ORIGINAL PAPER

Preparation of novel heterogeneous catalysts by adsorptionof a cationic tetrapyrrole on to bentonite: equilibrium, kinetics,and thermodynamics

Altug Mert Sevim • Rustam Hojiyev •

Ahmet Gul • Mehmet Sabri Celik

Received: 13 May 2010 / Accepted: 15 July 2011 / Published online: 19 August 2011

� Springer-Verlag 2011

Abstract Adsorption of an octacationic tetrapyrrole,

octakis(2-trimethylammoniumethylsulfanyl)porphyrazina-

tocobalt octaiodide (QCoPz), from aqueous solutions on to

negatively charged bentonite was investigated. Effects of

temperature, dye concentration, solid concentration, and

contact time on adsorption were determined. Zeta potential

and ion-release measurements were also used as supporting

experiments. Experimental data were analyzed using four

adsorption kinetic models; a pseudo second-order kinetic

model resulted in better correlation with experimental

results than the others. Experimental equilibrium data were

analyzed by non-linear regression using five adsorption

isotherm models with two, three, or four terms. Free

energies, enthalpies, and entropies for the adsorption pro-

cess were determined. The results indicated that adsorption

of QCoPz on bentonite was exothermic and spontaneous in

nature. FT-IR spectroscopy of the composite and its

desorptive behavior were also investigated to identify the

mechanism of adsorption. The novel QCoPz–bentonite

composites obtained are likely to be used in ‘‘green

chemistry’’ and in a wide range of optical and/or catalytic

applications, especially those crucially important in the

petroleum and pulp/paper industries for waste water

cleaning (destruction of mercaptans, sulfides, phenol, and

halogenated aromatics, etc.) and removal of bad odor.

Keywords Clays � Dyes � Intercalation compounds �IR spectroscopy � Porphyrazine

Introduction

Porphyrazines can be regarded as porphyrin analogues with

meso nitrogen atoms replacing the meso carbons. This

alteration results in significant structural and electronic

changes within the macrocycle [1]. Porphyrazines, how-

ever, have received much less synthetic interest than the

related porphyrins and phthalocyanines [2]. They have a

wide variety of redox properties [3]. Furthermore, because

it is possible to control their photochemical and electro-

chemical properties by modification of the substituents and

selection of the central metal, the importance of porphyr-

azines in biological, chemical, and physical research is

likely to increase. Intensive research interest in peripher-

ally functionalized porphyrazines during the last decade

has shown that these tetrapyrrole derivatives should be

regarded as alternatives to the phthalocyanines that have

found extensive applications in many fields, including

materials science, photodynamic therapy of tumors, and as

pigments and dyes [4].

Inorganic–organic hybrids are popular and new func-

tional materials and their importance from a nanostructural

perspective is increasing [5]. Not only is simple hybrid-

ization of inorganic materials and organic compounds

expected but also control of orientation at the molecular

level. Clay minerals are known to be multilayered mate-

rials that yield wide two-dimensional spaces with highly

ordered nanostructured environments for chemical reac-

tions. Synthetic clay minerals can be used as well-

characterized inorganic hosts and have attracted increasing

interest, especially for their applications in photochemical

A. M. Sevim � A. Gul (&)

Department of Chemistry, Istanbul Technical University,

34469 Maslak, Istanbul, Turkey

e-mail: ahmetg@itu.edu.tr

R. Hojiyev � M. S. Celik

Faculty of Mines, Department of Mineral and Coal Processing,

Istanbul Technical University, 34469 Maslak, Istanbul, Turkey

123

Monatsh Chem (2012) 143:385–400

DOI 10.1007/s00706-011-0599-y

and catalytic reactions [6]. Interest in intercalation of

compounds on to clay minerals can be attributed mainly to

the fact that host–guest interactions affect chemical, cata-

lytic, electronic, and optical properties of guest

components [7]. These composite systems not only have

improved catalytic activity, stability, and selectivity but

also enable easy recovery and reuse of the catalyst because

of the supporting environment [8]. Such processes are

currently demanded by environmentally friendly or

‘‘green’’ chemistry which takes three aspects unlimited

natural resources into account:

1. oxygen (from air) as oxidant;

2. often water as solvent; and

3. for photo-oxidations, use of visible light (solar

radiation) [9].

Adsorption of both phthalocyanines and porphyrins on

to clays are frequently investigated, along with the catalytic

and optical applications of the products. Examples include

biomimetic oxidation of hydrocarbons by clay-intercalated

metalloporphyrins and/or metallophthalocyanines [10], use

of molecular oxygen for catalytic oxidation of hydrocar-

bons with clay-intercalated porphyrins [11], photochemical

hole burning of cationic porphyrins intercalated in saponite

[12], removal and photodecomposition of n-nonylphenol

using clay with incorporated copper phthalocyanine [13].

There have been reports of unique hybrid materials in

which either 5,10,15,20-tetrakis(N-methylpyridinium-4-

yl)porphyrin or 5,10,15,20-tetrakis(N,N,N-trimethylanilin-

ium-4-yl)porphyrin (TMAP) cations are adsorbed on the

clay surface with high charge densities and without

aggregation completely neutralizing the surface negative

charges [14]. Despite a number of studies on porphyrin and

clay hybrids [15, 16], studies of clay–porphyrazine com-

posites and their properties are rare.

Sulfide, sulfite, thiosulfate, and mercaptans are by-

products of industrial processes and pollutants of waste and

natural water [17]. At low concentrations the toxic prop-

erties of aqueous solutions can be eliminated by bacterial

oxidation [18] but sulfide concentrations above

70–200 mg dm-3 inhibit bacterial metabolism. Waste

water from the oil industry contains sulfide up to

25 g dm-3. Complete oxidation of sulfur-containing to

non-toxic compounds, before discharging them into

waterways, is the necessary solution to this environmental

pollution problem [19]. Sulfide and other sulfur compounds

were effectively removed by strong oxidants (H2O2,

NaOCl) or catalytically and photocatalytically using metal

tetrapyrroles (especially phthalocyanines) in the presence

of air. Also toxic phenols and chlorinated phenols are

among the basic soluble pollutants of industrial and com-

munal waste water. Polychlorinated aromatics are

especially persistent in the environment, because of their

resistance to oxidation under aerobic conditions leading to

accumulation in the biosphere [20]. For destruction of

phenols and halogenated aromatics, a recently developed

method uses iron tetrasulfophthalocyanine as catalyst in the

presence of H2O2 as oxidant [21]. Another possibility for

waste water cleaning is photooxidation under visible light

irradiation. TiO2 has been used as a photoexcitable semi-

conductor but, because of the band gap of the colorless

TiO2, this process works in the UV region, where only 3%

of solar radiation is active. To overcome this problem of

UV absorption, photosensitizers such as phthalocyanines

and their derivatives, absorbing in the visible region of

light, must be used.

Bentonite is a smectite clay mineral of 2:1 layered sili-

cate that swells when treated with water. The inner layer is

composed of an octahedral sheet, which is situated between

two tetrahedral sheets. Substitution of Al3? for Si4? in the

tetrahedral layer and Mg2? or Zn2? for Al3? in the octa-

hedral layer result in a net negative surface charge in the

clay. The charge imbalance is offset by exchangeable cat-

ions, for example H?, Na?, or Ca2? on layer surfaces [22].

