Adsorption of methylene blue from aqueous solutions using activated carbon derived from coffee husks

JOURNAL OF SCIENCE OF HNUE DOI: 10.18173/2354-1059.2015-00076 Chemical and Biological Sci. 2015, Vol. 60, No. 9, pp. 32-43 This paper is available online at Received December 22, 2015. Accepted December 30, 2015. Contact Le Van Khu, e-mail address: lvkhu.hnue@gmail.com 32 ADSORPTION OF METHYLENE BLUE FROM AQUEOUS SOLUTIONS USING ACTIVATED CARBON DERIVED FROM COFFEE HUSKS Le Van Khu and Truong Duc Manh Faculty of Chemistry, Hanoi National University of Education Abstract. The adsorption of methylene blue (MB) onto activa ed arbon (AC) derived from coffee husk using zinc chloride activation at 600 oC was studied. The effects of various parameters such as contact time, initial MB concentration and temperature were investigated. The adsorption kinetics of pseudo first-order and pseudo second-order were used for the kinetic studies and were found to follow the pseudo second-order kinetic model. The experimental equilibrium data were analyzed using the adsorption isotherm models of Langmuir, Freundlich, Fritz-Schluender, Sips and Toth. The equilibrium data were best represented by the Sips isotherm. Thermodynamic parameters (ΔGo, ΔHo and ΔSo) were evaluated. The obtained results revealed the potential for use of activated carbon derived from coffee husk to remove methylene blue from aqueous solutions. Keywords: Activated carbon, methylene blue, adsorption, kinetics of adsorption, thermodynamics. 1. Introduction The discharge of colored wastes from textile, printing, paper, pharm c utical, plastics, leather and food processing industries poses a serious environmental problem [1, 2]. Many studies have been conducted on the toxicity of the various dyes and their impact on aquatic ecosystems, as well as environmentally friendly technologies that could reduce the dye content in wastewater [3]. Numerous physico-chemical methods, namely adsorption, chemical coagulation [4], filtration [5], ion exchange, electrolysis and biological [6] have been developed for removing dye pollutants. Of these methods, adsorption using activated carbon is most widely used due to its effectiveness, low cost and ease of operation [7]. The syn hesis of activated carbon from agricultural waste precursors is oftentimes a practical alternative to commercial activated carbon since its textural and chemical properties could be controlled by optimizing synthesis conditions (type of precursor and activation me hod) [8, 9]. It is estimated that about 1.75 million tons of coffee are produced annually in Vietnam. Only a small portion of the coffee husk produced is used as either microbial fertilizer or fuel, the rest being discarded as waste. With a cellulose content of over 68%, coffee husk was consequently selected as a raw material to produce activated carbon [10]. Adsorption of methylene blue from aqueous solutions using activated carbon derived 33 In the present study, the activated carbon prepared from coffee husk by zinc chloride activation was used as an adsorbent for the removal of methylene blue (MB) from aqueous solution. The adsorption kinetics, isotherm and thermodynamic properties were also explored. 2. Content 2.1. Experimental procedure 2.1.1. Preparation of activated carbon Coffee husk (Arabica coffee) obtained from a coffee mill in Chieng Ban - Son La Province was washed with water to remove dust and water soluble substances. It was then oven-dried at 110 oC for 12 h, grounded, and sieved to obtain particles that have anaverage size of 2.0 mm. The prepared raw material was carbonized at 450 oC under a nitrogen flow (300 mL.min-1) for 90 minutes. The resulting samples were impregnated with ZnCl2 for 48 h at room temperature and then dried at 120 oC for 12 h. The samples were then heated at 600 oC for 2 h in anitrogen atmosphere at a flow rate of 300 mL.min-1. Finally, the activated product was grounded, n utralized by 0.1 N HCl and washed several times with hot distilled water to a constant pH of ~ 6.0. The washed activated carbon samples were dried at 120 oC for 24 h and stored in a desiccator. 