The present study focused on the application of response surface methodology to optimize the fabrication of activated carbon (AC) from sugarcane bagasse for adsorption of Cu2+ ion. The AC was synthesized via chemical activation with ZnCl2 as the activating agent. The central composite design based experiments were performed to assess the individual and interactive effect of influential parameters, including activation temperature, ZnCl2 impregnation ratio and activation time on the AC yield and removal of Cu2+ ion from the aqueous environment. The statistically significant, well-fitting quadratic regression models were successfully developed as confirmed by high F- and low P-values (<0.0001), high correlation coefficients and lack-of-fit tests. Accordingly, the optimum AC yield and removal efficiency of Cu2+ were predicted, respectively, as 48.8% and 92.7% which were approximate to the actual values. By applying the predicted optimal parameters, the AC shows a surprisingly high surface area of around 1,500 m2/g accompanied by large pore volume and narrow micropore size at low fabrication temperature.

INTRODUCTION

Activated carbons (ACs) are normally fabricated from various carbonaceous sources through the oxidative reaction of the carbon atoms, suggesting a number of prominent characteristics such as large surface area, well-developed porosity, and tunable surface chemistry (Ahmad et al. 2007). The ACs have been thus widely used as efficient solid adsorbents for removal of environmental pollutants through the adsorption route. However, the application of commercial ACs is quite limited due to their high cost. To address this difficulty, fabrication of low-cost agricultural waste-based ACs have been developed (Yahya et al. 2015), adding more economic values to the abundant low-cost agricultural wastes which are normally burnt causing unfavorable effects to the environment. Like other agricultural residues, sugarcane bagasse contains several lignocellulosic compounds and its structure is built by cellulose molecules surrounded by hemicellulose, lignin and pectin components (Kalderis et al. 2008). Therefore, this material can be transformed into AC to increase adsorption capacities for removal of hazardous chemicals.

In order to fabricate AC, two methods have been commonly used including physical and chemical activation (Hadi et al. 2015). In physical activation, carbonization of raw material was conducted under inert atmosphere followed by controlled gasification at high temperature in the presence of CO2 or steam. In chemical activation, initial materials were soaked with strong dehydrating reagents. Then, chemical-impregnated precursors were pyrolyzed at high temperature. The violent reactions between carbonaceous components and reagent molecules occurred, resulting in the formation of a highly porous structure (Kalderis et al. 2008; Hadi et al. 2015; Yahya et al. 2015). Among well-known activation chemicals, ZnCl2 was proved as an efficient reagent regardless of several environmental disadvantages (Rozada et al. 2005; Zhang et al. 2005; Hadi et al. 2015). The properties of the resulting AC were found to depend primarily on the activation temperature, impregnation ratio (IR) between the activating agent and the precursor, and the activation time (Huang et al. 2015). There have been a number of attempts to determine the optimum fabrication conditions of ZnCl2-AC with the surface area ranging from around 300 m2/g to 900 m2/g using low or moderate activation temperature (Tsai et al. 2001; Kalderis et al. 2008). The main objectives of this study are to fabricate ZnCl2-ACs from sugarcane bagasse precursor and to assess effects of various influential parameters, i.e. activation temperature, IR and activation time, on formation yield of the AC and the removal efficiency of Cu2+ from aqueous solution using response surface methodology (RSM) involving central composite design (CCD) method.

EXPERIMENTAL PROCEDURE

Production of AC

Sugarcane bagasse samples were collected from the local juice stores in Ho Chi Minh City, Vietnam. The precursors were initially washed with distilled water for several times, dried under the sunshine and then ground to diameters of approximately 1.0 mm. The dried materials were soaked with a ZnCl2 solution for 24 h. IR between ZnCl2 and the precursor was calculated as: 
formula
1
where and are the weight of the anhydrous ZnCl2 (g) and the precursor (g), respectively.
Then, ZnCl2-impregnated precursors were placed in a heat-resistant glass vessel connected to an electric furnace. The pyrolysis was conducted under a steady stream of nitrogen (400 cm3/min) at different temperatures and time intervals. After that, the as-received samples were repeatedly washed with deionized water until approaching the neutral pH value. Finally, the synthesized ACs were dried at 105 °C. The AC yield was quantified using the following equation: 
formula
2
where and are the weight of AC (g) and dry weight of precursor (g), respectively.

