Desalination 275 (2011) 276–284

Contents lists available at ScienceDirect

Desalination
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / d e s a l

Optimization of chromium(VI) sorption potential using developed activated carbon

from sugarcane bagasse with chemical activation by zinc chloride
K.J. Cronje a, K. Chetty a, M. Carsky a, J.N. Sahu c,⁎, B.C. Meikap a,b,⁎
a
School of Chemical Engineering, Faculty of Engineering, University of KwaZulu-Natal, Howard College Campus, King George V. Avenue, Durban, Pin-4041, South Africa
b
Department of Chemical Engineering, Indian Institute of Technology (IIT), Kharagpur, P.O. Kharagpur Technology, West Bengal, Pin-721302, India
c
Department of Chemical Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur Pin-50603, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: An adsorbent prepared from sugarcane bagasse with chemical activation by zinc chloride was used to study its
Received 29 November 2010 sorption potential on removing chromium(VI). In this work the application of response surface methodology
Received in revised form 1 March 2011 (RSM) is presented for optimization of removal of Cr(VI) ions from aqua solutions using activated carbon as
Accepted 3 March 2011
adsorbent. While the goal of adsorption of chromium optimization was to improve adsorption conditions in
Available online 1 April 2011
batch process, i.e., to minimize the adsorbent dose and to increase the initial concentration of Cr(VI). A 24 full
Keywords:
factorial central composite design experimental design was employed. Analysis of variance (ANOVA) showed
Adsorption a high coefficient of determination value (R2 = 0.99) and satisfactory prediction second-order regression
Activated carbon model was derived. Maximum chromium removal efficiency was predicted and experimentally validated. The
Optimization optimum adsorbent dose, temperature, initial concentration of Cr(VI) and initial pH of the Cr(VI) solution
Chromium(VI) were found to be 6.85 g/L, 40 °C, 77.5 mg/l, and 8.58 respectively. Under optimal value of process parameters,
Wastewater treatment high removal ( N 87%) was obtained for Cr(VI).
© 2011 Elsevier B.V. All rights reserved.

1. Introduction Strong exposure of Cr(VI) causes cancer in digestive tract and lungs
[9] and may cause epigastric pain, nausea, vomiting, severe diarrhoea
Chromium, an element which has been used by man for years, can and hemorrhage [10]. Overexposure of chromium workers to
be regarded as a longstanding environmental contaminant. It chromium dusts and mists has been related to irritation and corrosion
constantly released into the environment in a number of ways, like of the skin and the respiratory tract and, probably, to lung carcinoma.
as leather tanning, electroplating, manufacturing of dye, paint and It has adverse impacts on aquatic species as it accumulates in fish
paper are just few examples [1–3] and also by natural processes, like tissues and causes reduction in fish production at higher concentra-
as mainly by volcanic activity and weathering of rocks [4]. Chromium tion [11] and [12].
is a highly reactive element and exists in six oxidation states. Its two Due to environmental concern, discharge limits of both Cr(III) and
most stable states are Cr(III) and Cr(VI). But Cr(VI) is 500 times more Cr(VI) have been instituted by most industrial countries. Their
toxic than the Cr(III) [5]. Furthermore, Cr(III) has limited hydroxide concentration in industrial wastewaters ranges from 0.5 to 270 mg/
solubility making it relatively immobile and less available for L [2]. The tolerance limit for Cr(VI) for discharge into inland surface
biological uptake. Chromium's threshold concentration, on inhibitory waters is 0.1 mg/Land in potable water is 0.05 mg/L [13,14].
effect on heterotrophic organisms, is 10 mg/L for Cr(III) and 1 mg/L for The pollution of chromium metal in aqueous has become severe
Cr(VI). Hexavalent chromium is considered powerful carcinogenic with the development of economy. People have paid attention to it
agent that modifies DNA transcription process causing important which will make us sick such as cancer or even make us to die. In order
chromosomic aberration by the International Agency for Research on to comply with this limit, it is essential that industries treat their
Cancer [6] and [7]. The hexavalent chromium has also been classified effluents to reduce the Cr(VI) concentration in water and wastewater
as a group A carcinogen by USEPA based on its chronic effects [8]. to acceptable levels before its transport and cycling into the natural
environment. Therefore, the level of chromium(VI) in final effluent
has to be reduced by the application of appropriate technology. There
⁎ Corresponding authors at: Department of Chemical Engineering, Faculty of are many methods about how to deal with the chromium metal
Engineering, University of Malaya, Kuala Lumpur Pin-50603, Malaysia School of pollution, such as chemical precipitation [15], ion exchange [16],
Chemical Engineering, Faculty of Engineering, University of KwaZulu-Natal, Howard electrochemical precipitation [17], reduction [18], adsorption [19],
College Campus, King George V. Avenue, Durban, Pin-4041, South Africa. Tel.: +60 3
7967 5295(Office); fax: +60 3 7967 5319.
solvent extraction [20], membrane separation [21], concentration
E-mail addresses: jnsahu@um.edu.my, jay_sahu@yahoo.co.in (J.N. Sahu), [22], evaporation, reverse osmosis and bio-sorption [23,24] and
meikap@ukzn.ac.za (B.C. Meikap). emulsion per traction technology [25].

