A comparative study on the removal efficiency of metal ions (Cu 2+,Ni 2+,and Pb 2+ ) using sugarcane bagasse-derived ZnCl 2 -activated carbon by the response surface methodology

Abstract

In the present article, the optimization of removal of Cu²⁺, Ni²⁺, and Pb²⁺ using sugarcane-based activated carbon was studied. The activated carbon synthesized from low-cost sugarcane bagasse via ZnCl₂ activation process was found to possess surprisingly high surface area (>1500 m²/g). In adsorption study, the individual and interactive effects of three critical parameters including initial concentration, pH of solution, and activated carbon dosage on removal efficiency of Cu²⁺, Ni²⁺, and Pb²⁺ were assessed by applying the response surface methodology combined with central composite design. The established second-order polynomial regression models provided an excellent interpretation of experimental data with acceptable coefficients of determination (R²) and analysis of variance demonstrated statistical significance of the model terms. In the examined region, the predicted maximum removal efficiency and adsorption capacity were found in the order: Ni²⁺ (66.4%, 2.99 mg/g) < Cu²⁺ (90%, 13.24 mg/g) < Pb²⁺ (99.9%, 19.3 mg/g), which were also confirmed in the verification experiments.

Keywords

Removal of metal ions sugarcane bagasse response surface methodology activated carbon adsorption isotherm

Introduction

Heavy metals have been demonstrated to be harmful for both human and environmental health owing to their acute toxicity and persistence in nature. Although the presence of trace amounts of several metals is beneficial for biological transportation in micro-organisms, excessive intakes of heavy metals, such as copper, nickel, and lead, may cause serious health issues such as neurological damage, paralysis, blindness, and chromosome breakage (Wasi et al., 2013). According to the World Health Organization (WHO), the permitted concentration limits of copper, nickel, and lead in drinking water are 1.0, 0.02, and 0.05 mg/L, respectively (De Zuane, 2007). Thus, the removal of toxic metal ions from contaminated water attracts particular concern from the scientists and governmentalists all over the world.

Many methods have been attempted to eliminate heavy metals from aqueous solution, such as coagulation (Charerntanyarak, 1999), filtration with coagulation (Al-Abri et al., 2010), precipitation (Matlock et al., 2002), ozonation (Prieto-Rodríguez et al., 2013), adsorption, ion exchange (Da¸browski et al., 2004), reverse osmosis (Cath et al., 2006), and advanced oxidation (Barakat, 2011). Among those, the adsorption of heavy metals using activated carbon (AC) holds a number of important advantages, such as high adsorption capacity and efficient recovery of heavy metals, good selectivity, sludge free operation, and cost effectiveness (Charerntanyarak, 1999; Kadirvelu et al., 2001; Park et al., 2007). However, the fabrication of commercial activated from costly and non-renewable precursors such as petroleum residues, wood, coal, peat, and lignite carbon has restricted its applications (Marsh and Rodríguez-Reinoso, 2006). In recent years, the use of abundant agricultural wastes as a renewable and inexpensive alternative source for the synthesis of ACs has been developed (Ioannidou and Zabaniotou, 2007). The sugarcane (saccharum officinarum) is one of the most popular tropical species widely cultivated in tropical countries (Brazil, India, etc.) and makes up a large proportion of the sugar industries in the world. The extraction of each ton of milled sugarcane releases 180–280 kg of sugarcane bagasse (Farahani et al., 2011). Similar to other agricultural wastes, the low-cost sugarcane bagasse mainly consists of carbonaceous constituents, i.e. 42% cellulose, 25% hemicellulose, and 20% lignin, thus being a great resource for fabrication of ACs (Amin, 2008; Farahani et al., 2011). In recent years, there have been a number of reports on the fabrication of cheap and efficient powder and granular AC from sugarcane bagasse and molasse via different chemical activation methods. The resulting ACs possessed high surface area (>1000 m²/g) with well nanoporous structure as promising for many important applications ranging from wastewater treatment, water purification to electrical technology (G da C Gonçalves et al., 2016; SPC Gonçalves et al., 2016; Jain and Tripathi, 2015; Sreńscek-Nazzal et al., 2013; Tao et al., 2015).

