Biochemical Engineering Journal 27 (2005) 83–93

Removal of methylene blue by mango seed kernel powder

K. Vasanth Kumar a,∗ , A. Kumaran b
a Department of Chemical Engineering, A.C. Tech, Anna University, Chennai, TN 600 025, India
b Department of Chemistry, Madras Christian College, Tambaram, Chennai, TN 600 059, India
Received 18 November 2004; received in revised form 19 June 2005; accepted 2 August 2005

Abstract
Batch experiments were carried out for the sorption of methylene blue onto mango seed kernel particles. The operating variables studied
were initial solution pH, temperature, adsorbent mass, initial dye concentration and contact time. Equilibrium data were fitted to Freundlich
and Langmuir isotherm equation and the equilibrium data were found to well represented by Langmuir isotherm equation. The monolayer sorp-
tion capacity of mango seed kernel for methylene blue sorption was found to be 142.857 mg/g at 303 K. The sorption kinetics was found to
follow pseudo first order kinetic model. The methylene blue uptake process was found to be controlled by both surface and pore diffusion
with surface diffusion at the earlier stages followed by pore diffusion at later stages. The average effective diffusion coefficient was calcu-
lated and found to be 5.66 × 10−4 cm2 /s. Analysis of sorption data using Boyd plot confirms that the external mass transfer as the rate limiting
step in the sorption process. Various thermodynamic parameters such as enthalpy of sorption
H, free energy change
G and entropy
S
were estimated. The positive value of
H and negative values of
G shows the sorption process is exothermic and spontaneous. The posi-
tive value of entropy
S shows the increased randomness at the solid–liquid interface during the sorption of dye ions onto mango seed kernel
particles.
© 2005 Elsevier B.V. All rights reserved.

Keywords: Sorption; Mango seed kernel; Methylene blue; Kinetics; Equilibrium; Mechanism; Thermodynamic

1. Introduction the removal of dye (methylene blue) from its aqueous solutions.
Methylene blue has wider applications, which include coloring
The removal of color from dye bearing effluents is one of the paper, temporary hair colorant, dyeing cottons, wools, coating
major problem due to the difficulty in treating such wastewa- for paper stock, etc. Though methylene blue is not strongly haz-
ters by conventional treatment methods. The most commonly ardous, it can cause some harmful effects. Acute exposure to
used method for color removal are biological oxidation and methylene blue will found cause increased heart rate, vomiting,
chemical precipitation. However, these process are effective and shock, Heinz body formation, cyanosis, jaundice, quadripelgia
economic only in the case where the solute concentration are rel- and tissue necrosis in humans. Previously several researchers
atively high. Currently sorption process is proved to be one of had proved several low cost materials such as pear millet husk
the effective and attractive processes for the treatment of these carbon [6]; Aspergillus niger [7]; rice husk [8]; hair [8]; cot-
dye-bearing wastewaters [1–5]. Also this method will become ton waste [8]; bark [8]; perlite [9]; carbonized press mud [10];
inexpensive, if the sorbent material used is of inexpensive mate- bagasse bottom ash [10]; raw kaolin [11]; pure kaolin [11]; cal-
rial and does not require any expensive additional pretreatment cined raw kaoline [11]; calcined pure kaoline [11]; NaOH treated
step. In the present study mango seed kernel, a waste material raw kaolin [11]; coir pith [12]; guava seeds activated carbon [13];
obtained from mango juice manufacturing industry have been iron humate [14]; neem sawdust [15]; clay [16]. In the present
used as adsorbent for the removal of methylene blue from its study mango seed kernel have been used as an adsorbent for
aqueous solution. Methylene blue is selected as a model com- the removal of methylene blue from its aqueous solutions. In
pound in order to evaluate the capacity of mango seed kernel for the present study the effect of initial methylene blue concentra-
tion, fly ash mass, initial dye solution pH and contact time on
the amount of color removal have been investigated. Also the
∗ Corresponding author. Tel.: +91 9444469876. sorption of methylene blue at solid–liquid interfaces have been
E-mail address: vasanth [email protected] (K.V. Kumar). studied extensively at equilibrium and various thermodynamic

1369-703X/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.bej.2005.08.004
84 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93

