- Email: [email protected]
Application of central composite design for simultaneous removal of methylene blue and Pb2+ ions by walnut wood activated carbon
Accepted Manuscript Application of central composite design for simultaneous removal of methylene blue and Pb2+ ions by walnut wood activated carbon M. Ghaedi, H. Mazaheri, S. Khodadoust, S. Hajati, M.K. Purkait PII: DOI: Reference:
S1386-1425(14)01037-3 http://dx.doi.org/10.1016/j.saa.2014.06.138 SAA 12397
To appear in:
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
Received Date: Revised Date: Accepted Date:
22 April 2014 16 June 2014 29 June 2014
Please cite this article as: M. Ghaedi, H. Mazaheri, S. Khodadoust, S. Hajati, M.K. Purkait, Application of central composite design for simultaneous removal of methylene blue and Pb2+ ions by walnut wood activated carbon, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy (2014), doi: http://dx.doi.org/10.1016/j.saa. 2014.06.138
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3
5
Application of central composite design for simultaneous removal of methylene blue and Pb2+ ions by walnut wood activated carbon
6
M. Ghaedia, H. Mazaheria, S. Khodadoust b, S. Hajatic, M. K. Purkaitd
4
a
7 b
8
Cellular and Molecular Research Center, Yasuj University of Medical Science, Yasuj, Iran c
9 10
Chemistry Department, Yasouj University, Yasouj 75918-74831, Iran
d
Physic Department, Yasouj University, Yasouj, Iran
Department of Chemical Engineering, Indian Institute of Technology Guwahati, Assam-7S1039, India
11
* Corresponding author: Email: [email protected]
12
------------------------------------------------------------------------------
13
Abstract
14
Activated carbon was prepared from walnut wood which was locally available, non-toxic,
15
abundant and cheap. This new adsorbent was characterized using BET, FTIR and SEM. Point of
16
zero charge (pHPZC) and oxygen containing functional groups were also determined. The
17
prepared adsorbent was applied for simultaneous removal of Pb2+ ions and methylene blue (MB)
18
dye from aqueous solution. The prominent effect and interaction of variables such as amount of
19
adsorbent, contact time, concentration of MB and Pb2+ ions was optimized by central composite
20
design. The equilibrium data obtained at optimum condition was fitted to conventional isotherm
21
models and found that Langmuir model was the best fitted isotherm. Kinetic data were fitted
22
using various models. It was revealed that the adsorption rate follows pseudo second order
23
kinetic model and intraparticle diffusion model.
24
Keywords: Central composite design; Simultaneous removal; Methylene blue; Pb2+ ions
25 26 27 1
1
1. Introduction
2
Heavy metal ions with considering their nature and levels enter to various aqueous
3
ecosystems. They may be considering as toxic or poisonous for human health and animals even
4
at low concentrations. On the other hand, their content strongly can be affected by the presence
5
of other compounds, especially dyes. Another category of hazardous compounds and materials
6
are dyes that may generate serious problems and hazards for human and other organism that
7
attributed to their high biotoxicity, mutagenic and carcinogenic effects [1-4]. On the other hand,
8
presence of dyes by hindering from light penetration that leads to reduce in photosynthesis and
9
generation of harmful hazards to other organisms. These associated hazards and problem make a
10
necessity to design and develop new waste water treatment approach to reduce their level below
11
threshold limit. One of the most well known and famous method for dye removal is adsorption
12
that is superior to other conventional technique including biological treatment, adsorption,
13
coagulation/flocculation, chemical oxidation, membrane separation and ion exchange for
14
removal of dyes and heavy metals have been developed [5,6]. Diverse materials such as activated
15
carbon, zeolite, clay, polymer, and nanomaterials have been extensively applied for pollutants
16
removal [7-9].
17
Methylene blue (MB) commonly applied as coloring dye, while it is also used for dying cotton
18
and silk [10]. The harmful impacts of such dye on water make an urgent task to remove them
19
from waste streams before discharge to water resources. Lead and its compound, especially as
20
Pb2+ ions are known as important pollutants even at low concentrations by affecting central
21
nervous system, kidneys, gastrointestinal system [11]. Therefore, the removal of these pollutants
22
as sole component or their combination from waste water is a challenging requirement and need
23
more attention.
24
Previous studies focused on single component removal, while limited reports for simultaneous
25
adsorption of metal ions and dyes on natural zeolite [12-16] and activated carbon [17, 18] have
26
been reported. In this paper, MB and Pb2+ ions was selected as model and their adsorption
27
behavior in binary systems (kinetics and equilibrium) was studied and compared.
28
Activated carbons (amorphous solids with large internal surface areas and pore volumes) simply
29
can be prepared from natural resource such as coal [19], wood [20], coconut shell [21], tea waste
30
[22] and rice husk [23]. This approach is green and environmental friendly behavior due to
31
conversion of unwanted, worthless agricultural waste to useful, low cost, cheep and high surface 2
1
area adsorbent which are able to remove organic chemicals and metals of environmental and /or
2
economic concern from various resources following their carbonization and activation [24, 25].
3
Walnut Tree (agricultural tree) grows rapidly in different regions, while its wood can be
4
applied in carpentry work and fuel. Storm, natural and/or agricultural activities such as pruning
5
lead to cutting and separation of the branches and trunks of many walnut trees. The high
6
abundant of waste wood (walnut tree) make an economic task the burning and putting this tree
7
waste into a sealed container that cause its simple conversion to carbon source. This behavior
8
reveals that the present homemade walnut carbon (HWC) is very cost effective and non-toxic
9
material with at least energy consumption. In this work, carbon following production was
10
activated by concentrated nitric acid was.
11
In addition to significant decrease in the number of experimental runs by central composite
12
design under response surface methodology (RSM), the main and interaction effect of variables
13
including contact time, adsorbent dosage, initial MB and Pb2+ ions concentration simply was
14
investigated and optimum value of each variable was specified.
15 16
2. Experimental
17
2.1. Instruments and reagents
18
Chemicals including NaOH, HCl, NaNO3, Pb(NO3). 2 H2O, Na2CO3 and NaHCO3 with
19
the highest purity available were purchased from Merck (Dermasdat, Germany). MB dye
20
(Sigma–Aldrich) has following information (a) color index number: 52.015, (b) molecular
21
weight: 319.86 g/mol, (c) empirical formula: C16H18N3SCl and (d) λmax: 664 nm. Analyte
22
solutions were prepared by dissolving their appropriate amount in double distilled water, while
23
the pH was adjusted using pH/Ion meter model-686. The absorbance spectra for MB was
24
recorded in the range of 300 nm to 750 nm using Jasco UV-Visible spectrophotometer model V-
25
530 with a fixed slit width of 2 nm and scan speed of 1000 nm/min. Pb2+ ions was determined by
26
atomic absorption spectrophotometer Varian model AA 220 at λ=217 nm. Fourier transform
27
infrared spectroscopy (FTIR in the range of 400–4000 cm−1) of the adsorbent was recorded using
28
FT-IR spectrophotometer (Model: FT-IR JASCO 460 Plus). Spectra obtained were analyzed.
29
The STATISTICA, a statistical package software version 7.0 (Stat Soft Inc., Tulsa, USA)
30
was used for experimental design analysis and their subsequent regression analysis. Statistical
31
analysis of the model was performed to evaluate the analysis of variance (ANOVA). The quality 3
1
of the polynomial model equation was judged statistically by the coefficient of determination R2
2
and its statistical significance was determined by F-test. P-values less than 0.05 were considered
3
to be statistically significant.
4 5
2.2. Multi-component adsorption of Pb2+ ions and MB on natural activated carbon
6
Binary system of MB – Pb2+ ions was used in the adsorption experiments. All
7
experiments were carried out using different amount of adsorbent in 50 ml beakers in a magnetic
8
stirrer at 400 rpm to obtain the optimum conditions (pH, contact time and initial dye and Pb2+
9
ions concentration), while all experiments conducted at room temperature. The effect of pH on
10
the simultaneous removal was investigated by conducting similar experiments as follow: 50 mL
11
of solution of 50 mg L-1 of Pb2+ ion and 15 mg L-1 of the MB was thoroughly mixed with 0.05
12
and 0.10 g of adsorbent at pH values of 2-6 until achievement of the equilibrium. Subsequent
13
optimization of pH, a central composite design (CCD) was applied for the investigation of the
14
influence of variables and their interaction. Adsorption capacities for Pb2+ ions and dye (qi, mg g-
15
1
) on adsorbent were calculated by a mass balance: (C 0, i - C f, i )V
(1)
16
qi =
17
where C0,i and Cf,i is the initial and final concentration (mg L-1) of pollutant i (i.e., dye or Pb2+
18
ions) in the binary solution, V is the solution volume (L) and m indicate the adsorbent amount
19
(g).
m
20 21
2.3. Preparation of activated carbon
22
Dry branches of walnut tree were grinded and cut into small pieces (lower than one
23
centimeter). They were washed with distilled water and triton X-100 to remove impurities and
24
subsequently were dried in outdoors for 72 hours. Then, 500 g of dried sample were placed into a
25
container with a small hole (size about 5 mm) and about 2 hours directly was heated in flame.
26
Obtained coal washed again with distilled water and placed at 105 ° C for 24 h to dry that lead to
27
conversion of raw material to carbon.
28
This carbon was milled and sieved in the mesh range of 50-60 and activated thoroughly by
29
addition of nitric acid solution as follows: 10 g of carbon was mixed with 150 mL of 5 mol L-1
30
nitric acid and the mixture was refluxed for 6 h at 105 ° C [26-28]. This HNO3-activated carbon 4
1
filtered and washed with deionized water at 50 °C until approximate neutral pH and finally dried
2
at 105 ° C for 24 h. This adsorbent was used for the simultaneous adsorption of MB and Pb2+
3
ions. The pH corresponding to the point of zero charge (pHZPC) of the adsorbent was determined
4
by the pH drift method reported elsewhere [29]. Determination of oxygen containing functional
5
groups was performed by the Boehm titration method [30]. In this method, 1.0 g of the activated
6
carbon (AC) were kept in contact with 15 mL solution of NaHCO3 (0.1M), Na2CO3 (0.05M) and
7
NaOH (0.1M) to identify its acidic groups and 0.1M HCl for basic groups /sites (in separate
8
experiments) respectively, at room temperature for more than 2 days. Subsequently, the aqueous
9
solutions were back titrated with HCl (0.1M) for acidic and NaOH (0.1M) for basic groups. The
10
number and type of acidic sites were calculated by considering that NaOH neutralizes
11
carboxylic, lactonic and phenolic groups; Na2CO3 neutralizes carboxylic and lactonic groups,
12
while NaHCO3 neutralizes only carboxylic groups. Carboxylic groups were therefore quantified
13
by direct titration with NaHCO3. The difference between the groups titrated with Na2CO3 and/or
14
NaHCO3 approximately indicate amount of lactones group and the difference between the groups
15
titrated with NaOH and Na2CO3 reveal the phenol content. Basic sites were determined by
16
titration with HCl. Neutralization points were known using potentiometric titration methods. In
17
order to neutralize basic groups /sites, remaining HCl in the solution was back titrated with 0.1M
18
NaOH.
19 20
2.4. Central composite design
21
Central composite design (CCD) was used to study the individual and synergetic effect of
22
the four factors towards two responses. This method can reduce the number of experimental
23
trials required to evaluate main effect of each parameter and their interactions [31]. In general,
24
CCD is characterized by three operations namely: 2n axial runs, 2n factorial runs and central
25
runs. The CCD in present research is designed based on carrying out 8 axial points, 16 factorial
26
points and 4 replicates at the center (28 experiments) according to value reported in Table 1.
27
Total number of experiments =2n +2n+nc
28
where n is the number of factors, nc is the number of center points. Alpha (α) approximately
29
show the distance of the axial from center point, which is rotatable and strongly depends on the
30
number of factorial points and can be calculated from following equation [32]:
(1)
5
1
α=Np1/4
2
Present experimental design is based on variables including amount of adsorbent (X1), contact
3
time (X2), concentration of Pb2+ ions (X3) and MB (X4) in randomized fashion to minimize the
4
effects of the uncontrolled factors [33].
5
The optimal conditions for the responses (MB and Pb2+ ions removal percentage) were
6
determined using the optimal model predictor quadratic equation given as:
7
Y = b0 + ∑ bi x i + ∑ bii xi2 + ∑
(2)
n
n
i =1
i =1
n
n
∑b
ij
(3)
xi x j
i =1 j = i +1
8
where Y is the predicted response (removal percentage); Xi’s are the independent variables that
9
are known for each experimental run. The parameter b0 is the model constant; bi is the linear
10
coefficient; bii are the quadratic coefficients and bij is the interaction coefficients. The regression
11
analysis is used to fit the equations for both responses to estimate the statistical significance of
12
the equation obtained following analysis of the experimental data was analyzed using the
13
STATISTICA software.
14 15
2.5. Desirability function (DF)
16
After generation of the polynomial equations based on explanation of mathematical
17
relationship between removal percentage and individual variables, the desirability function (DF)
18
was used to find the best optimum levels for each variable [34]. The DF value is in the range of
19
0, 0.5 and 1. Based on analysis of experimental response and their conversion to desirability
20
function (dfi), it was explored that has value in the range of 0 to 1. The 1 and 0 indicate the
21
maximum and minimum desirability, respectively based on the Derringer and Suich [34]
22
equation as follow:
23
df =
, α≤U≤β
(4)
df = 1 , U >
df = 0 , U <
24
In Eq. (4), α and β are the lowest and the highest values, respectively, (obtained for the response
25
i) and wi is the weight. The individual desirability scores for the predicted values for each
26
dependent variable are then combined into overall DF by computing their geometric mean of
27
different dfi values. 6
1
DF = df × df … × df !
"
, 0 ≤ v ≤ 1 (i = 1, 2 … n)
(5)
) v = 1 *
2
Where dfi indicate the desirability of the response Ui (i = 1, 2, 3, . . ., n) and vi represents
3
the importance of responses. Inspecting the desirability profiles, it determines which levels of the
4
predictor variables produce the most desirable predicted responses on the dependent variables
5
[35].