0011-9164/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.desal.2011.03.019
 K.J. Cronje et al. / Desalination 275 (2011) 276–284 277

Adsorption process has proved to be the most effective for the temperature under a nitrogen flow rate of 150 mL/min STP. Samples
removal of heavy metals from wastewaters [26–30]. The biggest were held at the final temperature (carbonization temperature) for
barrier in the application of this process by the industries is the high carbonization times of 40 min before cooling down under nitrogen.
cost of adsorbents presently available for commercial use. The The products were washed sequentially with 0.5 N HCl, hot water
cost of adsorption technology application can be reduced, if the and finally cold distilled water to remove residual organic and
adsorbent is inexpensive. So there is a need to develop low cost and mineral matters, and then dried at 110 °C. In all experiments,
easily available adsorbents for the removal of heavy metal ions from heating rate and nitrogen flow was kept constant. Then carbon was
the aqueous environment. For these reasons, activated carbon dried to 60 °C. After this, carbon was crushed in a small ball mill with
adsorption has been widely used for the treatment of chromium 50 small balls for one hour. The powder from ball mill is collected
containing wastewaters [19,31–33]. The materials developed for and dried to remove the moisture. Then this powder carbon was
this purpose range from agricultural waste products, biomass and kept in airtight packet for the experimental use.
various solid substances. Agricultural biomasses mainly consist of
lignin, cellulose, hemi-cellulose and some proteins which make 2.3. Batch adsorption experiment
them effective adsorbent for heavy metal cations. Sugarcane bagasse
is a soft, fibrous residue produced at the sugar mills as a by-product The experiments were performed in a thermal shaker at controlled
when the sugarcane is crushed for juice extraction, which is temperature for a period of 1 h at 120 rpm using 250 mL Erlenmeyer
available in large quantities at no cost and can form a good basis flasks containing 100 mL of different Cr(VI) concentrations at
for the development of adsorbent materials. Sugarcane bagasse different temperature(10–50 °C). The Jar tests were carried out to
contains carboxylic and hydroxyl groups and hence can be a cheap, study the effect of various operating variables on the adsorption rate.
attractive and effective adsorbent for the removal of heavy metals Adsorption of Cr(VI) on developed activated carbon was conducted
from wastewater [34]. containing different weighted amounts of each sample with 100 mL
Sugarcane bagasse, which is the main waste from sugarcane solution of different initial concentration of Cr(VI) ions. The batch
industry in South Africa, has been used as a raw material for the process was used so that there is no need for volume correction.
preparation of activated carbon in the present work. Activated carbon Samples were taken out at regular interval and the residual concen-
was prepared from sugarcane bagasse by chemical activation with tration in the solution was analyzed using Atomic Absorption
ZnCl2. In most of the earlier reported sorption of chromium ions Spectroscopy (AAS) after filtering the adsorbent with Whatmen filter
studies, effect of individual parameters has been reported maintaining paper to make it carbon free.
other process parameters constant at unspecified levels. This
approach does not depict the combined effect of all process 2.4. Experiment parameter
parameters. This approach is time consuming and requires a number
of experiments to determine optimum levels, which may be The effect of pH on the percentage removal was investigated over
unreliable. These limitations of a classical method can be eliminated the range of 2–12. The pH was adjusted by addition of 0.1 M HCl or
by optimizing all the process parameters collectively by statistical NaOH. The effect adsorbent dosage (1–10 g/L), initial Cr(VI) ions
experimental design such as Response Surface Methodology (RSM) concentration (10–100 mg/L) and temperature (10–50 °C) on the
[35–40]. This paper reports the combined effect of four process adsorption potential of Cr(VI) ions were studied. The Cr(VI) concen-
parameters, viz., adsorbent dose, pH, temperature and absorbent tration retained in the adsorbent phase was calculated according to:
concentration, on the removal of chromium from simulated solutions
by treated sugarcane bagasse using Central Composite Face-Centered ðC0 −Ce ÞV
Experimental Design in RSM. qe = ð1Þ
W

2. Materials and methods

where C0 and Ce are the initial and equilibrium concentrations (mg/L)
of chromium solution respectively; V is the volume (L); and W is the
2.1. Adsorbate: chromium(VI)
weight (g) of the adsorbent. Two replicates per sample were done and
the average results are presented.
A stock solution of Cr(VI) was prepared (1000 mg/L) by dissolving
By knowing the chromium concentration at initial concentrations
required amount of potassium dichromate (K2Cr2O7) in distilled
and equilibrium concentrations, the efficiency of adsorption of chro-
water. The stock solution was diluted with distilled water to obtain
mium by activated carbon can be calculated by using the following
desired concentration ranging from 10 to 100 mg/L.
equation for efficiency of adsorption chromium:
2.2. Adsorbent: sugarcane bagasse activated
C0 −Ce
The sugarcane bagasse were collected from Sugar Milling Adsorptionð%Þ = × 100 ð2Þ
C0
Research Institute (SMRI), Durban, South Africa and washed with
hot deionized water four to five times for removing dirt and dust
particles. The Sugarcane bagasse was solar dried over a period of 2.5. Analytical method
time. The dried sugarcane bagasse were then crushed in the ball mill
and sieved in 60–85 mesh (BSS) to obtain desired particle size. Atomic-absorption spectrophotometry utilizes the phenomenon
Chemical activation of the precursor was done with ZnCl2. 10 g of that atoms absorb radiation of particular wavelength. By atomic-
dried precursor was well mixed with distilled water so that 100 mL absorption spectrophotometer, the metals in water sample can be
concentrated solution contained 10 g of ZnCl2. The chemical ratio is analyzed. It consists of four basic structural elements; a light source
defined as the ratio of chemical activating agent (ZnCl2) to the (hollow cathode lamp), an atomizer section for atomizing the sample
precursor. The chemical ratio (activating agent/precursor) was 100% (burner for flame, graphite furnace for electrothermal atomization), a
in this case. The mixing was performed at 50 °C for 1 h. After mixing, monochrometer for selecting the analysis wavelength of the target
the slurry was subjected to vacuum drying at 100 °C for 24 h. The element, and a detector for converting the light into an electrical
resulting chemical-loaded samples were placed in a stainless-steel signal. It detect concentration of Cr(VI) in ppm level in the solution
tubular reactor and heated (5 °C/min) to the final carbonization and volume of sample required is only 1 mL for one analysis.
278 K.J. Cronje et al. / Desalination 275 (2011) 276–284