ACs are typically fabricated from carbonaceous materials by oxidative reaction of carbon atoms using the pyrolysis and/or chemical activation (Ioannidou and Zabaniotou, 2007). Structurally, adsorptive properties of ACs are determined not only by inherent features of material sources but also by the production methods. Two major methods are frequently utilized to synthesize the ACs consisting of physical and chemical activation. Physical activation is conducted in the presence of CO₂ or steam followed by gasification at high temperatures (700–900℃) (Sun and Jiang, 2010). In chemical activation, initial precursor or its char residue is soaked with strong dehydrating reagents, such as KOH and ZnCl₂, which is subsequently pyrolysed to facilitate the formation of the porous ACs (Hayashi et al., 2000). Although dehydrating reagent ZnCl₂ is found to be environmentally less advantageous than the other chemical activators, ZnCl₂ is still used as a very common chemical in the preparation of ACs (Caturla et al., 1991; Olivares-Marín et al., 2006). The activation using ZnCl₂ has been reported to offer the ACs with distinct characteristics like very high surface area (possibly up to 3000 m²/g) and enrichment of surface functional groups (Caturla et al., 1991). Owing to such advantages, the excellent performance of ZnCl₂-ACs in the elimination of toxic pollutants such as dyes, heavy metal ions have been found (Baccar et al., 2009; Caturla et al., 1991; Hsu and Teng, 2000; Lo et al., 2012; Olivares-Marín et al., 2006).

The present study focuses on the developing of potentially cost-effective and efficient sugarcane bagasse-derived AC for the removal of heavy metal ions (Cu²⁺, Ni²⁺, and Pb²⁺) using response surface methodology (RSM) and central composite design (CCD) as a strong mathematical tool for optimization of the adsorption process. In the fabrication process, the chemical activator ZnCl₂ was applied to form porous carbon structure as well as create and diverse functional groups on the surface. The removal tests were performed with three common heavy metals including Cu²⁺, Ni²⁺, and Pb²⁺ ions. The quadratic polynomial regression models involving three input variables, i.e. concentration of heavy metal ions, adsorbent dosage, pH, and taking adsorption capacity of the synthesize-AC as the main response were successfully developed. The goodness of fit of the model and the significance of model terms were assessed via the analysis of variance (ANOVA) and coefficient of regression (R²).

Experimental procedure

Fabrication of AC from sugarcane bagasse

Vietnamese sugarcane bagasse was collected and washed with distilled water for several times to remove all dust and residual sugar. Then, the precursor was dried under the sunshine and ground to the diameters of approximately 1.0 mm. The dried materials were soaked with a ZnCl₂ solution (ZnCl₂: precursor = 1.5:1 by wt.%) for 24 h. The ZnCl₂-impregnated bagasse was heated under nitrogen at 500℃ for 1 h. Finally, the receiving ZnCl₂-AC was repeatedly washed with deionized water until approaching a neutral solution and dried at 105℃ for 24 h.

Determination of pH point of zero charges (pHzpc)

To determine pH_zpc of sugarcane bagasse-derived AC, the pH value of KCl (0.1 mM.L⁻¹) solution was adjusted (pH = 2–12) using NaOH and HCl 1 M. First, 0.25 g of AC was added to flasks containing 50 mL of 0.1 ml.L⁻¹ KCl. The solutions were kept stably and their final pH was measured by using a pH meter. The curve was plotted via final pH against the initial pH. The pH point of zero charges was calculated at which the curve crosses pH initial = pH final.

Batch adsorption experiments

Adsorption experiments were carried out in the Erlenmeyer flasks containing 50 mL of aqueous solution of M (II) ions (M = Cu, Pb, and Ni) and a fixed amount of the synthesized sugarcane bagasse AC. Input factors were investigated including initial M (II) concentration (8.0–92.0 mg/L), adsorbent dosage (0.8–9.2 g/L), and pH of the solution (0.6–7.4). The mixture was mixed thoroughly until obtaining the adsorption equilibrium. The residual concentrations were confirmed by AAS analysis. The removal efficiency and the adsorption capacity were calculated as follows:

M (II) removal (%) = \frac{C_{o} - C_{f}}{C_{o}} . 100

(1)

where C_o and C_f are the initial and equilibrium M (II) concentrations (mg/L), respectively.

q_{f} (mg / g) = \frac{C_{o} - C_{f}}{W} . V = \frac{C_{o} - C_{f}}{d_{AC}}

(2)

where V (mL) is the volume of the M (II) solution, W (g) is the weight of the adsorbent, and d_AC (g/L) is the dosage of the AC.