behaviors have been investigated. Further the kinetics and the particles with 100 mL of dye solution of different initial dye con-
mechanistic steps involved in the sorption process were also centration 100, 80, 60, 50, 40, 30 and 20 mg/L.
evaluated.
2.2.5. Sorption kinetics
2. Experimental Sorption kinetics experiments were carried out using mag-
netic stirrers of 2 L capacity. Kinetic experiments were carried
2.1. Materials out by agitating 2 L of dye solution of initial concentration
100 mg/L with 2 g of mango seed kernel in 2.5 L beakers at
The ripened fruits of Mangifera indica (Mango) were col- a constant agitation speed of 300 RPM. 2.5 mL of samples were
lected from the local juice manufacturing industry and seeds pipetted out using 10 mL syringe at time intervals of 10, 20, 30,
were separated from the fruit. The hard seed coat was removed 40, 50, 60, 70, 120 and 240 min. The collected samples were
and the seeds were sun dried. These dried seeds were coarsely then centrifuged and the concentration in the supernatant solu-
powdered using domestic mixie and the particle size in the range tion were analyzed using UV spectrophotometer.
of 60–85 mesh (BSS) were used as adsorbents for the present
study. 3. Results and discussions
The dye used in all the experiments was methylene blue,
a basic (cationic) dye. The dye methylene blue was obtained 3.1. Effect of pH
from BDH chemicals. Synthetic dye solutions were prepared
by dissolving weighed amount methylene blue in 1 L of dou- Fig. 1 shows the effect of pH on amount of dye adsorbed qe
ble distilled water. The NaOH pellets and HCl solution used (mg/g). Where qe was obtained from the mass balance equation
for optimizing pH are obtained from Qualigens fine chemicals, given by:
Mumbai, India.
(C0 − Ce )V
qe = (1)
2.2. Sorption experiments M
where, C0 (mg/L) and Ce (mg/L) is the concentration in the
2.2.1. Effect of pH solution at time t = 0 and at equilibrium time, V is the volume of
The effect of pH on the amount of color removal was ana- solution treated and M is the amount of biomass added (g). From
lyzed over the pH range from 3 to 8. The pH was adjusted the Fig. 1, it was observed that the solution pH affects the amount
using 0.1 N NaOH and 0.1 N HCl solutions. In this study of dye adsorbed. The dye uptake was found to increase from with
30 mL of dye solution of 100 mg/L was agitated with 0.02 g increasing pH and it increases from 8.55 to 109.788 mg/g for an
of mango seed kernel for 48 h, which is more than sufficient increase in pH from 3 to 8. Several reasons may be attributed to
to reach equilibrium. The samples were then centrifuged and dye sorption behavior of the biosorbent relative to solution pH.
the left out concentration in the supernatant solution were ana- The surface of mango seed kernel may contain a large number
lyzed using UV Spectrophotometer (Deep Vision 301 E) by of active sites and the solute (dye ions) uptake can be related
monitoring the absorbance changes at a wavelength of max- to the active sites and also to the chemistry of the solute in the
imum absorbance (665 nm). For all the experiments, except
the adsorption kinetics, the contact was made using water
bath shakers at 303 K and at a constant agitation speed of
95 strokes/min.

2.2.2. Effect of adsorbent mass

The effect of algae mass on the amount of color adsorbed
was obtained by agitating 30 mL of dye solution of initial dye
concentration of 100 mg/L with weighed amount of mango
seed kernel (ranging from 0.01 to 0.06 g) till equilibrium is
reached.

2.2.3. Effect of initial dye concentration

The effect of initial dye concentration on the equilibrium
uptake was estimated by contacting 0.06 g of mango seed kernel
particles with 100 mL of dye solution of different initial concen-
tration ranging from 20 to 100 mg/L.

2.2.4. Sorption equilibrium

Equilibrium studies were carried out at three different solu-
tion temperatures of 303, 313 and 323 K. Equilibrium experi- Fig. 1. Effect of initial pH (C0 : 100 mg/L; M: 0.02 g/30 mL; temperature: 303 K;
ments were performed by contacting 0.06 g of mango seed kernel agitation speed: 95 stroke min−1 ).
 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93 85

solution. At lower pH, the surface charge may get positively a r2 value of 0.9787:
charged, thus making H+ ions compete effectively with dye 0.0003qe
cations causing a decrease in amount of dye adsorbed (mg/g). M= (2)
0.0242qe − 1
At higher pH the surface of mango seed kernel particles may
get negatively charged, which enhances the positively charged 3.3. Effect of initial dye concentration
dye cations through electrostatic force of attraction. A similar
observation was previously observed for methylene blue onto Fig. 3 shows the plot between amount of dye adsorbed qe
waste coir pith [12] and for malachite green onto neem saw versus initial dye concentration. From the figure it was observed
dust [15]. that the amount of dye adsorbed gets increased from 23.4 to
76.8 mg/g for an increase in initial dye concentration from 20 to
3.2. Effect of adsorbent mass 100 mg/L, whereas the percent color removal decreases from 78
to 51.2% for an increase in initial dye concentration from 20 to
Fig. 2 shows the plot between amount of dye adsorbed qe 100 mg/L. The relation between qe and initial dye concentration
against adsorbent concentration (g). From the figure it was fits the Eq. (3) with a coefficient of determination of 0.991 as
observed that, the amount of dye adsorbed gets varied with follows:
varying sorbent mass and it decreased with increase in adsor-
0.4167C02 + 4.0086C0
bent mass. The amount of dye adsorbed decreases from 184 qe = (3)
to 48.085 mg/g for an increase in adsorbent mass from 0.01 0.5833C0 − 4.0086
to 0.06 g. Whereas the percent color removal increased from
3.4. Sorption equilibrium
61.33 to 96.17% with an increase in adsorbent mass from 0.01
to 0.06 g. At higher biomass to solute concentration ratio, there
The analysis and design of sorption process requires the rele-
is a very fast superficial sorption onto the adsorbent surface that
vant adsorption equilibria, which is the most important piece of
produces a lower solute concentration in the solution than when
information in understanding an adsorption process [2]. Sorption
biomass to solute concentration ratio is lower. This is because
equilibria provide fundamental physiochemical data for evaluat-
a fixed mass of biomass can only adsorb a certain amount of
ing the applicability of sorption process as an unit operation. The
dye. Therefore, the more the adsorbent dosage, the larger is the
two most commonly used equilibrium relations are Freundlich
volume of effluent that a fixed mass of biomass can purify. The
[17] and Langmuir [18] isotherm equations. In the present inves-
decrease in amount of dye adsorbed qe (mg/g) with increasing
tigation the equilibrium data were analyzed using the Freundlich
adsorbent mass is due to the split in the flux or the concentration
and Langmuir isotherm expression given by Eqs. (4) and (5),
gradient between solute concentration in the solution and the
respectively:
solute concentration in the surface of the adsorbent. Thus with
1/n
increasing adsorbent mass, the amount of dye adsorbed onto Freundlich : qe = Kf Ce (4)
unit weight of adsorbent gets splitted thus causing a decrease
qm KL Ce
in qe value with increasing adsorbent mass concentration. The Langmuir : qe = (5)
relation between the qe and adsorbent mass fits the Eq. (2) with 1 + K L Ce
where Kf (mg/g) (L/g)n and n are Freundlich constants related
to sorption capacity and sorption intensity of adsorbents, qm is