Adsorption experiments

Input variables were investigated including activation temperature, IR and activation time. The samples of the synthesized AC were poured into an Erlenmeyer flask containing 50 mL of aqueous solution of Cu2+ 50 ppm. After obtaining absorption equilibrium in 60 minutes, the AC was removed from the mixture using filter paper. The residual ion concentrations were determined by atomic absorption spectroscopy and Cu2+ removal was calculated by the following equation: 
formula
3
where and are initial and final Cu2+ concentrations (ppm), respectively.
In addition, the Cu2+ uptake was calculated by the following equation: 
formula
4
where V is the volume of the Cu2+ solution (mL) and is the weight of AC adsorbent (g).

Instruments and techniques

The X-ray powder diffraction (XRD) of AC was implemented on a D8 Advance Bruker powder diffractometer with a Cu–Kα excitation source. The diffraction spectra were recorded with a scan rate of 0.02°/s. The angle range (2θ) was investigated between 0° and 50°. The morphological study of the material surface was identified by scanning electron microscope (SEM) technique on the instrument S4800, Japan. SEM morphology was recorded utilizing an accelerating voltage source of 10 kV with a magnification of 7,000 and various length equivalents from 5 μm to 1 mm. The Fourier transform infrared (FT-IR) spectra were recorded by using the Nicolet 6700 spectrophotometer instrument. The solid mixture of potassium bromide crystals and AC particles was ground to a fine powder and made into a pellet. The N2 adsorption/desorption isotherm was obtained using the Micromeritics 2020 volumetric adsorption analyzer system; before analysis the AC sample was degassed in a vacuum for 6 h at 150 °C.

Experimental design with RSM

Optimum conditions for the AC yield and removal efficiency of Cu(II) was determined by means of CCD under RSM, which is a collection of mathematical and statistical techniques that are useful for the optimization of chemical reactions and industrial processes (Rozada et al. 2005; Zhang et al. 2005; Ahmad et al. 2007; Kalderis et al. 2008; Hadi et al. 2015; Huang et al. 2015; Yahya et al. 2015). The concept of a response surface method involves several independent variables (xi, i = 1, 2, 3, 4….k) and the response variables (y). A second order model is utilized to establish the true functional relationship between y and the set of independent values ‘xi’. In the present study, the RSM approach was used to assess the reciprocal interaction between two responses (AC yield and Cu2+ removal efficiency) and three of the most significant parameters (activation temperature, IR and activation time), as well as to optimize the relevant conditions of variables to find out the best responses (Muhamad et al. 2013). Among the experimental matrix designs, CCD is one of the most popular techniques of quadratic designs, whereby random experimentation was established for testing a lack of fit without employing a large number of design points (Table 1). The center variables (encoded 0) are utilized to determine the experimental error and the reproducibility of the data. The margin points including the low (encoded −1), high (encoded +1) and rotatable (encoded ±) levels are also manipulated. In addition, the CCD matrix for three independent variables (k = 3) enumerates the 2k factorial experiments, 2 k axial experiments, and six replication experiments as in the following formula: 
formula
5
where N is defined as a total number of experiments for three independent variables (k = 3).
Table 1

Independent variables matrix and their encoded levels

No. Independent factors Code Levels
 
 −1 + 1 +  
Activation temperature (°C)  332 400 500 600 668 
IR (-)  0.16 0.5 1.0 1.5 1.84 
Activation time (min)  9.5 30 60 90 110.5 
No. Independent factors Code Levels
 
 −1 + 1 +  
Activation temperature (°C)  332 400 500 600 668 
IR (-)  0.16 0.5 1.0 1.5 1.84 
Activation time (min)  9.5 30 60 90 110.5 
The response value, metal ions elimination (y), deals with the mathematical correlation between various variables that is approximated by the quadratic polynomial regression equation as given by 
formula
6
where y is the predicted response, and are the independent variables. The parameter is the model constant; is the linear coefficient; is the second-order coefficient and is the interaction coefficient.

