of QGlu in Figure 2(b) show an irregular and porous surface, which is heterogeneously shaped and cracked. These morphological changes can be attributed to the cross-linking of chitosan with glutaraldehyde. Similar results were obtained by Fwu-Long et al. (2003). Figure 2(c) shows an irregular surface with some wrinkles that is less porous, which suggests that the pores could have acted as active sites in the dye adsorption.

Figure 2.

SEM micrographs of dry chitosan bead (a), dry chitosan–glutaraldehyde bead (b), and dry chitosan–glutaraldehyde bead after dye adsorption (c). SEM: scanning electron microscopy.

X-ray diffraction and NMR spectroscopy

The aim here is to associate peaks with the crystalline phases present in the sample.Comparing the X-ray diffraction patterns (Figure 3) for chitosan and QGlu, chitosan exhibits a very sharp peak at 2θ = 11° and a sharp crystalline peak at 2θ = 21°, whereas in the QGlu pattern, a reduced intensity of the peak at 2θ = 21 and the disappearance of the peak at 2θ = 11° are observed, indicating the loss of crystallinity due to cross-linking. Similar results were observed by Cestari et al. (2004) and Singh et al. (2009).

Figure 3.

Comparison of the X-ray diffractograms of chitosan and chitosan–glutaraldehyde.

Figure 4 shows the ¹H NMR spectra (750 MHz) for chitosan and QGlu in deuterated hydrochloride solvent.

Figure 4.

1H NMR spectra of (a) chitosan–glutaraldehyde (CH₂–Glu: CH₂ group of the glutaraldehyde) and (b) chitosan (H3–H6: d-glucosamine and N-acetyl glucosamine groups, H₂: d-glucosamine fraction). NMR: nuclear magnetic resonance.

Figure 4(b) gives the ¹H NMR spectrum obtained for chitosan. The signal at 3.4–3.8 ppm shows the d-glucosamine and N-acetyl glucosamine groups (H3–H6), the signal for d-glucosamine fraction (H2) appears at 3.0 ppm, and a singlet at 1.8 ppm is evident, corresponding to the protons of acetyl group of N-acetyl glucosamine. Moreover, in Figure 4(a) of the spectrum of QGlu, a reduction in the intensities of the peaks at 3.0–3.8 ppm indicates that the glucosamine groups were involved in the cross-linking process with the glutaraldehyde. Several new peaks, not well defined, appeared at 1.0–1.7 ppm. Their appearance is a consequence of the complex structure obtained from the QGlu reaction. These peaks can be attributed to three different chemical environments of the CH₂ group. The different environments are due to the contribution of the molecule of glutaraldehyde in the new cross-linked polymer. The signals of protons of aldehyde groups are observed at 9.6 ppm. Similar results were observed by Monteiro and Airoldi (1999) and Kildeeva et al. (2009).

Column performance at various operating conditions

The most important parameters for a breakthrough curve study are the bed height, flow rate, and initial inlet concentration. The effects of these parameters on the shape of the breakthrough curve and column performance were investigated.

Effect of bed depth on breakthrough curves

The breakthrough curves at different bed depths (3, 6, and 12 cm) are shown in Figure 5. From Figure 5, using a constant influent concentration of 15 mg l⁻¹ and a flow rate of 1 ml min⁻¹, we can see that the shape and slope of each curve are different with the variation of the bed depth. The breakthrough time increases at higher bed depth of adsorbent. As the bed height is increased from 3 to 12 cm, a decrease in the slope of breakthrough curve is observed, which resulted in a rapid mass transfer zone. In addition, as seen in Figure 5, the breakthrough curve does not follow a characteristic “S” shape profile produced in ideal adsorption systems. An “S” shape profile is associated with adsorbates with small molecular diameters and simple structures, which are not true for the DB71 dye (Walker and Weatherley, 1997).

Figure 5.

Breakthrough curves of DB71 removal using QGlu beads packed in columns of different bed depths (Co = 15 mg l⁻¹, Q = 1 ml min⁻¹, pH 4, 25℃). DB71: Direct Blue 71; QGlu: chitosan–glutaraldehyde.

The adsorption data, including the DB71 uptake and removal percentage, are presented in Table 1. According to Table 1, when the bed height is decreased, the removal percentage is decreased from 46.36 to 26.65%, and the total weight of DB71 adsorbed by QGlu in the column (q_total) is also decreased from 132.26 to 37.53 mg. A higher bed height indicates a larger amount of DB71 adsorbed due to an increase in the surface area of adsorbent, providing more binding sites for the adsorption. Similarly, removal percentage decreases with decreasing bed depth, the reason could be decreasing amount of the adsorbent (Ahmad and Hameed, 2010; Futalan et al., 2011; Han et al., 2009).

