Introduction

Water pollution by different types of pollutants is a topical issue (Ben Aissa 2012). Especially in the textile industry, the wastewaters are one of the most important sources of pollution of surface water and groundwater (Ben Hamissa et al. 2013).

Indeed, the world production of dyes is estimated at 800,000 tons/year (Palmieri et al. 2005). Some of these dyes, approximately 140,000 tons/year, are released during the process of dyeing textile fibers (Ben Mansour et al. 2009). The textile sector is one of six industries producing half of pollution flows (Ben Mansour et al. 2011). The effluents from this area are very colorful and difficult to treat (Ben Mansour et al. 2011). The presence of this dye wastewater is increasingly considered a major nuisance (Ben Mansour et al. 2011). Among the dyes used in most textile industries are metallic dyes (Lajili 2010). Textile industries, especially the phase of coloration and finishing, mainly use chemicals harmful to health and generate pollution of surface water and groundwater (Lajili 2010).

When water is polluted by toxic and persistent substances, adequate treatment is necessary for the protection of the environment. Several procedures used for dye removal from industrial effluents are essentially coagulation (Rodrigues et al. 2013), ultrafiltration (Căilean et al. 2009), reverse osmosis (Vijayageetha et al. 2014), the adsorption on activated carbon (Ncibi et al. 2008a), and biological treatment (Abidin and Rahmat 2010). Indeed, the physicochemical treatments require considerable amounts of oxidants (expensive) (Ncibi et al. 2006) and sometimes lead to the formation of undesirable intermediate products (Zawlotzki Guivarch 2004). Moreover, certain products are resistant to this type of treatment which can be toxic (Zawlotzki Guivarch 2004). Although biological treatments are massively used, they remain powerless to eliminate certain toxic and persistent compounds such as dyes, hence the need to search for better alternatives (Zoughuir et al. 1998).

Other alternatives to eliminate discharges of textile dyes are today the subject of study, especially biosorption which is a very interesting method using natural materials that are often little or poor recovered (Ben Hamissa et al. 2013; Ncibi et al. 2007). Among these natural materials, we can mention rice husk (Vadivelan and Kumar 2005), beech sawdust (Batzias and Sidiras 2004), agro-industry wastes (Garg et al. 2004), Posidonia oceanica fibers (Ncibi et al. 2006), and Agave americana (Ben Hamissa et al. 2010).

Luffa cylindrica belongs to the Cucurbitaceae family (Demir et al. 2008). It is generally grown in China, Japan, Asia, Africa, and Central and South America (Demir et al. 2008). The fibers are composed of 63 % cellulose, 18 % hemicellulose, and 11.5 % lignin and can be used by industry for many purposes, such as packaging, insulating, or filling materials (Mazali and Alves 2005). L. cylindrica fiber has a fibrous vascular system that can enable removal of water pollutants (Adie et al. 2013). Moreover, many environmentally conscious consumers appreciate that L. cylindrica fibers are biodegradable and natural and are renewable resources (Demir et al. 2008).

The main objective of the present work is to evaluate the biosorption capability of a low-cost and renewable biomass (Demir et al. 2008), L. cylindrica fibers, for the removal of alpacide blue as a model of compound for metallic dyes. In fact, this dye is widely used in the textile industry in Tunisia. The effects of pH, contact time, initial dye concentration, temperature, and biomass amount on the biosorption capacity were investigated. Moreover, mathematical modeling of the kinetics and adsorption isotherms was also performed.

Materials and methods

Biosorbent and sorbate preparation

The adsorbent used in this study is the L. cylindrica fibers. This biomass was purchased at the local market Sousse. First, the fibers were cut finely, washed several times with distilled water to remove all impurities, and then dried in an oven at 70 °C for 48 h at a constant weight.

Alpacide blue used in this study is a metal dye, obtained from a textile industry (Chimitex, Tunisia). The metal dye chemical structure is unknown. Stock solutions were prepared by dissolving accurately weighed amounts of dye (100 mg) in 1 L of distilled water. All working solutions were prepared by diluting the stock solution with distilled water to achieve the needed concentrations.

