Separation and Purification Technology 50 (2006) 388–397

Kaolinite, montmorillonite, and their modified derivatives as

adsorbents for removal of Cu(II) from aqueous solution
Krishna G. Bhattacharyya a,∗ , Susmita Sen Gupta b
a Department of Chemistry, Gauhati University, Guwahati 781014, Assam, India
b Department of Chemistry, BN College, Dhubri 783324, Assam, India

Received 15 October 2005; received in revised form 11 December 2005; accepted 15 December 2005

Abstract
Adsorption of metals by clay minerals is a complex process controlled by a number of environmental variables. The present work investigates the
removal of Cu(II) ions from an aqueous solution by kaolinite, montmorillonite, and their poly(oxo zirconium) and tetrabutylammonium derivatives.
The entry of ZrO and TBA into the layers of both kaolinite and montmorillonite was confirmed by XRD measurement. The specific surface areas
of kaolinite, ZrO-kaolinite, TBA-kaolinite, montmorillonite, ZrO-montmorillonite, TBA-montmorillonite were 3.8, 13.4, 14.0, 19.8, 35.8 and
42.2 m2 /g, respectively. The cation exchange capacity (CEC) was measured as 11.3, 10.2, 3.9, 153.0, 73.2 and 47.6 meq/100 g for kaolinite, ZrO-
kaolinite, TBA-kaolinite, montmorillonite, ZrO-montmorillonite, TBA-montmorillonite, respectively. Adsorption increased with pH till Cu(II)
ions became insoluble in alkaline medium. The kinetics of the interactions suggests that the interactions could be best represented by a mechanism
based on second order kinetics (k2 = 7.7 × 10−2 to 15.4 × 10−2 g mg−1 min−1 ). The adsorption followed Langmuir isotherm model with monolayer
adsorption capacity of 3.0–28.8 mg g−1 . The process was endothermic with
H in the range 29.2–50.7 kJ mol−1 accompanied by increase in entropy
and decrease in Gibbs energy. The results have shown that kaolinite, montmorillonite and their poly(oxo zirconium) and tetrabutyl-ammonium
derivatives could be used as adsorbents for separation of Cu(II) from aqueous solution.
© 2005 Elsevier B.V. All rights reserved.

Keywords: Adsorption isotherm; Kinetics; Cu(II); Kaolinite; Montmorillonite; Poly(oxo zirconium) clay; Tetrabutyl-ammonium clay

1. Introduction Clays are hydrous aluminosilicates broadly defined as those

minerals that make up the colloid fraction (<2 ␮m) of soils, sed-
The presence of heavy metals in the environment is a major iments, rocks and water [6]. The particles of clay minerals may
concern due to their acute toxicity. Aqueous effluents contain- be crystalline or amorphous, platy or fibrous and may vary from
ing heavy metal contaminants continue to be discharged into colloid dimensions to those above the limit of resolution of an
the environment from the industries. Natural weathering pro- ordinary microscope [7]. Clay structure is layered with inter-
cesses, waste emissions, atmospheric depositions and anthro- layer space and the layers may be electrically neutral or charged.
pogenic activities also contribute to the metal content of water The high specific surface area, chemical and mechanical stabil-
bodies. The problems associated with heavy metal pollution ity, variety of structural and surface properties, higher values
could be reduced or minimized by precipitation, ultra-filtration, of cation exchange capacities, etc., make the clays an excellent
electrode-deposition, solvent extraction [1,2], etc., but these group of adsorbents. Presence of both Bronsted and Lewis acid-
processes have the disadvantages of high cost, generation of ity [8] on clay surface further enhance their adsorption capacity.
secondary pollutants, poor removal efficiency and ineffective- Clay minerals in soil play the role of a natural scavenger of pol-
ness for low metal concentration. Adsorption has been found lutants from water through both ion exchange and adsorption
to be an effective and economic method with high potential for mechanism. As water flows over soil or penetrates underground,
the removal, recovery and recycling of metals from wastewater clays in soil take up various pollutants through ion exchange and
[3–5]. adsorption mechanisms.
Considering the favourable characteristics, adsorption of
∗ Corresponding author. Fax: +91 361 2700311. metal ions and other substances on clays has received consid-
E-mail address: krishna2604@sify.com (K.G. Bhattacharyya). erable attention. For example, illite has been shown to adsorb

1383-5866/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.seppur.2005.12.014
 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397 389