Table 1.

Experimental data of the column parameters determined at various inlet concentrations of the DB71 solution at pH 4 and 25℃.

C_o (mg l⁻¹)	Q (ml min⁻¹)	Z (cm)	q_total (mg)	q_eq(exp) (mg g⁻¹)	W_total (mg)	% removal
15	1	6	89.03	125.89	288.93	30.81
15	1	12	132.26	93.50	285.29	46.36
15	1	3	37.53	106.15	140.84	26.65
15	2	3	36.18	102.31	214.18	16.89
15	3	3	22.47	63.56	142.73	15.74
30	1	3	74.19	209.81	370.08	20.05
50	1	3	121.49	343.59	777.19	15.63

DB71: Direct Blue 71.

Effect of flow rate on the breakthrough curves

To investigate the effect of varying the flow rate (1, 2, and 3 ml min⁻¹) on breakthrough curves, a bed height of 3 cm and an initial DB71 concentration of 15 mg l⁻¹ are kept constant. The results are shown in Figure 6 and Table 1, indicating that breakthrough generally occurred faster with higher flow rate. At the lowest flow rate of 1 ml min⁻¹, the highest % removal (26.65%) and experimental adsorption capacity (106.15 mg g⁻¹) were obtained. Breakthrough time was increased with a decrease in the flow rate. This is because at a low rate of inlet DB71, QGlu had more time to contact with dye, which resulted in higher removal of DB71 ions in column. The variation in the slope of the breakthrough curve and adsorption capacity may be explained on the basis of mass transfer fundamentals. The reason is that at higher flow rate, the mass transfer rate gets increased, namely the amount of dye adsorbed onto unit bed height gets increased with increasing flow rate leading to faster saturation. This effect was also observed by Ahmad and Hameed (2010), Futalan et al. (2011), Han et al. (2008, 2009), and Singha and Sarkar (2015).

Figure 6.

Breakthrough curves of DB71 removal by QGlu beads packed in a column for different flow rates (Co = 15 mg l⁻¹, Z = 3 cm, pH 4, 25℃). DB71: Direct Blue 71; QGlu: chitosan–glutaraldehyde.

Effect of initial inlet concentration on breakthrough curves

The effect of the initial inlet concentration on breakthrough curves was evaluated using three initial feed solutions, i.e. 15, 30, and 50 mg l⁻¹ of DB71, while a bed height of 3 cm and a flow rate of 1 ml min⁻¹ were used. The results are shown in Figure 7 and Table 1. The breakthrough times of these curves decrease with increasing initial inlet concentration. At lower inlet concentrations, breakthrough curves were dispersed and breakthrough occurred slowly. This effect was also observed by Han et al. (2009). However, the values of q_total and q_e increase with the increase of the inlet dye concentration, and the % removal decreases from 26.65 to 15.63% with the increase of the DB71 concentration from 15 to 50 mg l⁻¹. This can be explained by the fact that a low influent concentration causes a slow transport of the DB71 dye to the surface of the QGlu adsorbent, which implies a decreased diffusion coefficient and a decreased mass transfer driving force. Therefore, the diffusion process is concentration dependent (Futalan et al., 2011; Han et al., 2008; Karimi et al., 2012).

Figure 7.

Breakthrough curves of DB71 removal by QGlu beads packed in a column for different initial inlet concentrations (Q = 1 ml min⁻¹, Z = 3 cm, pH 4, 25℃). QGlu: chitosan–glutaraldehyde. DB71: Direct Blue 71.

Breakthrough curve models

To describe the fixed-bed column behavior, two models (Adams–Bohart and Thomas) were used to obtain a kinetic model in the column and to estimate the breakthrough curves.

Adams–Bohart model

The Adams–Bohart adsorption model was applied to the experimental data for the description of the initial part of the breakthrough curve. Plotting C_i/C_o against t gave the breakthrough curves predicted by the Adams–Bohart model (Figures 5 to 7) and the experimental points. Using linear regression analysis on all breakthrough curves, the respective values of k_AB and N_o were calculated and are presented in Table 2. The values of the kinetic constant k_AB and the maximum adsorption capacity N_o at all flow rates show no significant difference. The values of k_AB and N_o decrease as the bed depth increases. Additionally, the value of k_AB decreases and N_o increases with the increase of the initial DB71 concentration. The Adams–Bohart model is based on surface reaction theory and assumes that equilibrium is not instantaneous; hence, the adsorption rate is proportional to both the residual capacity of the adsorbent and adsorbate concentration. Besides, the model is applied in regions of low concentration and when the adsorption rate is limited by mass transfer (Han et al., 2009; Karimi et al., 2012). Figures 5 to 7 show the comparison of the experimental curve and the predicted curve according to the Adams–Bohart model. The experimental breakthrough curves are not close to those predicted by this model, and the values of R² are lower according to Table 2. Hence, the Adams–Bohart model cannot be used to predict the experimental data in the range of conditions used and with this dye.