Fig. 2. Effect of adsorbent mass (C0 : 100 mg/L; V: 30 mL; pH: 8; temperature: Fig. 3. Effect of initial dye concentration (M: 0.2 g/30 mL; pH: 8; temperature:
303 K; agitation speed: 95 stroke min−1 ). 303 K; agitation speed: 95 stroke min−1 ).
86 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93

Table 1 Table 3
Equilibrium isotherm expressions for the sorption of methylene blue onto mango Low cost adsorbents and their sorption capacities for the removal of methylene
seed kernel particles blue from its aqueous solutions
Temperature Predicted Langmuir Predicted Freundlich Adsorbent Sorption capacity Reference
(K) isotherm isotherm
Pear millet husk carbon 66 mg/g [6]
303 qe = 28.085Ce /1 + 0.1966Ce qe = 43.9035Ce0.2856 Aspergillus niger 15.5 mg/g [7]
313 qe = 65.786Ce /1 + 0.4605Ce qe = 61.1927Ce0.2205 Rice husk 312.2570 mg/g [8]
323 qe = 126.569Ce /1 + 0.8227Ce qe = 85.7432Ce0.1553 Coal 323.6815 mg/g [8]
Hair 158.2312 mg/g [8]
Cotton waste 277.7757 mg/g [8]
the maximum sorption capacity of mango seed kernel to uptake Bark 914.5849 mg/g [8]
Carbonized press mud 50 mg/g [10]
methylene blue (mg/g) and KL is the Langmuir constant related
Bagasse bottom ash 142 mg/g [10]
to the energy of adsorption (L/g). The value of n falling in the Raw kaolin 27.49 mg/g [11]
range of 1–10 indicates favorable sorption. The linearised form Pure kaolin 91.87 mg/g [11]
of Freundlich and Langmuir can be written as follows: Calcined raw kaolin 13.44 mg/g [11]
Calcined pure kaolin 56.31 mg/g [11]
1
log qe = log Kf + (6) NaOH treated raw kaolin 204.00 mg/g [11]
n(log Ce ) NaOH treated pure kaolin 122.01 mg/g [11]
Guava seeds activated carbon 0.667–0.625 mg/g [13]
Ce 1 Ce Iron humate 0.0882 mmol/g [14]
= + (7)
qe K L qm Qo Neem saw dust 2.120–3.622 mg/g [15]
Clay 6.3–5 mg/g [16]
Thus the Freundlich constant Kf and n can be calculated from Mango seed kernel 142.857–153.846 mg/g This study
the intercept and slope of plot between log qe and log Ce . Simi-
larly the Langmuir constants qm and KL were calculated from the
slope of plot between Ce /qe versus Ce . The predicted Freundlich methylene blue and their sorption capacity values. Table 3 also
and Langmuir isotherm equation for methylene blue onto mango shows the sorption capacity of mango seed kernel powder to
seed kernel particles at different solution temperatures useful for uptake methylene blue. From the Table 3, it was observed that,
design calculations were given in Table 1. the sorption capacity of mango seed kernel determined from the
Fig. 4a–c shows the Freundlich and Langmuir curve for present study is comparable enough to some of the low cost
methylene blue onto mango seed kernel at a constant solution adsorbents previously reported for the uptake of methylene blue
temperature of 303, 313 and 323 K, respectively, along with from its aqueous solutions.
the experimental data. The Freundlich and Langmuir curves in The essential characteristics of the Langmuir isotherm can be
Fig. 4a–c were generated using the equations given in Table 1. expressed in terms of dimensionless constant separation factor
From these figures it was observed that the equilibrium data or equilibrium parameter, RL , given by [19]:
are very well represented by Langmuir isotherm equation when
compared to Freundlich equation. The sorption equilibrium data 1
RL = (8)
fit Freundlich and Langmuir equation with a coefficient of deter- 1 + K L C0
mination, r2 values ranging from 0.8388 to 0.9161 and 0.9993-
1, respectively. The best fit of equilibrium data in Langmuir The parameter RL indicated the shape of isotherm as follows:
isotherm expression at all three-solution temperatures confirms
Value of RL Type of isotherm
the monolayer coverage of methylene blue onto mango seed ker-
nel particles. The calculated Freundlich and Langmuir isotherm RL > 1 Unfavorable
constants, predicted Freundlich and Langmuir equations and RL = 1 Linear
0 < RL < 1 Favorable
their corresponding linear regression correlation coefficient val- RL = 0 Irreversible
ues at different temperatures were given in Table 2. From Table 2,
it was observed that the maximum sorption capacity, qm of The calculated RL values versus initial solute concentration
mango seed kernel particles for methylene blue was found to at three different temperatures were represented in Fig. 5. From
be in the range of 142.857–153.846 mg/g. Table 3 lists some Fig. 5 it was observed that at all temperature conditions, sorption
of the low cost materials used as adsorbents for the removal of was found to be more favorable at higher concentrations. Also

