- Email: [email protected]

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Adsorption of phosphate from aqueous solutions onto modiﬁed wheat residue: Characteristics, kinetic and column studies Xing Xu, Baoyu Gao ∗ , Wenyi Wang, Qinyan Yue, Yu Wang, Shouqing Ni School of Environmental Science and Engineering, Shandong University, Jinan 250100, PR China

a r t i c l e

i n f o

Article history: Received 10 September 2008 Received in revised form 25 November 2008 Accepted 3 December 2008 Available online 9 December 2008 Keywords: Modiﬁed wheat straw (MWS) Phosphate Adsorption kinetics Column adsorption Activation energy

a b s t r a c t Kinetic and column adsorption of phosphate from aqueous solution using modiﬁed wheat residue (MWS) as an adsorbent were studied in a batch reactor. The respective characteristic rate constants and activation energy were presented after linear and non-linear ﬁtting. In addition, the effects of inﬂuent concentration of phosphate and ﬂow rates on the column adsorption were also investigated. The results showed that the adsorption process could reach equilibrium in 10–15 min, and the pseudo-second-order equation generated the best agreement with experimental data for adsorption systems. The activation energy was 3.39 kJ mol−1 indicating that the synthesis process was a physical adsorption. In the column tests, the increase of inﬂuent concentration and ﬂow rate both decreased the breakthrough time, and the MWSpacked column exhibited excellent phosphate removal from aqueous solution. These results provide strong evidence of the potential of MWS for the technological applications of phosphate removal from aqueous solutions. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Phosphate released from wastewaters is undesirable to the environment because it provides an excessive of nutrients in the natural water body and accelerates eutrophication. Despite the environmental beneﬁts of limiting nutrient release, there is a constant need to supply phosphorus to agriculture and industry. Thus, development of treatment methods that facilitate the removal of phosphate from wastewaters prior to discharge into natural waters is required. In wastewater treatment technology, various techniques including chemical and biological methods have been successfully applied, such as reverse osmosis [1], biological de-nitriﬁcation [2], electrodialysis [3] and adsorption [4]. Further development of adsorbents has been investigated, which focuses on the research of adsorbents prepared from agricultural residues. Materials such as sugarcane bagasse [5], peanut hull [6], apple pomace [7], sawdust [8], coconut husk [9], orange peel [10], banana pith [11] and pine bark [5] have been utilized for this purpose. Wheat straw (WS) is one of the well-known ﬁbre crops, which industry potential is now being intensively reconsidered [13]. An appropriate chemical composition in WS with 16.4% lignin, 32.1% cellulose (as ␣-cellulose), 29.2% hemicelluloses and 22.3% extractives, suggests a broad potential application to adsorbent production; this is due to the large amount of easily available hydroxyl

groups existing in the cellulose, hemicelluloses and lignin, which can easily make a series of chemical reactions, such as condensation, etheriﬁcation and copolymerization [12]. This paper illustrates the preparation of an adsorbent from WS, and our main objective is to study the adsorption kinetics and column adsorption of phosphate from aqueous solutions using modiﬁed wheat straw (MWS) as adsorbent. Kinetic equations were designed to determine the adsorption kinetic rates. The effect of inﬂuent phosphate concentration and ﬂow rate on the penetrating time of column was identiﬁed after the experiments of column adsorption. The data analyzed by the adsorption kinetics and column adsorption can provide valuable information for the design and operation of waste water treatment plants, which represent great practical value for the technological applications of phosphate removal from aqueous solutions. 2. Materials and method 2.1. Materials WS, obtained from Liao Cheng, Shandong, China, was washed with water, dried at 60 ◦ C for 6 h and sieved into particles with diameters 100–250 m. 2.2. Preparation of MWS

∗ Corresponding author. Tel.: +86 531 88364832; fax: +86 531 88364513. E-mail address: [email protected] (B. Gao). 0927-7765/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2008.12.006