2.1.2. Characterization The textural characterization of the AC was based on the N2 adsorption isotherms, determined at 77 K using a Tri Star 3000 analyzer- Micromeritics. The specific surface area (SBET) was calculated by applying the BET equation to the adsorption data [11]. The micropore surface area (Smi) and the micropore volume (Vi) were evaluated using the t-plot method [12]. The mesopore volume (VBJH) was estimated using the Barrett-Joyn r-Halenda (BJH) method [13]. The average pore width (DBET) was calculated using following equation: mi BJH BET BET 4 V V D S ( )  (1) The chemical composition was determined by EDX using an EDX-LE VIOEL 6610 LV. 2.1.3. Methylene blue adsorption experiments Methylene blue (MB, CI = 52015; chemical formula: C16H18ClN3S; molecular weight = 319.86 g.mol-1, a cationic dye) supplied by Xilong Chemical Co. Ltd, China, was used as an adsorbate and was not purified prior to use. Double distilled water was used to prepare all of the solutions and reagents. The concentration of MB in the supernatant solution before and after adsorption were determined using a UV-Vis spectrophotometer (LIUV-310S) at 667 nm. Adsorption isotherms were performed at four different temperatures (10, 20, 30 and 40 oC). Typically about 100 mg of AC were added in 100 mL Erlenmeyer flasks containing 50 mL of MB solution with differing initial concentrations (50  250 mg.L-1). The flasks were kept in an isothermal shaker (120 rpm) for 48 h to reach solid – solution equilibrium. Then the samples were filtered prior to analysis in order to minimize interference of the carbon fines with the analysis. The adsorption capacity at equilibrium, e (mg.g -1), was calculated using the formula: 0 e e (C C )V q m   (2) Le Van Khu and Truong Duc Manh 34 where C0 and Ce (mg.L -1) are liquid-phase concentrations of MB at initial and equilibrium, respectively. V (L) is the volume of the solution and m (g) is the mass of dry adsorbent used. Kinetics experiments were carried out using 200 mL of solution of known initial MB concentration (150  300 mg.L-1) and 0.2 g of AC in a 250 mL flask which was kept in a temperature-controlled shaking water bath (10  40oC) and the aqueous solution-adsorbent mixtures were stirred at 150 rpm. At preset time intervals, 4 mL samples were pipetted out, filtered and the MB concentrations were measured. The amount of adsorption at time t, qt (mg g -1), was calculated using the formula: 0 t t (C C )V q m   (3) where C0 and Ct (mg.L -1) are the liquid-phase concentrations of dye initially and at any time t, respectively. V is the volume of the solution (L) and m (g) is the mass of dry adsorbent used. 2.2. Results and discussion 2.2.1. Characterization of adsorbents Physico-chemical and porosity characteristics of the activated carbon are presented in Table 1. Table 1. Physico-chemical and porosity characteristics of activated carbon Surface and Pore Texture SBET (m 2.g-1) Smi (m 2.g-1) SBJH (m 2.g-1) Vmi (cm 3.g-1) VBJH (cm 3.g-1) DBET (nm) 1376 1138 238 0.5300 0.4415 2.82 Atom Mass percent C (%) O (%) Cl (%) Fe (%) Zn (%) Total (%) 97.32 1.50 0.98 0.04 0.15 100 The as-prepared AC had a developed specific surface area (1376 m2.g-1) with high mesopore content (mesopore surface area ~ 238 m2.g-1, mesopore volume ~ 0.4415 cm3.g-1) with an average pore width of 2.82 nm. It is noticeable that the as-prepared AC has high C content (97.32%) with high purity (only 0.04% Fe and 0.15% Zn). The results also revealed that the AC prepared from coffee husk are environmental friendly materials and could be used for adsorption of large organic molecules. 