Experimental design with RSM

The optimization of the removal efficiency of Cu²⁺, Ni²⁺, and Pb²⁺ by sugarcane bagasse AC was determined using RSM involving CCD as a useful mathematical and statistical technique in evaluating the correlation between independent (x) and response (y) variables through experimental runs (Bezerra et al., 2008; Hanrahan and Lu, 2006; Lo et al., 2012). In the present study, initial concentration, adsorbent dosage, and pH of solution were taken as three independent variables (x_i) and the percentage removal of heavy metals was the main response (y). The mathematical relationship between the main response “y” and the set of independent values “x” is established according to the following second-order model:

y = f (x) = β_{o} + \sum_{i = 1}^{k} β_{i} x_{i} + \sum_{i = 1}^{k} \sum_{j = 1}^{k} β_{ij} x_{i} x_{j} + \sum_{i = 1}^{k} β_{ii} x_{i}^{2}

(3)

where y is the predicted response; x_i and x_j are the independent variables (i,j = 1, 2, 3, 4….k). The parameter β_o is the model constant; β_i is the linear coefficient; β_ii is the second–order coefficient; and β_ij is the interaction coefficient. Among the experimental matrix designs, CCD is one of the most popular techniques of quadratic designs, which 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 include the low (encoded −1), high (encoded +1), and rotatable (encoded ±α) levels. The CCD matrix for three independent variables (k = 3) enumerates the 2 ^k factorial experiments, 2k axial experiments, and six replication experiments with the total number of tests calculated as:

N = 2^{k} + 2 k + c = 2^{3} + 2.3 + 6 = 20

(4)

where N is the number of experiments required for three independent variables (k = 3). ANOVA was calculated using Design–Expert version 9.0.5.1 (DX9). ANOVA of the quadratic polynomial regression model was utilized to identify the signification of model variables and the relationship between the responses and influential factors.

Table 1.

Independent variables matrix and their encoded levels.

No	Independent factors	Code	Levels
No	Independent factors	Code	−α	−1	0	+1	+α
1	Initial concentration (mg/L)	x ₁	8.0	25	50	75	92.0
2	Adsorbent dosage (g/L)	x ₁	0.8	2.5	5	7.5	9.2
3	pH of solution (–)	x ₁	0.6	2	4	6	7.4

Instruments and techniques

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 7000 and various length equivalents from 5 µm to 1 mm. The solid mixture of potassium bromide crystals and AC particles were ground to fine powder and made a pellet. The N2 adsorption/desorption isotherm was obtained using the Micromeritics 2020 volumetric adsorption analyzer system. BET surface area was measured using the isotherm equation.

Results and discussion

Characterization of the AC

The AC was fabricated from the sugarcane bagasse using ZnCl₂ as the activating reagent following the protocol developed in our previous study. The detailed investigations about the effect of fabrication conditions as well as surface and structure properties have been already reported (Tran et al., 2016). Accordingly, the activation temperature of 500℃ was selected since it was able to provide the AC with high surface area and preferable surface chemistry. The typical characteristics of the synthesized AC are listed in Table 1. By measuring N₂ adsorption/desorption isotherm at −196℃, the BET surface area of the AC was calculated to be 1502.0 m²/g with the average pore radius at 0.85 nm and the micro–pore volume of 0.886 cm³/g (Tran et al., 2016). In comparison with other studies (Kalderis et al., 2008; Suescún-Mathieu et al., 2014; Tsai et al., 2001), the surprisingly high surface area of 1500 m²/g, which was reported to be only attainable at the temperature as high as 850℃, can be obtained at the moderate activation temperature (500℃) using this fabrication protocol. Moreover, the average pore size (0.85 nm) appeared lower than those in other studies (1.70–2.03 nm) (Kalderis et al., 2008; Suescún-Mathieu et al., 2014; Tsai et al., 2001), proposing that high micropore volume and narrow micropores of this AC could possibly support high adsorption capacities (Caturla et al., 1991; Olivares-Marín et al., 2006; Tongpoothorn et al., 2011).

The SEM photograph in Figure 1 shows that the AC exposes a rough texture surface, where a variety of pore sizes are randomly distributed. The formation of porous and defect structure carbon can be explained due to the release of non-carbon elements such as hydrogen, oxygen, and nitrogen from the surface of char during pyrolysis process to form rigid carbon skeleton with rudimentary pore structure. The pH point of zero charge of the AC was determined to be 6.6 according to the results shown in Figure 2 and Table 2.

Figure 1.

SEM micrograph of the AC.

Figure 2.

Measurement of pH_zpc: the initial versus final pH plot.

Table 2.

Properties of the AC.