6 7
3. Results and discussion
8
3.1. Adsorbent characterization
9
FTIR, pH determination at zero point of charge [29] (pHZPC) and Boehm titration method
10
[30] were used to characterize this new adsorbent. The characteristic functional groups of the
11
walnut activated carbon (WAC) were investigated using FTIR spectra. FTIR spectrum of WAC
12
(Fig. 1) represents the spectrum bands of ν(O-H) at 3435.56, ν(C-H) at 2925.48 cm-1, ν(C=O)
13
at 1716.34 cm-1, ν(C=C) at 1619.91 cm-1, ν(C-O) or C-C or C=O at 1230.27-921.82 cm-1, δ(C-
14
H) at 1456.66, and γ(C-H) at 708.92-527.70 cm-1. The band 3435.56 cm-1 is attributed to the
15
hydroxyl groups in phenolic and aliphatic structures. The band 2925.48 cm-1 is CH stretching in
16
aromatic methoxyl groups and in methyl and methylene groups of side chains. The band of the
17
carbonyl group in ketones, aldehydes or carboxyl appears at 1716.91 cm-1. The band at 1619.91
18
cm-1 may originate from carbohydrates. The bands at 1230.27 - 921.82 cm-1 is associated with C-
19
O, C-C stretching and C-OH bending in structure of activated carbon [36]. Some surface
20
properties such as pore size and volume were studied by surface area analysis are summarized in
21
Table 2. The porosity and surface area of the WAC show high surface area (1012.75 m2 g-1), that
22
may be attributed to activated carbon and presence of low sized WAC.
23 24
3. 2. pHzpc
25
Distribution of AC charge has significant and important role on the removal and
26
interaction of various compounds with adsorbent. The adsorbent charge (positive or negative)
27
influences the interaction of surface with solute ions or molecules by changing the mechanism
28
and correspond forces. The specified pH known as zero point of charge (ZPC), adsorbent surface 7
1
is neutral [37] and the prominent mechanism for solute transfer is the diffusion into the adsorbent
2
micro and meso-pores. At pH above this value, the surface becomes negative and the positive
3
ions attracted on the surface according to electrostatic attraction. The estimated value of pHzpc
4
(1.9) show he high tendency of these cationic species for strong adsorption on present adsorbent.
5 6 7 8
3. 3. Determination of oxygen containing functional groups
9
the Boehm titration technique [38] based on following principle: only strongly acidic carboxylic
10
groups are neutralized by sodium bicarbonates (Na2HCO3), lactonic and carboxylic group
11
neutralized by sodium carbonate (Na2CO3). The weakly acidic phenolic groups simply can be
12
determined following titration by sodium hydroxide. Neutralization with hydrochloric acid (HCl)
13
makes possible the amount of surface basic group’s of AC such as pyrones [39, 40].
14
Characterization of the prepared adsorbent was performed and the results are presented in Table
15
3. The high amount of various polarity functional groups creates various reactive sites for
16
adsorption of different polarity group with various sizes.
The type and amount of WAC surface functional groups can be determined routinely by
17 18 19 20
3.4. Effect of pH on removal efficiency
21
various pH over the range of 2-7 (Fig. 2). For Pb2+ ions, adsorption increases with increasing pH,
22
while at low pHs due to abundance of proton concentration a noticeable repulsive force and
23
competition between Pb2+ ions and proton for binding onto the surface was appeared that lead to
24
the decrease in both species removal percentage. Similar behavior and trend for MB with lower
25
slope was observed. It is well known that heavy and transition metal ions will react with
26
hydroxyl species in basic solution to for charged species including M(OH)+ or non- soluble
27
species such as M(OH)2 which strongly affect the nature, mechanism and adsorption efficiency.
28
MB (basic dye) form cation (C+) and reduced ions (CNH+) in solution in acidic pH that hinder
29
from its adsorption onto adsorbent surface. Another most probable mechanism for adsorption
30
may occur following complexation of metal ions with dye compounds. This reaction may lead to
31
formation of neutral species that are very high ability for adsorption onto adsorbent. This
32
pathway was reduced in acidic media and strongly enhanced following increasing the pH. On the
The effect of pH on the removal of MB and Pb2+ ions by WAC was investigated at
8
1
other hand (as it is known) by rising the pH, adsorption of both species (cations) on adsorbent
2
strongly facilated and their easily association and accumulation on adsorbent surface lead to
3
enhance in removal percentage. Some previous reports have shown the similar results for
4
Pb2+and MB [41, 42]. After investigation of pH effect later work was performed in natural pH of
5
solution (5.2 to 5.7).
6 7 8 9
3.5. Response surface methodology (RSM) Several classes of RSM such as central composite design (CCD), Box–Behnken design
10
and three-level factorial design have different properties and characteristics. Amongst the three,
11
CCD is a more popular technique applicable for parameter optimization [43], estimation and
12
evaluation of the main and interaction of variables with least number of experiments (Table 1).
13
The polynomial regression equation is most prominent relation for analysis of correlation
14
between variable and responses.
15
Tables 4 and 5 presents the results of the analysis of variance (ANOVA) and regression
16
coefficients of pb2+ ion and MB that reveal high contribution of the quadratic model for
17
explanation of adsorption behavior (significant with p < 0.05). The lack of fit (LOF) as a symbol
18
of data variation around the fitted model is criterion for judgment about adequacy of a model for
19
fitting the experimental result. The significance of LOF shows the non-suitability of model for
20
well fitting the experimental data. A p-value of LOF i.e. 0.143 (Table 4) indicates the good
21
ability of model and best selection of optimum conditions. The suitability of the polynomial
22
model equation for pb2+ ions fitting was expressed by the coefficient of determination (R2=0.
23
0.988 and adjusted R2=0.974), which their value reveal the amount of deviation around the mean.
24
The high adjusted R2 values indicate a good correlation and relationship between the
25
experimental data and the obtained model. ANOVA of MB data (Table 5) show a p-value of
26
LOF around 0.186 that confirm the high efficiency and suitability of model for fitting and
27
explanation of experimental data, while the coefficient of determination for its polynomial model
28
equation is R2=0. 0.972 and adjusted R2 of 0.942. The results were analyzed by regression
29
analysis of CCD (Tables 4 and 5) and make it possible to obtain the following equation for pb2+
30
ions and MB, respectively:
31
9
1
YPb 2 + = 93.11 + 10.58 X 1 + 1.17 X 2 − 7.18 X 3 − 2.65 X 4 − 4.04 X 12 − 1.76 X 32 − 0.93 X 42 + 4.63 X 1 X 3 + 2.52 X 1 X 4
(6)
2 3 4
YMB = 90.96 + 10.89 X 1 2.81X 2 − 4.85 X 3 − 3.53 X 4 − 4.24 X 12 − 2.17 X 1 X 2 + 3.45 X 1 X 3 + 3.47 X 1 X 4
(7)
5 6
The good and successful solving of above equations according to the desirability function (DF)
7
makes it possible to achieve an experimental equation that is able to interpret the behavior of the
8
proposed method for simultaneous removal of pb2+ ions and MB in aqueous samples with at least
9
experiments and high accuracy and repeatability. The plotting of the predicted versus the
10
observed response (R %) and the plot of the residuals versus the predicted response (removal
11
percentage) are shown in Figs. 3a and 3b. Both the figures reveal the presence of linear
12
relationship between them with high correlation coefficient that indicates normally distribution
13
of error around the mean and good applicability of model for explanation of experimental data.
14
These plots are required to check the normality assumption in fitted model.
15
Figures 4 show the most relevant fitted response surfaces for the design and depicts the
16
response surface plots of removal percentage (R%) versus significant variables, while the
17
curvatures of these plots confirm presence of their interaction. The surface plots show that
18
reduce in WAC mass and contact time lead to significant decrease in removal percentage that
19
may be attributed to lower diffusion and decrease in concentration gradient for both species.
20 21
3.6. Optimization of design
22
The profile for predicted values and desirability option in the STATISTICA software is
23
used for the optimization process. The scale in the range of 0.0 (undesirable) to 1.0 (very
24
desirable) must be maximized following best and efficient selection of variables. The CCD
25
optimization design matrix (Fig. 5) show that maximum MB removal percentage (100.0% and
26
desirability of 1.0) and Pb2+ ions removal percentage (99.1% with desirability of 1.0) was
27
achieved at following conditions: adsorbent mass (0.17 g), contact time (110 min), initial Pb2+
28
ions concentration (90 mg L-1) and initial MB concentration (20 mg L-1).
29 30 10
1
3.7. Kinetic study of binary system
2
It is important to predict the removal rate of pollutants from aqueous solutions for the
3
design and optimization of a wastewater treatment [44]. It was seen that the adsorption rate for
4
Pb2+ ions was significantly faster than MB dye, while similar increasing trend was observed for
5
both species (equilibrium was reached at about 110 min for MB and lower than this time was
6
achieved for Pb2+ ions). This phenomenon indicates the rapid external surface adsorption and
7
subsequent slower internal diffusion process which may be the rate-determining step. Different
8
kinetic models including pseudo-first and second order, intraparticle diffusion and elovich model
9
were used to assess the mechanism of adsorption and rate limiting steps such as chemical
10
reaction, diffusion control and mass transport processes. The pseudo-first order equation is given
11
as:
12
ln(qe−qt)=lnqe−K1t
13
where qe (mg/g) and qt (mg/g) are the adsorption capacities at equilibrium and at time t (min),
14
respectively; K1 is the pseudo first- order rate constant (1/min). The pseudo-second order rate
15
equation developed by Ho and McKay [45] assumes that the adsorption capacity depend on
16
surface active sites. The pseudo-second order equation is defined by:
17
t 1 t = + 2 Q t K 2Q e Qe
18
where K2 is the pseudo-second order rate constant (gmg-1 min-1). Due to porous nature of AC, in
19
the first stage it is necessary to investigate the diffusion process. Hence, to explore the behavior
20
of intraparticle diffusion as the rate limiting step in the adsorption process the intraparticle
21
diffusion equation was used. The intraparticle diffusion equation is given as [46]:
22
Qt=Kpt0.5 +C
23
Where Kp is the intraparticle diffusion rate constant (mg(g min1/2)-1) and C show the boundary
24
layer thickness.
25
The linear form of Elovich model equation is generally expressed as [47]:
26
Qt =
27
If the adsorption kinetic fits the Elovich model, a plot of Qt vs. Ln(t) should yield a linear
28
relationship with a slope of (1/β) and an intercept of (1/β) ln (αβ) (see Tables 6 and 7). The
29
pseudo-first order constants (Qe, K1 and R2) for MB and Pb2+ ions adsorption onto WAC were
1
β
ln(αβ ) +
(8)
(9)
(10)
1
β
(11)
ln( t )
11
1
presented in Tables 6 and 7. The theoretical Qe values calculated from the first-order kinetic
2
model did not so close to the experimental values and the correlation coefficients were also
3
found to be slightly lower (Tables 6 and 7). Similar pattern was found for both species. These
4
results indicated that the pseudo-first order kinetic model was not appropriate for modeling the
5
adsorption of MB and Pb2+ ions onto WAC.
6
The slope and intercept of the linear plot of t/Qt versus t give useful information and numerical
7
value of Qe and K2. Judgment based on presented results in Tables 6 and 7 show that the
8
obtained R2 values were higher than 0.99 and the theoretical Qe values were very close to the
9
experimental Qexp values, which confirm high efficiency of this model for explanation of this
10
model for explanation of data, probably through chemisorptions [48]. The h and K2 values
11
calculated from the pseudo-second-order kinetic model were higher for Pb2+ ions than for MB.
12
This result shows higher adsorption rate of Pb2+ ions toward MB for binding to reactive sites of
13
WAC. The not linear behavior of adsorption process over the whole time range show that
14
adsorption occur through more than one mechanism and stage. The initial external diffusion and
15
surface adsorption and following slow rate named as intraparticle diffusion, while the third
16
region attributed to final equilibrium stage. The lower value of R2 for Elovich model confirms
17
that this model is not appropriate to describe the behavior and rate of adsorption system (Tables
18
6 and 7).
19 20
3.8. Isotherm study of single and multi component system
21
Adsorption equilibrium isotherm is designed based on mathematical relation of the
22
amount of adsorbed target per gram of adsorbent (qe (mgg-1)) to the equilibrium non-adsorbed
23
amount of dye in solution (Ce (mgL-1)) at fixed temperature [49, 50]. Isotherm studies are
24
divided to well known models such as Langmuir, Freundlich and Tempkin based on well known
25
conditions. The Langmuir model is the most frequently employed model and given by following
26
equation [51]:
27
qe =
28
Where Ce, Qm and KL are the concentration of adsorbate at equilibrium (mgL-1), maximum mono
29
layer adsorption capacity (mgg-1) and Langmuir constant (Lmg-1), respectively. The suitability of
30
this model judged by equilibrium parameter (RL) defined by the following equation [52]:
Q m K LCe 1 + K LCe
(12)
12
1
RL =
1 1 + K L C0
(13)
2
Ce/qe was plotted against Ce and based on the slope and intercept of such lines, parameters
3
such as Qm, KL, RL and R2 was calculated and displayed in Tables 8 and 9. In this work, an
4
extended Langmuir model in Eq. (10) was employed to fit the experimental data in binary and
5
multi-component systems [53].
6
Q m,i K L, I C e, I
(14)
7
q e,I =
8
where KL, I is the adsorption equilibrium constant of dye or metal ion in mixed system. In
9
adsorption from binary solutions, the amounts of dye and metal ion adsorbed were expressed by
1 + ∑ K L, I C e, I
10
the following equation that (15) and (16) corresponds to the dye and metal ion respectively:
11
q e,1 =
12
q e;2 =
13
According to Eqs. (15) and (16), we have
14
K L,1Q 0,1C e,1
(15)
1 + K L,1C e,1 + K L,2 C e,2 K L,2 Q 0,2 C e,2
(16)
1 + K L,1C e,1 + K L,2 C e,2
K L,2 C e,2 K L,1C e,1
=
Q 0,1q e,2
(17)
q e,1Q 0,2
15
After rearrangement, a linear form of the expanded Langmuir model in binary system was
16
obtained.