Table 1 responses and the corresponding parameters are modeled and

Experimental range and levels of independent variables for adsorption of Cr(VI). optimized using ANOVA was used to estimate the statistical parame-
Variables Symbol −α −1 0 +1 +α ters by means of response surface methods.
Basically this optimization process involves three major steps,
Initial concentration (mg/L) X1 10 32.5 55 77.5 100
Adsorbent dose (g/L) X2 1 3.25 5.5 7.75 10 which are, performing the statistically designed experiments, esti-
Temperature (°C) X3 10 20 30 40 50 mating the coefficients in a mathematical model and predicting the
pH X4 2 4.5 7 9.5 12 response and checking the adequacy of the model.

Y = f ðX1 ; X2 ; X3 ; X4 ::::::Xn Þ ð3Þ

2.6. Multivariate experimental design
where Y is the response of the system, and Xi is the variables of action
called factors. The goal is to optimize the response variable (Y). It is
Four independent test variables were chosen for the statistical
assumed that the independent variables are continuous and control-
experiment design is as follows: initial feed concentrations of chro-
lable by experiments with negligible errors. It is required to find a
mium (X1, mg/L), adsorbent dose (X2, g/L), temperature (X3, °C), and
suitable approximation for the true functional relationship between
pH (X4). The range and the levels of the factors, which were varied
independent variables and the response surface [38].
according to the experimental design, are given in Table 1.
The true relationship between Y and Xn may be complicated and, in
A standard RSM design called central composite design (CCD) was
most cases, it is unknown. In this case, by assuming high-order
applied in this work to study the variables for adsorption of chromium
interactions negligible, main effects and low-order interactions may
from aqueous solution using prepared activated carbon in a batch
be estimated by fractional factorial designs. Individual second order
process. The central composite design was widely used for fitting a
effects cannot be estimated separately by 2n factorial designs. The first
second order model. By using this method, modeling is possible and it
order model is as follows:
requires only a minimum number of experiments. It is not necessary
in the modeling procedure to know the detailed reaction mechanism n n n
since the mathematical model is empirical. Generally, the CCD Y = b0 + ∑ bi Xi + ∑ ∑ bij Xi Xj ð4Þ
i=1 i j
consists of a 2n factorial runs with 2n axial runs and nc center runs
(six replicates). These designs consist of a 2n factorial or fractional
Meanwhile, it is also possible to test the overall curvature generated
(coded to the usual ± 1 notation) augmented by 2n axial points ( ± α
by pure second order effects by means of the following statistic:
,0,0…, 0), (0, ± α ,0…, 0),…, (0, 0…, ± α ), and nc center points (0, 0,
0,…,0) [41]. Each variable is investigated at two levels. Meanwhile, as    2
the number of factors, n, increases, the number of runs for a complete nc F Y1− Y0
LOFCURV = ð5Þ
replicate of the design increases rapidly. In this case, main effects and nc + F
interactions may be estimated by fractional factorial designs running
 
only a minimum number of experiments. Individual second-order where Y 0 is the mean of central replicates, Y 1 is the mean of factorial
effects cannot be estimated separately by 2n factorial designs. The experiments results and F is the number of experiments in factorial

Table 2
Experimental design matrix and results for adsorption of Cr(VI).

Run Actual level of factors Coded level of factors Removal efficiency (%)