In this study we synthesized a water-soluble quaternary

cobalt porphyrazine (QCoPz) and studied its adsorption

from aqueous solutions on to negatively charged bentonite.

Equilibrium adsorption isotherms were measured for the

single-component system and the experimental data were

analyzed by non-linear regression using five commonly

used models.

Experimental data were analyzed using four adsorption

kinetic models: pseudo first-order, pseudo second-order,

simple Elovich, and interparticle diffusion. Detailed error

analysis was undertaken to investigate the effect of using

different error criteria to obtain the best-fit isotherm and

isothermal data which describe the adsorption process. For

this purpose, three different error functions were used:

residual root mean square error (RMSE), average residuals

(eave), and mean relative deviation modulus (Ee). The data

obtained from adsorption isotherms at different tempera-

tures were fitted to different adsorption models to calculate

thermodynamic quantities such as the free energy of

adsorption, heat of adsorption, and entropy of adsorption.

Ion-release measurement by AAS (atomic absorption

spectroscopy), FTIR techniques, and desorption studies

were performed to understand the mechanism of adsorption

process.

Results and discussion

Effect of solids concentration

To determine the amount of solid to be used in adsorption

experiments, a series of tests was performed with different

386 A. M. Sevim et al.

123

percentage solid concentrations at an initial dye concen-

tration of 1 mmol dm-3. Examination of Fig. 1a reveals

that adsorption density decreases with increasing percent-

age solid concentration and then remains practically stable.

At 0.1% solid concentration, the adsorption density is

0.9560 mmol g-1 and then decreases to 0.0487 mmol g-1

at 2% solid concentration. Consequently, an optimum solid

concentration of 0.1% was selected for further testing.

Effect of mixing time

The adsorption of QCoPz on to bentonite is presented in

Fig. 1b as a function of contact time. More than 99% of

QCoPz was adsorbed in approximately 5 min followed by

a constant adsorption upon further increasing the contact

time. Although the equilibrium is achieved in a short time,

taking the effect of extreme conditions into account a

contact time of 2 h was selected for further testing.

Electrokinetic experiments

Because the isomorphous exchange among ions within the

clay layers, a deficiency in positive charge or, in other

terms, abundance of negative charge occurs. This abun-

dance is compensated by adsorption of cations, for example

Na? and Ca2? on to the layer surface. Bentonite has a

permanent negative charge that increases owing to iso-

morphous substitution of Al3? for Si4? in the tetrahedral

layer and Mg2? for Al3? in the octahedral layer. It is

known from the literature that there is a direct relationship

between adsorption of any adsorbate on to a solid surface

and the change of zeta potential of this solid surface [23]. A

series of measurements were thus conducted to determine

the relationship between adsorption mechanism and zeta

potential, as depicted in Fig. 1c. Evidently, pure clay has a

zeta potential of -35.8 mV, which means that the ben-

tonite used in this study has a negative surface charge.

Because of adsorption of the QCoPz on to the bentonite,

the zeta potential becomes positive, and a final zeta

potential of ?20.24 mV was observed. This change proves

the presence of cationic adsorbate on the negatively

charged clay surface.

Change of pH during adsorption

Bentonite clay and QCoPz have natural pH 8.23 and 8.00,

respectively, in distilled water. Different concentrations of

the dye yielded a constant value of pH 7.30 after the

adsorption process.

Kinetic modeling

Adsorption of organic compounds by natural materials in

aqueous solution is a phenomenon with often complex

kinetics, because of their heterogeneous reactive surfaces.

Because the rate of adsorption controls the residence time

of adsorbate at solid–liquid interfaces, it is an important

factor in determination of the performance of the adsorp-

tion process [24]. Prediction of the adsorption kinetics

provides the most vital information for designing adsorp-

tion systems. Adsorption kinetics include the search for the

best model that best represents the experimental data as a

function of environmental conditions.

In this study, in order to determine the efficiency of

bentonite for adsorption of QCoPz (initial concentra-

tion = 1 mmol dm-3) a series of batch adsorption kinetic

experiments were performed at natural pH (7–8) and 25 �C

for a 2-h contact period. Pseudo first and second-order,

simple Elovich, and intra-particle diffusion kinetic models

were then used to fit the experimental data.

: 25 °C

: 25 °C

: 25 °C

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0 0.5 1.0 1.5 2.0 2.5

adso

rpti

on d

ensi

ty/m

mol

g-1

solid concentration/%

: 25 °CCi : 1 mmol dm-3

pH: Natural (7-8) Mix. Time: 2h

0.930

0.935

0.940

0.945

0.950

0.955

0 100 200 300 400 500 600 700ad

sorp

tion

den

sity

/mm

ol g

-1

contact time/min

: 25 °CCi : 1 mmol dm-3

pH: Natural (7-8)Solid Conc. : 0.1%

-40

-30

-20

-10

0

10

20

30

0.00 0.01 0.02 0.03 0.04

zeta

pot

enti

al/m

V

Ce /mmol dm-3

: 25 °CpH: Natural (7-8)Solid Conc. : 0.1%

(A) (B) (C)

T

T

T

Fig. 1 Adsorption density of QCoPz versus solid concentration (a) and contact time (b) in the bentonite–QCoPz system; variation of zeta

potential with QCoPz concentrations (c)

Preparation of novel heterogeneous catalysts 387

123

The pseudo first-order Lagergren equation [25] is gen-

erally expressed as:

dqt=dt ¼ k1 qe � qtð Þ ð1Þ

where qe and qt (mmol g-1) are the amounts of adsorbed

dye on the adsorbent at equilibrium and at time t, and k1 is

the rate constant of the pseudo first-order kinetic model.

After integration and applying boundary conditions t = 0

to t and qt = 0 to qt, the integrated form of the equation

becomes:

log qe � qtð Þ ¼ log qe � k1=2:303ð Þt ð2Þ

The plots of log(qe - qt) against t (shown in Fig. 2a) for

the pseudo first-order equation give a linear relationship

and k1 and qe values can be determined from the slope and

intercept, respectively, of this equation. The values of the

first-order model equilibrium rate constant k1, which is the

slope of the line in Fig. 2a, and the corresponding

correlation coefficients, are summarized in Table 1

together with the kinetic constants of the other models

used.

The theoretical value qe from the first-order kinetic

model was significantly different from the experimental

value and the correlation coefficient was also found to be

slightly lower. The first-order kinetic model for this system

did not suit the data well.

The kinetic data were further analyzed using pseudo

second-order kinetics [26], represented as:

dqt=dt ¼ k2 qe � qtð Þ2 ð3Þ

where k2 is the pseudo second-order rate constant

(mmol g-1 min-1), qe and qt are the adsorption capacity

(mmol g-1) at equilibrium and at time t, respectively. For

the boundary conditions of t = 0 to t and qt = 0 to qt, the

integrated form of equation becomes:

t=qt ¼ 1= k2q2e

� �þ t=qe ð4Þ

where k2q2e represents the initial rate of adsorption. If

second-order kinetics are applicable, the plot of t/qt versus t

should be linear. The values of k2 and qe obtained from the

pseudo second-order kinetic model, the correlation coeffi-

cients R2, and the experimental qexp are given in Table 1.