Sorption tests and analytical methods

The biosorption experiments were carried out in batch reactor by adding 0.6 g dried L. cylindrica fibers in 75 mL of dye solution (i.e., a solid/liquid ratio of 8 g/L) with the desired concentration, pH, and temperature. All the tests were investigated at 20 °C, except the tests carried out to investigate the influence of temperature on biosorption. Temperature was controlled by a thermoregulated water bath, model Memmert BME1420 operating at 105 oscillations/min. The initial pH was adjusted by adding dilute solutions of HCl or NaOH. All experiments were conducted in triplicate, and the negative controls (without L. cylindrica) were simultaneously carried out to ensure that the biosorption capacity was due solely by the biomass and did not involve the walls of the container. After biosorption of alpacide blue, the residual concentrations were determined by spectrophotometry double faisceaux analysis (FT-IR) (Camspec M550) at λ max of 660 nm. Fourier transform infrared spectrometry (PerkinElmer Spectrum two) was used to analyze the functional groups in the adsorbent. The transmission spectrum was acquired with 4 cm−1 resolution and the spectrum was corrected for a KBr background. For BET specific surface area estimation, a porosimeter accelerated surface area and porosimetry system (ASAP 2010) was used.

Kinetics studies

For kinetic studies, 3 mL of the alpacide blue solution was taken periodically and then analyzed to assess the dye concentration. To evaluate the alpacide blue removal, the biosorption capacity at equilibrium time (Q e) or biosorption removal efficiency will be determined respectively according to the following equations:

$$ Q=\frac{\left({C}_0-{C}_i\right)\times V}{M}\left( mg/g\right) $$
(1)
$$ \%\;\mathrm{dye}\;\mathrm{removal}=\frac{\left({C}_0-{C}_i\right)\times 100}{C_0} $$
(2)

where C 0 is the initial alpacide blue concentration (mg/L), C i is the residual alpacide blue concentration at any time (mg/L), V is the volume of solution (L), and M is the mass of the sorbent (g). At equilibrium, C i is equal to C e and Q is equal to Q e.

The model of pseudo-first-order and the pseudo-second-order equation (Lagergren 1898; Ho 2004; Ho and McKay 1998) was used to fit the experimental data of the dye sorption before reaching equilibrium. The linearized form of the model of pseudo-first order is determined by expression 3:

$$ \log \left({Q}_{\mathrm{e}}-Q\right)= \log \left({Q}_{\mathrm{e}}\right)-\frac{k_1\cdot t}{2.303} $$
(3)

where t is the time (min) and k 1 is the equilibrium rate constant of pseudo-first-order sorption (min−1).

Unlike the pseudo-first-order model, the pseudo-second-order model is applicable to a larger interval of time (McKay and Ho 1999; Ncibi et al. 2006). The pseudo-second order is based on the sorption capacity of the solid phase and is described by Eq. 4:

$$ \frac{t}{Q}=\frac{1}{k_2\cdot {Q}_{\mathrm{e}}^2}+\frac{t}{Q_{\mathrm{e}}} $$
(4)

where k 2 is the equilibrium rate constant of pseudo-second-order sorption (g/mg.min).

Evaluation of thermodynamic parameters

Thermodynamic parameters such as free energy change (ΔG°), enthalpy change (ΔH°), and entropy change (ΔS°) for the sorption of alpacide blue on L. cylindrica were calculated according to the following equations (Ucun et al. 2008):

$$ \varDelta G{}^{\circ}=-RT\; ln\;{K}_C $$
(5)
$$ \varDelta G{}^{\circ}=\varDelta H{}^{\circ}-T\varDelta S{}^{\circ} $$
(6)

where R is the gas constant (8.314 J/mol/K), T is the absolute temperature (K), and K c is the apparent equilibrium constant K d, defined according Eq. 7 (Yurtsever and Sengil 2009):

$$ {K}_{\mathrm{C}}=\frac{C_{\mathrm{ad},\mathrm{eq}}}{C_{\mathrm{r},\mathrm{eq}}} $$
(7)

where C ad,eq and C r,eq being the concentration of adsorbed dye at equilibrium and remaining concentration dye at equilibrium, respectively (mg/L).