Two grams of WS were added with 20 mL of epichlorohydrin and 25 mL of N,N-dimethylformamide in a 250 mL three-neck round

X. Xu et al. / Colloids and Surfaces B: Biointerfaces 70 (2009) 46–52

47

Fig. 1. Synthesis of MWS (cellulose as example).

bottom ﬂask at 100 ◦ C for 1 h. Four mL of ethylenediamine was added and the solution was stirred for 1 h at 100 ◦ C, followed by adding 30 mL of 99% triethylamine (w/w) and the mixture was stirred for 3 h at 100 ◦ C. The product was washed with 500 mL of distilled water to remove residual chemicals, dried at 60 ◦ C for 12 h and sieved to obtain particles with diameter less than 250 m and then used in all the adsorption experiments [13,14]. The synthesis of MWS is shown in Fig. 1. N,N-dimethylformamide was used as an organic medium to enhance the susceptibility of the epoxide ring. Except for its weakbased catalysis in organic solvent, ethylenediamine can also be used as crosslinking agent for some synthesis of polymers between epichlorohydrin and amine [15,16], which will increase the attachment of triethylamine with epichlorohydrin. 2.3. Characteristics of MWS 2.3.1. Zeta potential (mV) The new anion exchanger prepared from WS is used for the removal of anionic pollutant. As a result, it is signiﬁcant to determine the change of surface charge of MWS in comparison with WS. Zeta potential of MWS and WS were determined by electro-kinetic analyzer (JS94H, Shanghai Zhongchen Digital Technical Apparatus Co., Ltd., China).

2.3.2. Infrared (IR) spectra analysis IR spectra were recorded on a PerkinElmer “Spectrum BX” spectrometer in the 4000–400 cm−1 region. 2.4. Batch adsorption Different concentrations of phosphate solutions were prepared by dissolving different qualities of KH2 PO4 in distilled water. To describe the effect of concentrations of phosphate on the adsorption kinetic curves, adsorption experiments were carried out by agitating 1 g of MWS with 500 mL of desired concentrations of phosphate solutions (10, 30, 40 and 50 mg(P) L−1 ), and at a temperature 20 ± 2 ◦ C in a stirrer operating at 120 rpm for 80 min. The effect of temperature on the adsorption curves was also investigated by agitating 1 g of MWS with 500 mL of 50 mg (P) L−1 phosphate solution at different temperatures (20, 40, 60 ◦ C) at 120 rpm for 80 min. Samples (1 mL) were taken at suitable time intervals and then ﬁltered to analyze residual phosphate concentrations in solutions with an UV–visible spectrophotometer (model UV754GD, Shanghai) at an absorbance wavelength of 700 nm. The equilibrium concentration in solid phase (qe ) was given as: qe =

(co − ce )V m

(1)

48

X. Xu et al. / Colloids and Surfaces B: Biointerfaces 70 (2009) 46–52

where qe is the amount of phosphate sorption per gram MWS at equilibrium, co and ce are the concentrations of phosphate at original and equilibrium, respectively. V is the volume of solution, and m is the amount of MWS (g). An aliquot of 1 g of MWS was added into an organic-glass column 200 mm in length and 12 mm in diameter, forming a section of column which would be used in the column tests. To determine the effect of concentrations of phosphate and ﬂow rates on the column adsorption, various concentrations of phosphate solution (100, 200, 300 mg (P) L−1 ) and ﬂow rates (1, 3, 5 and 10 mL min−1 ) were designed. The efﬂuent solutions were collected, and every 10 mL was selected as a sample to determine the concentrations of phosphate in the efﬂuent solutions. The column adsorption capacity (qed ) was calculated by the equation expressed as: qed =

(co Vo −

cn vn )

m

(2) Fig. 2. IR analysis of MWS and WS.

where qed is the amount of phosphate sorption per gram MWS at saturation (mg g−1 ), co is the original concentration of phosphate (mg L−1 ), Vo is the total volume of the inﬂuent solutions (L), cn is the concentration of sample n (mg L−1 ), and vn is the volume of sample n (L), m is the amount of MWS (g).