The values of pseudo second-order model variables qe and

k2 were determined from the slope and intercept of the t/qt

versus t plot (Fig. 2b).

According to the model, the calculated initial rate of

adsorption (k2q2e) of bentonite was 0.062 mmol g-1 min-1.

As can be seen from Table 1, the pseudo second-order

model adequately fits the data over the entire course of the

experiment with a high correlation coefficient ([0.98).

Moreover, the experimental adsorption capacities of the

adsorbent are close to the theoretical values estimated from

: 25 °C

: 25 °C

: 25 °C

: 25 °C

y = -0.1084x - 3.7866R² = 0.9591

-8

-7

-6

-5

-4

-3

-2

-1

0

0 5 10 15 20 25 30 35

log

(qe

-qt)

t / min

y = 1.0489x + 0.0617R² = 1

0

100

200

300

400

500

600

700

0 100 200 300 400 500 600 700

t /q t

t / min

y = 0.0082x + 0.926R² = 0.969

0.930

0.935

0.940

0.945

0.950

0.955

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

q t/m

mol

g-1

ln t

y = 0.0055x + 0.9266R² = 0.9417

0.930

0.935

0.940

0.945

0.950

0.955

0.960

0 1 2 3 4 5 6

q t/ m

mol

g-1

t0.5

(A)

(C) (D)

(B): 25 °C

: 25 °C

: 25 °C

: 25 °C

T

T

T

T

Fig. 2 Pseudo first-order (a),

pseudo second-order (b),

Elovich (c), and intraparticle

diffusion (d) plots for QCoPz

adsorption

388 A. M. Sevim et al.

123

the model. Hence, error analysis values are much lower

than those obtained by application of the pseudo first-order

model. For the bentonite–QCoPz system, these findings

revealed that the adsorption process follows the pseudo

second-order model.

The Elovich model, which is a general application of

chemisorption kinetics, assumes that the active sites of

adsorbents are heterogeneous and have different activation

energies for adsorption of organic compounds. The Elovich

model is expressed by the equation [27]:

dqt=dt ¼ a exp �b qtð Þ ð5Þ

For the boundary conditions t = 0 to t and qt = 0, the

integrated form of the equation becomes:

qt ¼ ð1=bÞ ½ln ðabÞ þ ln t� ð6Þ

Equation 6 is simplified as follows:

qt ¼ aþ b ln t ð7Þ

In accordance with the Elovich model the slope and

intercept of the plot of qt versus ln t, shown in Fig. 2c,

were used to calculate the constants a and b (Table 1).

The value of b (mmol g-1 min-1) is the initial adsorption

rate constant and a (mmol g-1) is related to the extent of

surface coverage and the energy of activation for

chemisorption. The value of the correlation coefficient

determined for the plot was 0.79. Moreover, the

calculated adsorption capacity values obtained from this

kinetic model do not fit the experimental adsorption

capacity values. The error analysis values are also

higher than those obtained from the pseudo second-order

model.

Last, adsorption kinetic data were analyzed to determine

whether intraparticle diffusion is the rate limiting step in

the adsorption process. The intraparticle diffusion approach

can be described by the equation [28]:

qt ¼ kpt0:5 ð8Þ

where qt (mmol g-1) is the concentration of QCoPz

adsorbed at time t and kp (mmol g-1 min-0.5) is the

intraparticle rate constant. This model also suggests that

the adsorption process is considered to be controlled by the

internal diffusion with a minor effect of external diffusion.

The plot of the fraction of QCoPz uptake against the square

root of contact time (t0.5) is shown in Fig. 2d. This figure

reveals that the straight line obtained for bentonite does not

pass through the origin. It has been suggested that this

occurs when external diffusion is dominant and

intraparticle diffusion is not a rate-limiting step. In this

case, the equation used to describe the model is:

qt ¼ kpt0:5 þ C ð9Þ

where C is the intercept and values of C give an idea about

the thickness of boundary layer (Table 1), i.e. the larger the

intercept the greater is the boundary layer effect.

Figure 3 compares the experimental and predicted qe

values for bentonite. Apparently, the pseudo second-order

model best fits bentonite for the whole contact time period

with the highest correlation constant and lowest error

analysis values.

Equilibrium modeling

The shape of an isotherm may be used to predict if an

adsorption system is ‘‘favorable’’ or ‘‘unfavorable’’. The

isotherm shape can also provide qualitative information on

the nature of the solute–surface interaction. In addition,

adsorption isotherms are developed to evaluate the capacity

of a material for the adsorption of a particular dye molecule

[29].

The adsorption isotherm of the bentonite–QCoPz system

at different temperatures is shown in Fig. 4. The adsorption

Table 1 Experimental and predicted kinetics of adsorption of QCoPz by bentonite

Model Constants Bentonite clay Model Constants Bentonite clay

Pseudo first-order qe/mmol g-1 2.27E-02 Simple Elovich a 9.26E-01

k1/min-1 1.08E-01 b 8.20E-03

R2 9.94E-01 R2 7.95E-01

Ee 9.29E-01 Ee 6.10E-03

eave 9.30E-01 eave -5.50E-03

RMSE 8.65E-01 RMSE 1.00E-04

Pseudo second-order qe/mol g-1 9.53E-01 Interparticle diffusion kp 5.50E-03

k2 6.79E-02 C 9.27E-01

R2 9.74E-01 R2 5.71E-01

Ee 1.00E-03 Ee 2.10E-02

eave -5.00E-04 eave -2.00E-02

RMSE 2.67E-06 RMSE 1.60E-03

Preparation of novel heterogeneous catalysts 389

123

isotherms are marked by two distinct regions of different

slope. Each region corresponds to a specific mechanism, as

will be described later. It was shown that the equilibrium

uptake increased with the increasing initial dye concen-

tration in the range of experimental concentrations used.

Each isotherm rises sharply in the initial stages for low Ce

and qe values, thus indicating that there are plenty of

readily accessible sites and a great affinity of the adsorbent

for the dye molecules. The adsorbent is saturated when the

plateau is reached. The decrease in the slope of the iso-

therm leads to a monolayer where the less active sites are

available to the adsorbate. Maximum adsorption density of

QCoPz (Cmax) on bentonite and corresponding equilibrium

concentration of QCoPz at the plateau were calculated as

4.01 mmol g-1 and 0.987 mmol dm-3, respectively, for

25 �C, 3.94 mmol g-1 and 4.06 mmol dm-3 for 40 �C,

and 3.72 mmol g-1 and 4.28 mmol dm-3 for 60 �C. Fig-

ure 4 shows that the adsorption capacity decreased with

increasing temperature, thus indicating that the adsorption

of QCoPz on bentonite was an exothermic and physical

process.

Several adsorption equilibrium theories are available in

the literature, for example the Langmuir, Freundlich,

Temkin, Redlich–Peterson, and Fritz–Schlunder models,

which can be used to describe equilibrium studies. In this

study, experimental data were compared by using these five

well known and widely applied isotherm equations to find

the best-fitting model for the data obtained. The different

equation variables and the underlying thermodynamic

hypotheses of these models often provide insight into the

adsorption mechanism, the surface properties, and the

affinity of the adsorbent [30]. The variables and correlation

coefficients of these models for the adsorption of QCoPz

were determined from non-linear regression (except for the

Temkin model) by using Wolfram Mathematica commer-

cial software, as summarized in Table 2.