Table 2.

Adams–Bohart parameters at different conditions using linear regression analysis.

C_o (mg l⁻¹)	Q (ml min⁻¹)	Z (cm)	k_AB (×10⁶) (l mg⁻¹min⁻¹)	N_o (×10⁻³)(mg l⁻¹)	R²
15	1	6	4.66	8.06	0.638
15	1	12	5.33	5.26	0.412
15	1	3	6.66	10.30	0.606
15	2	3	6.00	12.33	0.501
15	3	3	6.66	10.91	0.732
30	1	3	1.33	25.29	0.682
50	1	3	0.80	42.17	0.410

Concentrations of the DB71 solution at pH 4 and 25℃.

Thomas model

The linearized form of the Thomas model, i.e. ln (C_o/C_i − 1) versus t, was used to estimate the two unknown parameters, k_Th and q_o, of the Thomas equation. The estimated values of the Thomas parameters and the linear regression coefficients (R²) are presented in Table 3. A comparison between the experimental data and the predicted breakthrough curves obtained at different bed depths, flow rates, and initial concentrations is shown in Figures 5 to 7. From Table 3, in both cases as the influent DB71 concentration and the bed depth increase, the values of q_o increase, but the values of k_Th decrease, while that the values of q_o and k_Th increased with increasing the flow rate. Although, the fact that the Thomas model demonstrates some good adjustments for the adsorption conditions does not provide a good correlation in the prediction of the rupture curve. This is observed in differences found between the values of adsorption capacity obtained experimentally and calculated with the model. Therefore, although the Thomas model is one of the most used to describe the biosorption process behavior in fixed-bed columns, its main limitation is that its derivation is based on a second-order kinetics and considers that biosorption is not limited by the chemical reaction, but is controlled by the mass transfer at the interface. This discrepancy can lead to errors when this method is used to model biosorption processes under certain conditions (Aksu and Gönen, 2004).

Table 3.

Thomas parameters at different conditions using linear regression analysis.

C_o (mg l⁻¹)	Q (ml min⁻¹)	Z (cm)	k_Th (×10³) (ml mg⁻¹min⁻¹)	q_o (mg g⁻¹)	R²
15	1	6	13.33	79.08	0.832
15	1	11	6.66	126.67	0.503
15	1	3	20.00	76.69	0.818
15	2	3	20.00	102.06	0.793
15	3	3	33.33	208.91	0.838
30	1	3	6.66	146.39	0.897
50	1	3	4.00	290.01	0.824

Concentrations of the DB71 solution at pH 4 and 25℃.

BDST

The adsorption capacity $N_{o}'$ (mg l⁻¹) and rate constant k_a (l mg⁻¹min⁻¹) from BDST model were obtained from plotting the service time t (min) versus bed depth Z (cm) at C_i/C_o corresponding to 0.15, 0.51, and 0.65, as seen in Figure 8.

Figure 8.

Bed depth versus service time plot at different values of C_i/C_o in fixed-bed column for the adsorption of Direct Blue 71 dye.

The service time experimental (t) and predicted service time (t_p) for the varying bed depth (12, 6, and 3 cm) with flow rate of 1 ml min⁻¹ and an influent concentration of 15 mg l⁻¹ were calculated. These are shown in Table 4. The values of the service time predicted by the BDST model are in good agreement with the experimental values of service time, hence this suggested that the BDST model could be acceptable for this system; besides the high values of R² obtained from data (0.9844, 0.9944, and 0.9646) indicate the validity of the BDST model for the present column system.

Table 4.

Comparison between experimental service time and predicted service time.

C_i/C_o	N′_o (mg l⁻¹)	ka (×10⁶) (l mg⁻¹min⁻¹)	z (cm)	Service time experimental, t (min)	Service time predicted, t_p (min)
0.15	141.79	677.44	12	420	404.10
0.51	1167.21	7.40	6	2880	2725.74
0.65	5582.21	−10.37	3	2160	1337.10

Flow rate of 1 ml min⁻¹; inlet concentration of 15 mg l⁻¹, pH 4, 25℃.