Analysis of variance (ANOVA) was calculated using Design-Expert version 9.0.5.1 (DX9). ANOVA of the quadratic polynomial regression model was utilized to identify the significance of input variables and output variables as well as the relationship between the responses and the independent factors.

RESULTS AND DISCUSSION

Development of quadratic regression equation

Based on CCD, the experimental results of AC yield (y1) and Cu(II) removal (y2) are presented in Table 2. The range of the input variables was as follows: activation temperature from 332°C to 668°C, IR from 0.16 to 1.84 and activation time from 9.5 min to 110.5 min. The quadratic equations describing the correlation between the responses and independent variables are given as: 
formula
 
formula
Table 2

Matrix of observed and predicted values

Entry Independent factors (coded)
 
Experiment
 
Prediction
 
x1 (°C) x2 (–) x3 (min) y1 (%) y2 (%) y1 (%) y2 (%) 
400 0.5 30 51.1 40.8 50.0 36.6 
600 0.5 30 42.3 64.2 43.0 66.2 
400 1.5 30 49.0 91.2 49.3 89.1 
600 1.5 30 44.4 41.3 43.3 37.5 
400 0.5 90 46.8 46.8 47.3 49.6 
600 0.5 90 30.3 69.0 29.4 70.1 
400 1.5 90 47.0 88.4 45.7 85.4 
600 1.5 90 28.4 21.6 28.8 24.8 
332 1.0 60 50.3 53.5 50.9 56.9 
10 668 1.0 60 30.5 32.8 30.8 30.8 
11 500 0.16 60 44.0 64.0 44.2 62.5 
12 500 1.84 60 42.5 65.6 43.2 68.5 
13 500 1.0 9.5 48.9 62.0 49.3 66.3 
14 500 1.0 110.5 34.4 69.4 34.9 66.5 
15 500 1.0 60 45.2 82.2 43.8 78.0 
16 500 1.0 60 43.7 81.4 43.8 78.0 
17 500 1.0 60 44.0 76.6 43.8 78.0 
18 500 1.0 60 44.4 74.2 43.8 78.0 
19 500 1.0 60 43.1 74.4 43.8 78.0 
20 500 1.0 60 42.3 79.4 43.8 78.0 
Entry Independent factors (coded)
 
Experiment
 
Prediction
 
x1 (°C) x2 (–) x3 (min) y1 (%) y2 (%) y1 (%) y2 (%) 
400 0.5 30 51.1 40.8 50.0 36.6 
600 0.5 30 42.3 64.2 43.0 66.2 
400 1.5 30 49.0 91.2 49.3 89.1 
600 1.5 30 44.4 41.3 43.3 37.5 
400 0.5 90 46.8 46.8 47.3 49.6 
600 0.5 90 30.3 69.0 29.4 70.1 
400 1.5 90 47.0 88.4 45.7 85.4 
600 1.5 90 28.4 21.6 28.8 24.8 
332 1.0 60 50.3 53.5 50.9 56.9 
10 668 1.0 60 30.5 32.8 30.8 30.8 
11 500 0.16 60 44.0 64.0 44.2 62.5 
12 500 1.84 60 42.5 65.6 43.2 68.5 
13 500 1.0 9.5 48.9 62.0 49.3 66.3 
14 500 1.0 110.5 34.4 69.4 34.9 66.5 
15 500 1.0 60 45.2 82.2 43.8 78.0 
16 500 1.0 60 43.7 81.4 43.8 78.0 
17 500 1.0 60 44.0 76.6 43.8 78.0 
18 500 1.0 60 44.4 74.2 43.8 78.0 
19 500 1.0 60 43.1 74.4 43.8 78.0 
20 500 1.0 60 42.3 79.4 43.8 78.0 

x1: activation temperature, x2: impregnation ratio, x3: activation time.

y1: AC yield, y2: Cu2+ removal.