The Langmuir isotherm theory assumes monolayer

coverage of adsorbate over a homogenous adsorbent sur-

face [31]. KL (dm3 g-1) and aL (dm3 mmol-1) in Eq. 10

are the Langmuir isotherm constants; Ce and qe are the

liquid phase concentration and solid phase concentration of

the adsorbate at equilibrium.

qe ¼ KLCe= 1þ aLCeð Þ ð10Þ

The Langmuir constants KL and aL were evaluated by

linearization of Eq. 10:

Ce=qe ¼ 1=KL þ aL=KLCe ð11Þ

Hence by plotting Ce/qe against Ce it is possible to obtain

the value of KL from the intercept and the value of aL from the

slope. The theoretical monolayer capacity Qmax is numerically

equal to (KL/aL). Experimental data and data predicted by use

of the Langmuir model are compared in Fig. 5.

The Langmuir model can also be expressed by means of

a dimensionless constant RL, whose magnitude provides

information about whether the adsorption process is

spontaneous or non spontaneous. It can be calculated by

use of Eq. 12:

RL ¼ 1= 1þ aLCoð Þ ð12Þ

where Co is the initial dye concentration and aL is the

Langmuir constant. The value of RL indicates the adsorp-

tion process is irreversible when RL is 0, favorable when RL

is between 0 and 1, linear when RL is 1, and unfavorable

when RL is greater than 1 [32]. Values of RL calculated at

: 25 °C

0.92

0.94

0.96

0.98

1.00

1.02

1.04

1.06

1.08

1.10

0 100 200 300 400 500 600 700

q t/m

mol

g-1

t /min

ExperimentalSimple Elovichintraparticle diffusionPseudo-second order

: 25 °CT

Fig. 3 Comparison of experimental and estimated adsorption kinetics

of QCoPz

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0 1 2 3 4 5

adso

rpti

on d

ensi

ty/m

mol

g-1

Ce /mmol dm-3

25 ºC

40 ºC

60 ºC

Fig. 4 Adsorption isotherm for QCoPz by bentonite at three different

temperatures (conditions: contact time = 2 h; volume = 40 cm3;

clay mass = 40 mg; natural pH 7–8)

390 A. M. Sevim et al.

123

25, 40, and 60 �C were in range between 0 and 1 which

indicate that adsorption is favorable under the operation

conditions studied.

The Freundlich expression Eq. 13 is an exponential

equation and therefore assumes that as the adsorbate con-

centration increases so does the concentration of adsorbate

on the adsorbent surface. Theoretically, using this expres-

sion, an infinite amount of adsorption can occur [33].

qe ¼ KFC1=ne ð13Þ

In this equation KF and n are the Freundlich constants

related to adsorption capacity and adsorption intensity,

respectively. This expression is characterized by the

heterogeneity factor, 1/n, and so the Freundlich isotherm

may be used to describe heterogeneous systems.

The Freundlich constants are empirical constants which

depend on several environmental factors. The value of 1/n

ranges between 0 and 1, and indicates the degree of

non-linearity between solution concentration and adsorp-

tion as follows. If the value of 1/n is equal to unity, the

adsorption is linear; if the value is below unity, this implies

that the adsorption process is chemical; if the value is

above unity, adsorption is a favorable physical process;

the more heterogeneous the surface, the closer the value of

Table 2 Values of the variables obtained from the models, and the correlation coefficients

Model Variable Temperature

25 �C 40 �C 60 �C

Freundlich n 2.76E?00 1.65E?00 1.82E?00

KF/dm3 mmol-1 3.27E?00 1.78E?00 1.77E?00

R2 0.91 0.99 0.98

Ee 1.74E?00 5.57E-01 2.10E-01

eave -9.68E-02 -5.57E-01 -1.01E-02

RMSE 6.86E-01 5.91E-01 2.17E-01

Langmuir KL/dm3 g-1 2.42E?01 2.77E?00 2.87E?00

aL/dm3 mmol-1 5.06E?00 4.41E-01 5.22E?00

Qmax/mmol g-1 4.79E?00 6.27E?00 5.50E?00

R2 0.98 0.97 0.99

Ee 8.77E-01 8.65E-01 2.50E-01

eave -9.28E-02 -5.62E-01 4.43E-02

RMSE 2.98E-01 6.42E-01 1.61E-01

Temkin B1/J mmol-1 9.40E-01 8.13E-01 8.35E-01

KT/dm3 mmol-1 5.59E?01 1.53E?01 1.30E?01

R2 0.97 0.94 0.84

Ee 4.22E-01 7.74E-01 2.14E?01

eave 2.31E-05 5.01E-05 3.49E-01

RMSE 3.44E-01 4.55E-01 1.05E?00

Redlich–Peterson aRP/dm3 mmol-1 3.74E?00 1.04E-01 1.33E-01

B 1.16E?00 1.72E?00 1.62E?00

KRP/dm3 g-1 2.01E?01 2.10E?00 2.11E?00

R2 0.99 0.96 0.99

Ee 7.35E-01 7.26E-01 2.90E-01

eave -7.43E-02 -5.57E-01 7.28E-02

RMSE 2.45E-01 6.73E-01 1.50E-01

Fritz–Schlunder A 2.49E?02 1.84E-01 1.81E?00

a 1.84E?00 8.00E-01 7.46E-01

B 6.10E?01 2.13E-06 8.39E-07

b 1.80E?00 8.72E?00 9.05E?00

R2 0.99 0.97 0.99

Ee 2.42E-01 8.73E-01 5.60E-02

eave -1.11E-02 -6.80E-01 3.14E-02

RMSE 1.24E-01 7.83E-01 9.38E-02

Preparation of novel heterogeneous catalysts 391

123

1/n is to 0 [34]. This value is the adsorption density, and

because its value is between 1 and 10, the adsorption process

is considered to be satisfactory. If the value of n is less than 1,

the adsorption process is chemical in nature. The opposite

case indicates a physical adsorption process. In our study, the

values found for bentonite were around 1.65 and 2.75, which

proves that the adsorption is a spontaneous and physical

process. Experimental data and data predicted by the Lang-

muir model are compared in Fig. 5.

Redlich and Peterson [35] proposed an empirical equa-

tion incorporating three variables which may be used to

represent adsorption equilibria over a wide concentration

range and can be applied either in homogeneous or heter-

ogeneous systems because of its versatility. The Redlich–

Peterson equation is represented by Eq. 14:

qe ¼ KRPCe= 1þ aRPCbe

� �ð14Þ

where KRP (dm3 g-1) and aRP (dm3 mmol-1) are isotherm

constants, and b is an exponent which lies between 0 and 1.

This isotherm combines elements from both the Langmuir

and Freundlich equations, and the mechanism of adsorption

is a hybrid and does not follow ideal monolayer adsorption.

At high liquid-phase concentrations of the adsorbate

(which means the result of aRP 9 Cbe is bigger than 1),

Eq. 14 reduces to the Freundlich equation, i.e.:

qe ¼ KRP=aRPð ÞC1�be ð15Þ

where KRP/aRP and (1 - b) are the variables KF and 1/n,

respectively, of the Freundlich model. For b = 1, Eq. 14

reduces to the Langmuir isotherm. Experimental data and

data predicted by the Redlich–Peterson model are com-

pared in Fig. 5.