The BDST equation obtained at a flow rate of 1 ml min⁻¹ an inlet concentration of 15 mg l⁻¹ and C_i/C_o of 0.65 was used to predict the adsorbent performance at another concentration (50 mg l⁻¹). Based on equations (10) and (11), the values predicted for a′ and b′ were 565.71 cm min⁻¹ and −1993.47 min, respectively; the service time predicted was 503.65 min. Comparing the experimental service time (480 min) with the predicted service time (503.65 min), it can be seen that a good prediction was found for the case of changing the inlet concentration. These results indicate that the BDST model can enable prediction of adsorbent performance at other operating conditions. Similar results were observed by Han et al. (2009) and Singha and Sarkar (2015).

Conclusions

In the present study, QGlu beads were synthesized and then used in column experiments. The beads were characterized by SEM, X-ray diffraction, and ¹H NMR, which confirmed that chitosan reacted with glutaraldehyde to form hydrogels.

On the basis of the experimental results of the column experiments, the following conclusions can be drawn. First, QGlu beads can be used as an adsorbent to removal DB71 from solution. Furthermore, the adsorption was dependent on the bed depth, the influent DB71 concentration, and the flow rate. When the bed height is decreased, the removal percentage is decreased from 46.36 to 26.65%, and the total weight of DB71 adsorbed by QGlu in the column is also decreased from 132.26 to 37.53 mg. On the other hand, the maximum capacity of column was found to be about 343.59 mg DB71 per gram of QGlu adsorbent for flow rate of 1 ml min⁻¹, initial concentration of 50 mg l−1, and 3 cm bed height. Lastly, the behavior of the breakthrough curves can be defined by the Thomas model; this may mean that the internal and external diffusion is not the controlling step in the adsorption process of the column. The BDST model showed good agreement with the experimental data and the high values of correlation coefficients (R²≥ 0.9646) obtained indicate the validity of the BDST model for the present column system.

Footnotes

Acknowledgements

The corresponding author is grateful to CONACYT (333937).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was financed under Project no. 248160 from CONACYT-PN2014 and by project PROFAPI no. 2015-00381 from Instituto Tecnológico de Sonora.

References

Ahmad

Hameed

(2010) Fixed-bed adsorption of reactive azo dye onto granular activated carbon prepared from waste. Journal of Hazardous Materials 175(1–3): 298–303.

Aksu

Gönen

(2004) Biosorption of phenol by immobilized activated sludge in a continuous packed bed: Prediction of breakthrough curves. Process Biochemistry 39(5): 599–613.

Atar

Olgun

Wang

et al. (2011) Adsorption of anionic dyes on boron industry waste in single and binary solution using batch and fixed-bed systems. Journal of Chemical Engineering Data 56(3): 508–516.

Bulut

Gözübenli

Aydin

(2007) Equilibrium and kinetics studies for adsorption of direct blue 71 from aqueous solution by wheat shells. Journal of Hazardous Materials 144(1–2): 300–306.

Cestari

Vieira

E F S

Dos Santos

A G P

et al. (2004) Adsorption of anionic dyes on chitosan beads. 1. The influence of the chemical structures of dyes and temperature on the adsorption kinetics. Journal of Colloid and Interface Science 280(2): 380–386.

Chu

Chen

(2002) Reuse of activated sludge biomass: II. The rate processes for the adsorption of basic dyes on biomass. Process Biochemistry 37(10): 1129–1134.

Crini

Badot

(2008) Application of chitosan, a natural aminopolysaccharide, for dye removal from aqueous solutions by adsorption processes using batch studies: A review of recent literature. Progress in Polymer Science 33(4): 399–447.

Demarchi

Campos

Rodrigues

(2013) Adsorption of textile dye Reactive Red 120 by the chitosan-Fe (III)-crosslinked: Batch and fixed-bed studies. Journal of Environmental Chemical Engineering 1(4): 1350–1358.

Errais

Duplay

Darragi

et al. (2011) Efficient anionic dye adsorption on natural untreated clay: Kinetic study and thermodynamic parameters. Desalination 275(1–3): 74–81.

10.

Futalan

Kan

Dalida

et al. (2011) Fixed-bed column studies on the removal of copper using chitosan immobilized on bentonite. Carbohydrate Polymers 83(2): 697–704.

11.

Fwu-Long

Hsing-Wen

Shin-Shing

et al. (2003) Synthesis and characterization of biodegradable TPP/genipin co-crosslinked chitosan gel beads. Polymer 44(21): 6521–6530.

12.