Evaluation of equilibrium isotherms

The Langmuir, Freundlich, and Temkin adsorption isotherms were used to describe the sorption of a dye solution. Linear form of the Langmuir equation can be expressed by expression  8 (Langmuir 1916):

$$ \frac{C_{\mathrm{e}}}{Q_{\mathrm{e}}}=\frac{1}{Q_{\mathrm{m}}\cdot {K}_{\mathrm{L}}}+\frac{C_{\mathrm{e}}}{Q_{\mathrm{m}}} $$
(8)

where Q m is the maximum sorption capacity (mg/g) and K L the Langmuir equilibrium constant (L/mg).

After, other fittings were done by the linearized form of the isotherm of Freundlich (Freundlich 1906):

$$ \ln\;{Q}_{\mathrm{e}}= \ln {K}_{\mathrm{F}}+\left(\frac{1}{n}\right) \ln\;{C}_{\mathrm{e}} $$
(9)

where K F [(mg/g) × (L/mg)1/n] is Freundlich constant indicating the adsorption capacity and n (dimensionless) is Freundlich constant indicating the adsorption intensity.

The Temkin equation, based on the heterogeneous surface of the adsorbent, has the form (Temkin 1941)

$$ {Q}_{\mathrm{e}}=B\cdot \ln A+B\cdot \ln {C}_{\mathrm{e}} $$
(10)

where A (L/g) and B (J/mol) are Temkin constants, and B = RT/b. b is variation of adsorption energy. R is universal gas constant (8.314 J/mol/K) and T is temperature at 298 K (Dada et al. 2012).

Results and discussion

Characterization of L. cylindrica

Determining the fiber humidity rate allows to calculate the dry matter rate. This gives an idea on the fiber yield of plants. Humidity rate and dry matter rate of L. cylindrica fibers are 9.6 and 90.4 %, respectively. In general, these results are comparable to those of some lignocellulosic materials such as the A. americana (L.) fibers having a dry matter rate of order of 84 % (Ben Hamissa 2005).

To have an idea on the rate of minerals of L. cylindrica, we determined the rate of ash (1 %). The rate of ash of L. cylindrica fibers is comparable to that reported in the literature (0.4 %) (Satyanarayana et al. 2007). These results are comparable with those on the Posidonia with a rate of 1.8 % (Satyanarayana et al. 2007).

For determining the size of the surface of the studied fibers, we measured their specific surface. This measure is interesting when it is a surface phenomenon such as the biosorption. Specific surface area of the L. cylindrica is 0.49 m2/g. This result is low compared to those of commercially activated carbons whose specific surface varies from 500 to 1500 m2/g (Aznar 2011).

However, this value is comparable to certain lignocellulosic biomasses studied for biosorption dye textiles such as sugar cane bagasse dyes with a surface of 0.67 m2/g (Aznar 2011).

FT-IR characterization of L. cylindrica fibers

Figure 1 shows the FT-IR spectrum of L. cylindrica fibers. IR spectra of L. cylindrica present absorption peaks which demonstrate the complex nature of the L. cylindrica. Several intense bands can be attributed to functional groups present in cellulose and lignin which are the major chemical composition of L. cylindrica.

Fig. 1
figure 1

Fourier transform infrared (FT-IR) absorption spectrum of Luffa cylindrica fibers

Full size image

The spectrum obtained is comparable to other spectra of cellulose and lignin described in the literature (El Hage 2010). The absorptions at 3435 cm−1 are attributed to –OH and –NH groups (Ncibi et al. 2006). Two peaks present at 2922 and 2850 cm−1 correspond to symmetric –CH2 valence vibration and –CH stretching vibration, respectively, corresponding to the aliphatic fragments of the cellulose (Schwanninger et al. 2004). While bands at 2378 and 2344 cm−1 are indicative of amine and amide groups (Williams and Fleming 1995). Peaks at 1705 and 1639 cm−1 are an indication of the presence of unconjugated ester groups (Mahmoudi et al. 2015). In addition, the peak at 1444 cm−1 may be associated with phenolic –OH and C–O stretching. The peaks present at 1116 and 896 cm−1 originate from the elongation of the C–O–C (Liu et al. 2006).