2.2.2. Kinetics of MB adsorption * Effect of contact time The adsorption of MB on AC prepared from coffee husk as a function of time at 30 oC for different initial dye concentrations ranging from 150 to 300 mg.L-1 are illustrated in F gure 1. The adsorption capacity increased with the increase in contact time. It was found that rapid adsorption of MB occurred in the first 20 min and thereafter the rate of adsorption was slower. The rate of MB uptake was higher in the beginning due to the availability of active sites on the surface of the AC. After that, the adsorption of MB was controlled by electrostatic repulsion between the adsorbed positively charged sorb te species on the AC surface and the available cationic sorbate species in solution as well as the slow pore diffusion of the solute ion into the bulk of the adsorbent. Equilibrium was reached after about 5 h. Adsorption of methylene blue from aqueous solutions using activated carbon derived 35 t (min) 0 100 200 300 400 q t (m g g -1 ) 0 50 100 150 200 250 C o = 150 mg L -1 C o = 200 mg L -1 C o = 250 mg L -1 C o = 300 mg L -1 Figure 1. Effect of contact time on adsorption of different initial MB concentrations by AC at 30 o C * Adsorption kinetic study In order to analyze the adsorption kinetics of MB on AC, correlations between adsorbed amounts and time were investigated by testing different mathematical expressions corresponding to various models. In this study, pseudo first-order (Lagergren‟s equation) and pseudo second-order kinetic models were tested. a) t (min) 0 100 200 300 400 lo g (q e -q t) -1.0 0.0 1.0 2.0 3.0 150 mg/L, Experimental Pseudo first-order 200 mg/L, Experimental Pseudo first-order 250 mg/L, Experimental Pseudo first-order 300 mg/L, Experimental Pseudo first-order b) t (min) 0 100 200 300 400 t/ q t (m in g m g -1 ) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 150 mg/L, Experimental Pseudo second-order 200 mg/L, Experimental Pseudo second-order 250 mg/L, Experimental Pseudo second-order 300 mg/L, Experimental Pseudo second-order Figure 2. MB uptake by AC at different MB initial concentrations according to a) pseudo first-order kinetic model b) pseudo second-order kinetic model The pseudo first-order equation of Lagergren [14] based on solid capacity is given by: 1 e t e k t log(q q ) logq 2,303   (4) where k1 is the rate constant of pseudo first-order kinetics (min -1) and t is the contact time (min). The values of k1 or MB adsorption on AC were determined from the plot of log(qe - qt) against (Figure 2a). These values are shown in Table 2. The experimental data deviated greatly from the linearity, as was evidenced by the low qe and lo correlation values. Thus, the pseudo first-order model is inapplicable to this system. Similar results were observed for the adsorption of MB on bamboo-based activated carbon [15]. Le Van Khu and Truong Duc Manh 36 Table 2. The obtained parameters for the first- and second order kinetic models at 30 o C and different initial MB concentrations Co (mg.L -1 ) qe, exp (mg.g -1 ) Pseudo first-order Pseudo second-order qe, cal (mg.g -1 ) k1  10 2 (min -1 ) R 2 qe, cal (mg.g -1 ) k2  10 3 (g.mg -1 .min -1 ) R 2 150 148.98 13.27 0.92 0.6047 148.37 6.64 0.9999 200 195.28 53.35 1.28 0.7900 194.55 1.25 0.9996 250 215.93 37.59 1.00 0.8085 215.52 1.47 0.9997 300 226.73 51.81 0.96 0.8137 225.23 1.02 0.9994 A linear form of the pseudo second-order model [16] is given by: 2 t 2 e e t 1 1 t q k q q   (5) where k2 is the rate constant of the pseudo second- rder a sorption (g.m -1.min-1). The values of k2 and qe were calculated