S_BET (m²/g)	Average pore radius (nm)	Micropore volume (cm³/g)	pH_zpc
1502.0	0.85	0.886	6.6

Assessment of experimental results with Design–Expert

Twenty batch runs were performed according to the CCD design to visualize the effects of three independent influential factors on the percentage removal. The adsorption experiments were conducted in the initial concentration range of 8.0–92.0 mg/L, the adsorbent dosage of 0.8–9.2 g/L, and pH values of 0.6–7.4. The obtained experimental and predicted results for the percentage removal of Cu²⁺, Ni²⁺, and Pb²⁺ are displayed in Table 3. By applying RSM for the quadratic regression models, the empirical relationships between the removal efficiencies of Cu²⁺, Ni²⁺, and Pb²⁺, and three independent variables were described as following equations:

{Cu}^{2 +} removal (%) = 65.1 - 2.56 x_{1} + 10.68 x_{2} + 28.35 x_{3} - 3.49 x_{1} x_{2} + 4.64 x_{1} x_{3} - 6.51 x_{2} x_{3} + 0.41 x_{1}^{2} - 8.12 x_{2}^{2} - 5.41 x_{3}^{2}

(5)

{Ni}^{2 +} removal (%) = 65.1 - 2.56 x_{1} + 10.68 x_{2} + 28.35 x_{3} - 3.49 x_{1} x_{2} + 4.64 x_{1} x_{3} - 6.51 x_{2} x_{3} + 0.41 x_{1}^{2} - 8.12 x_{2}^{2} - 5.41 x_{3}^{2}

(6)

{Pb}^{2 +} removal (%) = 94.6 - 11.51 x_{1} + 12.31 x_{2} + 27.0 x_{3} - 6.04 x_{1} x_{2} + 7.54 x_{1} x_{3} - 13.01 x_{2} x_{3} - 7.22 x_{1}^{2} - 8.30 x_{2}^{2} - 17.81 x_{3}^{2}

(7)

Table 3.

Matrix of observed and predicted values.

Run	Independent factors			Experimental of metal removal (%)			Prediction of metal removal (%)
Run	x ₁	x ₂	x ₃	Cu²⁺	Ni²⁺	Pb²⁺	Cu²⁺	Ni²⁺	Pb²⁺
1	25	2.5	2	7.2	13.6	21.4	10.1	17.7	21.9
2	75	2.5	2	6.3	6.3	0.8	2.7	6.4	−4.1
3	25	7.5	2	47.4	24.0	84.6	51.5	24.1	84.7
4	75	7.5	2	28.0	15.2	27.9	30.1	16.8	34.5
5	25	2.5	6	72.0	46.8	96.4	70.6	51.0	86.9
6	75	2.5	6	85.1	26.4	94.0	81.7	32.1	91.0
7	25	7.5	6	81.6	50.0	95.6	85.9	55.7	97.6
8	75	7.5	6	85.3	39.2	81.0	83.1	40.8	77.5
9	8	5	4	76.1	55.5	90.8	70.5	49.9	93.5
10	92	5	4	57.4	30.4	53.4	61.9	27.9	54.8
11	50	0.8	4	20.6	20.8	41.8	24.2	15.2	50.4
12	50	9.2	4	64.7	30.4	96.3	60.1	27.9	91.8
13	50	5	0.6	5.1	9.6	1.6	2.1	8.9	−1.2
14	50	5	7.4	95.5	64.4	82.7	97.5	57.0	89.6
15	50	5	4	61.8	43.4	91.1	65.1	40.5	94.6
16	50	5	4	61.6	39.6	97.5	65.1	40.5	94.6
17	50	5	4	63.2	37.8	93.2	65.1	40.5	94.6
18	50	5	4	63.6	37.6	97.8	65.1	40.5	94.6
19	50	5	4	67.9	40.4	94.5	65.1	40.5	94.6
20	50	5	4	72.3	42.8	94.1	65.1	40.5	94.6

The statistical significance of the quadratic models are evaluated by the F-test ANOVA and P-values. In detail, if the P-values are less than 0.05 (95% confidence level), the influence of variables on the response is statistically significant. Furthermore, if the model fits experimental data well, lack of fit (LOF) value appears insignificant. ANOVA is assessed and exhibited in Table 4. Obviously, the large F-values of all three models with low probability value (P < 0.0001) suggest the statistical significance of the modeled regression models. The goodness of fit of the proposed models is evaluated via the correlation coefficients (R²). The more R² value approaches 1.0, the higher accuracy of the obtained models is obtained. Both the R² and adjusted R² were higher than 0.96 for Cu²⁺, Pb²⁺ while those for Ni²⁺ were lower, at 0.9424 and 0.8906, respectively. Considering LOF, the LOF appears insignificant with the model for Cu²⁺. However, the significant F-values with Prob. > F lower than 0.05 in the cases of Ni²⁺ and Pb²⁺ could possibly be resulted from the involvement of other influential factors. The graphs of experimental versus predicted data show a well-fitting, which demonstrates a high adequacy of the models (Figure 3).

Table 4.

ANOVA for response surface quadratic models.