Effect of contact time and initial alpacide blue concentration

The influence of contact time on the biosorption of alpacide blue dye using L. cylindrica fibers is presented in Fig. 2. The amounts of dye adsorbed at equilibrium under operating conditions (pH 2, T 20 °C, and particle size 250–500 μm) were 9.258 mg/g for the alpacide blue.

Fig. 2
figure 2

Influence contact time and of initial alpacide blue concentration on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (solid/liquid ratio = 8 g/L, pH = 2, T = 20 °C)

Full size image

Dye solutions with initial concentrations of 20, 30, 40, 60, and 80 mg/L were studied. The sorbed amount of dye progressively increased with their respective initial concentrations, either before or after the achievement of equilibrium (Fig. 2). Furthermore, raising the dye concentration from 20 to 80 mg/L allows the fibers to increase their biosorption capacities from 2.20 to 9.63 mg/g, respectively, at pH 2 and 20 °C. In every case, it was observed that dye uptake was rapid for the first 60 min and thereafter proceeded at a slower rate to finally attain saturation after 150 min. The higher biosorption rate during the initial period may be due to the high number of sites available for the biosorption, and no less than 80 % of total dye was removed. The higher biosorption rate during the initial period may be due to the high number of sites available for the sorption, and no less than 80 % of total dye was removed. After this period, the number of sites available for biosorption decreased so that the dye molecules required longer time periods in order to reach the least accessible sites (Ben Hamissa et al. 2010).

Effect of pH

Textile dyes are complex organic compounds which possess various aromatic rings and functional groups. As shown by the FT-IR spectrum (Fig. 1), the L. cylindrica fibers contain different functional groups such as carboxyl, hydroxyl, amine, and other groups. Firstly, the point of zero charge of L. cylindrica was determined (Fig. 3). The point of zero charge is a concept relating to the phenomenon of adsorption, expressing the pH at which the electrical charge density on a biomass surface is zero. The point of zero charge of L. cylindrica was found to be 7.85. This result demonstrates that the L. cylindrica surface has a positive charge at pH values less than the point of zero charge and thus should be able to absorb anions. The initial pH of the solution is an important parameter that must be taken into account in any biosorption study. The effect of this factor on the change of the biosorption capacity was analyzed on a range of pH from 2 to 10. The results presented in Fig. 4, show that the equilibrium biosorption capacity is maximum at pH 2 (5.80 mg/g) and decreased up to 6, and then remained nearly constant (3.65 mg/g) over the initial pH ranges of 6–10. At higher pH, the surface charge may get negatively charged, thus making (OH−) ions compete effectively with dye anions causing a decrease in the amount of dye adsorbed (Ben Hamissa et al. 2008; Ncibi et al. 2008b). At lower pH, the fiber biomass, mainly lignin and cellulose chains, may get positively charged, which enhances the negatively charged dye anions through electrostatic forces of attraction (Tunc et al. 2009; Wong et al. 2009; Ncibi et al. 2006; Mahmoudi et al. 2015). This may be attributed to hydrogen bonding between the anionic dye and the functional groups (mainly hydroxy and carboxy) on the surface of the biosorbent (Ncibi et al. 2006). In addition, the hydrogen ion can act as a bridging ligand between the L. cylindrica fibers and the alpacide blue (Ben Hamissa et al. 2010; Ncibi et al. 2006). However, the interaction of the alpacide blue with all functional groups is possible via many sorption mechanisms including ion exchange, van der Waals interaction, hydrogen bond (between cellulose-OH and the alpacide blue), and microprecipitation (Ben Douissa et al. 2013).