from the plot of t/qt versus t (Figure 2b). The values of qe and k2, along with the correlation coefficients for the pseudo second-order m dels, are presented in Table 2. The qe,exp and the qe,cal values from the pseudo second-order kinetic model are close to each other. The correlation coefficients are closer to unity for pseudo second-or er kinetics than for pseudo first-order kinetics. This suggests that the adsorption behavior of MB on AC derived from coffee husk can be well represented by the pseudo second-ord r model. t (min) 0 100 200 300 400 q t (m g g -1 ) 0 50 100 150 200 250 T = 10 o C T = 20 o C T = 30 o C T = 40 o C Figure 3. Effect of temperature on the adsorption of MB by AC, Co = 200 mg.L -1 Table 3. Calculated and experimental qe values and the pseudo second-order rate constants for the adsorption of MB by AC at different temperatures Co (mg.L -1 ) T ( o C) qe, exp (mg.g -1 ) qe, cal (mg.g -1 ) k2×10 3 (g.mg -1 .min -1 ) R 2 200 10 188.01 189.04 0.92 0.9996 20 193.77 194.17 1.35 0.9998 30 195.28 194.55 1.25 0.9996 40 196.96 196.46 1.85 0.9999 The effect of temperature on the adsorption of MB from aqueous solution is examined and presented in F gure3. The adsorption capacity increased slightly while temperature changed from 10 to 20 oC and was almost unchanged with further increase in temperature. Adsorption of methylene blue from aqueous solutions using activated carbon derived 37 The pseudo second-order model was also applied and the correlation coefficient values are close to unity. With the exception of 30 oC, an increase in temperature resulted in an increase of rate constant k2. 2.2.3. Adsorption equilibrium study The adsorption isotherm of MB on the AC sample at 30 oC is shown in Figure 4. As can be seen, MB-adsorbed amounts increase as MB concentration increases, rather sharply in the low-concentration region and more smoothly at higher concentration. T = 30 o C C e(mg L -1 ) 0 25 50 75 100 125 150 175 q e (m g g -1 ) 80 100 120 140 160 180 200 220 Experimental Langmuir Freundlich Fritz-Schluender Sips Toth Figure 4. Adsorption isotherms of MB on AC at 30 o C The curves were calculated using five models listed in Table 4 and by minimizing the HYBRID Table 4. Isotherms and the parameters involved in the different equilibrium adsorption isotherms Isotherm Expression Parameters Notes Langmuir   m L e e L e Q K C q 1 K C Qm, KL - Freundlich  F1/n e F e q K C KF , nF - Fritz-Schluender   FS m FS e e m FS e Q K C q 1 K C Qm, KFS, mFS KFS = KL; FS 1 m n  Sips       S S m S e e m m S e K C q Q 1 K C Qm, KS, mS KS = KL; S 1 m n  Toth     m e e 1/t t Th e Q C q K C Qm, KTh, t Th L 1 K K  ; 1 t n  To optimize adsorption for dye removal, it is important to establish the most appropriate correlation for the equilibrium curves. Various i otherm equations have been used to describe the equilibrium of the adsorption. In this study the correlation between the amount of adsorption at equilibrium and the liquid-phase concentration was tested using five isotherm equations given by Langmuir, Freundlich, Fritz-Schluender, Sips and Toth. The expressions of these isotherms are listed in Table 4. Le Van Khu and Truong Duc Manh 38 The parameters of these five isotherm equations were evaluated by minimizing the values of error function and using solver tools in Excel software version 2003 for Windows. For this purpose four different error functions were employed, including: The sum of the squares of the errors: N 2 e,pre e,mes i i 1 SSE (q q )    (6) The root mean square error: RMSE =     N 2 e,pre e,mes i i 1 1 q q N (7) The hybrid fractional error function: HYBRID =             2 N e,pre e,mes i 1 e,mes i q q100 N p q (8) Marquardt‟s percent standard deviation: MPSD =             2 N e,pre e,mes i 1 e,mes i q q1 100 N p q (9) where e,mesq and e,preq are the experimental and predicted adsorption capacities, respectively; N is the number of experimental data; p is the parameter of the isotherm equation. * Choosing the best isotherm model Average relative errors (ARE) are calculated according to equation (10) indicated the fit between the experimental a d predicted values for adsorption capacity used in plotting isotherm curves. The best fit isotherm was selected based on this model and the error functions that produced minimum value of ARE.           