Temkin and Pyzhev [36] considered the effects of

indirect adsorbate–adsorbate interactions on adsorption

isotherms and suggested that the heat of adsorption of all

molecules in the layer would decrease linearly with cov-

erage. The Temkin isotherm equation is given by the

equation:

qe ¼ ln KTCeð Þ ð16Þ

Equation 16 can be expressed in its linear form as:

qe ¼ B1 ln KT þ B1 ln Ce ð17Þ

where

B1 ¼ RT=b ð18Þ

The adsorption data can be analyzed according to

Eq. 17. A plot of qe versus ln Ce enables determination of

the isotherm constants KT and B1 (Fig. 5). The constant B1

is related to the heat of adsorption (J mmol-1).

An important four-variable equation of Langmuir–

Freundlich type was developed empirically by Fritz and

Schlunder [37]. It is expressed by the equation:

qe ¼ ACae= 1þ BCb

e

� �b and a\1 ð19Þ

where qe is the amount adsorbed at equilibrium

(mmol g-1), Ce the equilibrium concentration of the

adsorbate (mmol dm-3), A and B are the Fritz–Schlunder

parameters, and a and b are the Fritz–Schlunder equation

exponents. At high liquid-phase concentrations of the

adsorbate, Eq. 19 reduces to the Freundlich equation:

qe ¼ ACa�be =B ð20Þ

where A/B and (a - b) are the variables KF and 1/n,

respectively, of the Freundlich model. For a = b = 1,

Eq. 19 reduces to the Langmuir equation, with B = aL and

A = KL.

When the curves obtained for bentonite are compared

(Fig. 5), experimental values are almost identical with the

data obtained by use of the Fritz–Schlunder adsorption

model. Table 2 indicates that the correlation constants

obtained for all temperatures and systems are equal to 0.99

and that error analyses yield the lowest values. For all

temperatures, the experimental data almost correlate to the

four-variable Fritz–Schlunder isotherm equation. This

good correlation is ascribed to the four-variable nature of

the model.

Thermodynamic evaluation of the adsorption process

The adsorption mechanism (i.e., chemical or physical) is

often an important indicator describing the type and level of

interactions between the adsorbate and the adsorbent. If

adsorption decreases with increasing temperature, it may be

indicative of a physical process, whereas the reverse is

0

1

2

3

4

5

6

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

q e/m

mol

g-1

Ce /mmol dm-3

Experimental

Freundlich

Langmuir

Temkin

Redlich-Peterson

Fritz-Schlunder

Fig. 5 Comparison of experimental and predicted adsorption iso-

therms of QCoPz on bentonite for all the models analyzed

392 A. M. Sevim et al.

123

generally true for chemisorption. In the bentonite–QCoPz

system, the decrease in adsorption with increasing temper-

ature and fast adsorption kinetics may suggest the presence

of physical adsorption. Nevertheless, this alone is not suffi-

cient to determine the type of adsorption. The type of

adsorption may be determined by thermodynamic quantities,

for example free energy of adsorption (DG�ads) and the heat of

adsorption (DH�ads) both of which can be obtained from the

adsorption data given in Fig. 4. A general isotherm for

adsorption at the solid–liquid interface, taking into account

the effect of size ratio (n) and lateral interaction coefficient

(a) between adsorbed molecules, has the form [38]:

h eð�2ahÞ= 1� hð Þn¼ KCe ð21Þ

where

K ¼ eð�DG�ads=RTÞ=55:5 ð22Þ

is the adsorbability of the adsorbate molecule at infinitively

low coverage, Ce is the equilibrium concentration in

mol dm-3, h is the degree of surface coverage of the

mineral with the collector molecule at (C/Cmax), R is the

gas constant (8.1314 J mol-1 K-1), T is the temperature in

K, and Cmax is the adsorption density at the plateau. The

free energy of adsorption can be calculated from Eq. 22 as

a function of h.

Calculation of DG�ads was performed using five models:

the Flory–Huggins [39], Frumkin [40], modified Frumkin

[41], Dubinin–Radushkevich, and Langmuir [42] equa-

tions. If the adsorption data obey these equations, the

variables, i.e., n, a, and K, are plugged into Eqs. 21 and 22

and the free energy of adsorption is calculated. The value

of n for the Flory–Huggins and modified Frumkin equa-

tions depend on the size of the adsorbate. The values of

pairs a and n pairs for the Frumkin, modified Frumkin,

Flory–Huggins, and Langmuir equations are 1,1; 2,1; 2,0;

and 1,0, respectively.

For instance, by rearranging Eq. 21 and taking the log-

arithms for n = 2 and a = 1, the modified Frumkin

equation is obtained:

ln h=Ce 1� hð Þ2h i

¼ 2ahþ ln K ð23Þ

The results from use of all equations are shown in

Fig. 6. The slopes and intercepts of the straight lines in

Fig. 6 have been used to determine the value of DG�ads with

y(25 °C) = 2.4917x + 7.5734

y(40 °C) = 0.7458x + 6.6732

y(60 °C) = 0.7555x + 6.7864

y(25 °C) = 2.33x + 21.691

y(40 °C) = 1.3685x + 10.323

y(60 °C) = 1.488x + 11.125

y(25 °C)=1.6572x+14.648

y(40 °C)=1.0549x+7.3338

y(60 °C)=1.0562x+7.489

y(25 °C) = 2,4917x + 7,5734

y(40 °C) = 0,7458x + 6,6732

y(60 °C) = 0,7555x + 6,78640

2

4

6

8

10

12

0.0 0.2 0.4 0.6 0.8 1.0

ln(

/(C(1

-)

25ºC

40ºC

60ºC

y(25 °C) = 5,0022x + 7,3862

y(40 °C) = 2,6822x + 6,5538

y(60 °C) = 2,9062x + 6,59210

2

4

6

8

10

12

14

0.0 0.2 0.4 0.6 0.8 1.0

ln(

/(C(1

-)2 )

25ºC

40ºC

60ºC

y(25 °C) = 1,6572x + 14,648

y(40 °C) = 1,0549x + 7,3338

y(60 °C) = 1,0562x + 7,489

-5

-4

-3

-2

-1

0

1

2

3

4

5

-12 -11 -10 -9 -8 -7 -6 -5 -4

ln(

/(1-

))

ln C

25ºC

40ºC

60ºC

y(25 °C) = 2,33x + 21,691

y(40 °C) = 1,3685x + 10,323

y(60 °C) = 1,4488x + 11,125-8

-6

-4

-2

0

2

4

6

8

-12 -10 -8 -6 -4

ln(

/(1-

)2 )

ln C

25ºC

40ºC

60ºC

(B)

(D)(C)

(A)Fig. 6 Illustration of the

Langmuir (a), Flory–Huggins

(b), Frumkin (c), and modified

Frumkin (d) equations based on

the data in Fig. 5

Preparation of novel heterogeneous catalysts 393

123

the results presented in Table 3. DG�ads determines the

affinity of the mineral surface for the adsorbate molecules

at very low coverage. The term a, on the other hand,

represents the strength of lateral interactive forces among

QCoPz molecules which are adsorbed on the surface. The

magnitude of a increases with the magnitude of coating,

and indicates the intensity of the interaction. Negative

values of the a coefficient mean there is an interaction

between adsorbate molecules and, generally, polar moieties

of the adsorbate are interacting with each other. Positive

values of a indicate repulsion [43, 44]. As seen in Table 3,

the calculated a value is low and positive for bentonite–

QCoPz. The low value indicates that the interaction in the

adsorptive layer is of weak lateral nature. The positive

value, on the other hand, shows that polar moieties of the

adsorbate apply a repulsive force to each other. The

Frumkin equation and its modified form take into account

both the size ratio and the lateral interaction coefficient and

it seem to result in better predictions than the Langmuir

and Flory–Huggins equations.