Adsorption kinetics study is important to treatment of aqueous efﬂuents as it provides valuable information on the reaction pathway and the control mechanism of the adsorption process. To demonstrate the adsorption process of phosphate onto MWS, a kinetic investigation was conducted using four kinetic models. (1) Pseudo ﬁrst-order model A kinetic model for adsorption analysis is the pseudo ﬁrstorder rate expressed in the form [17]: (3)

After integration and applying the boundary conditions, for qt = 0 at t = 0 and qt = qt at t = t, the integrated form of Eq. (3) becomes: ln(qe − qt ) = ln qe − k1 t

(4)

where qe and qt are the amounts of phosphate adsorbed per gram MWS at equilibrium and time t (mg g−1 ), respectively; k1 is the rate constant of pseudo ﬁrst-order (min−1 ). (2) Modiﬁed pseudo ﬁrst-order model In this model, which has no theoretical derivation, the pseudo ﬁrst-order equation is modiﬁed through its rate constant by replacing k1 by K1 , where k1 = K1 (qe /qt ) which gives the rate equation [17]: qe dqt = K1 (qe − qt ) qt dt

(5)

After integration and applying the boundary conditions, for qt = 0 at t = 0 and qt = qt at t = t, the integrated form of Eq. (5) becomes: qt + ln(qe − qt ) = ln qe − K1 t qe

dqt = k2 (qe − qt )2 dt

(7)

Integrating this equation for the boundary conditions:

2.5. Kinetic models applied to the adsorption of phosphate

dqt = k1 (qe − qt ) dt

[17,18]:

(6)

It is the integrated law for the modiﬁed pseudo ﬁrst-order reaction, where K1 is the rate constant of modiﬁed pseudo ﬁrstorder (min−1 ). (3) Pseudo second-order model If the rate of adsorption is a pseudo second-order model, the pseudo second-order kinetic rate equation is given as follows

1 t t = + qt qe k2 q2e

(8)

where k2 is the equilibrium rate constant of pseudo secondorder (g(mg min)−1 ). (4) Intra-particle diffusion model The three models above cannot identify the diffusion mechanism, so the intra-particle diffusion model was proposed. The initial rate of intra-particle equation is as the following [19,20]: qt = kp t 0.5

(9) (g mg−1

min−0.5 ),

where kp is the intra-particle rate constant and the intra-particle rate constant kp is a function of equilibrium concentration in solid phase qe and intra-particle diffusivity D according to the equation expressed as: 6qe kp = R

D

(10)

where R is the particle radius and D is intra-particle diffusivity. 3. Results and discussion 3.1. Characteristics of MWS 3.1.1. Zeta potential (mV) analysis Sample’s zeta potentials were performed using electro-kinetic analyzer. Results show that the zeta potentials of MWS and WS are +32 and −35 mV, respectively, indicating the existence of positivecharge functional groups on the MWS structure. 3.1.2. IR analysis of MWS The IR spectral change of WS and MWS is shown in Fig. 2. The band intensity at 3380 cm−1 indicates a mass of hydroxyl groups; ketone groups are observed by the band intensity at 2920 cm−1 , and the band at 1625 cm−1 is associated with the special vibration of aromatic cyclic groups. IR analysis of MWS shows a change in the structure between MWS and WS. Fig. 2 shows chloric alky in MWS by the intense vibration of band at 623 cm−1 . The intense vibration at 1350 cm−1 indicates that a large number of amino groups have been grafted into the MWS structure. As a result, the adsorption of phosphate onto MWS can be ascribed to the signiﬁcant increase