The data of the ANOVA for regression equations are shown in Table 3. The significance of regression models was assessed by correlation coefficients (R2), and values of P and F. In general, the smaller value of P and the larger value of F indicate statistical significance of the models. A P-value less than 0.05 reveals the statistical significance of a factor effect (at 95% confidence level) (Ghorbani et al. 2008). For both AC yield and Cu(II) removal, the F-values of the two quadratic models were 1.4 and 2.01, respectively, with P < 0.0001 and the adequate precision (AP) ratios greater than 4.0, demonstrating the statistical significance of both models. Lack of fit appeared insignificant with P-values of 0.3616 and 0.2310 for AC yield and Cu2+ removal respectively. Moreover, the R2 and adjusted R2 obtained were higher than 0.97 for AC yield and higher than 0.995 for Cu2+ removal, showing a high goodness of fit (Esfandiar et al. 2014). The graphs of actual versus predicted data (Figure 1) shows the well-fitting, with high correlation coefficients suggesting a high adequacy of the models.
Table 3

ANOVA for response surface quadratic models

Response Source Sum of squares Degree of freedom Mean square F-value Prob > F Comment 
AC yield (%) Model 822.32 91.37 74.65 <0.0001 s SD = 1.11 
 489.95 489.95 400.31 <0.0001 s Mean = 42.63 
 1.31 1.31 1.07 0.3260 n CV = 2.6 
 252.18 252.18 206.05 <0.0001 s Press = 69.23 
 0.55 0.55 0.45 0.5173 n R2 = 0.9853 
 58.86 58.86 48.09 <0.0001 s R2(adj.) = 0.9721 
 0.36 0.36 0.30 0.5988 n AP = 28.199 
 15.41 15.41 12.59 0.0053 s  
 0.010 0.010 0.008 0.9294 n  
 5.05 5.05 4.13 0.0696 n  
Residuals 12.24 10 1.22 – –  
Lack of fit 7.13 1.43 1.4 0.3616 n  
Pure error 5.11 1.02 – –  
Cu2+ removal (%) Model 6686.71 742.97 40.80 <0.0001 s SD = 0.37 
 821.39 821.39 45.10 <0.0001 s Mean = 8.32 
 43.56 43.56 2.39 0.1530 n CV = 4.46 
 0.041 0.041 0.0022 0.9632 n Press = 10.12 
 3292.66 3292.66 180.80 <0.0001 s R2 = 0.9975 
 40.95 40.95 2.25 0.1646 n R2(adj.) = 0.9952 
 138.61 138.61 7.61 0.0202 s AP = 80.416 
 2099.17 2099.17 115.27 <0.0001 s  
 280.86 280.86 15.42 0.0028 s  
 241.83 241.83 13.28 0.0045 s  
Residuals 182.11 10 18.21 – –  
Lack of fit 121.60 24.32 2.01 0.2310 n  
Pure error 60.51 12.10 – –  
Response Source Sum of squares Degree of freedom Mean square F-value Prob > F Comment 
AC yield (%) Model 822.32 91.37 74.65 <0.0001 s SD = 1.11 
 489.95 489.95 400.31 <0.0001 s Mean = 42.63 
 1.31 1.31 1.07 0.3260 n CV = 2.6 
 252.18 252.18 206.05 <0.0001 s Press = 69.23 
 0.55 0.55 0.45 0.5173 n R2 = 0.9853 
 58.86 58.86 48.09 <0.0001 s R2(adj.) = 0.9721 
 0.36 0.36 0.30 0.5988 n AP = 28.199 
 15.41 15.41 12.59 0.0053 s  
 0.010 0.010 0.008 0.9294 n  
 5.05 5.05 4.13 0.0696 n  
Residuals 12.24 10 1.22 – –  
Lack of fit 7.13 1.43 1.4 0.3616 n  
Pure error 5.11 1.02 – –  
Cu2+ removal (%) Model 6686.71 742.97 40.80 <0.0001 s SD = 0.37 
 821.39 821.39 45.10 <0.0001 s Mean = 8.32 
 43.56 43.56 2.39 0.1530 n CV = 4.46 
 0.041 0.041 0.0022 0.9632 n Press = 10.12 
 3292.66 3292.66 180.80 <0.0001 s R2 = 0.9975 
 40.95 40.95 2.25 0.1646 n R2(adj.) = 0.9952 
 138.61 138.61 7.61 0.0202 s AP = 80.416 
 2099.17 2099.17 115.27 <0.0001 s  
 280.86 280.86 15.42 0.0028 s  
 241.83 241.83 13.28 0.0045 s  
Residuals 182.11 10 18.21 – –  
Lack of fit 121.60 24.32 2.01 0.2310 n  
Pure error 60.51 12.10 – –  