Cd(II) [9] and natural bentonite eliminates zinc from aqueous (C4 H9 )4 N+ Br− (CDH, Mumbai, India) were used. A stock solu-
solution [10]. Removal of Cr(III), Ni(II), Zn(II), Cu(II) and tion containing 1000 mg of Cu(II) per liter was prepared by
Cd(II) by natural and Na-exchanged bentonites is also reported dissolving CuSO4 ·5H2 O (Qualigens, Mumbai, India) in double
[11]. Strawn et al. [12] reported the adsorption of Cu(II) by smec- distilled water and was used to prepare the adsorbate solutions
tites (montmorillonite and bedellite), whereas Lin and Juang by appropriate dilution. The Cu(II) solution as prepared had a
[13] used surfactant modified montmorillonite for the removal pH of 5.7 which did not change significantly with dilution.
of Cu(II) and Zn(II).
Intercalation of clays cross-linked with inorganic or organic
clusters, known as pillared interlayered clays (PILC) or cross- 2.2. Clay adsorbents
linked clays (CLC) has received wide attention recently as
shape-selective catalysts, separating agents, supports, adsor- Kaolinite, KGa-1b (C1) and Montmorillonite, SWy-2 (C2)
bents, etc. These materials have a two-dimensional pore size were obtained from the University of Missouri-Columbia,
larger than conventional zeolites and exhibit specific properties Source clay Minerals Repository, USA.
depending on the nature of the pillar [14]. In particular, clays pil- Poly(hydroxy zirconium) kaolinite (C3) and montmorillonite
lared with metal oxides are of great importance because of their (C4) were prepared by Burch and Warburton [15] procedure,
high thermal stability, high surface area and intrinsic catalytic i.e. by making a suspension of 4 g of the clay in 100 mL of
activity. These materials are usually prepared by ion-exchanging double distilled water followed by slow addition of 100 mL of
cations in the interlayer region of swelling clays with bulky alky- 0.1 M solution of ZrOCl2 under constant stirring. Stirring was
lammonium ions, polynuclear complex ions bearing inorganic continued for 24 h after which the suspension was filtered and
ligands (hydroxo ligand, chloro ligand), large metal complex the modified clay was washed with water till it was free of Cl−
ions bearing organic ligands, etc. The intercalated species are ion and then dried in an air oven at 373 K for 30 min.
capable of preventing the collapse of the interlayer spaces, prop- For preparing TBA-clays [19] (TBA-kaolinite C5, TBA-
ping open the layers as pillars, and forming interlayer space. On montmorillonite C6) the clays were saturated with Na+ by stir-
heating, the intercalated inorganic species are converted to metal ring 10 g of the sample with 1 L of 1 M NaCl solution for 12 h
oxide clusters, generating a stable microporous structure with a and then allowing to settle. The supernatant liquid was discarded
high surface area. and the process was repeated with fresh 1 M NaCl solutions. This
Burch and Warburton [15] introduced Zr-tetramers from fresh was repeated twice. The clay was separated by centrifugation,
zirconyl chloride solution into montmorillonite. Ohtsuka et al. washed with water several times to make Cl− free. The Na+ -
[14] have shown that the zirconium species exist in a number saturated clay was then mixed with water and 300 mL aqueous
of polynuclear ionic species in zirconium oxychloride solution TBA-Br solution in a total volume of 1 L with TBA-Br content
and pillaring can give three kinds of microporous clays with 7-, five times the CEC of the clay. The mixture was centrifuged
12- and 14-Å interlayer spacings. The major species in zirco- and washed with water several times till it was free of Br− . The
nium oxychloride solutions at room temperature has been shown resulting organoclay was dried in an air oven at 373 K for 30 min.
as the zirconium tetramer, [Zr4 (OH)8+x (H2 O)16−x ](8−x)+ , giv- All the clays were calcined before using them as adsorbents
ing the intercalation compound with a 7-Å interlayer spacing. (kaolinite, montmorillonite and the ZrO-derivatives at 773 K and
Upon hydrolytic polymerization of the zirconium tetramer in the TBA-derivatives at 973 K for 10 h, the higher temperature
solution, the more highly polymerized zirconium species are in case of the TBA-derivatives was necessary to get rid of the
generated that can provide the intercalation compounds with organic template).
12- or 14-Å interlayer spacings. The two polymeric species
forming the intercalates with 12- and 14-Å interlayer spac- 2.3. XRD measurement
ings are composed of three-dimensionally polymerized clusters
based upon the tetramer. Pereira et al. [16] prepared pillared Zr- XRD measurements were done with Phillips Analytical X-
montmorillonite in a similar way. Preparation of pillared mont- ray spectrometer (PW 1710) using Cu K␣ radiations.
morillonite with TMA (tetramethylammonium), TMP (tetram-
ethylphosphonium) and TMPA (trimethyl-phenylammonium)
cations has also been reported [17,18]. 2.4. Surface area
The present study has been designed with an aim to investi-
gate the differences in behaviour of kaolinite, montmorillonite Sears’ method was chosen to estimate the surface areas of
and their poly(oxo zirconium) and tetrabutyl-ammonium deriva- the clay adsorbents [20]. 0.5 g of each clay was acidified with
tives towards adsorption of Cu(II) from aqueous solution. 0.1 M HCl to a pH 3–3.5. The volume was made up to 50 mL
with distilled water after addition of 10.0 g of NaCl. The titration
2. Materials and methods was carried out with standard 0.1 M NaOH in a thermostatic bath
at 298 ± 0.5 K to pH 4.0, and then to pH 9.0. The volume, V,
2.1. Chemicals required to raise the pH from 4.0 to 9.0 was noted and the surface
area was computed from the following equation:
Reagent grade chemicals, viz., ZrOCl2 ·8H2 O (Loba
Chemie, Mumbai, India) and tetrabutyl-ammonium bromide, S (m2 /g) = 32V − 25 (1)
390 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397

2.5. Cation exchange capacity (CEC) phases. The adsorption process is described by the Langmuir
isotherm [22]:
The copper bisethylenediamine complex method [21] was
used to determine the cation exchange capacity (CEC) of the Ce /qe = (1/bqm ) + (1/qm )Ce (3)
clays. Fifty milliliters of 1 M CuCl2 solution was mixed with
where b and qm are Langmuir coefficients, respectively, repre-
102 mL of 1 M ethylenediamine solution to allow for the for-
senting the adsorption equilibrium constant and the monolayer
mation of the [Cu(en)2 ]2+ complex. The slight excess of the
capacity, Ce is the concentration of Cu(II) at equilibrium. The
amine ensures complete formation of the complex. The solution
linear Langmuir plots are obtained by plotting Ce /qe versus Ce ,
is diluted with water to 1 L to obtain a 0.05 M solution of the
respectively, from which the adsorption coefficients are evalu-
complex. 0.5 g of dry clay sample was mixed with 5 mL of the
ated.
complex solution in an Erlenmeyer flask, diluted with distilled
The thermodynamic parameters,
H,
S and
G, for the
water to 25 mL and the mixture was shaken for 30 min in a ther-
adsorption process are calculated using the relation [23]:
mostatic water bath and centrifuged. The concentration of the
complex remaining in the supernatant is determined by iodomet- ln Kd =
S/R −
H/RT (4)
ric method. For this, 5 mL of the supernatant was mixed with
5 mL of 0.1 M HCl to destroy the [Cu(en)2 ]2+ complex and KI where Kd , the distribution coefficient of the adsorbate, is equal
salt was added at 0.5 g/mL of solution. The mixture was titrated to (qe /Ce ). The plot of ln Kd versus 1/T yields straight lines with
with 0.02 M Na2 S2 O3 solution with starch as indicator and the the slope and the intercept giving values of
H and
S. These
CEC was calculated from the formula: values could be used to compute
G from the Gibbs relation,