Another equation that has been used to determine the

possible adsorption mechanism is the Dubinin–Radushke-

vich equation, which assumes a constant sorption potential

[45]. The linear presentation of this equation is expressed

by:

ln qe ¼ ln Qm � Ke2 ð24Þe ¼ RT ln 1þ 1=Ceð Þ ð25Þ

where e is the Polanyi potential, Qm is the monolayer

capacity (mol g-1), Ce is the equilibrium concentration

(mol dm-3), and K is a constant related to the adsorption

energy (mol2 J-2). Qm and K can be obtained from the

intercept and slope of the plot of ln qe versus e2.

The mean free energy of sorption, E, is calculated by use

of the equation:

E ¼ 2Kð Þ�0:5 ð26Þ

The Dubinin–Radushkevich variables and mean free

energy are given in Table 3. The magnitude of E is useful

Table 3 Thermodynamic data for adsorption of QCoPz on bentonite

Model Variable Temperature

298 K 313 K 333 K

1,000/T 3.36E?00 3.20E?00 3.00E?00

Cmax/mol g-1 4.30E-03 3.90E-03 3.70E-03

Langmuir KL/dm3 mol-1 6.90E?03 1.05E?03 1.20E?03

ln KL 8.84E?00 6.95E?00 7.09E?00

DG�ads/kJ mol-1 -3.11E?01 -2.66E?01 -2.69E?01

DH�ads/kJ mol-1 -3.97E?01 -3.97E?01 -3.97E?01

TDS�ads/kJ mol-1 -8.58E?00 -1.32E?01 -1.28E?01

Flory–Huggins KFH/dm3 mol-1 1.10E?04 1.89E?03 2.16E?03

ln KFH 9.31E?00 7.54E?00 7.68E?00

DG�ads/kJ mol-1 -3.30E?01 -3.01E?01 -3.24E?01

DH�ads/kJ mol-1 -3.70E?01 -3.70E?01 -3.70E?01

TDS�ads/kJ mol-1 -4.01E?00 -6.95E?00 -4.65E?00

Frumkin a 1.25E?00 3.70E-01 3.80E-01

K/dm3 mol-1 1.95E?03 7.91E?02 8.86E?02

ln K 7.57E?00 6.67E?00 6.79E?00

DG�ads/kJ mol-1 -2.87E?01 -2.78E?01 -2.99E?01

DH�ads/kJ mol-1 -1.78E?01 -1.78E?01 -1.78E?01

TDS�ads/kJ mol-1 1.10E?01 1.00E?01 1.21E?01

Modified Frumkin a 2.50E?00 1.34E?00 1.45E?00

K/dm3 mol-1 1.61E?03 7.02E?02 7.29E?02

ln K 7.39E?00 6.55E?00 6.59E?00

DG�ads/kJ mol-1 -2.83E?01 -2.75E?01 -2.94E?01

DH�ads/kJ mol-1 -1.81E?01 -1.81E?01 -1.81E?01

TDS�ads/kJ mol-1 1.02E?01 9.44E?00 1.13E?01

Dubinin–Radushkevich Qm/mol g-1 3.69E-05 1.54E-02 5.10E-03

K/mol2 J-2 7.13E-09 6.63E-09 3.14E-09

E/kJ mol-1 8.37E?00 8.68E?00 1.26E?01

394 A. M. Sevim et al.

123

for estimating the type of sorption reaction; generally an

energy range from 1 to 16 kJ mol-1 indicates that the

adsorption process is physical and a range from 8 to

16 kJ mol-1 indicates an ion-exchange reaction [45].

Therefore, the E values obtained from the above equation

(8.4–12.6 kJ mol-1) are practically in the ion-exchange

energy range, supporting the idea that for adsorption of

QCoPz species on to bentonite the ion-exchange

mechanism is very important.

Another very important thermodynamic value in deter-

mining the type of adsorption is the heat of adsorption

(DH�ads). This can be obtained from Van’t Hoff’s equation

[46]:

dðln KÞ=d 1=Tð Þ ¼ �DHoads=R ð27Þ

The slopes of ln K versus 1/T are used in Eq. 27 to

calculate DH�ads for each model. The results are presented in

Table 3, with the entropy values calculated by use of

Eq. 28 given below:

DG ¼ DH � T DS ð28Þ

The adsorption process is composed of two contributions,

enthalpic and entropic, which determine whether or not the

reaction is spontaneous. The negative value of free energy of

adsorption calculated according to all models indicates that

adsorption of QCoPz ions on the bentonite surface and

between the layers is spontaneous. The negative value of the

enthalpy (Table 3) means that heat is released from

the adsorption process. Generally, an exothermic adsorption

process signifies either physisorption or chemisorption

whereas an endothermic process is attributable unequiv-

ocally to chemisorption. In an exothermic process,

physisorption is distinguished form chemisorption by

considering the absolute value of the adsorption enthalpy.

Typically, the enthalpy of a physisorption process is less than

41.86 kJ mol-1 whereas the enthalpy of a chemisorption

process approaches 100 kJ mol-1. In this work the absolute

values of enthalpy are relatively low, approaching those

typical of physisorption. This conclusion is also supported by

the kinetics of adsorption which were complete in less than 5

min with 99% yield. The positive value of DS�ads indicates

that there is an increase in the randomness of the solid–

solution interface of the system during the adsorption

process. In ion-exchange processes, especially, water

molecules released into the bulk solution from the solid

surface cause an increase of entropy [47].

Desorption studies

To investigate the stability of QCoPz adsorbed on to ben-

tonite, desorption studies were conducted in distilled water.

As mentioned in the previous section, Fig. 7a shows the

desorption behavior. Figure 7a indicates that the desorption

proceeds very quickly in the first step, then slows; after the

fifth step, the amount of QCoPz on the bentonite surface

decreases by ca. 13%.

Desorption shows that the adsorption process occurring

physically by electrostatic interaction and ion exchange is

reversible in nature. Positively charged polar ends of the

dye, which is adsorbed by the stern layer, interact with each

other and the interaction is mainly repulsion (the positive

value of the a constant, obtained from the Frumkin and

adapted Frumkin models, indicates repulsive behavior, as

mentioned previously). If there is a lack of negative ions to

compensate for the positive charges, the repulsive forces

among polar groups cause separation between QCoPz

molecules which thus return to the diffuse layer in the

solution.

As a result, it can be considered that QCoPz molecules

on the clay surface undergo desorption in distilled water. In

addition, the rapid nature of this desorptive process is an

indication of physical adsorption of this dye on bentonite

clay.