Response	Source	Sum of squares	Degree of freedom	Mean square	F-value	Prob. > F	Comment
Cu²⁺ removal (%)	Model	14,527.07	9	1614.12	60.57	<0.0001^a	SD = 5.16
	x ₁	89.44	1	89.44	3.36	0.0969^b	Mean = 56.14
	x ₂	1557.99	1	1557.99	58.46	<0.0001^a	CV = 9.2
	x ₃	10,974.18	1	10,974.18	411.80	<0.0001^a	Press = 1494.63
	$x_{1}^{2}$	97.30	1	97.30	3.65	0.0851^b	R²= 0.9820
	$x_{2}^{2}$	172.05	1	172.05	6.46	0.0293^a	R²_(adj.) = 0.9658
	$x_{3}^{2}$	339.30	1	339.30	12.73	0.0051^a	AP = 26.121
	$x_{1} x_{2}$	2.36	1	2.36	0.089	0.7722^b
	$x_{1} x_{3}$	949.28	1	949.28	35.62	0.0001^a
	$x_{2} x_{3}$	422.00	1	422.00	15.84	0.0026^a
	Residuals	266.49	10	26.65
	Lack of fit	177.82	5	35.56	2.01	0.2317^b
	Pure error	88.67	5	17.73
Ni²⁺ removal (%)	Model	4323.16	9	480.35	18.19	<0.0001^a	SD = 5.14
	x ₁	586.71	1	586.71	22.21	0.000^a	Mean = 33.71
	x ₂	193.79	1	193.79	7.34	0.0220^a	CV = 15.25
	x ₃	2797.53	1	2797.53	105.91	<0.0001^a	Press = 1838.49
	$x_{1}^{2}$	8.20	1	8.20	0.31	0.5896^b	R²= 0.9424
	$x_{2}^{2}$	28.50	1	28.50	1.08	0.3234^b	R²_(adj.) = 0.8906
	$x_{3}^{2}$	1.36	1	1.36	0.052	0.8250^b	AP = 13.931
	$x_{1} x_{2}$	4.65	1	4.65	0.18	0.6837^b
	$x_{1} x_{3}$	647.34	1	647.34	24.51	0.0006^a
	$x_{2} x_{3}$	102.87	1	102.87	3.89	0.0767^b
	Residuals	264.14	10	26.41
	Lack of fit	234.24	5	46.85	7.84	0.0206^a
	Pure Error	29.89	5	5.98
Pb²⁺ removal (%)	Model	21,472.41	9	2385.82	63.61	<0.0001^a	SD = 6.12
	x ₁	1809.46	1	1809.46	48.25	<0.0001^a	Mean = 71.83
	x ₂	2070.54	1	2070.54	55.21	<0.0001^a	CV = 8.53
	x ₃	9953.60	1	9953.60	265.40	<0.0001^a	Press = 2755.76
	$x_{1}^{2}$	291.61	1	291.61	7.78	0.0192^a	R²= 0.9828
	$x_{2}^{2}$	454.51	1	454.51	12.12	0.0059^a	R²_(adj.) = 0.9674
	$x_{3}^{2}$	1354.60	1	1354.60	36.12	0.0001^a	AP = 23.472
	$x_{1} x_{2}$	751.22	1	751.22	20.03	0.0012^a
	$x_{1} x_{3}$	992.38	1	992.38	26.46	0.0004^a
	$x_{2} x_{3}$	4570.61	1	4570.61	121.87	<0.0001^a
	Residuals	375.05	10	37.50
	Lack of fit	341.99	5	68.40	10.34	0.0114^a
	Pure error	33.06	5	6.61

Note: x₁,x₂, and x₃ are the main factors. $x_{1}^{2}$ , $x_{2}^{2}$ , and $x_{3}^{2}$ are the square factors. $x_{1} x_{2}$ , $x_{1} x_{3}$ , and $x_{2} x_{3}$ are the interaction factors.

Significant at p < 0.05.

Insignificant at p > 0.05.

Figure 3.

Experimental versus predicted plot of regression models for the removal of Cu²⁺ (a), Ni²⁺ (b), and Pb²⁺ (c).

Table 5.

Model confirmation.