N e,pre e,mes i 1 e,mes i q q100 ARE N q (10) Parameters as calculated for different isotherms for MB-AC systems are presented in Table 5. As can be seen from this table, the Sips model gave the best fit along the entire range of equilibrium concentration (the smallest of ARE), and the HYBRID function error gave the best regression. Figure 4 presents how well the five equations fit the data for the adsorption of MB on AC prepared from coffee husk by ZnCl2 activation. Adsorption of methylene blue from aqueous solutions using activated carbon derived 39 Table 5. Isotherm parameters for MB–AC systems at 30 oC Isotherm Parameters SSE RMSE HYBRID MPSD Langmuir Qm (mg.g -1) 187.54 187.54 191.98 182.65 KL (L.mg -1) 15.98 15.98 20.86 19.39 ARE (%) 7.65 7.65 6.92 7.90 Freundlich KF mg.g−1(L.mg−1)1/n 134.49 134.49 142.77 141.59 1/n 0.08 0.08 0.07 0.08 ARE (%) 7.93 7.93 5.92 6.71 Fritz - Schluender Qm (mg.g -1) 34.49 34.48 47.38 40.00 KFS (L.mg -1) 163.51 163.51 154.15 160.48 mFS 0.96 0.96 0.95 0.95 ARE (%) 3.06 3.07 2.48 2.91 Sips Qm (mg.g -1) 206.02 206.02 205.36 207.07 Ks (L.mg -1) 16.48 16.48 17.07 16.35 ms 0.43 0.43 0.44 0.41 ARE (%) 1.76 1.76 1.60 1.76 Toth KTh (L.mg -1) 0.11 0.11 0.10 0.11 Qm (mg.g -1) 208.68 208.68 206.79 210.44 t 0.36 0.36 0.37 0.34 ARE (%) 1.79 1.79 1.66 1.75 * Effect of temperature Temperature had a pronounced effect on the adsorption capacity of the adsorbents. Figure 5 shows the plots of the adsorption isotherms, qe versus Ce for MB-AC system at different temperatures ranging from 10 to 40 oC. It is seen that when the temperature increased, the adsorptivity of MB increased. It might be explained that, with an increase in temperature, the mobility of the MB increases and the retarding forces acting on the diffusing ions decrease, thereby increasing the adsorptive capacity of adsorbent. C e(mg L -1 ) 0 25 50 75 100 125 150 175 200 q e (m g g -1 ) 80 100 120 140 160 180 200 220 Experimental, 10 o C Sips, 10 o C Experimental, 20 o C Sips, 30 o C Experimental, 30 o C Sips, 30 o C Experimental, 40 o C Sips, 40 o C Figure 5. Adsorption isotherms of MB on AC at different temperatures The curves were calculated using the Sips equation and the HYBRID function error Le Van Khu and Truong Duc Manh 40 Table 6. Sips constants for adsorption of MB on AC at different temperatures T ( o C) Qm (mg.g -1 ) KS (L.mg -1 ) MS HYBRID ARE (%) Sips equation 10 185.18 3.223 0.715 4.7619 2.93       0.715 e e 0.715 e 3.223C q 185.18 1 3.223C 20 189.28 5.265 0.687 2.0914 1.31       0.687 e e 0.687 e 5,265C q 189.28 1 5,265C 30 205.36 17.074 0.440 2.5663 1.60       0.440 e e 0.440 e 17.074C q 205.36 1 17.074C 40 203.60 7.813 0.617 2.5139 1.57       0.617 e e 0.617 e 7.813C q 203.60 1 7.813C The application of five adsorption isotherm models based on minimizing the value of four error functions showed that from 10 to 40 oC, he adsorption of MB onto AC best fit the Sips isotherm using the HYBRID error function. The isotherm parameters as well as HYBRID and ARE values are shown in Table 6. As can be seen from Table 6, ARE values are quite small, in the range of 1.31  2.9 %. Qm increases moderately with the increasing in temperature and was in the range of 185.18  205.36 mg.g-1. The high adsorption