17
C e,1 q e,1
=
C e,1 q e,2 C e,1 1 + + K L,1Q 0,1 Q 0,1 q e,1Q 0,2
(18)
18
According to Eq. (18), the values of Ce,1/qe,1 has linear correlation with Ce,1 and Ce,1qe,2/qe,1
19
Q0,2 if the adsorption conformed the expanded Langmuir model by using Eq. (18). The isotherm
20
parameters of MB and Pb2+ ions in the binary solutions were estimated and listed in Tables 10
21
and 11, respectively. Their effect on behavior of adsorption of each other was analyzed by
22
considering P-Factor as correlative technique applicable for multi component systems [54] as
23
following:
13
1
qi =
(Q ) (Q )
m ,i multi solute
(19)
m ,i single solute
2
where (Qm,i ) single solute is the maximum monolayer adsorption capacity for pollutant i in the binary
3
solution and (Qm ,i ) multi solute is the maximum monolayer adsorption capacity of that pollutant with
4
the same initial concentration in a mono-component solution. +, value lower than 1 confirm that
5
the adsorption of pollutant i is accelerated by other pollutants, while the zero value show
6
independent adsorption rate. Finally, the +, value higher than one shows negative role of each
7
species on adsorption of other species [55–57].
8
Our results show that the increment of MB and Pb2+ ions concentrations significantly affects the
9
uptake of each other in the binary system. Note that the adsorption capacities for Pb2+ can be
10
decreased (i.e., +, < 1.0) by the presence of MB in binary solutions. On the other hand, the
11
decrease in uptake of MB in binary solution is higher than Pb2+ ions that confirm by lower value
12
of +-./012.3. 425. compare to +6478 (see Tables 10 and 11).
13
The Freundlich isotherm model [58-60] can be expressed by following equation:
14
qe = KFC1/n
15
where KF is adsorption capacity at unit concentration and 1/n is adsorption intensity. 1/n values
16
indicate the type of isotherm to be irreversible (1/n=0), favorable (0<1/n<1), unfavorable
17
(1/n>1). Eq. (21) can be modified to a linear form:
(20)
18 19
logqe = log KF +(1/n) log Ce
(21)
20 21
The applicability of this model for explanation of real experimental data can be judged by
22
plotting log qe versus log Ce. The values of KF, 1/n and R2 (correlation coefficient) of single and
23
binary system are shown in Tables 8-11. The linear form of Tempkin isotherm is given as:
24 25
qe= B1lnKT+ B1lnCe
(22)
26
where B1 =
27
(8.314 J mol-1 K−1) is the universal gas constant [61, 62]. A plot of qe versus lnCe enables the
28
determination of the isotherm constants B1 and KT from the slope and the intercept, respectively.
RT is related to the heat of adsorption, T is the absolute temperature in Kelvin and R b
14
1
KT is the equilibrium binding constant (Lmol-1) corresponding to the maximum binding energy
2
and constant B1is related to the heat of adsorption. The values of KT, B1 and R2 (correlation
3
coefficient) are shown in Tables 8-11.
4
The Dubinin–Radushkevich (D–R) isotherm model was applied to estimate the porosity, free
5
energy and the characteristics of adsorbents [63, 64]. The D–R isotherm is applicable to
6
homogeneous surfaces. Constant adsorption potential is calculated from the following linear
7
equation.
8
lnqe= lnQm− Bε2
9
where B is a constant related to the adsorption energy, Qm is the theoretical saturation capacity
(23)
10
and ε is the Polanyi potential which is generally calculated from Eq. (18).
11
ε = RT ln(1 + 1/Ce)
12
The slope of the plot of lnqe versus ε² gives B (mol2(kJ)-2) and the intercept yields the adsorption
13
capacity, (Qm (mgg-1). The mean free energy of adsorption (E) which concerns to transport of
14
energy of target molecule to the adsorbent surface which is generally evaluated using the
15
following equation:
16
E = (2B)-1/2
17
The calculated values of D–R parameters in single and binary system are shown in Tables 8 -11.
18
The correlation coefficient values (R2) show that the dye removal isotherm using WAC does not
19
follow the Freundlich, Tempkin and D-R isotherms (Table 8 and 9). The linear fit between the
20
Ce/qe versus Ce and calculated correlation coefficients (R2) for Langmuir isotherm model show
21
that the dye removal isotherm can be approximated as Langmuir model (Table 8 and 9). This
22
means that the adsorption of dyes takes place at specific homogeneous sites and a one layer
23
adsorption onto WAC surface.
(24)
(25)
24 25
4. Conclusion
26
In the present study, it was observed that the use of WAC is an efficient, fast and cheap
27
adsorption method for the removal of Pb2+ and MB. Adsorbent was prepared and characterized
28
for its surface properties. The influences of experimental parameters on the Pb2+ and MB
29
removal percentage were investigated by experimental design methodology. CCD method was
30
adopted and the parameters are optimized. The equilibrium and kinetic studies were investigated
15
1
for the adsorption process. The isotherm models such as Langmuir, Freundlich, and Tempkin
2
were evaluated and the equilibrium data were best described by the Langmuir model. The
3
process kinetics can be successfully fitted to the pseudo-second-order kinetic model.
4 5 6
16
1 2 3
References:
4
[2] E. Akyuz, M. Imamoglu, H. Altundag, Atom. Spectrosc. 34 (2013) 146.
5
[3] S.F. Li, Bioresour. Technol. 101 (2010) 2197.
6
[4] N. Bao, Y. Li, Z.T. Wei, G.B. Yin, J.J. Niu, J. Phys. Chem. C 115 (2011) 5708.
7
[5] M. Rafatullah, O. Sulaiman, R. Hashim, A. Ahmad, J. Hazard. Mater. 177 (2010) 70.
8
[6] F. Aydemir, H. Altundag, M. Imamoglu, Fresenius Environ. Bull. 21 (2012) 3589.
9
[7] M. Ghaedi, H. Tavallali, M. Montazerozohori, S.D. Mousavi, S. Khodadoust, M. Soylak,
[1] T. Terdkiatburana, S. Wang, M.O. Tadĕ, Chem. Eng. J. 139 (2008) 437.
10
Fresenius Environ. Bull. 20 (2011) 2783.
11
[8] S. Khodadoust, M. Ghaedi, Spectrochimi. Acta, Part A 133 (2014) 87.
12
[9] S. Khodadoust, N. Cham kouri, Spectrochimi. Acta, Part A 123 (2014) 85.
13
[10] I. Tan, B. Hameed, A. Ahmad, Chem. Eng. J. 127 (2007) 111.
14
[11] S.R. Shukla, R.S. Pai, J. Hazard. Mater. 125 (2005) 147.
15
[12] M.V. Mier, R.L. Callejas, R. Gehr, B.E.J. Cisneros, P.J.J. Alvarez, Water Res. 35 (2001)
16
373.
17
[13] M. Ghaedi, M. Montazerozohori, Z. Andikaey, A. Shokrollahi, S. Khodadoust, M. Behfar,
18
S. Sharifi, Int. J. Electrochem. Sci, 6 (2011) 4127.
19
[14] D.H. Lee, H. Moon, J. Chem. Eng. 18 (2001) 247.
20
[15] M. Ghaedi, M. Montazerozohori, M. Behfar, S. Khodadoust, Z. Andikaey, M. Nejati Biareh,
21
Sensor letters 9 (2011) 1718.
22
[16] S. Wang, E. Ariyanto, J. Colloid Interface Sci. 314 (2007) 25. [17] S.R. Shukla, R.S. Pai, Sep. Purif. Technol. 43 (2005) 1.
23
[18] M. Ghaedi, H. Tavallali, M. Montazerozohori, E. Zahedi, M. Amirineko, S. Khodadoust, G.
24
Karimipour, Environ. Monit. Assess 184 (2012) 6583.
25
[19] E.N.E. Qada, S.J. Allen, G.M. Walker, Chem. Eng. J. 124 (2006) 103.
26
[20] J. Acharya, J.N. Sahu, C.R. Mohanty, Chem. Eng. J. 149 (2009) 249.
27
[21] A.L. Cazetta, A.M.M. Vargas, E.M. Nogami, M.H. Kunita, M. R. Guilherme, A. C. Martins,
28
T.L. Silva, J.C.G. Moraes, V. C. Almeida, Chem. Eng. J. 174 (2011) 117.
29
[22] T. Madrakian, A. Afkhami, M. Ahmadi, Spectrochimi Acta Part A 99 (2012) 102.
30
[23] K.K Wong, C.K Lee, K.S Low, M.J Haron, Process Biochem. 39 (2003) 437.
31
[24] K. Sun, J.c. Jiang, Biomass Bioenergy 34 (2010) 539. 17
1
[25] M. Ghaedi, M. Pakniat, Z. Mahmoudi, S. Hajati, R. Sahraei, A. Daneshfar, Spectrochimi,
2
Acta, Part A 123 (2014) 402.
3
[26] J. Jaramillo, P.M. Álvarez, V.Gómez-Serrano, Fuel Process. Technol. 91 (2010) 1768.
4
[27] C. Moreno-Castilla, M.A. Ferro-Garcia, J.P. Joly, I. Bautista-Toledo, F. Carrasco-Marin, J.
5
Rivera-Utrilla, Langmuir 11 (1995) 4386.
6
[28] L. Liu, Q. Deng, Y. Liu, T. Ren, Z. Yuan, Catal. Commun. 16 (2011) 81.5
7
[29] M.I. Kandah , R. Shawabkeh , M.A. Al-Zboon, Appl. Surf. Sci. 253 (2006) 821.
8
[30] O.A. Ekpete, M. J.N.R. Horsfall, Res. J. Chem. Sc. 1 (2011) 10.
9
[31] S. Khodadoust, M.R. Hadjmohammadi, Anal. Chim. Acta 699 (2011) 113.
10
[32] S. Khodadoust, M. Ghaedi, J. Sep. Sci. 36 (2013) 1734.
11
[33] M. Roosta, M. Ghaedi, A. Daneshfar, R. Sahraei, A. Asghari, Ultrason. Sonochem. 21(1) (
12
2014) 242.
13
[34] G. Derringer, R. Suich, J. Qual. Tech. 12 (1980) 214.
14
[35] S. Khodadoust, M. Ghaedi, M.R. Hadjmohammadi, Talanta 116 (2013) 637.
15
[36] S. Mopoung, W.Nogklai, Int. J. Phys. Sci. 3 (2008) 234.
16
[37] S.Z.Mohammadi, M.A. Karimi, D. Afzali , F. Mansouri, Desalination 262 (2010) 86.
17
[38] I.I. Salame, T. J. Bandosz, J. Colloid Interface Sci. 240 (2001) 256.
18
[39] Jr. Strelko, D.J. Malik, M. Streat, Carbon 40 (2002) 95.
19
[40] Z.Jiang, Y. Liu, X.Sun, F. Tian, F. Sun, C.Liang, W. You, C. Han, C. Li, Langmuir 19
20
(2003) 7.
21
[41] M. Momčilović , M. Purenović, A. Bojić, A. Zarubica, M. Ranđelović, Desalination 276
22
(2011) 53.
23
[42] L. Aia, Ch. Zhangb, F. Liaoa, Y. Wanga, M. Li, L. Meng, J. Jiang, J. Hazard. Mater. 198
24
(2011) 282.
25
[43] A.A. Ahmad, B.H. Hameed, A.L. Ahmad, J. Hazard. Mater. 170 (2009) 612.
26
[44] X. Li, Y. Li, S. Zhanga, Z.Yeb, Chem. Eng. J. 183 (2012) 88.
27
[45] Y.S. Ho, G. McKay, Trans. Inst. Chem. Eng. 76 B (1998) 183.
28
[46] W.J. Weber, J.C. Morris, J. Sanit. Eng. Div. ASCE 89 (1963) 31.
29
[47 ] Y.S. Ho, G. McKay, Process Biochem. 34 (1999) 451.
30
[48] X. Hu, J. Wang, Y. Liu, X. Li, G. Zeng, Z. Bao, X. Zeng, A. Chen, F. Long, J. Hazard.
31
Mater. 185 (2011) 306. 18
1
[49] Ghaedi, M., Spectrochimi. Acta, Part A 94 (2012) 346.
2
[50] Y. S. Ho, G. Mckay, Wat. Res. 34 (2000) 735.
3
[51] I. Langmuir, J. Am. Chem. Soc. 38 (1916) 2221.
4
[52] S.K. Das, J. Bhowal, A.R. Das, A.R. Guha, Langmuir 22 (2006) 7265.
5
[53] K.K.H. Choy, J.F. Porter, G. McKay, Chem. Eng. J. 103 (2004) 133.
6
[54] M. Ghaedi, S. Hajati, B. Barazesh, F. Karimi, G. Ghezelbash, J. Ind. Eng. Chem. 19 (2013)
7
227.
8
[55] C.A. Basar, J. Hazard. Mater. 135 (2006) 232.
9
[56] Y.S. Ho, Wat. Res. 40 (2006) 119.
10
[57] S. Hajati, M. Ghaedi, F. Karimi, B. Barazesh, R. Sahraei, A. Daneshfar, J. Ind. Eng. Chem.
11
20 (2014) 564.
12
[58] H.M.F. Freundlich, Zeitschrift fur Physikalische Chemie 57 (1906) 385.
13
[59] E.R. Alley, Water Quality Control Handbook, McGraw-Hill Education, Europe, London,
14
2000, pp. 125–141.
15
[60] L.D. Benefield, J.F. Judkins, B.L. Weand, Process Chemistry for Water and Wastewater
16
Treatment, Prentice Hall, Englewood Cliffs, New Jersey, 1982, pp. 191–210.
17
[61] G. Akkaya, A. Ozer, Process Biochem. 40 (2005) 3559.
18
[62] Y.S. Ho, G. McKay, Chem. Eng. J. 70 (1998) 115.
19
[63] M.M. Dubinin, Chem. Rev. 60 (1960) 235.
20
[64] M.M. Dubinin, Zhurnal Fizicheshoi Khimii 39 (1965) 1305.
21 22 23 24
19
Figure captions:
1 2
Fig. 1. FT-IR spectrum of WAC.
3
Fig. 2. Effect of pH on removal of 15 mg L-1 MB and 50 mg L-1 Pb2+ in binary solution by 0.05
4
and 0.1 g WAC.
5
Fig. 3. a) The experimental data versus the predicted data of normalized removal of Pb2+, b) the
6
experimental data versus the predicted data of normalized removal of MB.
7
Fig. 4. Response surfaces for the 24 central composite designs (a) X 1– X 3 (Pb2+); (b) X 3– X
8
(Pb2+); (c) X 4– X 3 (Pb2+); (d) X 1– X 3 (MB) (e) X 2– X 3 (MB); (f) X 4– X 3 (MB).
9
Fig. 5. Profiles for predicated values and desirability function for removal percentage of Pb2+ and
10
MB. Dashed line indicated current values after optimization.