G =
H − T
S at constant temperature. In deriving the values
MSV (x − y)
CEC (meq/100 g) = (2) of the thermodynamic parameters, it is assumed that the enthalpy
1000m does not change with temperature.
where M is the molar mass of the complex, S the concentration The kinetics of the adsorption process was studied at constant
of the thio solution, V the volume (mL) of the complex taken temperature using pseudo second order model [24–26] for the
for iodometric titration, m the mass of adsorbent taken (g), x rate expression given by
the volume (mL) of thio required for blank titration (without
the adsorbent) and y is the volume (mL) of thio required for the dqt /dt = k2 (qe − qt )2 (5)
titration (with the adsorbent).
where k2 is the second order rate constant, qt and qe are the
amount of Cu(II) adsorbed at time (t) and at equilibrium, respec-
2.6. Adsorption experiments
tively. By integrating (5) and using the boundary conditions of
t = 0 to t = t and qt = 0 to qt = qt , the linear form of the equation
The adsorption experiments were carried out in 100 mL
is obtained as
Erlenmeyer flasks by mixing together a constant amount of
clay with a constant volume of the aqueous solution of Cu(II). t/qt = (1/k2 qe2 ) + (1/qe )t (6)
The contents in the flasks were agitated by placing them in a
constant temperature water bath thermostat for a known time The plot of t/qt versus t gives a linear plot, which allows com-
interval. The mixture was then centrifuged (Remi R 24) and putation of qe and k2 .
Cu(II) remaining unadsorbed in the supernatant liquid was deter-
mined with Atomic Absorption Spectroscopy (Varian SpectrAA 3. Results and discussion
220, air-acetylene oxidizing flame, Lamp current 4 mA, wave-
length 324.8 nm, slit width 0.5 nm, optimum working range 3.1. Adsorbent characterization
0.02–3.0 ␮g/mL). The following conditions were maintained for
different sets of experiments: 3.1.1. XRD study
The XRD spectra of the clays are given in Fig. 1 and a sum-
Effects of pH Clay 2 g/L; Cu(II) 50 mg/L; temperature 303 K; interac- mary of the results is given below:
tion time 360 min; pH 1.0–6.0 at unit intervals
Kinetics Clay 2 g/L; Cu(II) 50 mg/L; temperature 303 K; pH 5.7;
interaction time 15, 30, 45, 60, 90, 120, 150, 180, 240, (i) Intercalation followed by calcination showed a consider-
360 min able loss of crystallinity in all the clay minerals with a
Isotherms Clay 2 g/L; temperature 303 K; interaction time 360 min; concomitant loss in the number and intensity of the XRD
pH 5.7; Cu(II) 10, 20, 30, 40, 50 mg/L
peaks. The number of peaks in 2θ range 1–30◦ decreased
Thermodynamics Clay 2 g/L; interaction time 360 min; pH 5.7; temperature
303, 308, 313 K; Cu(II) 10, 20, 30, 40, 50 mg/L from 13 (pure, calcined montmorillonite) to 5 for ZrO-
montmorillonite, and to 9 for TBA-montmorillonite. In the
same range of 2θ, the number of peaks decreased from 7
2.7. Theoretical basis (pure, calcined kaolinite) to 5 for ZrO-kaolinite, and to just
1 for TBA-kaolinite.
The process of adsorption involves an array of phenomena (ii) A general widening of the XRD peaks and loss of inten-
that can alter the distribution of pollutants among the constituent sity was also a common feature. The intercalation of
 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397 391

Fig. 1. XRD patterns for (a) kaolinite (C1), (b) montmorillonite (C2), (c) ZrO-kaolinite (C3), (d) ZrO-montmorillonite (C4), (e) TBA-kaolinite (C5) and (f) TBA-
montmorillonite.