X. Xu et al. / Colloids and Surfaces B: Biointerfaces 70 (2009) 46–52

49

Fig. 4. Adsorption kinetics curves of various concentrations of phosphate solutions. Fig. 3. Effect of adsorbent dosage on adsorption of phosphate by MWS and WS.

of amino groups which are used as adsorption sites for phosphate removal. 3.2. Effect of adsorbent dosage The effect of adsorbent dosage on the adsorption of phosphate is shown in Fig. 3. The removal efﬁciency of phosphate increases signiﬁcantly from 52.1% to 92.5% with increase of MWS dosage from 0.5 to 2 g L−1 , and the adsorption is nearly constant when dosage exceeds 2 g L−1 . With increasing adsorbent dosage, more adsorption sites are available for adsorption and nearly all the phosphate is adsorbed when the adsorbent is increased to 2 g L−1 . So, it is reasonable that the phosphate removal could not increase much anymore

when the dosage of the adsorbent is higher than 2 g L−1 . Fig. 3 also shows the phosphate removal capacity of raw WS. The low phosphate removal (4.8%) of raw WS indicates that the adsorption of raw WS is only based on surface adsorption, and a chemical modiﬁcation is required to introduce some functional groups into the structure of WS for the phosphate removal [21]. 3.3. Effect of phosphate concentrations on adsorption kinetics Kinetics of an adsorption process is vital in wastewater treatment as it provides essential information on the reaction pathways and solute uptake rate. The effect of the initial concentrations of phosphate solutions on the adsorption curves is depicted in Fig. 4. The qe increases with the initial concentrations of phosphate from

Fig. 5. Kinetic equation for adsorption of various concentrations of phosphate onto MWS.

50

X. Xu et al. / Colloids and Surfaces B: Biointerfaces 70 (2009) 46–52

Table 1 Kinetic parameters for adsorption rate expressions. C0 /mg L−1

Pseudo ﬁrst-order −1

50 40 30 10

Modiﬁed pseudo ﬁrst-order −1

Pseudo second-order −1

−1

Intra-particle diffusion

k1 /min

R

K1 /min

kp /mg (g min)

k2 /g (mg min)

R

kp /mg (g min)−1

R2

0.3824 0.3123 0.2597 0.2654

0.989 0.978 0.956 0.933

0.2911 0.2645 0.2108 0.1929

0.996 0.997 0.991 0.988

0.0684 0.1342 0.1417 0.3208

0.9996 0.9999 0.9998 0.9999

9.4912 7.653 5.4534 1.8228

0.982 0.939 0.967 0.909

2

2

Table 2 Kinetic parameters at different temperature. T (◦ C)

qea (mg g−1 )

Pseudo ﬁrst-order −1

k1 (min 20 40 60

22.9 21.6 20.1

0.3824 0.2858 0.2023

)

qeb (mg g 20.12 14.06 10.09

Pseudo second-order −1

)

2

−1

R

k2 (g mg

0.989 0.912 0.874

0.0684 0.0755 0.0808

min

−1

)

Intra-particle diffusion −1

qeb (mg g 22.99 21.65 20.16

)

R

kp mg (g min)−1

R2

0.9996 0.9999 0.9998

9.4912 8.7432 8.5764

0.982 0.963 0.951

2

Where qea is experimental data, qeb is calculated data.