11 12
20
2
1 2 3
Table 1. Design matrix for the 24 central composite designs. Factors (X1) Amount of adsorbent (g) (X2) Contact time (min) (X3) pb2+ concentration (mol L-1) (X4) MB concentration (mol L-1)
4
Run (X1) (X2) 1 -1 -1 2 -1 -1 3 -1 -1 4 -1 -1 5 -1 +1 6 -1 +1 7 -1 +1 8 -1 +1 9 +1 -1 10 +1 -1 11 +1 -1 12 +1 -1 13 +1 +1 14 +1 +1 15 +1 +1 16 +1 +1 17 -2 0 18 +2 0 19 0 -2 20 0 +2 21 0 0 22 0 0 23 0 0 24 0 0 25 (C) 0 0 26 (C) 0 0 27 (C) 0 0 28 (C) 0 0 ER%: extraction recovery percentage
5
(C): Center point
Low (−1) 0.15 120 100 20
(X3) -1 -1 +1 +1 -1 -1 +1 +1 -1 -1 +1 +1 -1 -1 +1 +1 0 0 0 0 -2 +2 0 0 0 0 0 0
Levels Central(0) 0.11 90 80 15 (X4) -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 0 0 0 0 0 0 -2 +2 0 0 0 0
6 7 8 9 10 21
High(+1) 0.07 60 60 10
R metal % 93.40 80.79 66.6 60.66 95.6 84.25 68.48 60.12 98.52 100.00 92.33 92.53 98.90 100.00 94.64 93.9 53.50 100.00 88.01 96.49 98.38 73.35 96.08 82.32 92.74 94.58 93.42 91.70
Star point α= 2 -α +α 0.19 0.03 150 30 120 40 25 5 R dye % 90.89 70.42 76.15 55.37 98.54 87.66 75.56 70.73 99.67 99.82 94.65 95.37 99.81 99.81 98.54 96.21 48.28 99.7 88.12 99.59 99.88 83.65 99.34 86.13 90.81 90.63 93.65 88.75
1 2 3
Table 2. BET parameter for WAC. Parameter Surface Area Single point surface area at p/p° = 0.206301245 BET Surface Area: Langmuir Surface Area: t-Plot Micropore Area: t-Plot External Surface Area: BJH Adsorption cumulative surface area of pores between 17.000 Å and 3000.000 Å width: BJH Desorption cumulative surface area of pores between 17.000 Å and 3000.000 Å width: Pore Volume Single point adsorption total pore volume of pores less than 1256.713 Å width at p/p° = 0.984352481: t-Plot micropore volume: BJH Adsorption cumulative volume of pores between 17.000 Å and 3000.000 Å width: BJH Desorption cumulative volume of pores between 17.000 Å and 3000.000 Å width: Pore Size Adsorption average pore width (4V/A by BET): BJH Adsorption average pore width (4V/A): BJH Desorption average pore width (4V/A):
4 5 6 7 8 9 10 11 12 22
Amount 32.5869 m²/g 31.9078 m²/g 42.7704 m²/g 21.5314 m²/g 10.3764 m²/g 3.585 m²/g 1.9991 m²/g
0.029029 cm³/g
0.009946 cm³/g 0.018727 cm³/g 0.013268 cm³/g
36.3909 Å 208.962 Å 265.473 Å
1 2
Table 3. Results of Boehm titration method for WAC Parameter Acid soluble Water soluble
Amount NO NO 0.857 mmol g-1 0.3 mmol g-1 0.268 mmol g-1 NO
Carboxylic(acidic functions) Phenol Lactones Basic sites 3 4 5 6 7 8 9 10 11 12 13 14 15
23
1 2 3
Table 4.Analysis of variance (ANOVA) for central composite design of pb2+.
4 5
Source of Sum of square Dfa Mean square variation X1 2686.474 1 2686.474 X2 32.713 1 32.713 X3 1236.396 1 1236.396 X4 169.283 1 169.283 X12 392.406 1 392.406 X22 0.682 1 0.682 X32 74.748 1 74.748 X42 20.804 1 20.804 X1 X2 0.540 1 0.540 X1 X3 343.732 1 343.732 X1 X4 101.506 1 101.506 X2 X3 0.065 1 0.065 X2 X4 0.384 1 0.384 X3 X4 2.673 1 2.673 Lack of Fit 57.512 10 5.751 Pure Error 4.382 3 1.461 Total SS 5069.034 27 a Df: Degrees of freedom b Test forcomparing modelvariance with residual (error) variance
6 7 8 9 10 11 12 13 14 15 16 17 24
F-valueb
P value
1839.211 22.396 846.460 115.894 268.648 0.467 51.174 14.243 0.370 235.325 69.493 0.045 0.263 1.830 3.937
0.000028 0.017884 0.000089 0.001714 0.000494 0.543523 0.005626 0.032577 0.586038 0.000602 0.003618 0.846413 0.643339 0.269036 0.143212
1 2 3
Table 5.Analysis of variance (ANOVA) for central composite design of MB dye.
4 5
Source of Sum of square Dfa Mean square variation X1 2847.082 1 2847.082 X2 189.619 1 189.619 X3 565.510 1 565.510 X4 299.909 1 299.909 X12 432.353 1 432.353 X22 12.506 1 12.506 X32 0.954 1 0.954 X42 4.686 1 4.686 X1 X2 75.690 1 75.690 X1 X3 191.546 1 191.546 X1 X4 192.516 1 192.516 X2 X3 1.904 1 1.904 X2 X4 31.192 1 31.192 X3 X4 0.990 1 0.990 Lack of Fit 129.350 10 12.935 Pure Error 12.252 3 4.084 Total SS 5100.276 27 a Df: Degrees of freedom b Test forcomparing modelvariance with residual (error) variance.
6 7 8 9 10 11 12 13 14 15 16 17 25
F-valueb
P value
697.1534 46.4312 138.4743 73.4376 105.8686 3.0624 0.2336 1.1475 18.5339 46.9030 47.1405 0.4663 7.6379 0.2424 3.1673
0.000119 0.006465 0.001319 0.003340 0.001958 0.178430 0.661945 0.362598 0.023067 0.006372 0.006327 0.543691 0.069937 0.656241 0.186186
1 2 3
Table 6.kinetic parameters of Mb removal using (0.14, 0.17, 0.2 g) adsorbent over concentration in the range of 18-27 mg/L 0.17g
0.14g
0.2g
18.12mg/l
22mg/l
27.21mg/l
18.12mg/l
22mg/l
27.21mg/l
18.12mg/l
22mg/l
27.21mg/l
K1
0.0990
0.0599
0.0299
0.1013
0.0714
0.0714
0.1013
0.1957
0.0875
qe (cal)
7.430
10.447
7.430
2.223
3.491
10.139
0.475
1.574
3.319
R
0.981
0.906
0.981
0.930
0.957
0.960
0.834
0.81
0.907
Second order
K2
0.0239
0.0109
0.0068
0.0686
0.0343
0.0137
0.2395
0.1473
0.0302
kinetic model
qe (cal)
6.944
8.547
10.526
5.55
6.802
8.695
4.386
5.618
7.194
R2
0.998
0.995
0.991
0.999
0.998
0.997
0.999
0.998
0.997
H
1.152
0.798
0.755
2.118
1.587
1.037
4.608
4.651
1.565
Kdif
0.393
0.522
0.693
0.229
0.333
0.523
0.121
0.578
0.390
C
3.189
2.881
2.730
3.546
3.646
2.203
3.626
2.751
3.551
R2
0.878
0.961
0.976
0.695
0.783
0.901
0.477
0.692
0.751
β
0.945
0.707
0.537
1.513
0.971
0.686
2.688
1.064
0.875
α
6.708
3.389
2.575
38.632
11.080
4.130
1539.319
16.512
6.715
0.955
0.986
0.984
0.863
0.926
0.976
0.676
0.886
0.900
Models First order kinetic model:
Intraparticle diffusio
Elovich
Parameters
2
2
R
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 26
1 2 3
Table 7. kinetic parameters of Pb2+ removal using (0.14, 0.17 and 0.2 g) adsorbent and 80, 97 120 mg/L of MB.
4 0.17g
0.14g
Models First order kinetic
0.2g
Parameters
80 mg/l
97 mg/l
120 mg/l
80 mg/l
97 mg/l
120 mg/l
80 mg/l
97 mg/l
120 mg/l
K1
0.0368
0.0184
0.0138
0.0322
0.0276
0.0207
0.018
0.0184
0.0276
qe (cal)
4.305
6.426
7.362
2.213
5.164
6.252
1.288
2.307
3.908
model:
R
0.974
0.917
0.877
0.908
0.969
0.969
0.925
0.871
0.772
Second
K2
0.0231
0.0131
0.0116
0.0464
0.0161
0.0128
0.125
0.0390
0.0214
qe(cal)
28.571
32.258
38.461
23.809
28.571
33.33
20.000
24.390
29.412
order
2
kinetic
R
0.999
0.999
0.999
0.999
0.999
0.999
0.999
0.999
0.999
model:
H
18.868
13.699
17.241
26.316
13.158
14.286
50.000
23.256
18.518
Kdif
0.516
0.804
0.836
0.610
0.715
0.212
0.327
0.631
C
23.03
24.35
28.69
21.88
25.61
18.15
20.82
23.35
2
R
0.912
0.894
0.867
0.310 20.53 0.833
0.903
0.930
0.925
0.810
0.792
β
0.721
0.443
0.410
1.180
0.590
0.505
1.848
1.059
0.557
α
7576399
31637.84
89670.16
8.54E+09
204272.2
256045.8
7.55*10^9
9.72*10^9
228872.2
R2
0.980
0.982
0.965
0.933
0.973
0.987
0.903
0.941
0.890
Intraparticle diffusion
Elovich
2
5 6 7 8 9 10 11 12 13 14 15 16 17 27
1 2 3 4 5 6
Table 8. Parameters of different Isotherms for methylene blue in single component system. Models
Parameter
Langmuir Ce/qe= 1/KaQm + Ce/Qm
Qm (mg g1 ) Ka (L mg 1 ) RL
Freundlich Log qe = log KF+(1/n)log Ce
Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
R2 1/n KF (L mg 1 ) R2 Bl KT (L mg 1 ) R2 Qs (mg g-1) K R2
0.14 g
0.17 g
0.2 g
18.518
16.393
16.393
27.000
20.333
15.250
0.0019
0.0033
0.0121
-
-
-
0.6376 0.999 0.137
0.7484 0.999 0.168
0.6798 0.999 0.202
3.161
3.022
3.065
0.835 1.691
0.862 1.750
0.708 2.016
6870.61
2115.70
911.85
0.895
0.907
0.834
17.778 0.000000009 0.915
15.440 0.00000001 0.979
16.610 0.00000001 0.844
7 8 9 10 11 12 13 14 28
1 2 3 4 5 6 7
Table 9. Parameters of different Isotherms for Pb2+ ions in single component system
Models Langmuir Ce/qe= 1/KaQm + Ce/Qm
Freundlich Ln qe = lnKF+(1/n)lnCe Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
Parameter -1
Qm (mg g ) Ka (L mg -1) RL
R2 1/n KF (L mg -1) R2
Bl KT (L mg -1) R2 Qs (mg g-1) K R2
8 9 10 11 12 13 14 15 16 17 18
29
0.14 g 58.823 0.148 0.0263 0.0779 0.992 0.192 21.978 0.976
0.17 g 50.000 0.250 0.0157 0.0476 0.995 0.187 21.330 0.99
0.2 g 47.619 0.318 0.0124 0.0378 0.997 0.194 19.952 0.991
7.571 10.183 0.975 45.604 0.000002 0.801
6.459 19.302 0.996 39.805 2E-07 0.732
5.973 23.045 0.997 35.837 0.0000003 0.717
1 2 3 4 5 6
Table 10. Parametersof different Isotherms for methylene blue in multi-component system
Models
Parameter
Langmuir Ce/qe= 1/KaQm + Ce/Qm
RL
0.14 g 11.236 9.889 0.0040 0.6203
0.17 g 10.638 7.833 0.0062 0.9241
0.6067
0.6489
R 1/n KF (L mg -1) R2 Bl KT (L mg -1) R2 Qs (mg g-1)
0.999 0.086 2.596 0.842 0.764 169494.5 0.845 10.697
K
0.00000001
R2
0.933
0.998 0.078 2.545 0.904 0.620 1480758.0 0.944 9.964 0.00000000 5 0.976
Qm (mg g-1) Ka (L mg -1)
+,
2
Freundlich Ln qe = lnKF+(1/n)lnCe Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
7 8 9 10 11 12 13 14 15 16 17 30
0.2 g 10.309 7.461 0.0085 0.8467 0.6288 0.999 0.108 2.450 0.890 0.788 31785.0 0.933 9.516 0.0000000 08 0.907
1 2 3 4
Table 11. Parametersof different Isotherms for Pb2+ in multi-component system
Models
Parameter
Langmuir Ce/qe= 1/KaQm + Ce/Qm
RL
Qm (mg g-1) Ka (L mg -1)
+,
2
Freundlich Ln qe = lnKF+(1/n)lnCe Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
R 1/n KF (L mg -1) R2 Bl KT (L mg -1) R2 Qs (mg g-1) K R2
0.14 g 55.555 0.143 0.0272 0.0804 0.944 0.993 0.181 21.777 0.993 6.885 13.287 0.977 42.098 0.000001 0.602
5 6 7 8 9 10 11 12 13 14 15 16
31
0.17 g 45.454 0.293 0.0134 0.0409
0.2 g 41.666 0.453 0.0087 0.0268
0.910
0.875
0.995 0.146 22.594 0.988 4.784 88.561 0.989 37.003 0.0000003 0.694
0.997 0.148 21.478 0.975 4.360 127.857 0.989 33.448 0.0000001 0.660
1 2 3 4 5 6 7 8 9 10 11
12
13
14 15 16 17
Fig. 1.
18 19 20 21 22 23 32
1 2 3 4 5 6 7
120 100
Removal %
80 60 0.1g Pb
40
0.05g Pb 0.05g MB
20
0.1g MB
0 0
2
4 pH
8 9 10
Fig. 2.
11 12 13 14 15 33
6
8
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Fig. 3.
23 24 25 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Fig. 4 35
1 2 3 4 5
6 7 8
Fig. 5.
9 10
36
1
2 3
37
1 2
3
Highlight
4
5
•
Activated carbon was prepared from walnut wood.