montmorillonite with poly(hydroxo zirconium) salt fol- (iv) The tip width of the XRD peaks had also widened. The tip
lowed by calcination expanded the basal spacing from width of the 8.75◦ peak (2θ) in montmorillonite increased
10.09 to 10.50 Å (2θ = 8.75◦ ) which was accompanied by from 0.24 to 0.96 for poly(oxo zirconium) montmoril-
a decrease in intensity from 12.05 to 7.33%. For the TBA- lonite and of the 9.26◦ peak (2θ) from 0.20 to 0.56 for
montmorillonite, the expansion was from 9.54 to 9.62 Å TBA-montmorillonite. When pure, calcined kaolinite was
(2θ = 9.26◦ ) when the intensity decreased from 16.68 to converted to ZrO-kaolinite, the tip width of the 12.35◦ (2θ)
7.06%. In case of kaolinite and its derivatives, widening peak changed from 0.10 to 0.16. Similarly, the tip width of
of the basal spacings was not very prominent but the XRD the 21.28◦ (2θ) peak changed from 0.16 to 0.24 when pure,
peaks became less intense compared to those of the pure, calcined kaolinite was converted to TBA-kaolinite.
calcined kaolinite (e.g. 12.35◦ (2θ) peak had intensity of
only 73.4% for ZrO-kaolinite compared to the intensity of It has been well known that intercalation followed by cal-
99.6% for the pure, calcined kaolinite). Similar observa- cination reduces crystallinity [27,28] and the structure of the
tion was made with respect to all other XRD peaks, but the resultant clay becomes amorphous to XRD. The process of cal-
intensity decrease was much less in case of TBA-kaolinite. cination, necessary to stabilize the intercalated solids, leads to
(iii) Intercalation followed by calcination drastically reduced decrease in basal spacings sometimes accompanied by struc-
the number of low-angle peaks. The modified clays showed tural collapse. XRD of modified clays shows poor crystallinity,
no low-angle peaks in the XRD measurements. broad and less intense peaks compared to the parent clay mineral
392 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397

Table 1
Surface area and cation exchange capacity (CEC) of clay adsorbents
Parameters C1 C2 C3 C4 C5 C6

Surface area (m2 /g) 3.8 19.8 13.4 35.8 14.2 42.2
CEC (meq/100 g) 11.3 153.0 10.2 73.2 3.9 47.6

[29] due to the presence of layers intercalated with different

polycations, or an irregular stacking of intercalated and non-
intercalated layers. The parent clays are usually less ordered than
the intercalated clays, but the ordering can hardly withstand cal-
cination. The calcined clay gives less number of or no low-angle
peaks pointing to the disappearance of long-range ordering in c
direction. The results obtained in the present work are in agree-
ment with these observations. Further, the XRD measurements
indicated that ZrO and TBA have entered into the layers of both
kaolinite and montmorillonite. However, the extent of interca-
lation and pillaring was not measured in the present work.

3.1.2. Surface area

The surface area of a porous material is one of the most
useful microstructural parameters for defining its properties.
Fig. 2. Variation of amount of Cu(II) adsorbed per unit mass (qe ) of the clays at
It is the total internal boundary between the solid phase and
different pH at 303 K (clay 2 g/L; initial Cu(II) 50 mg/L; time 360 min).
the pore system [30]. The measured surface areas of the six
adsorbents are given in Table 1. The specific surface area of
kaolinite (C1) was only 3.8 m2 /g, but it increased to 13.4 m2 /g 3.2. Adsorption of Cu(II)
for the ZrO-derivative (C3) and 14.0 m2 /g for the TBA-derivative
(C5). Similarly, the montmorillonite sample (C2) had a specific 3.2.1. Effects of pH
surface area of 19.8 m2 /g, which is almost five times that of kaoli- Variation in the amount of adsorption per unit mass is shown
nite. On introduction of ZrO and TBA into montmorillonite, the in Fig. 2. The amount adsorbed (qe ) increased as the pH of
specific surface area increased to 35.8 m2 /g (C4) and 42.2 m2 /g the solution was changed to higher values. The experiments,
(C6), respectively. The surface areas of montmorillonite and Na- however, could not be done at pH > 6.0 due to low solubility of
montmorillonite were reported, respectively, as 19.0 m2 /g [31] Cu(II). The adsorption of Cu(II) apparently increased rapidly
and 27.3 m2 /g [32]. The surface area of kaolinite was reported after pH > 6.0 which was obviously due to onset of precipi-
as 5–25 m2 /g [33]. Similar results have been reported by other tation of the Cu(II) hydroxide [37,38]. Rapid decrease in the
authors [28,29] suggesting that intercalation creates a porous concentration of Cu(II) after pH > 6.0 was also observed in a sep-
framework increasing the surface area [34]. arate experiment in which Cu(II) concentration was monitored
at various pHs starting from an initial Cu(II) concentration of
3.1.3. Cation exchange capacity (CEC) 50 mg/L.
The CEC is the number of equivalents of exchangeable charge At low pH, the number of available hydrogen ions is high
per mass of clay, which is equivalent with the layer charge and Cu(II) ions have to compete with them for the adsorption
[35]. In this work, montmorillonite was found to have a very sites on the adsorbent surface. The active sites on the adsor-
large CEC (Table 1) compared to that of kaolinite, the values bent surface are weakly acidic in nature and with increase in
obtained being 11.3 meq/100 g (kaolinite) and 153.0 meq/100 g pH, they are gradually deprotonated favouring more and more
(montmorillonite) in agreement with the reported values [36]. Cu(II) uptake. Similar adsorption mechanism has been reported
Intercalation and calcination had a drastic effect on the CEC of by various authors [39–41].
the derivative clays resulting in a large decrease, particularly All the six clays showed an identical behaviour of increased
for the montmorillonite derivatives. ZrO-kaolinite and TBA- uptake of Cu(II) per unit mass with gradually increasing pH,
kaolinite thus had CEC of 10.2 and 3.9 meq/100 g while ZrO- but differences existed in the amounts of uptake. Montmoril-
montmorillonite and TBA-montmorillonite had CEC of 73.2 and lonite and TBA-montmorillonite exhibited highest adsorption
47.6 meq/100 g. Evidently, the large cations (ZrO2+ and TBA+ ) in that order followed by ZrO-montmorillonite, kaolinite, TBA-
mask some of the exchangeable cations in the interlayer space kaolinite and ZrO-kaolinite. The last two have very little differ-
leading to decrease in CEC. Similar results have been reported ence between them showing that introduction of TBA or ZrO into
recently [34]. The large decrease in CEC in case of montmo- kaolinite structure does not have a major impact with regard to
rillonite derivatives compared to the kaolinite derivatives is a Cu(II) uptake except for suppressing the same a little. In case of
clear indication that intercalation was more successful in case montmorillonite, while ZrO suppresses the adsorption of mont-
of montmorillonite. morillonite almost to the level of kaolinite, it is much less for
 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397 393

Fig. 3. Variation of amount of Cu(II) adsorbed per unit mass of clay (qe ) with
time at 303 K (clay 2 g/L; initial Cu(II) 50 mg/L; pH 5.7).