Note:s significant at P < 0.05, n insignificant at P > 0.05. SD: standard deviation; CV: coefficient of variation; AP: adequate precision.

Figure 1

Actual (left) versus predicted (right) data of quadratic models.

Figure 1

Actual (left) versus predicted (right) data of quadratic models.

Effect of process variables on AC yield and Cu2+ removal

Figure 2 describes the significant interaction between variables using three-dimensional response surface, in which a couple of factors were investigated at various fractional points and the other was fixed at a center point. At a center value of activation time (60 min), the effect of activation temperature (332°C–668°C) and IR (0.16–1.84) on AC yield and Cu(II) removal is presented in Figure 2(a) and 2(b). Obviously, the AC yield was dominantly influenced by activation temperature while the Cu(II) removal was significantly controlled by both activation temperature and IR. The AC yield showed negligible variation, despite a wide range of IR, on the constant activation temperature line. In contrast, the considerable loss of AC yield (about 20%) was observed when increasing the activation temperature from 332°C to 668°C. Such weight loss with increasing activation temperature can be attributed to the intensified devolatilization of the raw materials, which also induced the development of new pores (Adinata et al. 2007; Ahmad et al. 2009). Note that the Cu(II) removal efficiency increased when the activation temperature of ACs increased in the low range of IR (0.16–0.5) but a reverse trend occurs at higher IR. This is because increasing the amount of ZnCl2 may result in the enhanced evolution of new micropores, thus increasing the active sites available for Cu2+ adsorption (Wen et al. 2011; Sugumaran et al. 2012). Overall, both good AC yield and high Cu(II) removal efficiency can be achieved at the moderate activation temperature (400–500°C).
Figure 2

Surface response plot of AC yield (a) and Cu2+removal (b): effect of activation temperature, IR and activation time of 60 min. Surface response plot of AC yield (c) and Cu2+removal (d): effect of activation temperature, activation time and IR of 1.0. Surface response plot of AC yield (e) and Cu2+removal (f): effect of IR, activation time and activation temperature of 500 °C.

Figure 2

Surface response plot of AC yield (a) and Cu2+removal (b): effect of activation temperature, IR and activation time of 60 min. Surface response plot of AC yield (c) and Cu2+removal (d): effect of activation temperature, activation time and IR of 1.0. Surface response plot of AC yield (e) and Cu2+removal (f): effect of IR, activation time and activation temperature of 500 °C.

The effect of activation time (9.5 min–110.5 min) and activation temperature (332°C–668°C) on Cu2+ removal efficiency is illustrated in Figure 2(c) and 2(d). At the constant IR ratio, increasing both activation temperature and heating temperature affected negatively on the AC yield. As partly mentioned above, the intensified hydration and elimination reactions at higher temperature, and absolute reactions between the activating agent and the precursors at the longer heating time, could mainly contribute to the decrease of AC yield (Valix et al. 2004; Mui et al. 2010). For example, 51% of carbon yield was obtained at 400°C and 30 min of activation time; the increase of activation temperature to 668°C and activation time to 110.5 min reduced the AC yield to 16%. Meanwhile, the Cu2+ adsorption was inconsiderably influenced by the change of activation time of ACs at constant activation temperatures whereas the impact of activation temperature was obviously observed. The optimal removal efficiency was recorded with the activation temperature of around 400°C and activation time of 60 min. Figure 2(e) and 2(f) present the effect of IR (0.16–1.84) and activation time (9.5 min–110.5 min) on the AC yield and Cu2+ removal at the activation temperature of 500°C. Generally, AC yield decreased with increasing activation time at any IR whereas both IR and activation time remarkably affected the removal of Cu2+. The optimized parameters to achieve simultaneously highest AC yield and Cu2+ removal efficiency were determined as follows: activation temperature of 400°C, IR of 1.5 and activation time of 40.6 minutes.