6
•
The significant variables were optimized by using a CCD combined with DF.
7
•
The equilibrium and kinetic studies were investigated for the adsorption process.
8
38
S1386-1425(14)01037-3 http://dx.doi.org/10.1016/j.saa.2014.06.138 SAA 12397
To appear in:
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
Received Date: Revised Date: Accepted Date:
22 April 2014 16 June 2014 29 June 2014
Please cite this article as: M. Ghaedi, H. Mazaheri, S. Khodadoust, S. Hajati, M.K. Purkait, Application of central composite design for simultaneous removal of methylene blue and Pb2+ ions by walnut wood activated carbon, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy (2014), doi: http://dx.doi.org/10.1016/j.saa. 2014.06.138
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3
5
Application of central composite design for simultaneous removal of methylene blue and Pb2+ ions by walnut wood activated carbon
6
M. Ghaedia, H. Mazaheria, S. Khodadoust b, S. Hajatic, M. K. Purkaitd
4
a
7 b
8
Cellular and Molecular Research Center, Yasuj University of Medical Science, Yasuj, Iran c
9 10
Chemistry Department, Yasouj University, Yasouj 75918-74831, Iran
d
Physic Department, Yasouj University, Yasouj, Iran
Department of Chemical Engineering, Indian Institute of Technology Guwahati, Assam-7S1039, India
11
* Corresponding author: Email: [email protected]
12
------------------------------------------------------------------------------
13
Abstract
14
Activated carbon was prepared from walnut wood which was locally available, non-toxic,
15
abundant and cheap. This new adsorbent was characterized using BET, FTIR and SEM. Point of
16
zero charge (pHPZC) and oxygen containing functional groups were also determined. The
17
prepared adsorbent was applied for simultaneous removal of Pb2+ ions and methylene blue (MB)
18
dye from aqueous solution. The prominent effect and interaction of variables such as amount of
19
adsorbent, contact time, concentration of MB and Pb2+ ions was optimized by central composite
20
design. The equilibrium data obtained at optimum condition was fitted to conventional isotherm
21
models and found that Langmuir model was the best fitted isotherm. Kinetic data were fitted
22
using various models. It was revealed that the adsorption rate follows pseudo second order
23
kinetic model and intraparticle diffusion model.
24
Keywords: Central composite design; Simultaneous removal; Methylene blue; Pb2+ ions
25 26 27 1
1
1. Introduction
2
Heavy metal ions with considering their nature and levels enter to various aqueous
3
ecosystems. They may be considering as toxic or poisonous for human health and animals even
4
at low concentrations. On the other hand, their content strongly can be affected by the presence
5
of other compounds, especially dyes. Another category of hazardous compounds and materials
6
are dyes that may generate serious problems and hazards for human and other organism that
7
attributed to their high biotoxicity, mutagenic and carcinogenic effects [1-4]. On the other hand,
8
presence of dyes by hindering from light penetration that leads to reduce in photosynthesis and
9
generation of harmful hazards to other organisms. These associated hazards and problem make a
10
necessity to design and develop new waste water treatment approach to reduce their level below
11
threshold limit. One of the most well known and famous method for dye removal is adsorption
12
that is superior to other conventional technique including biological treatment, adsorption,
13
coagulation/flocculation, chemical oxidation, membrane separation and ion exchange for
14
removal of dyes and heavy metals have been developed [5,6]. Diverse materials such as activated
15
carbon, zeolite, clay, polymer, and nanomaterials have been extensively applied for pollutants
16
removal [7-9].
17
Methylene blue (MB) commonly applied as coloring dye, while it is also used for dying cotton
18
and silk [10]. The harmful impacts of such dye on water make an urgent task to remove them
19
from waste streams before discharge to water resources. Lead and its compound, especially as
20
Pb2+ ions are known as important pollutants even at low concentrations by affecting central
21
nervous system, kidneys, gastrointestinal system [11]. Therefore, the removal of these pollutants
22
as sole component or their combination from waste water is a challenging requirement and need
23
more attention.
24
Previous studies focused on single component removal, while limited reports for simultaneous
25
adsorption of metal ions and dyes on natural zeolite [12-16] and activated carbon [17, 18] have
26
been reported. In this paper, MB and Pb2+ ions was selected as model and their adsorption
27
behavior in binary systems (kinetics and equilibrium) was studied and compared.
28
Activated carbons (amorphous solids with large internal surface areas and pore volumes) simply
29
can be prepared from natural resource such as coal [19], wood [20], coconut shell [21], tea waste
30
[22] and rice husk [23]. This approach is green and environmental friendly behavior due to
31
conversion of unwanted, worthless agricultural waste to useful, low cost, cheep and high surface 2
1
area adsorbent which are able to remove organic chemicals and metals of environmental and /or
2
economic concern from various resources following their carbonization and activation [24, 25].
3
Walnut Tree (agricultural tree) grows rapidly in different regions, while its wood can be
4
applied in carpentry work and fuel. Storm, natural and/or agricultural activities such as pruning
5
lead to cutting and separation of the branches and trunks of many walnut trees. The high
6
abundant of waste wood (walnut tree) make an economic task the burning and putting this tree
7
waste into a sealed container that cause its simple conversion to carbon source. This behavior
8
reveals that the present homemade walnut carbon (HWC) is very cost effective and non-toxic
9
material with at least energy consumption. In this work, carbon following production was
10
activated by concentrated nitric acid was.
11
In addition to significant decrease in the number of experimental runs by central composite
12
design under response surface methodology (RSM), the main and interaction effect of variables
13
including contact time, adsorbent dosage, initial MB and Pb2+ ions concentration simply was
14
investigated and optimum value of each variable was specified.
15 16
2. Experimental
17
2.1. Instruments and reagents
18
Chemicals including NaOH, HCl, NaNO3, Pb(NO3). 2 H2O, Na2CO3 and NaHCO3 with
19
the highest purity available were purchased from Merck (Dermasdat, Germany). MB dye
20
(Sigma–Aldrich) has following information (a) color index number: 52.015, (b) molecular
21
weight: 319.86 g/mol, (c) empirical formula: C16H18N3SCl and (d) λmax: 664 nm. Analyte
22
solutions were prepared by dissolving their appropriate amount in double distilled water, while
23
the pH was adjusted using pH/Ion meter model-686. The absorbance spectra for MB was
24
recorded in the range of 300 nm to 750 nm using Jasco UV-Visible spectrophotometer model V-
25
530 with a fixed slit width of 2 nm and scan speed of 1000 nm/min. Pb2+ ions was determined by
26
atomic absorption spectrophotometer Varian model AA 220 at λ=217 nm. Fourier transform
27
infrared spectroscopy (FTIR in the range of 400–4000 cm−1) of the adsorbent was recorded using
28
FT-IR spectrophotometer (Model: FT-IR JASCO 460 Plus). Spectra obtained were analyzed.
29
The STATISTICA, a statistical package software version 7.0 (Stat Soft Inc., Tulsa, USA)
30
was used for experimental design analysis and their subsequent regression analysis. Statistical
31
analysis of the model was performed to evaluate the analysis of variance (ANOVA). The quality 3
1
of the polynomial model equation was judged statistically by the coefficient of determination R2
2
and its statistical significance was determined by F-test. P-values less than 0.05 were considered
3
to be statistically significant.
4 5
2.2. Multi-component adsorption of Pb2+ ions and MB on natural activated carbon
6
Binary system of MB – Pb2+ ions was used in the adsorption experiments. All
7
experiments were carried out using different amount of adsorbent in 50 ml beakers in a magnetic
8
stirrer at 400 rpm to obtain the optimum conditions (pH, contact time and initial dye and Pb2+
9
ions concentration), while all experiments conducted at room temperature. The effect of pH on
10
the simultaneous removal was investigated by conducting similar experiments as follow: 50 mL
11
of solution of 50 mg L-1 of Pb2+ ion and 15 mg L-1 of the MB was thoroughly mixed with 0.05
12
and 0.10 g of adsorbent at pH values of 2-6 until achievement of the equilibrium. Subsequent
13
optimization of pH, a central composite design (CCD) was applied for the investigation of the
14
influence of variables and their interaction. Adsorption capacities for Pb2+ ions and dye (qi, mg g-
15
1
) on adsorbent were calculated by a mass balance: (C 0, i - C f, i )V
(1)
16
qi =
17
where C0,i and Cf,i is the initial and final concentration (mg L-1) of pollutant i (i.e., dye or Pb2+
18
ions) in the binary solution, V is the solution volume (L) and m indicate the adsorbent amount
19
(g).
m
20 21
2.3. Preparation of activated carbon
22
Dry branches of walnut tree were grinded and cut into small pieces (lower than one
23
centimeter). They were washed with distilled water and triton X-100 to remove impurities and
24
subsequently were dried in outdoors for 72 hours. Then, 500 g of dried sample were placed into a
25
container with a small hole (size about 5 mm) and about 2 hours directly was heated in flame.
26
Obtained coal washed again with distilled water and placed at 105 ° C for 24 h to dry that lead to
27
conversion of raw material to carbon.
28
This carbon was milled and sieved in the mesh range of 50-60 and activated thoroughly by
29
addition of nitric acid solution as follows: 10 g of carbon was mixed with 150 mL of 5 mol L-1
30
nitric acid and the mixture was refluxed for 6 h at 105 ° C [26-28]. This HNO3-activated carbon 4
1
filtered and washed with deionized water at 50 °C until approximate neutral pH and finally dried
2
at 105 ° C for 24 h. This adsorbent was used for the simultaneous adsorption of MB and Pb2+
3
ions. The pH corresponding to the point of zero charge (pHZPC) of the adsorbent was determined
4
by the pH drift method reported elsewhere [29]. Determination of oxygen containing functional
5
groups was performed by the Boehm titration method [30]. In this method, 1.0 g of the activated
6
carbon (AC) were kept in contact with 15 mL solution of NaHCO3 (0.1M), Na2CO3 (0.05M) and
7
NaOH (0.1M) to identify its acidic groups and 0.1M HCl for basic groups /sites (in separate
8
experiments) respectively, at room temperature for more than 2 days. Subsequently, the aqueous
9
solutions were back titrated with HCl (0.1M) for acidic and NaOH (0.1M) for basic groups. The
10
number and type of acidic sites were calculated by considering that NaOH neutralizes
11
carboxylic, lactonic and phenolic groups; Na2CO3 neutralizes carboxylic and lactonic groups,
12
while NaHCO3 neutralizes only carboxylic groups. Carboxylic groups were therefore quantified
13
by direct titration with NaHCO3. The difference between the groups titrated with Na2CO3 and/or
14
NaHCO3 approximately indicate amount of lactones group and the difference between the groups
15
titrated with NaOH and Na2CO3 reveal the phenol content. Basic sites were determined by
16
titration with HCl. Neutralization points were known using potentiometric titration methods. In
17
order to neutralize basic groups /sites, remaining HCl in the solution was back titrated with 0.1M
18
NaOH.
19 20
2.4. Central composite design
21
Central composite design (CCD) was used to study the individual and synergetic effect of
22
the four factors towards two responses. This method can reduce the number of experimental
23
trials required to evaluate main effect of each parameter and their interactions [31]. In general,
24
CCD is characterized by three operations namely: 2n axial runs, 2n factorial runs and central
25
runs. The CCD in present research is designed based on carrying out 8 axial points, 16 factorial
26
points and 4 replicates at the center (28 experiments) according to value reported in Table 1.
27
Total number of experiments =2n +2n+nc
28
where n is the number of factors, nc is the number of center points. Alpha (α) approximately
29
show the distance of the axial from center point, which is rotatable and strongly depends on the
30
number of factorial points and can be calculated from following equation [32]:
(1)
5
1
α=Np1/4
2
Present experimental design is based on variables including amount of adsorbent (X1), contact
3
time (X2), concentration of Pb2+ ions (X3) and MB (X4) in randomized fashion to minimize the
4
effects of the uncontrolled factors [33].
5
The optimal conditions for the responses (MB and Pb2+ ions removal percentage) were
6
determined using the optimal model predictor quadratic equation given as:
7
Y = b0 + ∑ bi x i + ∑ bii xi2 + ∑
(2)
n
n
i =1
i =1
n
n
∑b
ij
(3)
xi x j
i =1 j = i +1
8
where Y is the predicted response (removal percentage); Xi’s are the independent variables that
9
are known for each experimental run. The parameter b0 is the model constant; bi is the linear
10
coefficient; bii are the quadratic coefficients and bij is the interaction coefficients. The regression
11
analysis is used to fit the equations for both responses to estimate the statistical significance of
12
the equation obtained following analysis of the experimental data was analyzed using the
13
STATISTICA software.
14 15
2.5. Desirability function (DF)
16
After generation of the polynomial equations based on explanation of mathematical
17
relationship between removal percentage and individual variables, the desirability function (DF)
18
was used to find the best optimum levels for each variable [34]. The DF value is in the range of
19
0, 0.5 and 1. Based on analysis of experimental response and their conversion to desirability
20
function (dfi), it was explored that has value in the range of 0 to 1. The 1 and 0 indicate the
21
maximum and minimum desirability, respectively based on the Derringer and Suich [34]
22
equation as follow:
23
df =
, α≤U≤β
(4)
df = 1 , U >
df = 0 , U <
24
In Eq. (4), α and β are the lowest and the highest values, respectively, (obtained for the response
25
i) and wi is the weight. The individual desirability scores for the predicted values for each
26
dependent variable are then combined into overall DF by computing their geometric mean of
27
different dfi values. 6
1
DF = df × df … × df !
"
, 0 ≤ v ≤ 1 (i = 1, 2 … n)
(5)
) v = 1 *
2
Where dfi indicate the desirability of the response Ui (i = 1, 2, 3, . . ., n) and vi represents
3
the importance of responses. Inspecting the desirability profiles, it determines which levels of the
4
predictor variables produce the most desirable predicted responses on the dependent variables
5
[35].