Table 2
Equilibrium isotherm constants for the sorption of methylene blue onto mango seed kernel particles
Temperature (K) Langmuir isotherm parameters Freundlich isotherm parameters

qm (mg/g) KL (L/mg) r2 Kf (mg/g) (L/g) n r2

303 142.857 0.1966 0.9993 43.9035 3.501 0.9161

313 142.857 0.4605 0.9998 61.1927 4.535 0.8998
323 153.846 0.8227 1 85.7432 6.4391 0.8388
 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93 87

Fig. 4. (a) Equilibrium curve for the sorption of methylene blue onto mango seed kernel at 303 K (M: 0.2 g/30 mL; pH: 8; agitation speed: 95 stroke min−1 ). (b)
Equilibrium curve for the sorption of methylene blue onto mango seed kernel at 313 K (M: 0.2 g/30 mL; pH: 8; agitation speed: 95 stroke min−1 ). (c) Equilibrium
curve for the sorption of methylene blue onto mango seed kernel at 323 K (M: 0.2 g/30 mL; pH: 8; agitation speed: 95 stroke min−1 ).

the value of RL in the range of 0–1 at all initial dye concentration centration. From the Fig. 6, it was observed that for an initial
and at all the three-solution temperature conditions confirms the dye concentration of 100 mg/L, the maximum amount (approx-
favorable uptake of methylene blue process. imately 79.5% of total amount of dye removed) of dye was
adsorbed within the first 30 min at an average sorption rate of
3.5. Sorption kinetics 2.0429 mg/g min (approximately 12.5% of total amount of dye
adsorbed) and thereafter the sorption rate tends to decrease and
The prediction of batch sorption kinetics is necessary for the proceeds at an average sorption rate of 0.1631 mg/g min. A sim-
design of sorption industrial columns [2]. The nature of sorption ilar trend was observed for the remaining range of initial dye
process will depend on physical or chemical characteristics of concentrations (125–200 mg/L) studied. The initial rapid phase
the adsorbent system and also on the system conditions. Fig. 6 may be due to increased number of vacant sites available at the
shows the plot between the amount of dye adsorbed (mg/g) initial stage, as a result there exist increased concentration gradi-
versus contact time for different initial methylene blue con- ent between adsorbate in solution and adsorbate in the adsorbent
88 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93

Table 4
Pseudo first order kinetic constants for different methylene blue concentrations
Initial dye concentration (mg/L) K1 (min−1 )

100 0.0476
125 0.05031
150 0.0456
175 0.0461

with respect to boundary conditions q = 0 at t = 0 and q = q at

t = t, then Eq. (9) simplifies to:


qe
ln = K1 t (10)
qe − q

Thus the rate constant K1 (min−1 ) can be calculated from

the plot of ln(qe /qe − q) versus time t. The calculated K1 values
and the corresponding coefficient of determination r2 values at
different initial dye concentrations are shown in Table 4. From
Table 4, the higher r2 values in the range of 0.9462–0.979 sug-
Fig. 5. Separation factor for methylene blue onto mango seed kernel. gests that the dye sorption process follow pseudo first order
kinetics. The calculated correlation coefficient values for first
order (0.9779) shows the applicability of the Lagergren kinetic
surface. This increase in concentration gradient tends to increase models for methylene blue onto mango seed kernel particles.
in dye sorption rate at initial stages. As time proceeds this con- In the present study based on the kinetic constants evaluated,
centration gets reduced due to the accumulation of dye particles the first order kinetic expression was tested for its consistency
in the vacant sites, leading to decrease in sorption rate at later in predicting the amount of dye adsorbed at any time. The q
stages from 30 to 120 min. The transient behavior of dye sorption values were predicted by applying the calculated kinetic con-
process following pseudo first order process can be represented stants in their corresponding kinetic expressions. Thus applying
by [20]: the pseudo first order rate constant in Eq. (10), the pseudo first
dq order kinetic rate expression for methylene blue onto mango
= K1 (qe − q) (9) seed particles can be obtained. The predicted Lagergren pseudo
dt
first order kinetic expression for methylene blue/mango seed
where, q and qe represents the amount of dye adsorbed (mg/g) kernel systems at different initial dye concentration are given
at any time t and at equilibrium time respectively and K1 rep- in Table 5. The q values predicted from the equations shown in
resents the sorption rate constant (min−1 ). Integrating Eq. (9) Table 5 were compared with experimental kinetic data as shown
in Fig. 6. From Fig. 6, it was observed that for the entire sorption
period, for all the range of initial dye concentration studied, the
pseudo first order expression better predicts the sorption kinet-
ics of methylene blue onto mango seed kernel. The generalized
Lagergren first order kinetic model for methylene blue/mango
seed kernel powder system is given by:
0.0258C0 + 1
q = (0.5128C0 + 19.89) − −8 −6
(11)
(−4x10 C03 +9x10 C02 )t
e
Eq. (11) can used to predict the q value at any time for any
of the initial dye concentration required. The predicted q value

Table 5
Lagergren pseudo first order kinetic expression for methylene blue onto mango
seed kernel
Initial dye concentration (mg/L) Predicted equation

100 q = 73.2 − 73.2/exp0.046t

125 q = 83 − 83/exp0.05031t
150 q = 92.7 − 92.7/exp0.0456t
Fig. 6. Adsorption kinetics for methylene blue onto mango seed kernel (C0 : 175 q = 112.7 − 112.7/exp0.0461t
100 mg/L; M: 2 g/2 L; pH: 8; agitation speed: 300 rpm).
 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93 89

Fig. 7. Predicted q values for different initial dye concentrations (pH: 8; agitation
speed: 800 rpm; temperature, 303 K; M: 2 g/2 L). Fig. 8. Intraparticle diffusion plot for methylene blue onto mango seed kernel
(C0 : 100 mg/L; M: 2 g/2 L; pH: 8; temperature: 303 K; agitation speed: 300 rpm).

for different initial dye concentrations using Eq. (9) for a fixed
operating line V/X of 2 L/2 g is shown in Fig. 7. the sorption process for the systems that have poor mixing, dilute
concentration of adsorbate, small particle sizes of adsorbent and
3.6. Sorption mechanisms higher affinity of adsorbate for adsorbent. Whereas, the intra-
particle diffusion will control the sorption process for a system
Prediction of the rate-limiting step is an important factor with good mixing, large particle sizes of adsorbent, high concen-
to be considered in sorption process. For solid–liquid sorption tration of adsorbate and low affinity of adsorbate for adsorbent.
process, the solute transfer process was usually characterized In general, external mass transfer is characterized by the ini-
by either external mass transfer (boundary layer diffusion) or tial solute uptake [4,5] and can be calculated from the slope of
intraparticle diffusion or both. The sorption dynamics can be plot between C/C0 versus time. The slope of these plots can
described by the following three consecutive steps which are as be calculated either by assuming polynomial relation between
follows: C/C0 and time or it can be calculated based on the assump-
tion that the relationship was linear for the first initial rapid
• Transport of the solute from bulk solution through liquid film phase (in the present study first 20 min). The later technique
to the adsorbent exterior surface. was used by Fernadez et al. [21] for lauryl benzyl sulphonate
• Solute diffusion into the pore of adsorbent except for a small onto algal residue and by Kumar et al. [22] for methylene blue
quantity of sorption on the external surface. Parallel to this is onto fly ash particles. In the present study the second technique
the intraparticle transport mechanism of the surface diffusion. was used by assuming the external mass transfer occurs in the
• Sorption of solute on the interior surfaces of the pores and first 20 min. Thus the initial sorption rates Ks (min−1 ) were
capillary spaces of the adsorbent. quantified as (C20 min /C0 )/20. The calculated Ks values were
found to be 0.0276, 0.0276, 0.028667 and 0.02886 min−1 for
The last step is considered to be an equilibrium reaction. Out an initial dye concentration of 100, 125, 150 and 175 mg/L,
of three steps, the third step is assumed to be rapid and considered respectively.
to be negligible. The overall rate of sorption will be controlled The intraparticle diffusion coefficient for the sorption of
by the slowest step, which would be either film diffusion or pore methylene blue was calculated from the slope of the plot
diffusion. However, the controlling step might be distributed between square root of time (min0.5 ) versus amount of dye
between intraparticle and external transport mechanism. What- adsorbed (mg/g). Previous studies by various researchers
ever may be the case external diffusion will be involved in the showed that the plot between q versus t0.5 represents multi
sorption process. The sorption of methylene blue onto mango linearity, which characterizes the two or more steps involved in
seed particles may be controlled due to film diffusion at earlier sorption process [23,24]. Fig. 8 shows the plot between q versus
stages (assumed to be for the first 20 min) and as the adsorbent t0.5 for methylene blue onto mango seed kernel particles. From
particles gets loaded with dye ions, the sorption process may be the Fig. 8, at all initial dye concentrations, the sorption process
controlled due to intraparticle diffusion. But for design purpose it tends to be followed by two phases. It was found that an initial
is necessary to calculate the slowest step involved in the sorption linear portion ended with a smooth curve followed by second
process. For sorption process, the external mass transfer controls linear portion. The two phases in the intraparticle diffusion plot
90 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93