Sample	Concentration (mg/L)	Dosage (g/L)	pH	Removal (%)		Capacity (mg/g)
Sample	Concentration (mg/L)	Dosage (g/L)	pH	Predicted	Tested	Predicted	Tested
AC–Cu	75.0	5.1	6.0	90.5	90.0	13.31	13.24
AC–Ni	22.5	5.0	6.8	65.0	66.4	2.93	2.99
AC–Pb	65.7	3.4	6.5	98.0	99.9	18.94	19.30

Optimization of the removal efficiency of Cu²⁺, Ni²⁺, and Pb²⁺ using the RSM

The surface responses of the quadratic polynomial models were described using three-dimensional (3D) curve plots, in which one variable was maintained at the zero levels while the others fluctuated in the examined ranges. The effects of input variables (initial concentration, adsorbent dosage, and pH of solution) on the removal efficiency of metal ions (Cu²⁺, Ni²⁺, and Pb²⁺) can be observed in Figures 4 –6 and Table 5. In detail, Figure 4 shows the obtained removal percentages while varying the two factors, initial concentration (892 mg/L) and adsorbent dosage (0.8–9.2 mg/L), at the constant value of pH 4.0. According to the observation in Figure 4(a) and (b), the effect of initial concentration on the removal efficiency of Cu²⁺ and Ni²⁺ was unremarkable while the adsorbent dosage plays a more controlling role. For example, the removal of metal ions increased to the maximum percentage removal of approximately 80% for Ni²⁺ and 50% for Cu²⁺ when increasing the AC dosage from 0.8 to 7.5 g/L and from 0.8 to 5 g/L, respectively. Further addition of adsorbent dosage resulted in drops of removal efficiencies. Both two variables including initial concentration and AC dosage affected strongly the removal of Pb²⁺ from aqueous solution. On the other hand, based on the results shown in Figure 6(c), up to 100% removal was obtained in the low initial concentration range (8–25 mg/L) with high values of adsorbent dosage (75–92 g/L).The former enhancement in adsorption can be explained due to the decrease of metal density in the solution while the latter likely resulted from the increased number of active sites as the adsorbent dosage increased.

Figure 4.

3D plots of the removal of Cu²⁺ (a), Ni²⁺ (b), and Pb²⁺ (c): Effect of initial concentration and AC dosage at pH 4.

The effects of initial concentration (8–92 mg/L) and pH of the solution (0.6–7.4) on the removal efficiency for Cu²⁺, Ni²⁺, and Pb²⁺ ions at the adsorbent dosage of 5.0 g/L are illustrated in Figure 5. The removal of metal ions hardly occurred in strongly acidic solution (pH < 2) regardless of initial concentration whereas a neutral environment (pH = 5.5–7.4) facilitated the adsorption of Cu²⁺, Ni²⁺, and Pb²⁺ on the AC surface. Accordingly, the optimal pH values were found at 6.0, 6.8, and 6.5 for Cu²⁺, Ni²⁺, and Pb²⁺, respectively. The adsorption of metal ions tents to increase with increasing pH as resulted from a decrease in electrostatic repulsion between the cations and the positively charged surface of the AC. As pH decreased, the more competition between H⁺ ions in solution and metal ions at the surface adsorption sites caused a decrease in adsorption of metal ions. In this study, the optimal pH-values for the three metals were reasonably close to the pH_zpc value (6.6).

Figure 5.

3D plots of the removal of Cu²⁺ (a), Ni²⁺ (b), and Pb²⁺ (c): Effect of initial concentration, pH of the solution, and at the AC dosage of 5 g/L.

Figure 6.

3D plots of the removal of Cu²⁺ (a), Ni²⁺ (b), and Pb²⁺ (c): Effect of AC dosage, pH of the solution, and at the initial concentration of 50 mg/L.

Note that effect of initial concentration in the optimal pH range was quite divergent. For instance, the highest removal percentage of Cu²⁺ (approximately 100%) was achieved at the concentration as high as 92 mg/L while that appeared at low concentrations (<24.8 mg/L) and in a wide intermediate concentration range (25–75 mg/L) for Ni²⁺ and Pb²⁺, respectively.

The effects of adsorbent dosage (0.8–9.2 g/L) and pH of the solution (0.6–7.4) on the removal efficiency of Cu²⁺, Ni²⁺, and Pb²⁺ ions at the constant initial concentration of 50 mg/L are shown in Figure 6. Again, the inefficient removal of metal ions at the low pH values (0.8–2.0) was observed in spite of the increase in AC dosage. The optimal pH value appeared to depend on the range. However, in the higher pH range (5.0–7.0), the change of pH value has more significant effect in the presence of low AC dosage; the larger amount of adsorbent led to better removal efficiency of metal ions at a higher value of pH (5.0–7.0).

According to the constructed model, the predicted maximum removal efficiencies follow the order: Ni²⁺ (65%) < Cu²⁺ (90%) < Pb²⁺ (99.9%). In verification experiments, the values obtained at optimum conditions were 66.4%, 90.0%, and 99.9% for Ni²⁺, Cu²⁺, and Pb²⁺, respectively, which approached the predicted values, indicating the suitability of the suggested models. In comparison, the order of adsorption capacity of the sugarcane bagasse-derived AC for the three metals: Ni²⁺ (2.99 mg/g) < Cu²⁺ (13.24 mg/g) < Pb²⁺ (19.3 mg/g) is a good agreement with the previous publications (Faur-Brasquet et al., 2002; Kadirvelu et al., 2000; Kongsuwan et al., 2009).