6 7
3. Results and discussion
8
3.1. Adsorbent characterization
9
FTIR, pH determination at zero point of charge [29] (pHZPC) and Boehm titration method
10
[30] were used to characterize this new adsorbent. The characteristic functional groups of the
11
walnut activated carbon (WAC) were investigated using FTIR spectra. FTIR spectrum of WAC
12
(Fig. 1) represents the spectrum bands of ν(O-H) at 3435.56, ν(C-H) at 2925.48 cm-1, ν(C=O)
13
at 1716.34 cm-1, ν(C=C) at 1619.91 cm-1, ν(C-O) or C-C or C=O at 1230.27-921.82 cm-1, δ(C-
14
H) at 1456.66, and γ(C-H) at 708.92-527.70 cm-1. The band 3435.56 cm-1 is attributed to the
15
hydroxyl groups in phenolic and aliphatic structures. The band 2925.48 cm-1 is CH stretching in
16
aromatic methoxyl groups and in methyl and methylene groups of side chains. The band of the
17
carbonyl group in ketones, aldehydes or carboxyl appears at 1716.91 cm-1. The band at 1619.91
18
cm-1 may originate from carbohydrates. The bands at 1230.27 - 921.82 cm-1 is associated with C-
19
O, C-C stretching and C-OH bending in structure of activated carbon [36]. Some surface
20
properties such as pore size and volume were studied by surface area analysis are summarized in
21
Table 2. The porosity and surface area of the WAC show high surface area (1012.75 m2 g-1), that
22
may be attributed to activated carbon and presence of low sized WAC.
23 24
3. 2. pHzpc
25
Distribution of AC charge has significant and important role on the removal and
26
interaction of various compounds with adsorbent. The adsorbent charge (positive or negative)
27
influences the interaction of surface with solute ions or molecules by changing the mechanism
28
and correspond forces. The specified pH known as zero point of charge (ZPC), adsorbent surface 7
1
is neutral [37] and the prominent mechanism for solute transfer is the diffusion into the adsorbent
2
micro and meso-pores. At pH above this value, the surface becomes negative and the positive
3
ions attracted on the surface according to electrostatic attraction. The estimated value of pHzpc
4
(1.9) show he high tendency of these cationic species for strong adsorption on present adsorbent.
5 6 7 8
3. 3. Determination of oxygen containing functional groups
9
the Boehm titration technique [38] based on following principle: only strongly acidic carboxylic
10
groups are neutralized by sodium bicarbonates (Na2HCO3), lactonic and carboxylic group
11
neutralized by sodium carbonate (Na2CO3). The weakly acidic phenolic groups simply can be
12
determined following titration by sodium hydroxide. Neutralization with hydrochloric acid (HCl)
13
makes possible the amount of surface basic group’s of AC such as pyrones [39, 40].
14
Characterization of the prepared adsorbent was performed and the results are presented in Table
15
3. The high amount of various polarity functional groups creates various reactive sites for
16
adsorption of different polarity group with various sizes.
The type and amount of WAC surface functional groups can be determined routinely by
17 18 19 20
3.4. Effect of pH on removal efficiency
21
various pH over the range of 2-7 (Fig. 2). For Pb2+ ions, adsorption increases with increasing pH,
22
while at low pHs due to abundance of proton concentration a noticeable repulsive force and
23
competition between Pb2+ ions and proton for binding onto the surface was appeared that lead to
24
the decrease in both species removal percentage. Similar behavior and trend for MB with lower
25
slope was observed. It is well known that heavy and transition metal ions will react with
26
hydroxyl species in basic solution to for charged species including M(OH)+ or non- soluble
27
species such as M(OH)2 which strongly affect the nature, mechanism and adsorption efficiency.
28
MB (basic dye) form cation (C+) and reduced ions (CNH+) in solution in acidic pH that hinder
29
from its adsorption onto adsorbent surface. Another most probable mechanism for adsorption
30
may occur following complexation of metal ions with dye compounds. This reaction may lead to
31
formation of neutral species that are very high ability for adsorption onto adsorbent. This
32
pathway was reduced in acidic media and strongly enhanced following increasing the pH. On the
The effect of pH on the removal of MB and Pb2+ ions by WAC was investigated at
8
1
other hand (as it is known) by rising the pH, adsorption of both species (cations) on adsorbent
2
strongly facilated and their easily association and accumulation on adsorbent surface lead to
3
enhance in removal percentage. Some previous reports have shown the similar results for
4
Pb2+and MB [41, 42]. After investigation of pH effect later work was performed in natural pH of
5
solution (5.2 to 5.7).
6 7 8 9
3.5. Response surface methodology (RSM) Several classes of RSM such as central composite design (CCD), Box–Behnken design
10
and three-level factorial design have different properties and characteristics. Amongst the three,
11
CCD is a more popular technique applicable for parameter optimization [43], estimation and
12
evaluation of the main and interaction of variables with least number of experiments (Table 1).
13
The polynomial regression equation is most prominent relation for analysis of correlation
14
between variable and responses.
15
Tables 4 and 5 presents the results of the analysis of variance (ANOVA) and regression
16
coefficients of pb2+ ion and MB that reveal high contribution of the quadratic model for
17
explanation of adsorption behavior (significant with p < 0.05). The lack of fit (LOF) as a symbol
18
of data variation around the fitted model is criterion for judgment about adequacy of a model for
19
fitting the experimental result. The significance of LOF shows the non-suitability of model for
20
well fitting the experimental data. A p-value of LOF i.e. 0.143 (Table 4) indicates the good
21
ability of model and best selection of optimum conditions. The suitability of the polynomial
22
model equation for pb2+ ions fitting was expressed by the coefficient of determination (R2=0.
23
0.988 and adjusted R2=0.974), which their value reveal the amount of deviation around the mean.
24
The high adjusted R2 values indicate a good correlation and relationship between the
25
experimental data and the obtained model. ANOVA of MB data (Table 5) show a p-value of
26
LOF around 0.186 that confirm the high efficiency and suitability of model for fitting and
27
explanation of experimental data, while the coefficient of determination for its polynomial model
28
equation is R2=0. 0.972 and adjusted R2 of 0.942. The results were analyzed by regression
29
analysis of CCD (Tables 4 and 5) and make it possible to obtain the following equation for pb2+
30
ions and MB, respectively:
31
9
1
YPb 2 + = 93.11 + 10.58 X 1 + 1.17 X 2 − 7.18 X 3 − 2.65 X 4 − 4.04 X 12 − 1.76 X 32 − 0.93 X 42 + 4.63 X 1 X 3 + 2.52 X 1 X 4
(6)
2 3 4
YMB = 90.96 + 10.89 X 1 2.81X 2 − 4.85 X 3 − 3.53 X 4 − 4.24 X 12 − 2.17 X 1 X 2 + 3.45 X 1 X 3 + 3.47 X 1 X 4
(7)
5 6
The good and successful solving of above equations according to the desirability function (DF)
7
makes it possible to achieve an experimental equation that is able to interpret the behavior of the
8
proposed method for simultaneous removal of pb2+ ions and MB in aqueous samples with at least
9
experiments and high accuracy and repeatability. The plotting of the predicted versus the
10
observed response (R %) and the plot of the residuals versus the predicted response (removal
11
percentage) are shown in Figs. 3a and 3b. Both the figures reveal the presence of linear
12
relationship between them with high correlation coefficient that indicates normally distribution
13
of error around the mean and good applicability of model for explanation of experimental data.
14
These plots are required to check the normality assumption in fitted model.
15
Figures 4 show the most relevant fitted response surfaces for the design and depicts the
16
response surface plots of removal percentage (R%) versus significant variables, while the
17
curvatures of these plots confirm presence of their interaction. The surface plots show that
18
reduce in WAC mass and contact time lead to significant decrease in removal percentage that
19
may be attributed to lower diffusion and decrease in concentration gradient for both species.
20 21
3.6. Optimization of design
22
The profile for predicted values and desirability option in the STATISTICA software is
23
used for the optimization process. The scale in the range of 0.0 (undesirable) to 1.0 (very
24
desirable) must be maximized following best and efficient selection of variables. The CCD
25
optimization design matrix (Fig. 5) show that maximum MB removal percentage (100.0% and
26
desirability of 1.0) and Pb2+ ions removal percentage (99.1% with desirability of 1.0) was
27
achieved at following conditions: adsorbent mass (0.17 g), contact time (110 min), initial Pb2+
28
ions concentration (90 mg L-1) and initial MB concentration (20 mg L-1).
29 30 10
1
3.7. Kinetic study of binary system
2
It is important to predict the removal rate of pollutants from aqueous solutions for the
3
design and optimization of a wastewater treatment [44]. It was seen that the adsorption rate for
4
Pb2+ ions was significantly faster than MB dye, while similar increasing trend was observed for
5
both species (equilibrium was reached at about 110 min for MB and lower than this time was
6
achieved for Pb2+ ions). This phenomenon indicates the rapid external surface adsorption and
7
subsequent slower internal diffusion process which may be the rate-determining step. Different
8
kinetic models including pseudo-first and second order, intraparticle diffusion and elovich model
9
were used to assess the mechanism of adsorption and rate limiting steps such as chemical
10
reaction, diffusion control and mass transport processes. The pseudo-first order equation is given
11
as:
12
ln(qe−qt)=lnqe−K1t
13
where qe (mg/g) and qt (mg/g) are the adsorption capacities at equilibrium and at time t (min),
14
respectively; K1 is the pseudo first- order rate constant (1/min). The pseudo-second order rate
15
equation developed by Ho and McKay [45] assumes that the adsorption capacity depend on
16
surface active sites. The pseudo-second order equation is defined by:
17
t 1 t = + 2 Q t K 2Q e Qe
18
where K2 is the pseudo-second order rate constant (gmg-1 min-1). Due to porous nature of AC, in
19
the first stage it is necessary to investigate the diffusion process. Hence, to explore the behavior
20
of intraparticle diffusion as the rate limiting step in the adsorption process the intraparticle
21
diffusion equation was used. The intraparticle diffusion equation is given as [46]:
22
Qt=Kpt0.5 +C
23
Where Kp is the intraparticle diffusion rate constant (mg(g min1/2)-1) and C show the boundary
24
layer thickness.
25
The linear form of Elovich model equation is generally expressed as [47]:
26
Qt =
27
If the adsorption kinetic fits the Elovich model, a plot of Qt vs. Ln(t) should yield a linear
28
relationship with a slope of (1/β) and an intercept of (1/β) ln (αβ) (see Tables 6 and 7). The
29
pseudo-first order constants (Qe, K1 and R2) for MB and Pb2+ ions adsorption onto WAC were
1
β
ln(αβ ) +
(8)
(9)
(10)
1
β
(11)
ln( t )
11
1
presented in Tables 6 and 7. The theoretical Qe values calculated from the first-order kinetic
2
model did not so close to the experimental values and the correlation coefficients were also
3
found to be slightly lower (Tables 6 and 7). Similar pattern was found for both species. These
4
results indicated that the pseudo-first order kinetic model was not appropriate for modeling the
5
adsorption of MB and Pb2+ ions onto WAC.
6
The slope and intercept of the linear plot of t/Qt versus t give useful information and numerical
7
value of Qe and K2. Judgment based on presented results in Tables 6 and 7 show that the
8
obtained R2 values were higher than 0.99 and the theoretical Qe values were very close to the
9
experimental Qexp values, which confirm high efficiency of this model for explanation of this
10
model for explanation of data, probably through chemisorptions [48]. The h and K2 values
11
calculated from the pseudo-second-order kinetic model were higher for Pb2+ ions than for MB.
12
This result shows higher adsorption rate of Pb2+ ions toward MB for binding to reactive sites of
13
WAC. The not linear behavior of adsorption process over the whole time range show that
14
adsorption occur through more than one mechanism and stage. The initial external diffusion and
15
surface adsorption and following slow rate named as intraparticle diffusion, while the third
16
region attributed to final equilibrium stage. The lower value of R2 for Elovich model confirms
17
that this model is not appropriate to describe the behavior and rate of adsorption system (Tables
18
6 and 7).
19 20
3.8. Isotherm study of single and multi component system
21
Adsorption equilibrium isotherm is designed based on mathematical relation of the
22
amount of adsorbed target per gram of adsorbent (qe (mgg-1)) to the equilibrium non-adsorbed
23
amount of dye in solution (Ce (mgL-1)) at fixed temperature [49, 50]. Isotherm studies are
24
divided to well known models such as Langmuir, Freundlich and Tempkin based on well known
25
conditions. The Langmuir model is the most frequently employed model and given by following
26
equation [51]:
27
qe =
28
Where Ce, Qm and KL are the concentration of adsorbate at equilibrium (mgL-1), maximum mono
29
layer adsorption capacity (mgg-1) and Langmuir constant (Lmg-1), respectively. The suitability of
30
this model judged by equilibrium parameter (RL) defined by the following equation [52]:
Q m K LCe 1 + K LCe
(12)
12
1
RL =
1 1 + K L C0
(13)
2
Ce/qe was plotted against Ce and based on the slope and intercept of such lines, parameters
3
such as Qm, KL, RL and R2 was calculated and displayed in Tables 8 and 9. In this work, an
4
extended Langmuir model in Eq. (10) was employed to fit the experimental data in binary and
5
multi-component systems [53].
6
Q m,i K L, I C e, I
(14)
7
q e,I =
8
where KL, I is the adsorption equilibrium constant of dye or metal ion in mixed system. In
9
adsorption from binary solutions, the amounts of dye and metal ion adsorbed were expressed by
1 + ∑ K L, I C e, I
10
the following equation that (15) and (16) corresponds to the dye and metal ion respectively:
11
q e,1 =
12
q e;2 =
13
According to Eqs. (15) and (16), we have
14
K L,1Q 0,1C e,1
(15)
1 + K L,1C e,1 + K L,2 C e,2 K L,2 Q 0,2 C e,2
(16)
1 + K L,1C e,1 + K L,2 C e,2
K L,2 C e,2 K L,1C e,1
=
Q 0,1q e,2
(17)
q e,1Q 0,2
15
After rearrangement, a linear form of the expanded Langmuir model in binary system was
16
obtained.
17
C e,1 q e,1
=
C e,1 q e,2 C e,1 1 + + K L,1Q 0,1 Q 0,1 q e,1Q 0,2
(18)
18
According to Eq. (18), the values of Ce,1/qe,1 has linear correlation with Ce,1 and Ce,1qe,2/qe,1
19
Q0,2 if the adsorption conformed the expanded Langmuir model by using Eq. (18). The isotherm
20
parameters of MB and Pb2+ ions in the binary solutions were estimated and listed in Tables 10
21
and 11, respectively. Their effect on behavior of adsorption of each other was analyzed by
22
considering P-Factor as correlative technique applicable for multi component systems [54] as
23
following:
13
1
qi =
(Q ) (Q )
m ,i multi solute
(19)
m ,i single solute
2
where (Qm,i ) single solute is the maximum monolayer adsorption capacity for pollutant i in the binary
3
solution and (Qm ,i ) multi solute is the maximum monolayer adsorption capacity of that pollutant with
4
the same initial concentration in a mono-component solution. +, value lower than 1 confirm that
5
the adsorption of pollutant i is accelerated by other pollutants, while the zero value show
6
independent adsorption rate. Finally, the +, value higher than one shows negative role of each
7
species on adsorption of other species [55–57].