X1 (mg/L) X2 (g/L) X3 (°C) X4 X1 X2 X3 X4 Experimental Predicted

1 32.5 3.25 20 4.5 −1 −1 −1 −1 55.302 57.22

2 77.5 3.25 20 4.5 +1 −1 −1 −1 65.557 65.72
3 32.5 7.75 20 4.5 −1 +1 −1 −1 72.188 73.37
4 77.5 7.75 20 4.5 +1 +1 −1 −1 66.460 67.52
5 32.5 3.25 40 4.5 −1 −1 +1 −1 68.509 66.49
6 77.5 3.25 40 4.5 +1 −1 +1 −1 72.903 74.19
7 32.5 7.75 40 4.5 −1 +1 +1 −1 85.082 86.62
8 77.5 7.75 40 4.5 +1 +1 +1 −1 79.878 79.97
9 32.5 3.25 20 9.5 −1 −1 −1 +1 42.811 42.79
10 77.5 3.25 20 9.5 +1 −1 −1 +1 55.283 54.49
11 32.5 7.75 20 9.5 −1 +1 −1 +1 65.941 65.40
12 77.5 7.75 20 9.5 +1 +1 −1 +1 60.659 62.75
13 32.5 3.25 40 9.5 −1 −1 +1 +1 63.421 63.11
14 77.5 3.25 40 9.5 +1 −1 +1 +1 75.121 74.01
15 32.5 7.75 40 9.5 −1 +1 +1 +1 89.766 89.68
16 77.5 7.75 40 9.5 +1 +1 +1 +1 87.401 86.24
17 10 5.5 30 7 0 0 0 −α 75.172 74.75
18 100 5.5 30 7 0 0 0 +α 80.206 79.80
19 55 1 30 7 0 0 −α 0 37.247 38.10
20 55 10 30 7 0 0 +α 0 68.169 66.49
21 55 5.5 10 7 0 −α 0 0 70.458 68.34
22 55 5.5 50 7 0 +α 0 0 99.810 101.11
23 55 5.5 30 2 −α 0 0 0 69.382 67.18
24 55 5.5 30 12 +α 0 0 0 57.636 59.02
25 55 5.5 30 7 0 0 0 0 70.846 70.86
26 55 5.5 30 7 0 0 0 0 70.909 70.86
27 55 5.5 30 7 0 0 0 0 70.866 70.86
28 55 5.5 30 7 0 0 0 0 70.831 70.86
29 55 5.5 30 7 0 0 0 0 70.797 70.86
30 55 5.5 30 7 0 0 0 0 70.927 70.86
 K.J. Cronje et al. / Desalination 275 (2011) 276–284 279

design. If the variance analysis indicates that overall curvature effect is

significant, auxiliary experiments are carried out to develop a second
order model. The second order model is defined as follows so as to 99
facilitate calculations:
95
n  n
2  n n 90

Normal % Probability
2
Y = b0 + ∑ bi Xi + ∑ bii Xi − X i + ∑ ∑ bij Xi Xj ð6Þ
i=1 i=1 i=1 j N 1 80
70
where 50

2 30
2 1 N 2 F + 2α
X i = ∑ Xi = ð7Þ 20
N i=1 N
10
5
by defining:

n 1
0 2
bi = b0 − ∑ bii X i ð8Þ
i=1

Eq. (6) may be written in usual form as -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00
Internally Studentized Residuals
0
n n n n
2
Y = b0 + ∑ bi Xi + ∑ bii Xi + ∑ ∑ bij Xi Xj ð9Þ
i=1 i=1 i = 1 j N1 Fig. 1. The Studentized residuals and normal percentage probability plot of adsorption
of Cr(VI).

The four independent variables X1, X2, X3 and X4 and the mathe-
matical relationship of the response Y on these variables can be ap- significant and the models were not aliased. The quadratic model was
proximated by quadratic/(second degree) polynomial Eq. (10) as selected as suggested by the software. Experiments were planned to
shown below: obtain a quadratic model consisting of 24 trials plus a star configu-
ration (α = ±2) and there replicates at the center point. The design of
Y = b0 + b1 X1 + b2 X2 + b3 X3 + b4 X4 + b12 X1 X2 + b13 X1 X3 + b14 X1 X4
this experiment is given in Table 2, together with the experimental
+ b23 X2 X3 + b24 X2 X4 + b34 X3 X4 + b11 X12 + b22 X22 + b33 X32 + b44 X42 and predicted results. The maximum adsorption of Cr(VI) was found
ð10Þ to be N 99.9%. Regression analysis was performed to fit the response
function of adsorption of Cr(VI) (%). The model expressed by Eq. (10),
Where b0 the constant, b1, b2, b3, and b4 are the linear coefficients, where the variables take their coded values, represents adsorption
b12, b13, b14, b23, b24, and b34 are the cross product coefficients, and b11, efficiency (Y) as a function of initial feed concentrations of chromium
b22, b33, and b44 are the quadratic coefficients. (X1), adsorbent dose (X2), temperature (X3) and pH (X4). The final
The number of tests required for the CCD includes the standard 2n empirical models in terms of coded factors for adsorption of Cr(VI)
factorial with its origin at the center, 2n points fixed axially at a efficiency (Y) is shown in Eq. (12):
distance, say α from the center to generate the quadratic terms, and
replicate tests at the center; where n is the number of variables. The Y = 70:86 + 1:26X1 + 7:10X2 + 8:19X3 −2:04X4 −3:59X1 X2 −0:2X1 X3 + 0:8X1 X4
axial points are chosen such that they allow rotatability [42], which 2 2 2 2
+ 0:99X2 X3 + 1:61X2 X4 + 2:76X3 X4 + 1:6X1 −4:64X2 + 3:46X3 −1:94X4
ensures that the variance of the model prediction is constant at all
ð12Þ
points equidistant from the design center. Replicates of the test at the
center are very important as they provide an independent estimate of
the experimental error. For four variables, the recommended number
of tests at the center is six [43]. Hence, the total number of tests (N)
required for the four independent variables is: 3.00

n 4
N = 2 + 2n + nc = 2 + ð2 × 4Þ + 6 = 30 ð11Þ
Internally Studentized Residuals

2.00

Once the desired ranges of values of the variables are defined, they
are coded to lie at ± 1 for the factorial points, 0 for the center points 1.00
and ± α for the axial points. The codes are calculated as functions of
the range of interest of each factor followed by Mahalik et al., and Sahu
0.00
et al., work [37–40].