The batch experiments at the optimal conditions were further conducted to verify the suitability of the proposed models. According to the results shown in Table 4, the experimental results for AC yield and Cu2+ removal were obtained as 48.8% and 92.7%, which is close to the predicted values of 49% and 90.8%, respectively. This demonstrates the high compatibility of the models with the experimental data.

Table 4

Model confirmation

Sample T (°C) IR (−) t (min) AC yield (%)
 
Removal (%)
 
Predicted Tested Predicted Tested 
AC400 400 1.5 40.6 49.0 48.8 90.8 92.7 
Sample T (°C) IR (−) t (min) AC yield (%)
 
Removal (%)
 
Predicted Tested Predicted Tested 
AC400 400 1.5 40.6 49.0 48.8 90.8 92.7 

Characterization of the optimized sugarcane AC

In the next step, the AC was synthesized using the predicted optimum conditions, i.e. IR of 1.5 and activation temperature of 400°C for 40 min. The structure of the optimized activated carbon was first characterized using XRD analysis. According to the observation in Figure 3(a), ZnCl2-AC with a broad peak in the 2-theta range of 20°–30° revealed the dominant amorphous structure, which is a typical property of porous AC (Zhao et al. 2009; Isahak et al. 2013). Furthermore, the existence of several sharp peaks may be due to residual ash and ZnCl2 (Ketcha et al. 2012; Ma et al. 2015). To explore the surface chemistry of the ZnCl2-AC, the FT-IR spectrum was recorded in the scanning range of 4,000–400 cm−1 and presented in Figure 3(b). In general, the as-synthesized AC was found to possess a number of important functional groups for the adsorption process. The spectrum revealed a broad band around 3,400 cm−1 ascribed to -OH stretching vibration in hydroxyl groups. The adsorption peaks around 1,685 cm−1 and 1,541 cm−1 indicate the presence of the C=O group in the structure of α,β–unsaturated aldehydes/ketones and O–N asymmetric stretch, respectively. The shape of the peak at 1,616 cm−1 can be attributed either to C=C stretching or to the asymmetric and symmetric stretching vibrations of C-O. Moreover, those detected at 2,343 cm−1 and 2,914 cm−1 are due to the presence of C≡C and C–H, respectively (Yagmur et al. 2008; Chandra et al. 2009; Sugumaran et al. 2012; Billy et al. 2013). Figure 3(c) presents the porous and defect structure of the as-synthesized AC as provided by SEM analysis. As predicted, the removal of non-carbon elements such as hydrogen, oxygen, and nitrogen released from the surface of char during the pyrolysis process resulted in the formation of a rigid carbon skeleton with rudimentary pore structure (Kamaruddin et al. 2011). By measuring the N2 adsorption/desorption isotherm at −196 °C (Figure 3(d) and 3(e)), the BET (Brunauer–Emmett–Teller) surface area of the AC was calculated to be 1,495.57 m2/g with the average pore radius of 8.5 Å and the micropore volume of 0.886 cm3/g. The radius of the AC ranged from around 5 Å to 20 Å, indicating the dominant presence of micropores and mesopores. Compared with other studies (Table 5), the low-temperature protocol applied in this study afforded the AC with a surprisingly high surface area of around 1,500 m2/g at 400 °C–500 °C while the previous studies showed that only the temperature as high as 850 °C afforded the equivalent values. Moreover, the obtained average pore size fluctuated just around 8.4 Å–8.5 Å, which is much lower than other studies (17.0 Å–20.3 Å). It is proposed that the high micropore volume and narrow micropores of the as-synthesized AC made an important contribution to its high adsorption capacity (Williams & Reed 2006; Yahaya et al. 2010). These results prove the success of applying RSM involving CCD to determine the optimal reaction conditions. Thereby, the AC with desired structure and performance could be fabricated with less time and with a significantly reduced number of experiments.
Table 5