8
Our results show that the increment of MB and Pb2+ ions concentrations significantly affects the
9
uptake of each other in the binary system. Note that the adsorption capacities for Pb2+ can be
10
decreased (i.e., +, < 1.0) by the presence of MB in binary solutions. On the other hand, the
11
decrease in uptake of MB in binary solution is higher than Pb2+ ions that confirm by lower value
12
of +-./012.3. 425. compare to +6478 (see Tables 10 and 11).
13
The Freundlich isotherm model [58-60] can be expressed by following equation:
14
qe = KFC1/n
15
where KF is adsorption capacity at unit concentration and 1/n is adsorption intensity. 1/n values
16
indicate the type of isotherm to be irreversible (1/n=0), favorable (0<1/n<1), unfavorable
17
(1/n>1). Eq. (21) can be modified to a linear form:
(20)
18 19
logqe = log KF +(1/n) log Ce
(21)
20 21
The applicability of this model for explanation of real experimental data can be judged by
22
plotting log qe versus log Ce. The values of KF, 1/n and R2 (correlation coefficient) of single and
23
binary system are shown in Tables 8-11. The linear form of Tempkin isotherm is given as:
24 25
qe= B1lnKT+ B1lnCe
(22)
26
where B1 =
27
(8.314 J mol-1 K−1) is the universal gas constant [61, 62]. A plot of qe versus lnCe enables the
28
determination of the isotherm constants B1 and KT from the slope and the intercept, respectively.
RT is related to the heat of adsorption, T is the absolute temperature in Kelvin and R b
14
1
KT is the equilibrium binding constant (Lmol-1) corresponding to the maximum binding energy
2
and constant B1is related to the heat of adsorption. The values of KT, B1 and R2 (correlation
3
coefficient) are shown in Tables 8-11.
4
The Dubinin–Radushkevich (D–R) isotherm model was applied to estimate the porosity, free
5
energy and the characteristics of adsorbents [63, 64]. The D–R isotherm is applicable to
6
homogeneous surfaces. Constant adsorption potential is calculated from the following linear
7
equation.
8
lnqe= lnQm− Bε2
9
where B is a constant related to the adsorption energy, Qm is the theoretical saturation capacity
(23)
10
and ε is the Polanyi potential which is generally calculated from Eq. (18).
11
ε = RT ln(1 + 1/Ce)
12
The slope of the plot of lnqe versus ε² gives B (mol2(kJ)-2) and the intercept yields the adsorption
13
capacity, (Qm (mgg-1). The mean free energy of adsorption (E) which concerns to transport of
14
energy of target molecule to the adsorbent surface which is generally evaluated using the
15
following equation:
16
E = (2B)-1/2
17
The calculated values of D–R parameters in single and binary system are shown in Tables 8 -11.
18
The correlation coefficient values (R2) show that the dye removal isotherm using WAC does not
19
follow the Freundlich, Tempkin and D-R isotherms (Table 8 and 9). The linear fit between the
20
Ce/qe versus Ce and calculated correlation coefficients (R2) for Langmuir isotherm model show
21
that the dye removal isotherm can be approximated as Langmuir model (Table 8 and 9). This
22
means that the adsorption of dyes takes place at specific homogeneous sites and a one layer
23
adsorption onto WAC surface.
(24)
(25)
24 25
4. Conclusion
26
In the present study, it was observed that the use of WAC is an efficient, fast and cheap
27
adsorption method for the removal of Pb2+ and MB. Adsorbent was prepared and characterized
28
for its surface properties. The influences of experimental parameters on the Pb2+ and MB
29
removal percentage were investigated by experimental design methodology. CCD method was
30
adopted and the parameters are optimized. The equilibrium and kinetic studies were investigated
15
1
for the adsorption process. The isotherm models such as Langmuir, Freundlich, and Tempkin
2
were evaluated and the equilibrium data were best described by the Langmuir model. The
3
process kinetics can be successfully fitted to the pseudo-second-order kinetic model.
4 5 6
16
1 2 3
References:
4
[2] E. Akyuz, M. Imamoglu, H. Altundag, Atom. Spectrosc. 34 (2013) 146.
5
[3] S.F. Li, Bioresour. Technol. 101 (2010) 2197.
6
[4] N. Bao, Y. Li, Z.T. Wei, G.B. Yin, J.J. Niu, J. Phys. Chem. C 115 (2011) 5708.
7
[5] M. Rafatullah, O. Sulaiman, R. Hashim, A. Ahmad, J. Hazard. Mater. 177 (2010) 70.
8
[6] F. Aydemir, H. Altundag, M. Imamoglu, Fresenius Environ. Bull. 21 (2012) 3589.
9
[7] M. Ghaedi, H. Tavallali, M. Montazerozohori, S.D. Mousavi, S. Khodadoust, M. Soylak,
[1] T. Terdkiatburana, S. Wang, M.O. Tadĕ, Chem. Eng. J. 139 (2008) 437.
10
Fresenius Environ. Bull. 20 (2011) 2783.
11
[8] S. Khodadoust, M. Ghaedi, Spectrochimi. Acta, Part A 133 (2014) 87.
12
[9] S. Khodadoust, N. Cham kouri, Spectrochimi. Acta, Part A 123 (2014) 85.
13
[10] I. Tan, B. Hameed, A. Ahmad, Chem. Eng. J. 127 (2007) 111.
14
[11] S.R. Shukla, R.S. Pai, J. Hazard. Mater. 125 (2005) 147.
15
[12] M.V. Mier, R.L. Callejas, R. Gehr, B.E.J. Cisneros, P.J.J. Alvarez, Water Res. 35 (2001)
16
373.
17
[13] M. Ghaedi, M. Montazerozohori, Z. Andikaey, A. Shokrollahi, S. Khodadoust, M. Behfar,
18
S. Sharifi, Int. J. Electrochem. Sci, 6 (2011) 4127.
19
[14] D.H. Lee, H. Moon, J. Chem. Eng. 18 (2001) 247.
20
[15] M. Ghaedi, M. Montazerozohori, M. Behfar, S. Khodadoust, Z. Andikaey, M. Nejati Biareh,
21
Sensor letters 9 (2011) 1718.
22
[16] S. Wang, E. Ariyanto, J. Colloid Interface Sci. 314 (2007) 25. [17] S.R. Shukla, R.S. Pai, Sep. Purif. Technol. 43 (2005) 1.
23
[18] M. Ghaedi, H. Tavallali, M. Montazerozohori, E. Zahedi, M. Amirineko, S. Khodadoust, G.
24
Karimipour, Environ. Monit. Assess 184 (2012) 6583.
25
[19] E.N.E. Qada, S.J. Allen, G.M. Walker, Chem. Eng. J. 124 (2006) 103.
26
[20] J. Acharya, J.N. Sahu, C.R. Mohanty, Chem. Eng. J. 149 (2009) 249.
27
[21] A.L. Cazetta, A.M.M. Vargas, E.M. Nogami, M.H. Kunita, M. R. Guilherme, A. C. Martins,
28
T.L. Silva, J.C.G. Moraes, V. C. Almeida, Chem. Eng. J. 174 (2011) 117.
29
[22] T. Madrakian, A. Afkhami, M. Ahmadi, Spectrochimi Acta Part A 99 (2012) 102.
30
[23] K.K Wong, C.K Lee, K.S Low, M.J Haron, Process Biochem. 39 (2003) 437.
31
[24] K. Sun, J.c. Jiang, Biomass Bioenergy 34 (2010) 539. 17
1
[25] M. Ghaedi, M. Pakniat, Z. Mahmoudi, S. Hajati, R. Sahraei, A. Daneshfar, Spectrochimi,
2
Acta, Part A 123 (2014) 402.
3
[26] J. Jaramillo, P.M. Álvarez, V.Gómez-Serrano, Fuel Process. Technol. 91 (2010) 1768.
4
[27] C. Moreno-Castilla, M.A. Ferro-Garcia, J.P. Joly, I. Bautista-Toledo, F. Carrasco-Marin, J.
5
Rivera-Utrilla, Langmuir 11 (1995) 4386.
6
[28] L. Liu, Q. Deng, Y. Liu, T. Ren, Z. Yuan, Catal. Commun. 16 (2011) 81.5
7
[29] M.I. Kandah , R. Shawabkeh , M.A. Al-Zboon, Appl. Surf. Sci. 253 (2006) 821.
8
[30] O.A. Ekpete, M. J.N.R. Horsfall, Res. J. Chem. Sc. 1 (2011) 10.
9
[31] S. Khodadoust, M.R. Hadjmohammadi, Anal. Chim. Acta 699 (2011) 113.
10
[32] S. Khodadoust, M. Ghaedi, J. Sep. Sci. 36 (2013) 1734.
11
[33] M. Roosta, M. Ghaedi, A. Daneshfar, R. Sahraei, A. Asghari, Ultrason. Sonochem. 21(1) (
12
2014) 242.
13
[34] G. Derringer, R. Suich, J. Qual. Tech. 12 (1980) 214.
14
[35] S. Khodadoust, M. Ghaedi, M.R. Hadjmohammadi, Talanta 116 (2013) 637.
15
[36] S. Mopoung, W.Nogklai, Int. J. Phys. Sci. 3 (2008) 234.
16
[37] S.Z.Mohammadi, M.A. Karimi, D. Afzali , F. Mansouri, Desalination 262 (2010) 86.
17
[38] I.I. Salame, T. J. Bandosz, J. Colloid Interface Sci. 240 (2001) 256.
18
[39] Jr. Strelko, D.J. Malik, M. Streat, Carbon 40 (2002) 95.
19
[40] Z.Jiang, Y. Liu, X.Sun, F. Tian, F. Sun, C.Liang, W. You, C. Han, C. Li, Langmuir 19
20
(2003) 7.
21
[41] M. Momčilović , M. Purenović, A. Bojić, A. Zarubica, M. Ranđelović, Desalination 276
22
(2011) 53.
23
[42] L. Aia, Ch. Zhangb, F. Liaoa, Y. Wanga, M. Li, L. Meng, J. Jiang, J. Hazard. Mater. 198
24
(2011) 282.
25
[43] A.A. Ahmad, B.H. Hameed, A.L. Ahmad, J. Hazard. Mater. 170 (2009) 612.
26
[44] X. Li, Y. Li, S. Zhanga, Z.Yeb, Chem. Eng. J. 183 (2012) 88.
27
[45] Y.S. Ho, G. McKay, Trans. Inst. Chem. Eng. 76 B (1998) 183.
28
[46] W.J. Weber, J.C. Morris, J. Sanit. Eng. Div. ASCE 89 (1963) 31.
29
[47 ] Y.S. Ho, G. McKay, Process Biochem. 34 (1999) 451.
30
[48] X. Hu, J. Wang, Y. Liu, X. Li, G. Zeng, Z. Bao, X. Zeng, A. Chen, F. Long, J. Hazard.
31
Mater. 185 (2011) 306. 18
1
[49] Ghaedi, M., Spectrochimi. Acta, Part A 94 (2012) 346.
2
[50] Y. S. Ho, G. Mckay, Wat. Res. 34 (2000) 735.
3
[51] I. Langmuir, J. Am. Chem. Soc. 38 (1916) 2221.
4
[52] S.K. Das, J. Bhowal, A.R. Das, A.R. Guha, Langmuir 22 (2006) 7265.
5
[53] K.K.H. Choy, J.F. Porter, G. McKay, Chem. Eng. J. 103 (2004) 133.
6
[54] M. Ghaedi, S. Hajati, B. Barazesh, F. Karimi, G. Ghezelbash, J. Ind. Eng. Chem. 19 (2013)
7
227.
8
[55] C.A. Basar, J. Hazard. Mater. 135 (2006) 232.
9
[56] Y.S. Ho, Wat. Res. 40 (2006) 119.
10
[57] S. Hajati, M. Ghaedi, F. Karimi, B. Barazesh, R. Sahraei, A. Daneshfar, J. Ind. Eng. Chem.
11
20 (2014) 564.
12
[58] H.M.F. Freundlich, Zeitschrift fur Physikalische Chemie 57 (1906) 385.
13
[59] E.R. Alley, Water Quality Control Handbook, McGraw-Hill Education, Europe, London,
14
2000, pp. 125–141.
15
[60] L.D. Benefield, J.F. Judkins, B.L. Weand, Process Chemistry for Water and Wastewater
16
Treatment, Prentice Hall, Englewood Cliffs, New Jersey, 1982, pp. 191–210.
17
[61] G. Akkaya, A. Ozer, Process Biochem. 40 (2005) 3559.
18
[62] Y.S. Ho, G. McKay, Chem. Eng. J. 70 (1998) 115.
19
[63] M.M. Dubinin, Chem. Rev. 60 (1960) 235.
20
[64] M.M. Dubinin, Zhurnal Fizicheshoi Khimii 39 (1965) 1305.
21 22 23 24
19
Figure captions:
1 2
Fig. 1. FT-IR spectrum of WAC.
3
Fig. 2. Effect of pH on removal of 15 mg L-1 MB and 50 mg L-1 Pb2+ in binary solution by 0.05
4
and 0.1 g WAC.
5
Fig. 3. a) The experimental data versus the predicted data of normalized removal of Pb2+, b) the
6
experimental data versus the predicted data of normalized removal of MB.
7
Fig. 4. Response surfaces for the 24 central composite designs (a) X 1– X 3 (Pb2+); (b) X 3– X
8
(Pb2+); (c) X 4– X 3 (Pb2+); (d) X 1– X 3 (MB) (e) X 2– X 3 (MB); (f) X 4– X 3 (MB).
9
Fig. 5. Profiles for predicated values and desirability function for removal percentage of Pb2+ and
10
MB. Dashed line indicated current values after optimization.