Conclusion

The results of this study show that the AC produced from sugarcane bagasse by the suggested ZnCl₂-activation protocol is an efficient adsorbent for Cu²⁺, Ni²⁺, and Pb²⁺ ions. By applying RSM involving CCD, the influence of three independent variables, including metal ion concentration, pH, and AC dosage on the removal efficiency of Cu²⁺, Ni²⁺, and Pb²⁺ was interpreted via the quadratic regression equations with sufficient statistical significance. Accordingly, the maximum adsorption capacities and removal efficiencies for Cu²⁺, Ni²⁺, and Pb²⁺ were obtained, respectively, at a concentration of 75.0 mg/L, 22.5 mg/L, and 65.7 mg/L; AC dosage of 5.1 g/L, 5.0 g/L, and 3.4 g/L; and pH values of 6.0, 6.8, and 6.5. Based on the experimental and modeling data obtained in the present study, the potentially cost-effective sugarcane bagasse-derived ZnCl₂-AC holds a great promise for application in the elimination of hazardous metal ions from industrial effluents.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is funded by Foundation for Science and Technology Development Nguyen Tat Thanh University, Ho Chi Minh City, Vietnam.

References

Al-Abri

Dakheel

Tizaoui

(2010) Combined humic substance and heavy metals coagulation, and membrane filtration under saline conditions. Desalination 253(1–3): 46–50.

Amin

(2008) Removal of reactive dye from aqueous solutions by adsorption onto activated carbons prepared from sugarcane bagasse pith. Desalination 223(1–3): 152–161.

Baccar

Bouzid

Feki

(2009) Preparation of activated carbon from Tunisian olive-waste cakes and its application for adsorption of heavy metal ions. Journal of Hazardous Materials 162(2–3): 1522–1529.

Barakat

(2011) New trends in removing heavy metals from industrial wastewater. Arabian Journal of Chemistry 4(4): 361–377.

Bezerra

Santelli

Oliveira

(2008) Response surface methodology (RSM) as a tool for optimization in analytical chemistry. Talanta 76(5): 965–977.

Caturla

Molina-Sabio

Rodríguez-Reinoso

(1991) Preparation of activated carbon by chemical activation with ZnCl2. Carbon 29(7): 999–1007.

Cath

Childress

Elimelech

(2006) Forward osmosis: Principles, applications, and recent developments. Journal of Membrane Science 281(1–2): 70–87.

Charerntanyarak

(1999) Heavy metals removal by chemical coagulation and precipitation. Water Science and Technology 39(10–11): 135–138.

Da¸browski

Hubicki

Podkościelny

(2004) Selective removal of the heavy metal ions from waters and industrial wastewaters by ion-exchange method. Chemosphere 56(2): 91–106.

10.

De Zuane

(2007) Chemical parameters—Inorganics. In: Handbook of Drinking Water Quality, Hoboken, NJ: John Wiley & Sons, Inc., pp. 49–148.

11.

Farahani

Abdullah

S R S

Hosseini

(2011) Adsorption-based cationic dyes using the carbon active sugarcane bagasse. Procedia Environmental Sciences 10(Part A). 203–208.

12.

Faur-Brasquet

Reddad

Kadirvelu

(2002) Modeling the adsorption of metal ions (Cu²⁺, Ni²⁺, Pb²⁺) onto ACCs using surface complexation models. Applied Surface Science 196(1–4): 356–365.

13.

Gonçalves

Pereira

Veit

(2016) Production of bio-oil and activated carbon from sugarcane bagasse and molasses. Biomass and Bioenergy 85: 178–186.

14.

Gonçalves

S P C

Strauss

Delite

(2016) Activated carbon from pyrolysed sugarcane bagasse: Silver nanoparticle modification and ecotoxicity assessment. The Science of the Total Environment 565: 833–840.

15.

Hanrahan

(2006) Application of factorial and response surface methodology in modern experimental design and optimization. Critical Reviews in Analytical Chemistry 36(3–4): 141–151.

16.

Hayashi

Kazehaya

Muroyama

(2000) Preparation of activated carbon from lignin by chemical activation. Carbon 38(13): 1873–1878.

17.

Hsu

L-Y

Teng

(2000) Influence of different chemical reagents on the preparation of activated carbons from bituminous coal. Fuel Processing Technology 64(1–3): 155–166.

18.

Ioannidou

Zabaniotou

(2007) Agricultural residues as precursors for activated carbon production—A review. Renewable and Sustainable Energy Reviews 11(9): 1966–2005.