3. Results and discussion -1.00

3.1. Development of regression model equation

-2.00

CCD was used to develop correlation between the adsorption of Cr

(VI) from aqueous solution variables to the adsorption of Cr(VI) in -3.00
activated carbon. The complete experimental range and levels of
independent variables are given in Table 1. Runs 25–30 at the center 20.00 40.00 60.00 80.00 100.00 120.00
point were used to determine the experimental error. According to Predicted (%)
the sequential model sum of squares, the models were selected based
on the highest order polynomials where the additional terms were Fig. 2. The predicted adsorption of Cr(VI) and Studentized residuals plot.
280 K.J. Cronje et al. / Desalination 275 (2011) 276–284

of the response depends on the mean level of Y, then this plot often
120.00
exhibits a funnel-shaped pattern [38–40,45]. This is also an indication
that there was no need for transformation of the response variable.
The actual and the predicted adsorption of chromium(VI) percent
100.00
are shown in Fig. 3. Actual values are the measured adsorption
response data for a particular run, and the predicted values are
evaluated from the adsorption model and are generated by using the
Predicted (%)

80.00
approximating functions. In Fig. 3, the values of R2 and Radj
2
were found
to be 0.991 and 0.983, respectively. The fair correlation coefficients
might have resulted by the insignificant terms in Table 3, and most
60.00 likely due to four different variables selected in wide ranges with a
limited number of experiments as well as the nonlinear influence of
the investigated parameters on process response.
40.00
3.3. Adsorption capacity of chromium(VI)

20.00 To investigate the effects of the four factors on the adsorption of

chromium, the response surface methodology was used, and three-
30.00 40.00 50.00 60.00 70.00 80.00 90.00 100.00 dimensional plots were drawn. Based on the ANOVA results obtained,
Experimental (%) adsorbent doses, initial concentration, temperature, and pH were
found to have significant effects on the adsorption of chromium and
Fig. 3. The actual and predicted plot of adsorption of Cr(VI).
the quadratic effects are significant. The interaction effects between
initial concentrations of Cr(VI) and temperature were considered
3.2. Statistical analysis moderate. The adsorption of chromium percent response surface
graphs is shown in Figs. 4–9.
Eq. (12) has been used to visualize the effects of experimental factors It can be clear from the Fig. 4, the combined effect of initial feed
on conversion percentage response in Figs. 1–9. The model adequacy concentration of Cr(VI) and adsorbent dose on adsorption of
check is an important part of the data analysis procedure as the chromium(VI) at constant pH (7) and temperature (30 °C). It can be
approximating model would give poor or misleading results if it were an seen that the maximum adsorption of chromium N 74% was deter-
inadequate fit. This is done by looking at the residual plots which are mined at constant pH (7) and temperature (30 °C). Fig. 5 shows the
examined for the approximating model [44]. The normal probability and interactive effect of temperature and initial concentration of Cr(VI) of
Studentized residuals plot is shown in Fig. 1 for adsorption of Cr(VI) the solution on percent adsorption of Cr(VI) onto activated carbon. It
efficiency percent. In Fig. 1, residuals show how well the model satisfies shows a maximum adsorption of chromium 100% was determined at
the assumptions of the analysis of variance (ANOVA) whereas the constant pH (7) and adsorbent dose (5.5 g/L).
Studentized residuals measure the number of standard deviations The three dimensional response surfaces of the combined effect of
separating the actual and predicted values. Fig. 1 shows that neither initial concentration of Cr(VI) and pH on adsorption of chromium(VI)
response transformation was needed nor there was any apparent at constant adsorbent dose (5.5 g/L) and temperature (30 °C) is shown
problem with normality. Fig. 2 shows the Studentized residuals versus in Fig. 6. A maximum adsorption of chromiumN 73% was determined
predicted conversion percent. The general impression is that the plot at constant adsorbent dose (5.5 g/L) and temperature (30 °C). The
should be a random scatter, suggesting the variance of original combined effect of temperature and adsorbent dose on adsorption of
observations is constant for all values of the response. If the variance chromium(VI) at constant pH (7) and initial feed concentration of Cr(VI)

Fig. 4. The combined effect of initial concentration of Cr(VI) and adsorbent dose on adsorption of Cr(VI) at constant pH (7) and temperature (30 °C).
 K.J. Cronje et al. / Desalination 275 (2011) 276–284 281

Fig. 5. The combined effect of temperature and initial concentration of Cr(VI) on adsorption of Cr(VI) at constant pH (7) and adsorbent dose (5.5 g/L).