Fig. 4. Second order plots for adsorption of Cu(II) at 303 K (clay 2 g/L; initial
the TBA-derivative. It is seen that at any pH, the order of Cu(II) Cu(II) 50 mg/L; pH 5.7).
adsorption is C2 > C6 > C4 > C1 > C5 > C3.
mental results of kinetic study. On comparison, it was found that
3.2.2. Kinetics of adsorption the second order kinetics based on plotting t/qe versus t (Fig. 4)
Cu(II) uptake by adsorption demonstrated a continuous yielded the best results.
increase till 240 min after which the increase was insignificant Very good linear plots for Cu(II)–clay systems (R: 0.99) were
(Fig. 3). obtained with second order rate constant, k2 from 7.7 × 10−2
In the initial bare surface, the sticking probability is high to 15.4 × 10−2 g mg−1 min−1 (Table 2). This also yielded good
and consequently, adsorption proceeded with a high rate. With agreement between the experimentally obtained qe values and
increasing coverage, the number of sites become less and Cu(II) those obtained from the slopes of the second order plots (devi-
ions have to fiercely compete among themselves for getting ations +1.2 to +11.5%, Table 2). The small deviations might be
adsorbed. Together with this, the rate at which Cu(II) ions due to the uncertainty inherent in obtaining the experimental
are transported from the bulk of the solution to the clay sur- qe values. The overall mechanism of Cu(II)–clay interactions
face significantly affect the kinetics [42]. In the present work, therefore predominantly follow second order kinetics in which
various kinetic models starting from Lagergren’s pseudo first both Cu(II) ions and the substrate (clay surface) are effective
order equation, second order equation, Elovich equation, Liquid participants. It is to be noted that although the six clays have
film diffusion model, and intra-particle diffusion equation were different uptakes of Cu(II) ions, the mechanism of interaction
tested to find out which model is in agreement with the experi- essentially remains the same.

Table 2
Second order rate constant (g mg−1 min−1 ) and qe values (experimental and computed) from the plots at 303 K (clay 2 g/L; initial Cu(II) 50 mg/L; pH 5.7)
Adsorbents k2 (×102 g mg−1 min−1 ) R qe (mg g−1 ) Deviation (%)
Experimental Second order plot

C1 9.4 0.99 3.9 4.3 10.2

C2 15.4 0.99 21.0 21.3 1.2
C3 14.3 0.99 2.6 2.9 11.5
C4 7.7 0.99 5.3 5.8 9.4
C5 14.4 0.99 2.8 3.1 10.7
C6 14.2 0.99 15.1 15.4 1.9
394 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397

tively. This decrease was 75.4 and 5.4%, respectively, for ZrO-
montmorillonite (C4) and TBA-montmorillonite (C6) in relation
to that of montmorillonite (C2). Introduction of ZrO into mont-
morillonite apparently blocked a large number of adsorption
sites and calcination did not make these sites free. TBA-groups
might have blocked adsorption sites similarly, but during cal-
cination, the pillaring TBA-cations must have broken down
leaving the sites free with a small decrease in qm .
The Langmuir equilibrium coefficient, b, has values of 18.9,
30.1, 34.6, 6.6, 25.1 and 5.8 L g−1 , respectively, for the adsor-
bents C1–C6. The two clays, ZrO-montmorillonite (C4) and
TBA-montmorillonite (C6), have lower ‘b’ values compared to
the others indicating that modification of montmorillonite with
ZrO and TBA has influenced the adsorption equilibrium more
than the others.
Pan et al. [43] studied the adsorption of Cu(II) on two types of
sewage sludge ash and reported Langmuir monolayer capacity,
qm , of 3.282 and 4.139 mg g−1 while, Ho et al. [5] have found
qm of 7.58 mg g−1 for adsorption of Cu(II) on tree fern. These
values are quite close to the values found in the present work.

Fig. 5. Langmuir isotherms for Cu(II) adsorption at 303 K (clay 2 g/L; initial 3.2.4. Thermodynamic studies
Cu(II) 10, 20, 30, 40, 50 mg/L; pH 5.7; time 360 min). Amount of Cu(II) adsorbed at three different temperatures of
303, 308 and 313 K showed an increasing trend with increase
Table 3 in temperature suggesting endothermic interactions. Typical
Langmuir adsorption parameters at 303 K (clay 2 g/L; initial Cu(II) 10, 20, 30, results are shown in Fig. 6 for adsorption of Cu(II) on montmo-
40, 50 mg/L; pH 5.7; time 360 min)
rillonite (C2). It is likely that adsorption of Cu(II) on clay surface
Adsorbents Langmuir coefficients requires an activation energy and rise in temperature helps more
R b (L g−1 ) qm (mg g−1 )