Fig. 3
figure 3

Influence point of zero charge of Luffa cylindrica fibers

Full size image
Fig. 4
figure 4

Influence of initial pH on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (C = 20 and 50 mg/L, solid/liquid ratio = 8 g/L, temperature = 20 °C)

Full size image

Effect of temperature

The dye removal ability of the L. cylindrica was also investigated at different temperatures (293, 308, and 323 K). The study of the influence of temperature on the biosorption of the alpacide blue by L. cylindrica fibers, for initial concentrations of 20, 50, and 70 mg/L and a pH = 2, is shown in Fig. 5. It was found that alpacide blue biosorption decreases from 2.46 to 2.05 mg/g, 5.80 to 5.26 mg/g, and 7.8 to 7.38 mg/g for initial concentrations of 20, 50, and 70 mg/L, respectively, with increasing temperature from 293 to 323 K. This phenomenon may be due to weakening of the bonds between the dye molecules and active sites of adsorbents for high temperature (Ben Hamissa et al. 2013; Fakhry 2013). This result suggests that the process can be considered as an exothermic process (Atmani et al. 2009). Similar results were obtained in other studies for the biosorption of methylene blue onto sawdust (Ben Hamissa et al. 2013).

Fig. 5
figure 5

Influence of temperature on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (solid/liquid ratio = 8 g/L, pH = 2)

Full size image

Thermodynamic analysis

To estimate the effect of temperature on the alpacide blue biosorption by L. cylindrica fibers at different dye concentrations (20, 50, and 70 mg/L), the thermodynamic was studied at 293, 308, and 323 K, and the thermodynamic parameters such as Gibbs free energy ΔG° (kJ/mol), enthalpy ΔH° (kJ/mol), and entropy ΔS° (J/mol/K) were estimated as described in “Effect of pH” section and listed in Table 1. The negative value of ΔG° indicates the spontaneous nature of biosorption of alpacide blue by L. cylindrica for all studied concentrations and temperatures (Ben Hamissa et al. 2013; Bouhamed et al. 2012). The negative value of ΔH° confirms the exothermic process, while the negative value of ΔS° suggested that this biosorption would lead to decreasing randomness (Atmani et al. 2009; Ncibi et al. 2007).

Table 1 Thermodynamic parameters estimated for alpacide blue biosorption by Luffa cylindrica at initial concentrations (C 0) of 20, 50, and 70 mg/L
Full size table

Effect of the biomass amount

Solid/liquid ratio of 5, 8, 10, 14, and 20 g/L of biomass were separately tested in a dye solution of 20 mg/L at pH = 2. Figure 6 shows that the amount of dye adsorbed to the equilibrium has increased significantly in the weight range examined. Furthermore, it is observed that the maximum retention is obtained with a mass of 20 g/L. For this quantity, the capacity of biosorption registers its highest value (2.03 mg/g, i.e., 79 % retention rate). The increase of the rate of dye retention in function of the increase of the fiber mass is due mainly to the high number of sites available for biosorption on the surface of the biomass sites. A similar trend has been observed for the biosorption of Cr(VI) onto P. oceanica (L.) biomass (Ncibi et al. 2008b).

Fig. 6
figure 6

Influence of biomass amount on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (C 0 = 20 g/L, pH = 2, T = 20 °C)

Full size image

Adsorption kinetic modeling

To predict the mechanism during the alpacide blue biosorption process, two kinetic models were used to fit the experimental data, namely the pseudo-first-order and the pseudo-second-order models. The best-fit model was selected based on both linear regression correlation coefficient (R 2) and the calculated Q e values as described in “Effect of contact time and initial alpacide blue concentration” section and listed in Table 2. The correlation coefficients obtained by the pseudo-first-order model at all initial alpacide blue concentrations explored were found to be less than 0.985, whereas those of the second-order model were equal to 0.999. In addition, the calculated Q e with the pseudo-first-order (Q e1) did not give acceptable values when compared to the experimental ones, whereas the second-order model (Q e2) gives acceptable values. However, the pseudo-second-order model shows the best fit to the experimental data related to the biosorption of alpacide blue by L. cylindrica fibers. The value of rate constant (k 2) were found to decrease with increasing concentration. A similar result has been observed by other authors (Ben Hamissa et al. 2008; Ncibi et al. 2009).