(55 mg/L) is shown in Fig. 7, the three dimensional response surfaces. A (VI) at constant initial feed concentration of Cr(VI) (55 mg/L) and
maximum adsorption of chromium 100% was determined at constant adsorbent dose (5.5 g/L). Increasing the temperature from 25 to 45 °C
pH (7) and adsorbent dose (5.5 g/L). facilitated the removal of Cr(VI) ions. The increase in metal uptake
In Fig. 8 it shows the three dimensional response surfaces which with increasing temperature may be due to either higher affinity of
were constructed to show the most important two variables (pH and sites for metal or an increase in number of binding sited on activated
adsorbent dose) on the adsorption of chromium(VI) at constant carbon. It is clear from this figure (Fig. 4) that the percent adsorption
temperature (30 °C) and initial concentration of Cr(VI) (55 mg/L). It of Cr(VI) decreases with increase in pH from 6.5 to 9.0. It is important
can be seen from the figure that initially the percentage removal that the maximum adsorption at all the concentrations takes place at
increases very sharply with the increase in adsorbent dosage but pH 7.0 (natural pH). A maximum adsorption of chromium 99.9% was
beyond a certain value 3.5–6.5 g/L, the percentage removal reaches determined at constant initial feed concentration of Cr(VI) (55 mg/L)
almost a constant value. This trend is expected because as the and adsorbent dose (5.5 g/L).
adsorbent dose increases the number adsorbent particles increases
and thus more Cr(VI) was attached to their surfaces. A maximum 3.4. Optimization by response surface modeling
adsorption of chromium N 74% was determined at constant temper-
ature (30 °C) and initial concentration of Cr(VI) (55 mg/L). One of the main aims of this study was to find the optimum process
In Fig. 9 shows the three dimensional response surfaces, the parameters to maximize the adsorption of chromium, from the
combined effect of temperature and pH on adsorption of chromium mathematical model equations developed in this study. The quadratic

Fig. 6. The combined effect of pH and initial concentration of Cr(VI) on adsorption of Cr (VI) at constant adsorbent dose (5.5 g/L) and temperature (30 °C).
282 K.J. Cronje et al. / Desalination 275 (2011) 276–284

Fig. 7. The combined effect of temperature and adsorbent dose on adsorption of Cr(VI) at constant pH (7) and initial concentration of Cr(VI) (55 mg/L).

model equation was optimized using quadratic programming (QP) to can be used to optimize the Cr (VI) adsorption process on modified
maximize of adsorption of chromium within the experimental range Sugarcane bagasse activated carbon. However, in this study, only
studied. The optimum adsorption conditions (Table 4) were determined selected critical process parameters were studied, leaving behind other
as temperature 40 °C, initial feed concentration of Cr(VI) 77.5 mg/L, pH parameters such as contact time that may have significant effect on the
8.58 and adsorbent dose 6.85 g/L have been determined as optimum adsorption process. Parameters for the adsorption model are estimated
levels of the process parameters to achieve the maximum adsorption of by statistical regression equations based on sorption data for various
chromium of 87.01%, compared to 86.92% which was maximum feedstocks. A second-order polynomial equation was adequate to
adsorption of chromium in the tests conducted. predict the permeability as the response. An R2 value of 0.991 ensures
a sufficient adjustment of the model with the experimental data. The
4. Conclusions model and results can be used to estimate the optimal adsorbents based
on feed concentration, pH or termperature of waste water. The statistical
Based on the design of experiment approach over conventional analysis shows tha adsorbent doses, initial feed concentration, temper-
strategy, it was found that interaction between the process parameters ature, and pH were found to have significant effects on the adsorption
significantly affect the adsorption of Cr(VI) on modified Sugarcane of chromium. The optimal adsorption of chromium was obtained
bagasse activated carbon. Once the interaction effects were identified, it adsorbent dose, temperature, initial concentration of Cr(VI) and initial

Fig. 8. The combined effect of pH and adsorbent dose on adsorption of Cr(VI) at constant temperature (30 °C) and initial concentration of Cr(VI) (55 mg/L).
 K.J. Cronje et al. / Desalination 275 (2011) 276–284 283

Fig. 9. The combined effect of pH and temperature on adsorption of Cr(VI) at constant initial concentration of Cr(VI) (55 mg/L) and adsorbent dose (5.5 g/L).