Table 6
Intraparticle diffusion coefficient for methylene blue onto mango seed kernel

Initial dye concentration (mg/L) Ki (mg/g min0.5 )

100 7.1918
125 4.9226
150 4.5582
175 8.7655

suggests that the sorption process proceeds by surface sorption

and the intraparticle diffusion. The initial curved portion of the
plot indicates boundary layer effect while the second linear por-
tion is due to intraparticle or pore diffusion. The slope of second
linear portion of the plot has been defined as the intraparticle
diffusion parameter Ki (mg/g min). On the other hand, the
intercept of the plot reflects the boundary layer effect. Larger the
intercept, greater is the contribution of the surface sorption in the Fig. 9. Boyd’s plot for methylene blue onto mango seed kernel.
rate-limiting step. The calculated intraparticle diffusion coeffi-
cient Ki value at different initial dye concentrations was shown
Di value at different initial dye concentration was shown in
in Table 6. Since the double nature of intraparticle diffusion plot
Table 7.
confirms the presence of both film and pore diffusion, in order
to predict the actual slow step involved, the kinetic data were
further analyzed using Boyd kinetic expression which is given 3.7. Sorption thermodynamics
by [25]:
For designing sorption column or batch sorption systems,
6 the designer should be able to understand the following: what
F =1− exp(−Bt) (12)
π2 changes can be expected to occur and how fast will they take
and place. The fast of the reaction can be calculated from the knowl-
q edge of kinetic studies. But the change in reaction that can be
F= (13) expected during the process require the brief idea of thermo-
q0
dynamic parameters. The concept of thermodynamic assumes
where, q0 is the amount of methylene blue adsorbed at infinite that in an isolated system where energy cannot be gained or
time (mg/g) and q represents the amount of dye adsorbed at any lost, the entropy change is the driving force [26]. The ther-
time t (min), F represents the fraction of solute adsorbed at any modynamic parameters that must be considered to determine
time t and Bt is a mathematical function of F. the process are enthalpy of sorption (
H), free energy change
Eq. (13) can be simplified to: (
G) and entropy change (
S) due to transfer of unit mole of
6 solute from solution onto the solid–liquid interface. The impor-
1−F = exp(−Bt) (14) tant thermodynamic function
H is very useful whenever there
π2
is a differential change occurs in the system. Enthalpy is an addi-
or
tive property that is its value is additive. The negative value of
Bt = −0.4977 − ln(1 − F ) (15)
H indicates the exothermic process and positive value indicates
the endothermic process. The other important thermodynamical
The Bt values at different contact time can be calculated using parameter is the change in entropy
S. The parameter
S is used
Eq. (15). The calculated Bt values were plotted against time t to identify the spontaneity in the sorption process. The value of
as shown in Fig. 9. Fig. 9 is used to identify whether external
H and
S were computed using the equation as follows:
transport or intraparticle transport controls the rate of sorption
[25]. From the Fig. 9, it was observed that the plots were linear
S
H
ln Kd = − (17)
but does not pass through the origin confirming that, for the R RT
studied initial dye concentration, external mass transport mainly
governs the sorption process [25]. The calculated B values were Table 7
used to calculate the effective diffusion coefficient, Di (cm2 /s) Diffusion coefficient for methylene blue onto mango seed kernel
using the relation [25]:
C0 (mg/L) Di (cm2 /s)
Di
B = Π2 (16) 100 5.50E−04
r2 125 6.64E−04
150 5.01E−04
where, r represents the radius of the particle calculated by sieve 175 5.48E−04
analysis and by assuming as spherical particles. The calculated
 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93 91

Fig. 11. Single stage batch adsorber.

Fig. 10. Plot between ln Kd vs. 1/T (M: 0.06 g/30 mL; pH: 8; agitation speed:
95 strokes min−1 ). tion with a high preference of methylene blue on mango seed
kernel.