C1 0.99 18.9 4.4

C2 0.99 30.1 28.8
C3 0.99 34.6 3.0
C4 0.99 6.6 7.1
C5 0.99 25.1 3.2
C6 0.99 5.8 27.3

3.2.3. Adsorption isotherm

That the interactions were chemical in nature were tested
by applying the Langmuir isotherm to the data obtained for
Cu(II)–clay interactions at constant temperature. The Langmuir
plots obtained by plotting Ce /qe versus Ce (Fig. 5) are linear
for all the six clays (R: 0.99, Table 3). These plots yielded two
important parameters: Langmuir monolayer capacity, qm , giving
the amount of Cu(II) required to occupy all the available sites
in unit mass of the clay and Langmuir equilibrium parameter, b,
related to the equilibrium constant of the process:
Cu(II) + clay = Cu(II) · · · clay (surface adsorption complex).
The six clays (C1–C6) have qm values of 4.4, 28.8, 3.0, 7.1, 3.2,
27.3 mg g−1 , respectively, following the order of adsorption as
C2 > C6 > C4 > C1 > C5 > C1. While introduction of both ZrO
and TBA has resulted in a reduction of the available adsorp-
tion sites for Cu(II), the reduction is very much drastic with
ZrO-montmorillonite (C4). Thus, compared to kaolinite (C1),
the Langmuir monolayer capacity (qm ) of ZrO-kaolinite (C3) Fig. 6. Plots of Ce vs. qe for montmorillonite (C2) at three different temperatures
and TBA-kaolinite (C5) decreased by 33.6 and 29.5%, respec- (clay 2 g/L; initial Cu(II) 10, 20, 30, 40, 50 mg/L; pH 5.7; time 360 min).
 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397 395

Table 4 Entropy increased for all the clays in the temperature range
Thermodynamic data for adsorption of Cu(II) (clay 2 g/L; pH 5.7; time 360 min) 303–313 K and
S varied from 86.8 to 180.7 J K−1 mol−1 for
Adsorbents
H (kJ mol−1 )
S (J K−1 mol−1 ) −
G (kJ mol−1 ) the six clay adsorbents. The adsorptive interactions normally
303 K 308 K 313 K
do not bring about any major change in surface configuration,
yet considerable entropy increase signifies an increased state of
C1 30.7 86.8 26.2 26.7 27.1 randomness at the solid–solution interface following adsorption.
C2 50.7 180.7 54.7 55.6 56.0
C3 50.5 148.3 44.9 45.6 46.4
These are most likely to be due to structural changes and adjust-
C4 36.0 105.3 31.9 32.4 33.0 ments in the adsorbate as well as the adsorbent as the adsorption
C5 49.8 148.8 45.0 45.8 46.5 proceeds. The structural changes arise from the release of ions
C6 29.2 96.5 29.2 29.7 30.2 like H+ and K+ from the clay surface into the solution and also
from partial desolvation of the metal ions [48]. The inference
from these results is that the endothermic adsorption interac-
Cu(II) ions to overcome this energy barrier and get attached to tions are entropically driven.
the surface [44]. Another possibility is the creation of additional Spontaneity of the adsorption process is established by
adsorption sites on clay surface arising due to dissociation of decrease in Gibbs energy,
G, which varied from −26.2 to
some of the surface components of clay [45] at a temperature −27.1 kJ mol−1 (C1), −54.7 to −56.0 kJ mol−1 (C2), −44.9
higher than the ambient which increases Cu(II) uptake. These to −46.4 kJ mol−1 (C3), −31.9 to −33.0 kJ mol−1 (C4), −45.0
two influences may operate independently of one another or may to −46.5 kJ mol−1 (C5) and −29.2 to −30.2 kJ mol−1 (C6) in
be operational simultaneously. the temperature range 303–313 K. The decrease was more at
The thermodynamic parameters,
H,
S and
G, for the higher temperature such that the interactions were favoured by
adsorption process, are computed from the plots of ln Kd ver- an increase in temperature in agreement with the endothermic
sus 1/T (Table 4). Typical ln Kd versus 1/T plots are shown nature of Cu(II) adsorption on the clays.
in Fig. 7. The adsorption enthalpy,
H, had the lowest value Thermodynamic data on metal adsorption on clays are
of 29.2 kJ mol−1 for TBA-montmorillonite (C6) and the high- scarce.
H,
S and
G for Cu(II) adsorption on surfac-
est values of 50.5 and 50.7 kJ mol−1 , respectively, for ZrO- tant modified montmorillonite were reported as 7.05 kJ mol−1 ,
kaolinite (C3) and montmorillonite (C2) in the temperature 9.09 J K−1 mol−1 and −9.66 kJ mol−1 , respectively [13]. Yuvaz
range 303–313 K. The values of
H are high enough to ensure et al. [47] have found that
H,
S and
G for adsorp-
strong interaction between the metal ions and the clay. Endother- tion of Cu(II) on Turkish kaolinite are 39.52 kJ mol−1 ,
mic adsorption of Cu(II) was also reported by other authors 11.7 J K−1 mol−1 and 4.61 kJ mol−1 , respectively. Echeverria et
[46,47]. al. [49] have found that
H,
S and
G for adsorption of Ni(II)
on illite have values of +16.8 kJ mol−1 , 58 J mol−1 K−1 and
−1.04 kJ mol−1 , respectively. These values are comparatively
smaller than the values obtained in the present study (except

H value of Cu(II) adsorption on Turkish kaolinite) indicating
the Cu(II)–clay adsorption complex to be much more stable with
the clays holding the Cu(II) ions strongly to the surface.