pH of the Cr(VI) solution were found to be 6.85 g/L, 40 °C, 77.5 mg/L Acknowledgements
and 8.58 respectively resulting in 87% of adsorption of chromium.
Furthermore, this study revealed that owing to the high initial feed The authors would like to gratefully acknowledge to the Sugar
concentration of Cr(VI), the optimum adsorption is achieved at different Milling Research Institute (SMRI), South Africa for supplying the
process parameters. However, more investigations are needed on dif- sugarcane bagasse as raw materials.
ferent types of industrial wastewaters, continuous process, and different
operating conditions before such conclusions can be generalized.
References
[1] M.J. Udy, Chromium, Reinhold Publishing Corporation, New York, 1956.
Table 3 [2] L.J. Casarett, J. Doul, Toxicology, the Basic Science of Poisons, Macmillan, New
Analysis of variance (ANOVA) for response surface quadratic model for adsorption of York, 1980.
Cr(VI). [3] J.O. Nriagu, E. Nieboer, Chromium in the Natural and Human Environment, Wiley,
New York, 1988.
Source Sum of Degree of Mean F value Prob N F Remarks [4] N. Serpone, E. Borgarello, E. Pelizzeti, E. Schiavello (Eds.), Photocatalysis and
squares freedom square Environment, Kluwer Academic, The Netherlands, 1988.
[5] Z. Kowalski, Treatment of chromic tannery wastes, Journal of Hazardous Materials
Model 4622.15 14 330.15 126.48 b 0.0001 Significant 37 (1994) 137–144.
X1 38.28 1 38.28 14.67 0.0016 Significant [6] US Department of Health, Services Human, Toxicological Profile for Chromium,
X2 1208.59 1 1208.59 463.01 b 0.0001 Significant Public Health Services Agency for Toxic substances and Diseases Registry,
X3 1610.24 1 1610.24 616.89 b 0.0001 Significant Washington, DC, 1991.
X4 99.90 1 99.90 38.27 b 0.0001 Significant [7] M. Cieslak-Golonka, Toxic and mutagenic effects of chromium (VI), A review,
X1 X2 205.91 1 205.91 78.89 b 0.0001 Significant Polyhedron 15 (1995) 3667–3689.
X1 X3 0.64 1 0.64 0.24 0.6285 [8] E.M. Nkhalambayausi-Chirwa, Y.T. Wang, Simultaneous chromium (VI) reduction
X1 X4 10.25 1 10.25 3.93 0.0661 and phenol degradation in a fixed film co-culture bioreactor: reactor perfor-
X2 X3 15.76 1 15.76 6.04 0.0267 Significant mance, Water Research 35 (2001) 19221–21932.
X2 X4 41.85 1 41.58 15.93 0.0012 Significant [9] D.B. Kaufaman, Acute potassium dichromate poisoning in man, American Journal
of Diseases of Children 119 (1970) 374–381.
X3 X4 121.82 1 121.82 46.67 b 0.0001 Significant
[10] E. Browning, Chromium in Toxicity of Industrial Metals, 2nd ed, Butterworths and
X21 70.54 1 70.54 27.02 0.0001 Significant
Co, London, 1969, pp. 76–96.
X22 590.94 1 590.94 226.39 b 0.0001 Significant
[11] G.A. Cutter, Trace elements in estuarine and coastal waters, Reviews of Geophysics
X23 329.29 1 329.29 126.15 b 0.0001 Significant S 29 (1991) 639–644.
X24 103.38 1 103.38 39.60 b 0.0001 Significant [12] K.K. Krishnani, I.S. Azad, M. Kailasam, A.R. Thirunavukkarasu, B.P. Gupta, K.O.
Residual 39.15 15 2.61 – – Joseph, M. Murlidhar, M. Abharam, Acute toxicity of some heavy metals to lates
Lack of fit 39.14 10 3.91 1647.45 b 0.0001 Significant calcarifer fry with a note on its histopathological manifestations, Journal of
Pure error 0.012 5 2.37E− 003 – – Environmental Science and Health 38 (4) (2003) 645–655.
Correlation 4661.30 29 – – – [13] World Health Organization, Guidelines for Drinking-Water Quality, 3rd ed, World
total Health Organization, Geneva, Switzerland, 2006.
[14] EPA, Environmental Pollution Control Alternatives, EPA/625/5-90/025,
EPA/625/4-89/023, Environmental Protection Agency, Cincinnati, US, 1990.
[15] X. Zhou, T. Korenaga, T. Takahashi, T. Moriwake, S. Shinoda, A process monitoring/
controlling system for the treatment of wastewater containing chromium(VI),
Table 4 Water Research 27 (1993) 1049–1054.
Model validation for adsorption of Cr(VI). [16] G. Tiravanti, D. Petruzzelli, R. Passiono, Pretreatment of tannery wastewaters by
an ion exchange process for Cr(III) removal and recovery, Water Science and
Initial feed Adsorbent Temperature, pH, X4 Adsorption of Cr(VI), Technology 36 (1997) 197–207.
concentration, dose, X3 (°C) Y (%) [17] N. Kongsricharoern, C. Polprasert, Chromium removal by a bipolar electrochem-
X1 (mg/l) X2 (g/l) ical precipitation process, Water Science and Technology 34 (1996) 109–116.
Predicted Experimental [18] J.C. Seaman, P.M. Bertsch, L. Schwallie, In situ Cr(VI) reduction within coarse-
textured, oxide-coated soil and aquifer systems using Fe(II) solutions, Environ-
77.50 6.85 40.0 8.58 87.01 86.92
mental Science & Technology 33 (1999) 938–944.
284 K.J. Cronje et al. / Desalination 275 (2011) 276–284