Guibal

McCarrick

Tobin

(2003) Comparison of the sorption of anionic dyes on activated carbon and chitosan derivatives from dilute solutions. Separation Science Technology 38(12–13): 3049–3073.

13.

Gupta

Suhas (2009) Application of low-cost adsorbents for dye removal – A review. Journal of Environmental Management 90(8): 2313–2342.

14.

Han

Ding

et al. (2008) Use of rice husk for the adsorption of Congo red from aqueous solution in column mode. Bioresource Technology 99(8): 2938–2946.

15.

Han

Wang

Zhao

et al. (2009) Adsorption of methylene blue by phoenix tree leaf powder in a fixed-bed column: Experiments and prediction of breakthrough curves. Journal Desalination 245(1–3): 284–297.

16.

McKay

(1998) Sorption of dye from aqueous solution by peat. Chemical Engineering Journal 70: 115–124.

17.

Karimi

Shojaei

Nematollahzadeh

et al. (2012) Column study of Cr(VI) adsorption onto modified silica-polyacrylamide microspheres composite. Chemical Engineering Journal 210: 280–288.

18.

Kildeeva

Perminov

Vladimirov

et al. (2009) About mechanism of chitosan cross-linking with glutaraldehyde. Russian Journal of Bioorganic Chemistry 35(3): 360–369.

19.

Liu

Kang

Huang

et al. (2012) Cu⁺² removal from aqueous solution by modified chitosan hydrogels. Journal of Chemical Technology and Biotechnology 87(7): 1010–1016.

20.

Monteiro

O A C

Jr Airoldi

(1999) Some studies of crosslinking chitosan-glutaraldehyde interaction in a homogeneous system. International Journal of Biological Macromolecules 26(2–3): 119–128.

21.

Sánchez-Duarte RG, López-Cervantes J, Sánchez-Machado DI, et al. (2013) Chitosan-based adsorbents gels for the removal of tris-azo dye: Isotherms and kinetics studies. Environmental Engineering and Management. 15(11): 2469.

22.

Sánchez-Duarte

Sánchez-Machado

López-Cervantes

et al. (2012) Adsorption of allura red dye by cross-linked chitosan from shrimp waste. Water Science and Technology 65(4): 618–623.

23.

Sartape AS, Mandhare AM, Jadhav VV, et al. (2013) Removal of malachite green dye from aqueous solution with adsorption technique using Limonia acidissima (wood apple) shell as low cost adsorbent. Arabian Journal Chemistry. 10(2): S3229–S3238.

24.

Shi

Sun

(1999) Chemically modified sunflower stalks as adsorbents for color removal from textile wastewater. Journal of Applied Polymer Science 71(11): 1841–1850.

25.

Singh

Sharma

Tripathi

et al. (2009) Poly(methylmethacrylate) grafted chitosan: An efficient adsorbent for anionic azo dyes. Journal of Hazardous Materials 161(2–3): 955–966.

26.

Singha

Sarkar

(2015) Analysis of the dynamic column using semi-empirical models: Case studies with removal of hexavalent chromium from effluent wastewater. Korean Journal of Chemical Engineering 32(1): 20–29.

27.

UN-Water (2008) Status report on IWRM and water efficiency plans for CSD16. Available at: http://www.unwater.org/downloads/UNW_Status_Report_IWRM.pdf (accessed June 2016).

28.

Vieira

M L G

Esquerdo

Nobre

et al. (2014) Glass beads coated with chitosan for the food azo dyes adsorption in a fixed bed column. Journal of Industrial and Engineering Chemistry 20(5): 3387–3393.

29.

Walker

Weatherley

(1997) Adsorption of acid dyes on to granular activated carbon in fixed beds. Water Research 31(8): 2093–2101.

Study of a fixed-bed column in the adsorption of an azo dye from an aqueous medium using a chitosan–glutaraldehyde biosorbent

Abstract

Keywords

Introduction

Materials and methods

Chemicals (materials)

Preparation of chitosan hydrogels

Characterization techniques of chitosan hydrogels

Scanning electron microscopy (SEM)

X-ray diffraction

NMR spectroscopy

Column setup

Column experiments

Effect of the bed height

Effect of initial influent concentration

Effect of flow rate

Modeling of the breakthrough curves

Adams–Bohart model

The Thomas model

BDST

Error analysis

Results and discussion

SEM

X-ray diffraction and NMR spectroscopy

Column performance at various operating conditions

Effect of bed depth on breakthrough curves

Effect of flow rate on the breakthrough curves

Effect of initial inlet concentration on breakthrough curves

Breakthrough curve models

Adams–Bohart model

Thomas model

BDST

Conclusions

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

References