19.

Jain

Tripathi

(2015) Nano-porous activated carbon from sugarcane waste for supercapacitor application. Journal of Energy Storage 4: 121–127.

20.

Kadirvelu

Faur-Brasquet

Le Cloirec

(2000) Removal of Cu(II), Pb(II), and Ni(II) by adsorption onto activated carbon cloths. Langmuir 16(22): 8404–8409.

21.

Kadirvelu

Thamaraiselvi

Namasivayam

(2001) Removal of heavy metals from industrial wastewaters by adsorption onto activated carbon prepared from an agricultural solid waste. Bioresource Technology 76(1): 63–65.

22.

Kalderis

Bethanis

Paraskeva

(2008) Production of activated carbon from bagasse and rice husk by a single-stage chemical activation method at low retention times. Bioresource Technology 99(15): 6809–6816.

23.

Kongsuwan

Patnukao

Pavasant

(2009) Binary component sorption of Cu(II) and Pb(II) with activated carbon from Eucalyptus camaldulensis Dehn bark. Journal of Industrial and Engineering Chemistry 15(4): 465–470.

24.

Wang

Tsai

(2012) Adsorption capacity and removal efficiency of heavy metal ions by Moso and Ma bamboo activated carbons. Chemical Engineering Research and Design 90(9): 1397–1406.

25.

Marsh

Rodríguez-Reinoso

(2006) Characterization of Activated Carbon. Activated Carbon, Amsterdam, Netherlands: Elsevier, pp. 143–242.

26.

Matlock

Howerton

Atwood

(2002) Chemical precipitation of heavy metals from acid mine drainage. Water Research 36(19): 4757–4764.

27.

Olivares-Marín

Fernández-González

Macías-García

(2006) Preparation of activated carbon from cherry stones by chemical activation with ZnCl₂. Applied Surface Science 252(17): 5967–5971.

28.

Park

Kim

Chae

(2007) Activated carbon-containing alginate adsorbent for the simultaneous removal of heavy metals and toxic organics. Process Biochemistry 42(10): 1371–1377.

29.

Prieto-Rodríguez

Oller

Klamerth

(2013) Application of solar AOPs and ozonation for elimination of micropollutants in municipal wastewater treatment plant effluents. Water Research 47(4): 1521–1528.

30.

Sreńscek-Nazzal

Kamińska

Michalkiewicz

(2013) Production, characterization and methane storage potential of KOH-activated carbon from sugarcane molasses. Industrial Crops and Products 47: 153–159.

31.

Suescún-Mathieu

Bautista-Carrizosa

Sierra

(2014) Carboxylic acid recovery from aqueous solutions by activated carbon produced from sugarcane bagasse. Adsorption 20(8): 935–943.

32.

Sun

Jiang

(2010) Preparation and characterization of activated carbon from rubber-seed shell by physical activation with steam. Biomass and Bioenergy 34(4): 539–544.

33.

Tao

H-C

Zhang

H-R

J-B

(2015) Biomass based activated carbon obtained from sludge and sugarcane bagasse for removing lead ion from wastewater. Bioresource Technology 192: 611–617.

34.

Tongpoothorn

Sriuttha

Homchan

(2011) Preparation of activated carbon derived from Jatropha curcas fruit shell by simple thermo-chemical activation and characterization of their physico-chemical properties. Chemical Engineering Research and Design 89(3): 335–340.

35.

Tsai

Chang

Lin

(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.

36.

Tran

(2016) Production of activated carbon from sugarcane bagasse by chemical activation with ZnCl₂: Preparation and characterization study. Research Journal of Chemical Sciences 6(5): 42–47.

37.

Wasi

Tabrez

Ahmad

(2013) Toxicological effects of major environmental pollutants: an overview. Environmental Monitoring and Assessment 185(3): 2585–2593.

A comparative study on the removal efficiency of metal ions (Cu 2+,Ni 2+,and Pb 2+ ) using sugarcane bagasse-derived ZnCl 2 -activated carbon by the response surface methodology

Abstract

Keywords

Introduction

Experimental procedure

Fabrication of AC from sugarcane bagasse

Determination of pH point of zero charges (pHzpc)

Batch adsorption experiments

Experimental design with RSM

Instruments and techniques

Results and discussion

Characterization of the AC

Assessment of experimental results with Design–Expert

Optimization of the removal efficiency of Cu2+, Ni2+, and Pb2+ using the RSM

Conclusion

Footnotes

Declaration of Conflicting Interests

Funding

References

Optimization of the removal efficiency of Cu²⁺, Ni²⁺, and Pb²⁺ using the RSM