[19] N. Calace, D.A. Muro, E. Nardi, B.M. Petronio, M. Pietroletti, Adsorption isotherms [32] C. Selomulya, V. Meeyoo, R. Amal, Mechanisms of Cr(VI) removal from water by
for describing heavy metal retention in paper mill sludges, Industrial and various types of activated carbons, Journal of Chemical Technology and
Engineering Chemistry Research 41 (2002) 5491–5497. Biotechnology 74 (1994) 111–122.
[20] K. Pagilla, L.W. Canter, Laboratory studies on remediation of chromium [33] J. Acharya, J.N. Sahu, B.K. Sahoo, C.R. Mohanty, B.C. Meikap, Removal of chromium
contaminated soils, Journal of Environmental Engineering 125 (1999) 243–248. (VI) from wastewater by activated carbon developed from Tamarind wood
[21] A.K. Chakravarti, S.B. Chowdhury, S. Chakrabarty, T. Chakrabarty, D.C. Mukherjee, activated with zinc chloride, Chemical Engineering Journal 150 (2009) 25–39.
Liquid membrane multiple emulsion process of chromium (VI) separation from [34] U.K. Garg, M.P. Kaur, D. Sud, V.K. Garg, Removal of hexavalent chromium from
wastewaters, Colloids and Surfaces. A: Physicochemical and Engineering Aspects aqueous solution by adsorption on treated sugarcane bagasse using response
103 (1995) 59–71. surface methodological approach, Desalination 249 (2) (2009) 475–479.
[22] C.F. Lin, W. Rou, K.S. Lo, Treatment strategy for Cr(VI) bearing wastes, Water [35] M.Z. Alam, S.A. Muyibi, J. Toramae, Statistical optimization of adsorption processes
Science and Technology 26 (1992) 2301–2304. for removal of 2,4-dichlorophenol by activated carbon derived from oil palm
[23] Z. Aksu, T. Kutsal, A comparative study for biosorption characteristics of heavy empty fruit bunches, J of Environmental Science 19 (2007) 674–677.
metal ions with C. vulgaris, Environmental Technology 11 (1990) 979–987. [36] J.N. Sahu, J. Acharya, B.C. Meikap, Optimization of production conditions for
[24] D. Mohan, K.P. Singh, V.K. Singh, Removal of hexavalent chromium from aqueous activated carbons from Tamarind wood by zinc chloride using response surface
solution using low-cost activated carbons derived from agricultural waste methodology, Bioresource Technology 101 (2010) 1974–1982.
materials and activated carbon fabric cloth, Industrial and Engineering Chemistry [37] J.N. Sahu, A.V. Patwardhan, B.C. Meikap, Response surface modeling and
Research 44 (4) (2005) 1027–1042. optimization for production of ammonia from urea in a batch reactor, Asia-
[25] I. Ortiz, M.F.S. Roman, S.M. Corvalan, A.M. Eliceche, Modeling and optimization of Pacific Journal of Chemical Engineering 4 (2009) 462–470.
an emulsion pertraction process for removal and concentration of Cr(VI), [38] K. Mahalik, J.N. Sahu, A.V. Patwardhan, B.C. Meikap, Statistical modelling and
Industrial and Engineering Chemistry Research 42 (2003) 5891–5899. optimization of hydrolysis of urea to generate ammonia for flue gas conditioning,
[26] J.N. Sahu, S. Agarwal, B.C. Meikap, M.N. Biswas, Performance of a modified multi- Journal of Hazardous Materials 182 (1–3) (2010) 603–610.
stage bubble column reactor for lead(II) and biological oxygen demand removal [39] J.N. Sahu, A.V. Patwardhan, B.C. Meikap, Optimization for production of ammonia
from wastewater using activated rice husk, Journal of Hazardous Materials 161 from urea in a semi-batch reactor for safe feedstock in power plants:
(2009) 317–324. experimental and statistical studies, CLEAN - Soil, Air, Water 38 (5–6) (2010)
[27] J. Acharya, J.N. Sahu, C.R. Mohanty, B.C. Meikap, Removal of lead(II) from 533–542.
wastewater by activated carbon developed from Tamarind wood by zinc chloride [40] J.N. Sahu, J. Acharya, B.C. Meikap, Response surface modeling and optimization of
activation, Chemical Engineering Journal 149 (2009) 249–262. chromium(VI) removal from aqueous solution using Tamarind wood activated
[28] C.P. Dwivedi, J.N. Sahu, C.R. Mohanty, B. Raj Mohan, B.C. Meikap, Column carbon in batch process, Journal of Hazardous Materials 172 (2009) 818–825.
performance of granular activated carbon packed bed for Pb(II) removal, Journal [41] R.H. Myers, Response Surface Methodology, Allyn and Bacon, New York, 1971.
of Hazardous Materials 156 (1–3) (2008) 596–603. [42] G.E.P. Box, J.S. Hunter, Multi-factor experimental design for exploring response
[29] C.K. Singh, J.N. Sahu, K.K. Mahalik, C.R. Mohanty, B. Raj Mohan, B.C. Meikap, surfaces, Ann Math Stat. 28 (1957) 195–241.
Studies on the removal of Pb(II) from wastewater by activated carbon developed [43] G.E.P. Box, W.G. Hunter, The 2k − p fractional factorial designs, parts I and II, J
from Tamarind wood activated with sulphuric acid, Journal of Hazardous Technometr 3 (1961) 311–458.
Materials 153 (1–2) (2008) 221–228. [44] G.E.P. Box, W.G. Hunter, Statistics for Experiments: An Introduction to Design,
[30] K. Rastogi, J.N. Sahu, B.C. Meikap, M.N. Biswas, Removal of methylene blue from Data Analysis and Model Building, Wiley Interscience, 1987.
wastewater using fly ash as an adsorbent by hydrocyclone, Journal of Hazardous [45] R.H. Myers, D.C. Montgomery, Response surface methodology: process and
Materials 158 (2008) 531–540. product optimization using designed experiments, 2nd editionJohn Wiley and
[31] E. Demirbas, M. Kobya, E. Senturk, T. Ozkan, Adsorption kinetics for the removal of Sons, USA, 2002.
chromium(VI) from aqueous solutions on the activated carbons prepared from
agricultural waste, Water SA 30 (2004) 533–539.