11 12
20
2
1 2 3
Table 1. Design matrix for the 24 central composite designs. Factors (X1) Amount of adsorbent (g) (X2) Contact time (min) (X3) pb2+ concentration (mol L-1) (X4) MB concentration (mol L-1)
4
Run (X1) (X2) 1 -1 -1 2 -1 -1 3 -1 -1 4 -1 -1 5 -1 +1 6 -1 +1 7 -1 +1 8 -1 +1 9 +1 -1 10 +1 -1 11 +1 -1 12 +1 -1 13 +1 +1 14 +1 +1 15 +1 +1 16 +1 +1 17 -2 0 18 +2 0 19 0 -2 20 0 +2 21 0 0 22 0 0 23 0 0 24 0 0 25 (C) 0 0 26 (C) 0 0 27 (C) 0 0 28 (C) 0 0 ER%: extraction recovery percentage
5
(C): Center point
Low (−1) 0.15 120 100 20
(X3) -1 -1 +1 +1 -1 -1 +1 +1 -1 -1 +1 +1 -1 -1 +1 +1 0 0 0 0 -2 +2 0 0 0 0 0 0
Levels Central(0) 0.11 90 80 15 (X4) -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 0 0 0 0 0 0 -2 +2 0 0 0 0
6 7 8 9 10 21
High(+1) 0.07 60 60 10
R metal % 93.40 80.79 66.6 60.66 95.6 84.25 68.48 60.12 98.52 100.00 92.33 92.53 98.90 100.00 94.64 93.9 53.50 100.00 88.01 96.49 98.38 73.35 96.08 82.32 92.74 94.58 93.42 91.70
Star point α= 2 -α +α 0.19 0.03 150 30 120 40 25 5 R dye % 90.89 70.42 76.15 55.37 98.54 87.66 75.56 70.73 99.67 99.82 94.65 95.37 99.81 99.81 98.54 96.21 48.28 99.7 88.12 99.59 99.88 83.65 99.34 86.13 90.81 90.63 93.65 88.75
1 2 3
Table 2. BET parameter for WAC. Parameter Surface Area Single point surface area at p/p° = 0.206301245 BET Surface Area: Langmuir Surface Area: t-Plot Micropore Area: t-Plot External Surface Area: BJH Adsorption cumulative surface area of pores between 17.000 Å and 3000.000 Å width: BJH Desorption cumulative surface area of pores between 17.000 Å and 3000.000 Å width: Pore Volume Single point adsorption total pore volume of pores less than 1256.713 Å width at p/p° = 0.984352481: t-Plot micropore volume: BJH Adsorption cumulative volume of pores between 17.000 Å and 3000.000 Å width: BJH Desorption cumulative volume of pores between 17.000 Å and 3000.000 Å width: Pore Size Adsorption average pore width (4V/A by BET): BJH Adsorption average pore width (4V/A): BJH Desorption average pore width (4V/A):
4 5 6 7 8 9 10 11 12 22
Amount 32.5869 m²/g 31.9078 m²/g 42.7704 m²/g 21.5314 m²/g 10.3764 m²/g 3.585 m²/g 1.9991 m²/g
0.029029 cm³/g
0.009946 cm³/g 0.018727 cm³/g 0.013268 cm³/g
36.3909 Å 208.962 Å 265.473 Å
1 2
Table 3. Results of Boehm titration method for WAC Parameter Acid soluble Water soluble
Amount NO NO 0.857 mmol g-1 0.3 mmol g-1 0.268 mmol g-1 NO
Carboxylic(acidic functions) Phenol Lactones Basic sites 3 4 5 6 7 8 9 10 11 12 13 14 15
23
1 2 3
Table 4.Analysis of variance (ANOVA) for central composite design of pb2+.
4 5
Source of Sum of square Dfa Mean square variation X1 2686.474 1 2686.474 X2 32.713 1 32.713 X3 1236.396 1 1236.396 X4 169.283 1 169.283 X12 392.406 1 392.406 X22 0.682 1 0.682 X32 74.748 1 74.748 X42 20.804 1 20.804 X1 X2 0.540 1 0.540 X1 X3 343.732 1 343.732 X1 X4 101.506 1 101.506 X2 X3 0.065 1 0.065 X2 X4 0.384 1 0.384 X3 X4 2.673 1 2.673 Lack of Fit 57.512 10 5.751 Pure Error 4.382 3 1.461 Total SS 5069.034 27 a Df: Degrees of freedom b Test forcomparing modelvariance with residual (error) variance
6 7 8 9 10 11 12 13 14 15 16 17 24
F-valueb
P value
1839.211 22.396 846.460 115.894 268.648 0.467 51.174 14.243 0.370 235.325 69.493 0.045 0.263 1.830 3.937
0.000028 0.017884 0.000089 0.001714 0.000494 0.543523 0.005626 0.032577 0.586038 0.000602 0.003618 0.846413 0.643339 0.269036 0.143212
1 2 3
Table 5.Analysis of variance (ANOVA) for central composite design of MB dye.
4 5
Source of Sum of square Dfa Mean square variation X1 2847.082 1 2847.082 X2 189.619 1 189.619 X3 565.510 1 565.510 X4 299.909 1 299.909 X12 432.353 1 432.353 X22 12.506 1 12.506 X32 0.954 1 0.954 X42 4.686 1 4.686 X1 X2 75.690 1 75.690 X1 X3 191.546 1 191.546 X1 X4 192.516 1 192.516 X2 X3 1.904 1 1.904 X2 X4 31.192 1 31.192 X3 X4 0.990 1 0.990 Lack of Fit 129.350 10 12.935 Pure Error 12.252 3 4.084 Total SS 5100.276 27 a Df: Degrees of freedom b Test forcomparing modelvariance with residual (error) variance.
6 7 8 9 10 11 12 13 14 15 16 17 25
F-valueb
P value
697.1534 46.4312 138.4743 73.4376 105.8686 3.0624 0.2336 1.1475 18.5339 46.9030 47.1405 0.4663 7.6379 0.2424 3.1673
0.000119 0.006465 0.001319 0.003340 0.001958 0.178430 0.661945 0.362598 0.023067 0.006372 0.006327 0.543691 0.069937 0.656241 0.186186
1 2 3
Table 6.kinetic parameters of Mb removal using (0.14, 0.17, 0.2 g) adsorbent over concentration in the range of 18-27 mg/L 0.17g
0.14g
0.2g
18.12mg/l
22mg/l
27.21mg/l
18.12mg/l
22mg/l
27.21mg/l
18.12mg/l
22mg/l
27.21mg/l
K1
0.0990
0.0599
0.0299
0.1013
0.0714
0.0714
0.1013
0.1957
0.0875
qe (cal)
7.430
10.447
7.430
2.223
3.491
10.139
0.475
1.574
3.319
R
0.981
0.906
0.981
0.930
0.957
0.960
0.834
0.81
0.907
Second order
K2
0.0239
0.0109
0.0068
0.0686
0.0343
0.0137
0.2395
0.1473
0.0302
kinetic model
qe (cal)
6.944
8.547
10.526
5.55
6.802
8.695
4.386
5.618
7.194
R2
0.998
0.995
0.991
0.999
0.998
0.997
0.999
0.998
0.997
H
1.152
0.798
0.755
2.118
1.587
1.037
4.608
4.651
1.565
Kdif
0.393
0.522
0.693
0.229
0.333
0.523
0.121
0.578
0.390
C
3.189
2.881
2.730
3.546
3.646
2.203
3.626
2.751
3.551
R2
0.878
0.961
0.976
0.695
0.783
0.901
0.477
0.692
0.751
β
0.945
0.707
0.537
1.513
0.971
0.686
2.688
1.064
0.875
α
6.708
3.389
2.575
38.632
11.080
4.130
1539.319
16.512
6.715
0.955
0.986
0.984
0.863
0.926
0.976
0.676
0.886
0.900
Models First order kinetic model:
Intraparticle diffusio
Elovich
Parameters
2
2
R
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 26
1 2 3
Table 7. kinetic parameters of Pb2+ removal using (0.14, 0.17 and 0.2 g) adsorbent and 80, 97 120 mg/L of MB.
4 0.17g
0.14g
Models First order kinetic
0.2g
Parameters
80 mg/l
97 mg/l
120 mg/l
80 mg/l
97 mg/l
120 mg/l
80 mg/l
97 mg/l
120 mg/l
K1
0.0368
0.0184
0.0138
0.0322
0.0276
0.0207
0.018
0.0184
0.0276
qe (cal)
4.305
6.426
7.362
2.213
5.164
6.252
1.288
2.307
3.908
model:
R
0.974
0.917
0.877
0.908
0.969
0.969
0.925
0.871
0.772
Second
K2
0.0231
0.0131
0.0116
0.0464
0.0161
0.0128
0.125
0.0390
0.0214
qe(cal)
28.571
32.258
38.461
23.809
28.571
33.33
20.000
24.390
29.412
order
2
kinetic
R
0.999
0.999
0.999
0.999
0.999
0.999
0.999
0.999
0.999
model:
H
18.868
13.699
17.241
26.316
13.158
14.286
50.000
23.256
18.518
Kdif
0.516
0.804
0.836
0.610
0.715
0.212
0.327
0.631
C
23.03
24.35
28.69
21.88
25.61
18.15
20.82
23.35
2
R
0.912
0.894
0.867
0.310 20.53 0.833
0.903
0.930
0.925
0.810
0.792
β
0.721
0.443
0.410
1.180
0.590
0.505
1.848
1.059
0.557
α
7576399
31637.84
89670.16
8.54E+09
204272.2
256045.8
7.55*10^9
9.72*10^9
228872.2
R2
0.980
0.982
0.965
0.933
0.973
0.987
0.903
0.941
0.890
Intraparticle diffusion
Elovich
2
5 6 7 8 9 10 11 12 13 14 15 16 17 27
1 2 3 4 5 6
Table 8. Parameters of different Isotherms for methylene blue in single component system. Models
Parameter
Langmuir Ce/qe= 1/KaQm + Ce/Qm
Qm (mg g1 ) Ka (L mg 1 ) RL
Freundlich Log qe = log KF+(1/n)log Ce
Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
R2 1/n KF (L mg 1 ) R2 Bl KT (L mg 1 ) R2 Qs (mg g-1) K R2
0.14 g
0.17 g
0.2 g
18.518
16.393
16.393
27.000
20.333
15.250
0.0019
0.0033
0.0121
-
-
-
0.6376 0.999 0.137
0.7484 0.999 0.168
0.6798 0.999 0.202
3.161
3.022
3.065
0.835 1.691
0.862 1.750
0.708 2.016
6870.61
2115.70
911.85
0.895
0.907
0.834
17.778 0.000000009 0.915
15.440 0.00000001 0.979
16.610 0.00000001 0.844
7 8 9 10 11 12 13 14 28
1 2 3 4 5 6 7
Table 9. Parameters of different Isotherms for Pb2+ ions in single component system
Models Langmuir Ce/qe= 1/KaQm + Ce/Qm
Freundlich Ln qe = lnKF+(1/n)lnCe Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
Parameter -1
Qm (mg g ) Ka (L mg -1) RL
R2 1/n KF (L mg -1) R2
Bl KT (L mg -1) R2 Qs (mg g-1) K R2
8 9 10 11 12 13 14 15 16 17 18
29
0.14 g 58.823 0.148 0.0263 0.0779 0.992 0.192 21.978 0.976
0.17 g 50.000 0.250 0.0157 0.0476 0.995 0.187 21.330 0.99
0.2 g 47.619 0.318 0.0124 0.0378 0.997 0.194 19.952 0.991
7.571 10.183 0.975 45.604 0.000002 0.801
6.459 19.302 0.996 39.805 2E-07 0.732
5.973 23.045 0.997 35.837 0.0000003 0.717
1 2 3 4 5 6
Table 10. Parametersof different Isotherms for methylene blue in multi-component system
Models
Parameter
Langmuir Ce/qe= 1/KaQm + Ce/Qm
RL
0.14 g 11.236 9.889 0.0040 0.6203
0.17 g 10.638 7.833 0.0062 0.9241
0.6067
0.6489
R 1/n KF (L mg -1) R2 Bl KT (L mg -1) R2 Qs (mg g-1)
0.999 0.086 2.596 0.842 0.764 169494.5 0.845 10.697
K
0.00000001
R2
0.933
0.998 0.078 2.545 0.904 0.620 1480758.0 0.944 9.964 0.00000000 5 0.976
Qm (mg g-1) Ka (L mg -1)
+,
2
Freundlich Ln qe = lnKF+(1/n)lnCe Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
7 8 9 10 11 12 13 14 15 16 17 30
0.2 g 10.309 7.461 0.0085 0.8467 0.6288 0.999 0.108 2.450 0.890 0.788 31785.0 0.933 9.516 0.0000000 08 0.907
1 2 3 4
Table 11. Parametersof different Isotherms for Pb2+ in multi-component system
Models
Parameter
Langmuir Ce/qe= 1/KaQm + Ce/Qm
RL
Qm (mg g-1) Ka (L mg -1)
+,
2
Freundlich Ln qe = lnKF+(1/n)lnCe Tempkin qe = Blln KT + BllnCe Dubinin and Radushkevich Ln qe= ln Qs – Kε2
R 1/n KF (L mg -1) R2 Bl KT (L mg -1) R2 Qs (mg g-1) K R2
0.14 g 55.555 0.143 0.0272 0.0804 0.944 0.993 0.181 21.777 0.993 6.885 13.287 0.977 42.098 0.000001 0.602
5 6 7 8 9 10 11 12 13 14 15 16
31
0.17 g 45.454 0.293 0.0134 0.0409
0.2 g 41.666 0.453 0.0087 0.0268
0.910
0.875
0.995 0.146 22.594 0.988 4.784 88.561 0.989 37.003 0.0000003 0.694
0.997 0.148 21.478 0.975 4.360 127.857 0.989 33.448 0.0000001 0.660
1 2 3 4 5 6 7 8 9 10 11
12
13
14 15 16 17
Fig. 1.
18 19 20 21 22 23 32
1 2 3 4 5 6 7
120 100
Removal %
80 60 0.1g Pb
40
0.05g Pb 0.05g MB
20
0.1g MB
0 0
2
4 pH
8 9 10
Fig. 2.
11 12 13 14 15 33
6
8
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Fig. 3.
23 24 25 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Fig. 4 35
1 2 3 4 5
6 7 8
Fig. 5.
9 10
36
1
2 3
37
1 2
3
Highlight
4
5
•
Activated carbon was prepared from walnut wood.
6
•
The significant variables were optimized by using a CCD combined with DF.
7
•
The equilibrium and kinetic studies were investigated for the adsorption process.
8
38