T: 25 °C

wavenumber/cm-1

(B)(A)

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0.0035

0.0040

0.0045

0 0.001 0.002 0.003 0.004

a

b

c

d

q e/m

mol

g-1

Ce /mmol dm-3

Adsorption isotherm

Desorption

:25oC

Fig. 7 Desorption plot for the

QCoPz–bentonite system at

25 �C (a) and infrared spectra of

bentonite in its original form or

treated with QCoPz solutions at

concentrations of 0.4, 3,

5 mmol dm-3 (b)

Preparation of novel heterogeneous catalysts 395

123

FT-IR spectral studies of QCoPz adsorbed

on to bentonite samples

FT-IR spectrometry is a useful tool for determining whe-

ther the adsorptive interaction between QCoPz and

bentonite is a physical or chemical process; FT-IR spectra

of dried composites can give an idea about this interaction.

Figure 7b shows the FTIR spectrum of bentonite. The

stretching vibrations of structural –OH groups were

observed at 3,627 cm-1, structural Si–O at 991 cm-1, and

stretching vibrations of structural AlAlOH, AlFeOH, or

AlMgOH groups generally between 915 and 880 cm-1.

Water in bentonite results in a broad stretching vibration

band at 3,417 cm-1. This peak generally has an overtone

shoulder peak at ca. 3,330 cm-1, because of the bending

vibration peak at 1,634 cm-1 [48].

FT-IR spectra of bentonite with different adsorbed

amounts of QCoPz are given in Fig. 7b. These spectra

show that low concentrations of the dye did not have an

effect; however, an initial concentration of 5 mmol dm-3

yielded signals attributable to the dye molecule at

2,921–2,850 and 1,473 cm-1. These signals prove the

adsorption of the dye molecule. In addition, this spectrum

indicates that reduced intensities of the bound water mol-

ecules at 3,417 and 3,627 cm-1 are because of exchange of

QCoPz with inorganic cations and bound water molecules.

Deformation of the band at 1,634 cm-1 (water of hydra-

tion) is evident after the adsorption process. Last, water

peaks are broadened, which indicates entrance of water

molecules to the structure. While QCoPz is being adsorbed

between layers, its adsorption is in aggregates or clusters,

thereby forming voids for water molecules [49]. Alterna-

tively, water molecules are coordinated at the axial

positions of the cobalt metal center during the adsorption

process.

IR spectra given in Fig. 7b clearly indicate that none of

the dye-containing clay samples furnished any peaks aris-

ing because of chemical (covalent) bonding. This is also an

indication of the physical adsorption behavior of the dye

molecule.

Mechanism of the adsorption process

The adsorption isotherms in Fig. 4 also contain two distinct

regions, each characterized by different adsorption rates

and mechanisms. In the first stage, adsorption takes place at

a lower rate and is governed by an ion-exchange process

and electrostatic interactions. However, the adsorption

continues to take place with increasing QCoPz concentra-

tion. In the second stage, the adsorption is mainly

characterized by strong p–p aggregates and electrostatic

interactions, but the ion-exchange process still continues to

take place at a lower rate as the release of sodium and

calcium ions (Fig. 8a, b) continues even in the plateau

region.

The free energy of adsorption (DG�ads) is composed of a

number of contributions. Among these, DG�elec;DG�solv;

DG�ion�exchange; and DG�agg are important [50]. Here DG�elec

is the electrostatic contribution to the total energy of the

interaction which occurs between quaternary ammonium

groups of the QCoPz and negatively charged bentonite

surfaces and layers, DG�agg represents the interaction

because of the association of porphyrazine molecules

(aggregation) at the interface, DG�solv is the contribution of

the solvation effects on the polar head of the adsorbate, and

DG�ion�exchange represents the exchange of similar ions (e.g.,

0.0

0.2

0.4

0.6

0.8

1.0

1.2

(A)

0.0

0.2

0.4

0.6

0.0 5.0 10.0 0.0 5.0 10.0

(B)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0.0 1.0 2.0 3.0 4.0

Total Adsorption

Ion Exchange

Ion Adsorption

0.2

0.3

0.4

0.5

0.0 2.0 4.0

adso

rpti

on/m

mol

g-1

Ce / mmol L-1

Ion Exchange

(C)so

dium

ion

conc

entr

atio

n/m

mol

dm

-3

cals

ium

ion

conc

entr

atio

n/m

mol

dm

-3

Ci /mmol dm-3 Ci /mmol dm-3Ce /mmol dm-3

Ce /mmol dm-3

adso

rpti

on/m

mol

g-1

Fig. 8 Released sodium (a) and calcium (b) ion concentrations as a function of initial QCoPz concentration under the adsorption conditions of

Fig. 5, and ion exchange, electrostatic (ion adsorption), and total adsorption isotherms of the QCoPz–bentonite system (c)

396 A. M. Sevim et al.

123

cations or anions) on the surface with those in the bulk and

ion exchange between QCoPz and ions which are between

bentonite layers and on the surface.

In our system, DG�ion�exchange; DG�elec; and DG�agg are

responsible for the adsorption reactions in the first and

second regions of the adsorption isotherm. Ion exchange

occurs mainly among Na? and Ca2? ions on the surface,

between layers and positively charged porphyrazine

ammonium groups. Formation of hydrogen bonds between

zeolite water and bound water has been proposed for a

number of cationic adsorbate systems, for example primary

amines. However, the quaternary ammonium groups,

which are surrounded by electron-donating CH3 substitu-

ents, and the absence of vacant electrons make hydrogen

bond formation impossible.

The significant increase in sodium and calcium ion

concentrations in Fig. 8a and b and the ion-exchange iso-

therm in Fig. 8c clearly shows that an ion-exchange

mechanism is responsible for the adsorption until nearly

0.372 mmol dm-3 equilibrium concentration. At this point,

calculations using Eq. 30 revealed the mechanism was

68% ion exchange and 32% electrostatic interaction [51].

However, electrostatic adsorption increased with increas-

ing equilibrium concentrations, it was, for example, 65%

for 0.468 mmol dm-3 and 86% for 0.171 mmol dm-3.

These results indicate that ion the exchange mechanism

proceeds very rapidly at the beginning of adsorption pro-

cess, and then electrostatic interactions become more

important than ion exchange through the plateau region

(Fig. 8c). The significant increase in the slope of the

adsorption isotherm in the second region is ascribed to the

aggregation p–p interactions among the adsorbed QCoPz

molecules.

%Ion exchange ¼ ZNa � Amount of sodium in solution

ZQCoPz � Amount of adsorbed QCoPz

� 100 ðZ ¼ Valence of ionÞ ð29Þ

The results in Fig. 8c are indicative of good agreement

with desorption behavior of QCoPz. Although low in

amount, fast desorption of QCoPz from the bentonite

surface into distilled water requires a substantial amount of

adsorbate molecules to be attached to clay surface by

electrostatic interactions (ion adsorption); calculations

performed using Eq. 29 and Fig. 8c indicate that

electrostatic interaction is a very important adsorption

mechanism.

It is known from the literature that negatively charged

smectite clays form very stable complexes with cationic

guests [52]. As a result, porphyrazine derivatives having

cationic moieties such as ammonium groups can easily be

intercalated into bentonite interlayers by means of elec-

trostatic interactions.