10 to 50 mg (P) L−1 . For all the tested concentrations there is a steep ascending trend of qt at the beginning of the adsorption process (0–5 min). A small increase of phosphate adsorption is observed in the curves between 5 and 15 min, and then qt is stabilized and the adsorption reached equilibrium. The experimental data were discussed and analyzed using pseudo ﬁrst-order equation, modiﬁed pseudo ﬁrst-order equation, pseudo second-order equation and intra-particle diffusion equation, respectively. Kinetic parameters for the four kinetic models and correlation coefﬁcients are evaluated as shown in Fig. 5 and Table 1. Results shown in Fig. 5 indicate that pseudo ﬁrst-order equation, modiﬁed pseudo ﬁrst-order equation and intra-particle diffusion equation can only describe the adsorption process before reaching equilibrium (0–15 min), while pseudo second-order model can be applied for the entire adsorption process as shown in Fig. 5(c). The values of R2 (Table 1) of pseudo second-order model for phosphate adsorption are satisfactory (>0.9996) and followed by those of modiﬁed pseudo ﬁrst-order equation, pseudo ﬁrst-order equation, intra-particle diffusion equation, respectively; this result indicates that the adsorption of phosphate onto MWS can be well represented by the pseudo-second-order kinetic model [17,18]. Similar kinetic results have also been reported for the adsorption of nitrite onto modiﬁed wheat residue (MWR) [21]. According to Eq. (9), if the plot of qt vs. t0.5 passes through the origin, intra-particle diffusion will be the rate-controlling step [17,19,20]. The lines shown in Fig. 5(d) do not pass through the origin, and therefore, intra-particle diffusion is not the only ratelimiting step; other kinetic processes occur simultaneously and contribute to the adsorption mechanism. The values of k1 and K1 in the ﬁrst-order equation and modiﬁed pseudo ﬁrst-order equation both increase with the increase of initial phosphate concentrations (Table 2), which reveals the fact that adsorption is faster in higher initial concentration of phosphate [17]. The kp in the intra-particle diffusion equation also increases with the increase of initial concentrations of phosphate. As seen from Eq. (10), the intra-particle rate constant kp is proportional to both qe and D. With increasing initial phosphate concentration, qe and D will increase accordingly [17,22], and therefore, the increase of kp with increasing initial concentrations of phosphate can be attributed to the increase of qe and D. Subsequently, the intra-particle diffusion starts to slow down due to the low adsorbate concentrations in solution and the ﬁnal equilibrium stage was then reached. However, the equilibrium rate constant of pseudo second-order k2 decreases with the increase of various initial concentrations of phosphate, which shows seemingly contradictory trends with the rate constants of k1 , K1 and kp . The Eq. (8) can be changed into Eq.

(11) expressed as: k2 =

F qe (1 − F)t

(11)

where F is the fraction of qt /qe , and t is the time to reach qt , so k2 can be expressed as a function of uptake fraction F (qt /qe ) and t. Eq. (11) indicates that k2 is inversely related to t, and therefore, the decrease of k2 with increasing initial concentrations of phosphate only suggests that a longer period of time will be needed for a higher concentration phosphate to realize the speciﬁc uptake fraction, while the adsorption rate (dqt /dt) at contact time t (t < 10 min) still increases as the initial concentration of phosphate increases. So the analysis results that show the seemingly contradictory trends in Table 1 are in fact compatible with each other [17]. 3.4. Effect of temperature on adsorption kinetics 3.4.1. Effect of temperature on adsorption kinetics The effect of temperature on the phosphate adsorption is shown in Fig. 6. The adsorption of phosphate onto MWS followed at three temperatures, 20, 40 and 60 ◦ C, shows a fast phosphate uptake and a decrease in amount of adsorption when increasing the temperature. The observed decrease in the adsorption capacity with an increase of the temperature from 20 to 60 ◦ C indicates that lower temperatures are favorable to phosphate removal by adsorption onto MWS; this may be due to a tendency for the phosphate ions to escape from the solid phase to the bulk phase with an increase in the temperature of the solution [23,24]. The experi-

Fig. 6. Effect of temperature on the kinetics.