4. Conclusion

Kaolinite, montmorillonite and their poly(oxo zirconium)

and tetrabutyl-ammonium derivatives interact with Cu(II) in
aqueous solution with sufficient uptake of the metal ions. Inter-
calation of both the clays with Zr(hydr)oxide or TBA does not
improve the adsorption capacity possibly due to the blocking of
the negatively charged sites and the pores by the poly(hydroxo
zirconium) or TBA ions. Although intercalation or pillaring
improves the adsorptive properties in some cases, the results
in the present set of experiments show that ZrO-polymers or
TBA ions are not the appropriate pillaring agents for raising the
adsorption capacity of either kaolinite or montmorillonite with
respect to Cu(II).
Adsorption of Cu(II) on the clays is influenced by pH of the
medium and it is seen that more and more Cu(II) ions are taken up
by the clays as pH is increased, the only limiting factor being the
precipitation of Cu(II) as the hydroxide at pHs > 6.0. The kinet-
Fig. 7. Plots of ln Kd vs. 1/T for clay adsorbents (clay 2 g/L; initial Cu(II) ics of the interactions is not straightforward, but as a whole, the
50 mg/L; pH 5.7; time 360 min). interactions could be better represented by a second order mech-
396 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397

anism. The adsorption data follow Langmuir isotherm and all [20] G. Sears, Determination of specific surface area of colloidal silica by
six adsorbents have reasonable Langmuir monolayer capacity. titration with sodium hydroxide, Anal. Chem. 28 (1956) 1981–1983.
Thermodynamically, the interactions are endothermic in nature, [21] F. Bergaya, M. Vayer, CEC of clays: measurement by adsorption of a
copper ethylenediamine complex, Appl. Clay Sci. 12 (1997) 275–280.
being driven by entropy increase and Gibbs energy decrease. [22] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica
and platinum, J. Am. Chem. Soc. 40 (1981) 1361–1403.
Acknowledgements [23] W. John Thomas, B. Crittenden, Adsorption Technology and Design,
Butterworth/Heinemann, Oxford, 1998, pp. 38–50.
The authors are very grateful to the three reviewers for [24] M.M. Abou-Mesalam, Sorption kinetics of copper, zinc, cadmium and
nickel ions on synthesized silico-antimonate ion exchanger, Colloids
carefully going through the manuscript and pointing out var- Surf. A: Physicochem. Eng. Aspects 225 (2003) 85–94.
ious errors and omissions along with very helpful comments. [25] C.T. Tien, C.P. Huang, Formation of surface complexes between heavy
One of the authors (SSG) is grateful to the University Grants metals and sludge particles, in: J.P. Vernet (Ed.), Trace Metals in the
Commission, India for providing assistance under the Faculty Environmental, 1. Heavy Metals in the Environment, Elsevier, Amster-
Improvement Programme for this work. dam, 1991, pp. 295–311.
[26] Y.S. Ho, G. McKay, Batch lead (II) removal from aqueous solution by
peat: equilibrium and kinetics, Trans. I Chem. E 77B (1999) 165–173.
References [27] L. Li, X. Liu, Y. Ge, R. Xu, Structural studies of pillared saponite, J.
Phys. Chem. 97 (1993) 10389–10393.
[1] N. Chiron, R. Guilet, E. Deydier, Adsorption of Cu(II) and Pb(II) onto [28] T. Grzybek, J. Klinik, D. Olszewska, H. Papp, J. Smarzowski, The Influ-
a grafted silica: isotherms and kinetic models, Water Res. 37 (2003) ence of montmorillonite treatment on structure, sorption properties and
3079–3086. catalytic behaviour. Part I. Zirconia pillared clays modified with man-
[2] K. Kadirvelu, C. Namasivayam, Activated carbon from coconut coirpith ganese as denox catalysts, Pol. J. Chem. 75 (2001) 857–868.
as metal adsorbent: adsorption of Cd(II) from aqueous solution, Adv. [29] J.L. Valverde, P. Sanchez, F. Dorado, I. Asencio, A. Romero, Preparation
Environ. Res. 7 (2003) 471–478. and characterization of Ti-pillared clays using Ti alkoxide. Influence of
[3] H.H. Tran, F.A. Roddick, J.A. O’Donell, Comparison of chromatography the synthesis parameters, Clays Clay Miner. 51 (2003) 41–51.
and desiccant silica gels for the adsorption of metal ions. I. Adsorption [30] J.J. Thomas, H.M. Jennings, A.J. Allen, The surface area of hardened
and kinetics, Water Res. 33 (1999) 2992–3000. cement paste as measured by various techniques, Concrete Sci. Eng. 11
[4] T.W. Chang, M.K. Wang, C. Lin, Adsorption of copper in the different (1999) 45–64.
sorbent/water ratios of soil system, Water Air Soil Pollut. 138 (2002) [31] J. Ravichandran, B. Sivasankar, Properties and catalytic activity of acid-
199–209. modified montmorillonite and vermiculite, Clays Clay Miner. 45 (1997)
[5] Y.S. Ho, C.T. Huang, H.W. Huang, Equilibrium sorption isotherm for 854–858.
metal ions on tree fern, Process Biochem. 37 (2002) 1421–1430. [32] T. Kasama, Y. Watanabea, H. Yamada, T. Murakami, Sorption of phos-