where, R is universal gas constant and equals to 8.314 J/mol K; 3.8. Design of batch sorption from isotherm data
T the absolute solution temperature and Kd is the distribution
coefficient which can be calculated using Eq. (17): Generally sorption process proceeds through varied mecha-
CAe nisms such as external mass transfer of solute onto sorbent fol-
Kd = (18) lowed by intraparticle diffusion. Unless extensive experimental
Ce
data are available concerning the specific sorption application,
The value of
S and
H can be calculated from the slope determining the rate-controlling step is impossible. Therefore,
from and intercept of plot between ln Kd versus 1/T as shown empirical design procedures based on sorption equilibrium con-
in Fig. 10. The calculated
H and
S at different solution con- ditions are the most common method to predict the adsorber
centrations were shown in Table 8. The positive value of
H size and performance. Sorption equilibrium is a dynamic con-
indicates the endothermic nature of the sorption process. The cept achieved when the rate at which molecules adsorb onto a
positive value of
S reflects the affinity of the mango seed ker- surface is equal to the rate at which they desorb. The physical
nel particles for methylene blue and the increasing randomness chemistry involved may be complex and no single theory of
at the solid–liquid interface during the sorption of methylene sorption has been put forward to explain all the systems. Fortu-
blue onto mango seed kernel particles. nately, engineer requires only the data at equilibrium conditions.
Another most important thermodynamic parameter involved Previously sorption isotherm relations have been used to predict
in the sorption process is the free energy change and can be the design of single stage batch sorption systems [2,27–29]. A
calculated using the relation: schematic diagram is shown in Fig. 11. The design objective is

G = −RT ln Kd (19) to reduce the dye solution of volume V (L) from an initial con-
centration of C0 to C1 (mg/L). The amount of adsorbent is M
The calculated values at different solute concentrations were and the solute loading changes from q0 (mg/g) to q1 (mg/g). At
shown in Table 8. The negative value of
G confirms the fea- time t = 0, q0 = 0 and as time proceeds the mass balance equates
sibility of the process and the spontaneous nature of the sorp- the dye removed from the liquid to that picked up by the solid.
The mass balance equation for the sorption system in Fig. 11
can be written as:
Table 8
Thermodynamical parameters for methylene blue onto Mango seed kernel V (C0 − C1 ) = M(q0 − q1 ) = Mq1 (20)
particles
C0 (mg/L)
H (J/mol)
S (J/mol) −
G (J/mol) At equilibrium conditions,
303 K 313 K 323 K
C1 → Ce and q1 → qe (21)
50 66106.276 233.365 5100.72 6878.745 9298.115
75 63725.147 221.734 3942.511 5604.143 7929.761 Since the sorption isotherm studies confirms that the equilib-
100 32552.635 115.406 2649.042 3565.446 4726.136 rium data for methylene blue onto mango seed kernel particle
125 17166.747 60.474 1326.11 1670.603 2409.703
150 13668.216 45.635 2836.624 3216.774 4408.334
fitted well in Langmuir isotherm, Langmuir isotherm equation
can be used for q1 in equation batch adsorber design.
92 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93

Acknowledgements

Thanks are due to the reviewers for their inspiring sugges-

tions.