X. Xu et al. / Colloids and Surfaces B: Biointerfaces 70 (2009) 46–52

mental data are ﬁtted with pseudo ﬁrst-order equation, pseudo second-order equation and intra-particle diffusion equation, and the results shown in Table 2 indicate that pseudo second-order equation can well describe the adsorption process affected by the temperature. With the increase of temperature, the qeb calculated by the pseudo second-order equation decrease, and the data agree well with the experimental qea values as shown in Table 2. Similar results were reported for the adsorption of acid dye onto activated palm ash [25]. 3.4.2. Determination of activation energy Ea The pseudo-second-order rate constant k2 is expressed as a function of temperature by the Arrhenius type relationship [26]: ln k = ln k2 −

Ea RT

(12)

Fig. 7 is the linear of Arrhenius equation. As is shown in Fig. 7, −Ea /R = −407.4, and Ea = 3.39 kJ mol−1 , which is in the range of physical adsorption (0–40 kJ mol−1 ); this indicates that the adsorption

Fig. 7. Linear of Arrhenius equation.

Fig. 8. Effect of various inﬂuent concentrations on the breakthrough curves.

Fig. 9. Breakthrough curves of various ﬂow rates.

51

52

X. Xu et al. / Colloids and Surfaces B: Biointerfaces 70 (2009) 46–52

process of phosphate by MWS is a physical adsorption process [25,27]. 3.5. Column adsorption of phosphate onto MWS 3.5.1. Effect of the inﬂuent concentrations on the breakthrough curves To evaluate the practical application of MWS to the continuous removal phosphate from solution, we conducted a column adsorption test. In this work, three initial concentrations were used to evaluate the performance in the continuous system (100, 200 and 300 mg (P) L−1 ). Fig. 8 shows the effect of the inﬂuent concentrations of phosphate on the shape of breakthrough curves at the same ﬂow rate (5 mL min−1 ). Results shown in Fig. 8 indicate that breakthrough time decreases as phosphate concentration increases, with the curve showing an S-type mode. The column adsorption capacity qed for phosphate by MWS are 67.1, 66.3 and 78.1 mg g−1 when inﬂuent concentrations are 100, 200 and 300 mg L−1 , respectively. The qed at 300 mg L−1 is higher than those at 100 and 200 mg L−1 ; this can be attributed to the concentration gradient which will enhance the adsorption process [28–30]. 3.5.2. Effect of ﬂow rates on the breakthrough curves To investigate the effect of ﬂow rates on the breakthrough curves of phosphate adsorption, 200 mg L−1 of phosphate is ﬁxed as the inﬂuent concentration, and the ﬂow rates are 1, 3, 5 and 10 mL min−1 , respectively. The breakthrough curves are shown in Fig. 9. As shown in Fig. 9, the breakthrough curve generally occurs faster with higher ﬂow rate at 10 mL min−1 , compared to those of the ﬂow rates at 1, 3 and 5 mL min−1 . The breakthrough time increases with the decrease of ﬂow rate, and therefore, the adsorption of phosphate is given more time to contact with MWS at a lower ﬂow rate which may result in the higher adsorption efﬁciency of phosphate in the column [30]. The results are veriﬁed by the qed at various ﬂow rates (1, 3, 5 and 10 mL min−1 ), which are 67.4, 65.8, 66.3 and 64.4 mg g−1 , respectively. Based on the results above, the effect of ﬂow rate can be ignored when it is lower than 5 mL min−1 . Similar results were reported for the adsorption chromium (VI) onto coconut coir pith [31]. 4. Conclusion Adsorption kinetics of phosphate onto MWS was studied by pseudo ﬁrst-order equation, modiﬁed pseudo ﬁrst-order equation, pseudo second-order equation and intra-particle diffusion equation. The equilibrium concentration in solid phase qe increased with the initial concentrations of phosphate, and a fast adsorption equilibrium was reached at 10–15 min. The pseudo second-order model

provided the best correlation of the experimental data. The temperature was inversely related to the qe , and the activation energy (3.39 kJ mol−1 ) of phosphate adsorption indicated that the adsorption process was a physical adsorption. On the basis of the experimental results of column adsorption, the increase of inﬂuent concentration and ﬂow rate decreased the breakthrough time, and an acceptable level of phosphate removal was achieved using the MWS-packed column. Acknowledgments The research was supported by the National Natural Science Foundation of China (50878121), the Science and Technology Development Key Program of Shandong Province, China (No. 2006GG2206007) and the Science and Technology Key Research Project of Environmental Protection of Shandong Province, China (No. [2006] 005). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31]