Comparison of the textural properties of sugarcane bagasse-derived ZnCl2-ACs

T (°C) IR (−) t (min) SBET (m2/g) vmicro (cm3/g) Pore size (Å) Reference 
400 1.5 40.6 1,495 0.88 8.5 This work 
500 1.5 40.6 1,502 0.86 8.5 This work 
600 1.5 40.6 1,143 0.63 8.4 This work 
500 1.25 30 905 0.33 19.5 Ghorbani et al. (2008)  
600 0.75 30 674 0.35 20.3 Kalderis et al. (2008)  
850 0.6 240 1,997 0.75 17.2 Gurten et al. (2012)  
850 0.8 240 1,617 0.91 17.0 Gurten et al. (2012)  
T (°C) IR (−) t (min) SBET (m2/g) vmicro (cm3/g) Pore size (Å) Reference 
400 1.5 40.6 1,495 0.88 8.5 This work 
500 1.5 40.6 1,502 0.86 8.5 This work 
600 1.5 40.6 1,143 0.63 8.4 This work 
500 1.25 30 905 0.33 19.5 Ghorbani et al. (2008)  
600 0.75 30 674 0.35 20.3 Kalderis et al. (2008)  
850 0.6 240 1,997 0.75 17.2 Gurten et al. (2012)  
850 0.8 240 1,617 0.91 17.0 Gurten et al. (2012)  
Figure 3

(a) XRD spectra, (b) FT-IR spectra, (c) SEM micrograph, (d) N2 adsorption isotherm and (e) pore size distribution of the AC.

Figure 3

(a) XRD spectra, (b) FT-IR spectra, (c) SEM micrograph, (d) N2 adsorption isotherm and (e) pore size distribution of the AC.

CONCLUSION

The present study has successfully optimized the fabrication of ZnCl2-activated AC from sugarcane bagasse for removal of Cu2+ ion from aqueous solution using RSM. The correlative effect of three factors – activation temperature, IR and activation time – on, simultaneously, the AC yield and the removal efficiency of Cu2+ was assessed. According to the ANOVA, the statistical significance of the two models was demonstrated through high F-values with P < 0.0001, correlation coefficients R2 > 0.9 and AP ratios >4.0. The fabrication conditions simultaneously optimized for both AC yield and Cu2+ adsorption were determined as follows: activation temperature of 400 °C, IR of 1.5 and activation time of 40.6 min. The predicted optimum AC yield and Cu2+ removal were 48.8% and 92.7% respectively, which were also confirmed by verification experiments. The optimized AC was found to offer a number of advantages such as high surface area but low fabrication temperature, large pore volume, narrow micropores and enrichment of surface functional groups. It is suggested that the low-cost AC synthesized from sugarcane bagasse using the RSM-derived protocol holds a great potential for application in removal of environmental pollutants.

ACKNOWLEDGEMENTS

This research is funded by Foundation for Science and Technology Development, Nguyen Tat Thanh University, Ho Chi Minh City, Vietnam; and a grant by the Korea Institute of Science and Technology (KIST) Institutional Program (Project No. 2Z04820-16-090), Seoul, Korea.