Table 2 Kinetic rate constants related to the biosorption of alpacide blue onto Luffa cylindrica fibers
Full size table

Adsorption isotherm modeling

To optimize biosorption system to remove dyes, it is important to establish the most appropriate correlation for the equilibrium curves. The isotherm data were analyzed using three equilibrium models, Langmuir (Fig. 7), Freundlich (Fig. 8), and Temkin (Fig. 9) isotherm models. The mathematical expressions are given in “Effect of temperature” section and listed in Table 3. The best-fit equilibrium model was determined based on the linear-squared regression correlation coefficient R 2. The equilibrium biosorption data were very well represented by both Freundlich and Temkin isotherms with high correlation coefficients of 0.994 and 0.978, respectively. The fit to experimental data (R 2 = 0.963) in Table 3 shows that the Langmuir isotherm is less adequate to explain the biosorption of alpacide blue into L. cylindrica, as compared to the Freundlich and Temkin isotherms, thus confirming the poor applicability of Langmuir isotherm for this type of biosorption process. The Freundlich isotherm may be used for heterogeneous systems with interaction between the molecules adsorbed. On the basis of these results, one should infer the occurrence of multilayer coverage of alpacide blue on L. cylindrica predicted by the Freundlich isotherm.

Fig. 7
figure 7

Linear plot of Langmuir isotherm on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (solid/liquid ratio = 8 g/L, pH = 2, T = 20 °C)

Full size image
Fig. 8
figure 8

Linear plot of Freundlich isotherm on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (solid/liquid ratio = 8 g/L, pH = 2, T = 20 °C)

Full size image
Fig. 9
figure 9

Linear plot of Temkin isotherm on the biosorption capacity of alpacide blue by Luffa cylindrica fibers (solid/liquid ratio = 8 g/L, pH = 2, T = 20 °C)

Full size image
Table 3 Langmuir, Freundlich, and Temkin model constants and correlation coefficients for the biosorption of alpacide blue by Luffa cylindrica fibers (pH = 2, T = 20 °C)
Full size table

As the Freundlich isotherm, the Temkin isotherm considers the interactions between adsorbates assuming that the biosorption heat of all molecules decreases linearly when the layer is covered and that the adsorption has a maximum energy distribution of uniform bond (Vargas et al. 2011). The constant b is related to the heat of biosorption. The positive value of b found (b = 185.507) confirms an exothermic process. The n values ranging between 1 and 10 is a measure of adsorption intensity or surface heterogeneity, becoming more heterogeneous as its value gets closer to 10. The calculated n values were less than 1 (0.399), which represent a very low heterogeneity of the surface. A similar trend has been observed for the biosorption of reactive dyes onto apple pomace (Robinson et al. 2002).

Comparison of dye biosorption by L. cylindrica fibers and other low-cost adsorbents

To establish the relative efficiency of L. cylindrica relative to others used to remove textile dye from aqueous solutions, a comparison was made on the basis of the biosorption capacity (Q m). Dealing with the comparative study, the related results (Table 4) revealed that the L. cylindrica fibers could be considered as a promising biomaterial to remove textile dyes when compared to orange peel (Namasivayam et al. 1996), apple pomace (Robinsonet al. 2002), alunite (Ozacar and Sengi 2003), P. oceanica (Ncibi et al. 2006), and coir pith (Namasivayam and Kavitha 2002). Despite its low biosorption capacity, the L. cylindrica had a better potential of biosorption relative to other cheap biological materials.

Table 4 Comparison of biosorption capacities between Luffa cylindrica fibers and other studied sorbents for textile dye
Full size table

Conclusion

The present work shows that the L. cylindrica fibers were applied successfully for the biosorption of alpacide blue textile dye. The pH value of the aqueous solution played a significant role in affecting the biosorption capacity of alpacide blue. The highest amount adsorbed for dye removal was attained at pH 2. In addition, the increase in the temperature of solution decreases the adsorbed quantity. While the increase of the biomass amount, increases the retention rate of dye by the L. cylindrica. The calculated thermodynamic parameters showed that the biosorption of alpacide blue onto L. cylindrica is an exothermic and spontaneous process. The biosorption isotherm studies showed that the Freundlich model presents a good fit to the experimental data (R 2 = 0.994). The mechanism follows a pseudo-second-order reaction model (R 2 = 0.999).