[6] T.J. Pinnavaia, Intercalated clay catalysts, Science 220 (1983) 365–371. phates on Al-pillared smectites and mica at acidic to neutral pH, Appl.
[7] W.A. Deer, R.A. Howie, J. Zussman, An Introduction to the Rock- Clay Sci. 25 (2004) 167–177.
forming Minerals, ELBS Longman, Essex, England, 1985. [33] C. Volzone, J.G. Thompson, A. Melnitchenko, J. Ortiga, S.R. Palethorpe,
[8] K. Tanabe, Catalysis Science and Technology, Springer Verlag, New Selective gas adsorption by amorphous clay-mineral derivatives, Clays
York, 1981. Clay Miner. 5 (1999) 647–657.
[9] J.C. Echeverria, E. Churio, J. Garrido, Retention mechanisms of Cd on [34] H. Jobstmann, B. Singh, Cadmium sorption by hydroxy-aluminium
illite, Clays Clay Miner. 50 (2002) 614–623. interlayered montmorillonite, Water Air Soil Pollut. 131 (2001) 203–
[10] A. Mellah, S. Chegrouche, The removal of zinc from aqueous solutions 215.
by natural bentonite, Water Res. 31 (1997) 621–629. [35] H. van Olphen, J.J. Fripiat (Eds.), Data Handbook for Clay Mate-
[11] E. Alvarez-Ayuso, A. Garcia-Sanchez, Removal of heavy metals from rials and other Non-metallic Minerals, Pergamon Press, New York,
wastewaters by natural and Na-exchanged bentonites, Clays Clay Miner. 1979.
51 (2003) 475–480. [36] R.E. Grim, Clay Mineralogy, McGraw-Hill, New York, 1968.
[12] D.G. Strawn, N.E. Palmer, L.J. Furnare, C. Goodell, J.E. Amonette, [37] S.T. Bosso, J. Enzweiler, Evaluation of heavy metal removal from aque-
R.K. Kukkadapu, Copper sorption mechanisms on smectites, Clays Clay ous solution onto scoecite, Water Res. 36 (2002) 4795–4800.
Miner. 52 (2004) 321–333. [38] G. Bayramoglu, S. Bektas, M.Y. Arica, Biosorption of heavy metal ions
[13] S.-H. Lin, R.-S. Juang, Heavy metal removal from water by sorption on immobilized white-rot fungus Trametes versicolor, J. Hazard. Mater.
using surfactant-modified montmorillonite, J. Hazard. Mater. B 92 (2002) B 101 (2003) 285–300.
315–326. [39] K.H. Chu, Removal of copper from aqueous solution by chitosan prawn
[14] K. Ohtsuka, Y. Hayashi, M. Suda, Microporous ZrO2 pillared clays shell: adsorption equilibrium and kinetics, J. Hazard. Mater. B 90 (2002)
derived from three kinds of Zr polynuclear ionic species, Chem. Mater. 77–95.
5 (1993) 1823–1829. [40] F. Pagnanelli, A. Esposito, L. Toro, F. Veglio, Metal speciation and
[15] R. Burch, C.I. Warburton, Zr containing pillared interlayered clays. 1. pH effect on Pb, Cu, Zn and Cd biosorption onto Sphaerotilus natans:
Preparation and structural characterization, J. Catal. 97 (1986) 503–510. Langmuir-type empirical model, Water Res. 37 (2003) 627–633.
[16] P.R. Pereira, J. Pires, M.B. de Carvalho, Zirconium pillared clays for [41] V. Boonamnuayvitaya, C. Chaiya, W. Tanthapanichakoon, S.
carbon dioxide/methane separation. 1. Preparation of adsorbent materials Jarudilokkul, Removal of heavy metals by adsorbents prepared from
and pure gas adsorption, Langmuir 14 (1998) 4584–4588. pyrolysed coffee residues and clay, Sep. Purif. Technol. 35 (2004) 11–22.
[17] M.A.M. Lawrence, R.K. Kukkadapu, S.A. Boyd, Adsorption [42] B. Yu, Y. Zhang, A. Shukla, S.S. Shukla, K.L. Dorris, The removal of
of phenol and chlorinated phenols from aqueous solution by heavy metal from aqueous solutions by sawdust adsorption—removal of
tetramethylammonium- and tetramethylphosphonium-exchanged mont- copper, J. Hazard. Mater. B 80 (2000) 33–42.
morillonite, Appl. Clay Sci. 13 (1998) 13–20. [43] S.-C. Pan, C.-C. Lin, D.-H. Tseng, Reusing sewage sludge ash as adsor-
[18] J.J. Stevens, S.J. Anderson, An FTIR study of water sorption on TMA- bent for copper removal from wastewater, Resour. Conserv. Recycl. 39
and TMPA-montmorillonite, Clays Clay Miner. 44 (1996) 142–150. (2003) 79–90.
[19] M.M. Mortland, S. Shaobai, S.A. Boyd, Clay–organic complexes as [44] A. Shukla, Y.-H. Zhang, P. Dubey, J.L. Margrave, S.S. Shukla, The role
adsorbents for phenol and chlorophenols, Clays Clay Miner. 34 (1986) of sawdust in the removal of unwanted materials from water, J. Hazard.
581–585. Mater. 95 (2002) 137–152.
 K.G. Bhattacharyya, S.S. Gupta / Separation and Purification Technology 50 (2006) 388–397 397

[45] S. Al-Ashem, Z. Duvnjak, Sorption of heavy metals by canola meal, [48] M.J. Angove, B.B. Johnson, J.D. Wells, The influence of temperature
Water Air Soil Pollut. 114 (1999) 251–276. on the adsorption of cadmium(II) and cobalt(II) on kaolinite, J. Colloid
[46] Y.-S. Ho, Removal of copper ions from aqueous solution by tree fern, Interf. Sci. 204 (1998) 93–103.
Water Res. 37 (2003) 2323–2330. [49] J. Echeverria, J. Indurain, E. Churio, J. Garrido, Simultaneous effect of
[47] O. Yuvuz, Y. Altunkayank, F. Guzel, Removal of copper, nickel, cobalt pH, temperature, ionic strength, and initial concentration on the retention
and manganese from aqueous solution by kaolinite, Water Res. 37 (2003) of Ni on illite, Colloids Surf. A: Physicochem. Eng. Aspects 218 (2003)
948–952. 175–187.