REFERENCES

REFERENCES
Ahmad
A. A.
Hameed
B. H.
Ahmad
A. L.
2009
Removal of disperse dye from aqueous solution using waste-derived activated carbon: optimization study
.
Journal of Hazardous Materials
170
(
2–3
),
612
619
.
Billy
T. H.
Guan
P. A. L.
Yap
T. Y. H.
2013
Physical preparation of activated carbon from sugarcane bagasse and corn husk and its physical and chemical characteristics
.
International Journal of Engineering Research and Science & Technology
2
(
3
),
1
14
.
Chandra
T. C.
Mirna
M. M.
Sunarso
J.
Sudaryanto
Y.
Ismadji
S.
2009
Activated carbon from durian shell: preparation and characterization
.
Journal of the Taiwan Institute of Chemical Engineers
40
(
4
),
457
462
.
Ghorbani
F.
Younesi
H.
Ghasempouri
S. M.
Zinatizadeh
A. A.
Amini
M.
Daneshi
A.
2008
Application of response surface methodology for optimization of cadmium biosorption in an aqueous solution by Saccharomyces cerevisiae
.
Chemical Engineering Journal
145
(
2
),
267
275
.
Gurten
I. I.
Ozmak
M.
Yagmur
E.
Aktas
Z.
2012
Preparation and characterisation of activated carbon from waste tea using, K2CO3
.
Biomass and Bioenergy
37
,
73
81
.
Huang
Y. P.
Hou
C. H.
Hsi
H. C.
Wu
J. W.
2015
Optimization of highly microporous activated carbon preparation from Moso bamboo using central composite design approach
.
Journal of the Taiwan Institute of Chemical Engineers
50
,
266
275
.
Isahak
W. N. R. W.
Hisham
M. W. M.
Yarmo
M. A.
2013
Highly porous carbon materials from biomass by chemical and carbonization method: a comparison study
.
Journal of Chemistry
2013
,
1
6
.
Kamaruddin
M. A.
Yusoff
M. S.
Ahmad
M. A.
2011
Optimization of durian peel based activated carbon preparation conditions for ammoniacal nitrogen removal from semi-aerobic landfill leachate
.
Journal of Scientific and Industrial Research
70
(
7
),
554
560
.
Ketcha
J. M.
Dina
D. J. D.
Ngomo
H. M.
Ndi
N. J.
2012
Preparation and characterization of activated carbons obtained from maize cobs by zinc chloride activation
.
American Chemical Science Journal
2
(
4
),
136
160
.
Mui
E. L. K.
Cheung
W. H.
Valix
M.
McKay
G.
2010
Activated carbons from bamboo scaffolding using acid activation
.
Separation and Purification Technology
74
(
2
),
213
218
.
Rozada
F.
Otero
M.
Parra
J. B.
Moran
A.
Garcia
A. I.
2005
Producing adsorbents from sewage sludge and discarded tyres: characterization and utilization for the removal of pollutants from water
.
Chemical Engineering Journal
114
(
1–3
),
161
169
.
Sugumaran
P.
Susan
V. P.
Ravichandran
P.
Seshadri
S.
2012
Production and characterization of activated carbon from banana empty fruit bunch and Delonix regia fruit pod
.
Journal of Sustainable Energy & Environment
3
,
125
132
.
Tsai
W. T.
Chang
C. Y.
Lin
M. C.
Chien
S. F.
Sun
H. F.
Hsieh
M. F.
2001
Characterization of activated carbons prepared from sugarcane bagasse by ZnCl2 activation
.
Journal of Environmental Science and Health, Part B: Pesticides, Food Contaminants, and Agricultural Wastes
36
(
3
),
365
378
.
Wen
Q.
Li
C.
Cai
Z.
Zhang
W.
Gao
H.
Chen
L.
Zeng
G.
Shua
X.
Zhao
Y.
2011
Study on activated carbon derived from sewage sludge for adsorption of gaseous formaldehyde
.
Bioresource Technology
102
(
2
),
942
947
.
Yahaya
N. K. E. M.
Pakir
M. F. M. L.
Abustan
I.
Ahmad
M. A.
2010
Effect of preparation conditions of activated carbon prepared from rice husk by ZnCl2 activation for removal of Cu(II) from aqueous solution
.
International Journal of Engineering & Technology IJET-IJENS
10
(
6
),
2
6
.
Yahya
M. A.
Al-Qodah
Z.
Ngah
C. W. Z.
2015
Agricultural bio-waste materials as potential sustainable precursors used for activated carbon production: a review
.
Renewable and Sustainable Energy Reviews
46
,
218
235
.