References

[1] V.K. Gupta, I. Ali, M.D. Suhas, Equilibrium uptake and sorption dynam-
ics for the removal of a basic dye (basic red) using low-cost adsorbents,
J. Colloid Interface Sci. 265 (2003) 257–264.
[2] Y.S. Ho, G. McKay, The kinetics of sorption of basic dyes from aque-
ous solutions by sphagnum moss peat, Can. J. Chem. Eng. 76 (1998)
822–826.
[3] G.M. Walker, L. Hansen, J.A. Hanna, S.J. Allen, Kinetics of a reactive
dye adsorption onto dolomite sorbents, Water Res. 37 (2003) 2081–
2089.
[4] G. McKay, S.J. Allen, I.F. McConvey, M.S. Otterburn, Transport process
in the sorption of colored ions by peat particles, J. Colloid Interface Sci.
80 (2) (1981) 323–339.
[5] G. McKay, S.J. Allen, Surface mass transfer process using peat as an
adsorbent for dyestuffs, Can. J. Chem. Eng. 58 (1980) 521–526.
Fig. 12. Adsorbent mass (M) against volume of solution treated (L). [6] B.S. Inbaraj, K. Selvarani, N. Sulochana, Evaluation of a carbonaceous
sorbent prepared from Pearl Millet Husk for its removal of basic dyes,
J. Sci. Ind. Res. 61 (2002) 971–978.
Eq. (21) can be rearranged as [7] Y. Fu, T. Viraraghavan, Column studies for biosorption of dyes from
aqueous solutions on immobilised Aspergillus niger fungal biomass,
Water SA 29 (4) (2003) 465–472.
M C0 − Ce C0 − C e C0 − Ce
= = = (22) [8] G. McKay, J.F. Porter, G.R. Prasad, The removal of basic dyes aqueous
V q1 qe q0 KL Ce /(1 + KL Ce ) solution by adsorption on low-cost materials, Water Air Soil Pollut. 114
(1999) 423–438.
Fig. 12 shows the plot between the predicted amount of [9] M. Dogan, M. Alkan, A. Turkyilmaz, Y. Ozdemir, Kinetics and mecha-
nism of removal of methylene blue by adsorption onto perlite, J. Hazard.
mango seed kernel particles required to remove dye solution
Mater. 109 (2004) 141–148.
of initial concentration 100 mg/L for 90, 80, 70 and 60% color [10] K.V. Kumar, Adsorption isotherms for basic dyes onto low cot adsor-
removal at different solution volumes (1, 2, 3, 4, 5, 6 and 7 L). bents, Res. J. Chem. Environ. 6 (2002) 61–65.
For a single stage batch sorption system, the design procedure is [11] D. Gosh, G. Bhattacharya, Adsorption of methylene blue on kaolinite,
outlined. For example, the amount of mango seed kernel required Appl. Clay Sci. 20 (2002) 295–300.
[12] C. Namasiayam, M.D. Kumar, K. Selvi, R.A. Begum, T. Vanathi, R.T.
for the 90% removal of methylene blue solution of concentra-
Yamuna, ‘Waste’ coir pith–a potential biomass for the treatment of dye-
tion 100 mg/L was 0.950, 1.900, 2.85 and 3.80 g for dye solution ing wastewaters, Biomass Bioenergy 21 (2001) 477–483.
volume of 1, 2, 3 and 4 L, respectively. [13] I.A. Rahman, B. Saad, Utilization of guava seeds as a source of activated
carbon for removal of methylene blue from aqueous solution, Malays.
J. Chem. 5 (2003) 8–14.
4. Conclusions [14] P. Janoš, Sorption of basic dyes onto iron humate, Environ. Sci. Technol.
37 (2003) 5792–5798.
The present study shows that the mango seed kernel, a [15] S.D. Khattri, M.K. Singh, Color removal from synthetic dye wastewater
using a biosorbent, Water Air Soil Pollut. 120 (2000) 283–294.
waste biomaterial generated in juice factories can be used as an [16] A. Gürses, S. Karaca, Ć. Doğar, R. Bayrak, A. Acikyildiz, M. Yalcin,
adsorbent for the removal of methylene blue from its aqueous Determination of adsorptive properties of clay/water system: methylene
solutions. The amount of dye sorbed was found to vary with ini- blue sorption, J. Colloid Interface Sci. 269 (2004) 310–314.
tial solution pH, adsorbent dose, contact time and temperature. [17] H.M.F. Freundlich, Über die adsorption in lösungen, Zeitschrift für
The amount of dye uptake (mg/g) was found to increase with Physikalische Chemie (Leipzig) 57 (1906) 385–470.
[18] I. Langmuir, The constitution and fundamental properties of solids and
increase in solution temperature and contact time and found to liquids, J. Am. Chem. Soc. 38 (1916) 2221–2295.
decrease with increase in adsorbent dosage. The sorption data [19] K.R. Hall, L.C. Eagleton, A. Acrivos, T. Vermeulen, Pore and solid dif-
were found to follow pseudo first order kinetics. Equilibrium fusion kinetics in fixed-bed adsorption under constant pattern conditions,
data were fitted very well in Langmuir isotherm equation con- Ind. Eng. Chem. Fundam. 5 (1966) 212–223.
firming the monolayer sorption capacity of methylene blue onto [20] S. Lagergen, Zur theorie der sogenannten adsorption gelöster stoffe,
Kunglinga Svenska Vetenskapsakademiens, Handlingar. 24 (1898)
mango seed kernel particles with a monolayer sorption capac- 1–39.
ity of 142.857 mg/g. The dye uptake process was found to be [21] N.A. Fernadez, E. Chacin, E. Gutierrez, N. Alastre, Adsorption of lauryl
controlled by external mass transfer at earlier stages and by benzyl suplhonate on algae, Bioresour. Technol. 54 (1995) 111–115.
intraparticle diffusion at later stages. Boyd plot confirms the [22] K.V. Kumar, V. Ramamurthi, S. Sivanesan, Modeling the mechanism
external mass transfer as the slowest step involved in the sorp- involved during the sorption of methylene blue onto fly ash, J. Colloid
Interface Sci. 284 (2005) 14–21.
tion process. Thermodynamic parameter shows that the process [23] Q. Sun, L. Yang, The adsorption of basic dyes from aqueous solutions
is endothermic, feasible and spontaneous. on modified peat–resin particle, Water Res. 37 (2003) 1535–1544.
 K.V. Kumar, A. Kumaran / Biochemical Engineering Journal 27 (2005) 83–93 93

[24] R. Sivaraj, C. Namasivayam, K. Kadirvelu, Orange peel as an adsorbent [27] M. Dogan, M. Alkan, Removal of methyl violet from aqueous solutions
in the removal of acid violet 17 (acid ye) from aqueous solutions, Waste by perlite, J. Colloid Interface Sci. 267 (2003) 32–41.
Manage. 21 (2001) 105–110. [28] M. Özacar, I.A. Şengil, Equilibrium data and process design for adsorp-
[25] G.E. Boyd, A.W. Adamson, L.S. Myers Jr., J. Am. Chem. Soc. 69 (1947) tion of disperse dyes onto alunite, Environ. Geol. 45 (2004) 762–768.
2836. [29] Z. Aksu, T. Kutsal, A bioseparation process for removing lead (II) ions
[26] Y.S. Ho, Removal of copper ions from aqueous solution by tree fern, from wastewater by using C. vulgaris, J. Chem. Technol. Biotechnol.
Water Res. 37 (2003) 2323–2330. 52 (1991) 109–118.