L. Thörneby, K. Persson, J. Agric. Eng. Res. 74 (1999) 159–170. A. Uygur, F. Karg, J. Environ. Manage. 71 (2004) 9–14. A.M.O. Mohamed, J. Hazard. Mater. 90 (2002) 297–310. I. Fernández-Olmo, J.L. Fernández, A. Irabien, Sep. Purif. Technol. 56 (2007) 118–125. U.S. Orlando, T. Okuda, W. Nishijima, React. Funct. Polym. 55 (2003) 311–318. R.M. Gong, Y.Y. Ding, M. Li, Dyes Pigments 64 (2005) 187–192. T. Robinson, B. Chandran, P. Nigam, Water Res. 36 (2002) 2824–2830. M. Ajmal, A.H. Khan, A. Ahmad, Water Res. 32 (10) (1998) 3085–3091. G.N. Manju, C. Raji, T.S. Anirudhan, Water Res. 32 (1998) 3070–3602. C. Namasivayam, N. Muniasamy, K. Gayatri, Bioresour. Technol. 57 (1996) 37–43. C. Namasivayam, D. Prabha, M. Kumutha, Bioresour. Technol. 64 (1998) 77–79. S. Nakamura, Y.A. Masato, S. Toshihiko, J. Appl. Polym. Sci. 4 (1992) 265–271. U.S. Orlando, A.U. Baes, W. Nishijima, Bioresour. Technol. 83 (2002) 195–198. U.S. Orlando, A.U. Baes, W. Nishijima, Chemosphere 48 (2002) 1041–1046. N. Bic¸ak, B.F. S¸enkal, React. Funct. Polym. 36 (1998) 71–77. R. Molinari, P. Argurio, T. Poerio, Desalination 162 (2004) 217–228. X.Y. Yang, A.D. Bushra, J. Colloid Interface Sci. 287 (2005) 25–34. M.T. Sulak, E. Demirbas, M. Kobya, Bioresour. Technol. 98 (13) (2007) 2590–2598. S.V. Mohana, N.C. Raka, J. Karthikeyan, J. Hazard. Mater. B90 (2002) 189–204. T. Fukuda, A. Aramata, J. Electroanal. Chem. 440 (1997) 153–161. Y. Wang, B.Y. Gao, W.W. Yue, Q.Y. Yue, Colloids Surf. A 308 (2007) 1–5. N.L. Filippova, J. Colloid Interface Sci. 206 (1998) 592–602. S. Karaca, A. Gürses, M. Ejder, M. Ac¸ıkyıldız, J. Colloid Interface Sci. 277 (2004) 257–263. Y.S. Ho, G. Mc Kay, Chem. Eng. J. 70 (1998) 115–124. B.H. Hameed, A.A. Ahmad, N. Aziz, Chem. Eng. J. 133 (2007) 195–203. N. Yeddou, A. Bensmaili, Desalination 185 (2005) 499–508. H. Nollet, M. Roels, P. Lutgen, P. Van der Meeren, Chemosphere 53 (2003) 655–665. R.P. Han, J.H. Zhang, W.H. Zou, H.J. Xiao, J. Shi, H.M. Liu, J. Hazard. Mater. 133 (2006) 262–268. R.P. Han, Y. Wang, W.H. Zou, Y.F. Wang, J. Shi, J. Hazard. Mater. 145 (2007) 331–335. J. Goel, K. Kadirvelu, C. Rajagopal, V.K. Gary, J. Hazard. Mater. 125 (2005) 211–220. P. Suksabye, P. Thiravetyan, W. Nakbanpote, S. Chayabuta, J. Hazard. Mater. 160 (2008) 56–62.