It is known that the properties of host materials and guest

molecules affect the adsorption structure of QCoPz in the

inter-layer space. The adsorption structure of the porphyrin

molecule, which is similar to our QCoPz, on the clay surface

or within the inter-layer space was investigated by X-ray

diffraction, electron spin resonance (ESR), dichroic absorp-

tion measurements, and dichroic absorption measurements

on a waveguide. According to X-ray diffraction (XRD)

measurements, tetrapyrrole derivatives are adsorbed on the

clay surface with orientation parallel to the clay surface [52].

Generally, the angle of orientation of the porphyrin relative to

the clay surface tends to increase as the charge density of the

clay increases [53]. Another clay mineral, saponite, which

has almost the same CEC value (0.997 meq g-1) as our

bentonite sample (0.987 meq g-1), was found to host tetra-

pyrrole molecules in an orientation parallel to the surface.

ESR experiments indicate that the orientation can also change

in response to atmospheric changes, for example humidity.

The structure of the porphyrazine affects the adsorption

structure of the porphyrazine in the clay complex. Meso-

tetrakis(5-trimethylammoniopentyl) porphyrin, which has a

flexible alkyl chain in the cationic portion, was used as

the cationic adsorbate. It seems that porphyrin adsorbs on

the clay with a non-parallel (tilted) orientation relative to

the surface [52].

The mean interlayer space value of bentonite used in

this study is 12.7 A; however, because the average length

of our QCoPz molecule is 18 A, according the literature, it

is concluded that our molecules might be adsorbed with an

angular or parallel orientation rather than being perpen-

dicular to the interlayers. This is depicted in Fig. 9.

Experimental

Materials

The Na-bentonite sample used as an adsorbent in this study

was obtained from the Resadiye–Tokat region (middle–

northeastern Anatolia) of Turkey. The Na-bentonite was

upgraded by the multi-stage hydrocyclone method, and the

purified sample was subjected to chemical and mineralogi-

cal analysis; the results are shown in Tables 4 and 5. It is

known that the adsorption capacity for reactive dyes

increases with decreasing particle size. The sample was

therefore ground to less than 150 lm to produce an average

particle size (d90) of 5.6 lm for adsorption experiments. The

purified bentonite sample, mainly Na-montmorillonite, has

a cation exchange capacity (CEC) of 0.98 meq g-1 [54].

The detailed procedure for synthesizing octakis(2-

dimethylaminoethylsulfanyl)porphyrazinatocobalt (CoPz)

has been described elsewhere [55, 56]. Cobalt porphyrazine

Preparation of novel heterogeneous catalysts 397

123

was converted into the corresponding quaternized product

(QCoPz) by reaction with iodomethane in dichlorometh-

ane. CoPz (120 mg, 0.1 mmol) was dissolved in 10 cm3

dichloromethane and 1.6 cm3 iodomethane was added.

This mixture was stirred in the dark under nitrogen atmo-

sphere at room temperature for two days. The dark blue

product was precipitated by centrifugation and filtered,

then washed with dichloromethane (2–3 times) and

diethyl ether, and finally dried under vacuum for 12 h. The

yield of the water-soluble product (C56H104CoI8N16S8,

2,332 g mol-1) was 105 mg (45%). IR:�v ¼ 2; 960;

2; 930; 1; 620; 1; 480; 1; 330; 1; 215; 1; 150; 1; 010; 980; 910;

800; 750 cm�1: The UV–visible spectrum shows the main

absorbance bands in water (kmax) at 356 and 638 nm. The

chemical structure of the adsorbate is given in Fig. 10.

Distilled and deionized water with a conductivity of

2 9 10-6 mho cm-1 was used in all experiments. Exper-

iments were conducted at 25 �C unless otherwise specified.

Fig. 9 Arrangement of QCoPz

in the interlayer space of

bentonite: (a) tilted and

(b) parallel orientation

(T tetrahedral layer,

O octahedral layer,

R trimethylammonium-

ethylsulfanyl)

Table 4 Results from chemical analysis of the bentonite sample used in this study

Component Na2O MgO Al2O3 SiO2 P2O5 K2O CaO MnO TiO2 Fe2O3 LOI

% by weight 2.50 2.30 17.30 59.30 0.10 0.30 0.80 0.02 0.30 3.60 13.50

Table 5 Results from mineralogical analysis of the bentonite sample

used in this study

Mineral name Formula

Montmorillonite Na0.3(Al, Mg)2Si4O10(OH)2�4H2O

Orthoclase KAlSi3O8

Biotite KMg2.5Fe0.52?AlSi3O10(OH)1.75F0.25

Opal-CT SiO2 1.5(H2O)

Quartz SiO2

Calcite CaCO3

Amorphous –

398 A. M. Sevim et al.

123

Methods

The electrokinetic properties of bentonite were determined

by use of a Zeta Meter 3.0? equipped with a micropro-

cessor unit. The samples were conditioned under

adsorptive testing conditions. A sample of bentonite

(50 mg in 50 cm3 solution) was shaken for 2 h followed by

centrifugation for 15 min. The suspension was kept still for

5 min to let larger particles settle. Approximately 25 cm3

clear supernatant was removed from the adsorption test vial

and introduced into the electrophoretic cell. An appropriate

amount of bentonite particles was pipetted out of the

bentonite bed and placed in the cell. Each data point is the

average from approximately ten measurements.

Adsorption tests were conducted in 10, 20, or 40 cm3

glass vials. A bentonite sample (10 mg) was mixed in

10 cm3 or its multiples with a solid-to-liquid ratio of 0.1%.

The vials were shaken for 2 h in an Edmund Buhler KL2

shaker and centrifuged for 15 min. The equilibrium con-

centrations of dyes were determined at 638 nm using a

WTW PhotoLab visible spectrophotometer. The adsorption

density was calculated by use of the formula given below

[57, 58]:

C ¼ Ci � Crð ÞV=m ð30Þ

where Ci and Cr represent the initial and residual concen-

trations in mol dm-3, m the mass of solid in grams, V the

volume of the solution in dm3, and C the adsorption density

in mol g-1.

The organobentonite formed by adsorption of QCoPz on

to bentonite was used in desorption studies to see whether

there is desorption in aqueous media and its amount. The

following procedure describes the experimental treatment:

20 mg bentonite was placed in a 20 cm3 flask, and 20 cm3

4 mmol dm-3 QCoPz solution was added; the flask was

then shaken at 420 rpm for 2 h. The resulting suspension

was centrifuged at 3,500 rpm for 15 min, and a 4 cm3

(20%) aliquot was taken for analyses. The supernatant was

compensated with the same amount of distilled water, and

the flask was shaken again. After centrifugation for 15 min,

4 cm3 was taken again, and the amount of QCoPz in the

solution was calculated. This in turn was used to obtain the

amount of QCoPz adsorbed on the clay. Distilled water

(4 cm3) was added to the flask and the above procedure

was repeated until the residual concentration was practi-

cally negligible.

The infrared spectra of natural bentonite and bentonite

with adsorbed QCoPz were analyzed using a Perkin–Elmer

Spectrum One Fourier transform infrared (FTIR) spec-

trometer to identify the functional groups responsible for

the adsorption. The change of Na? and Ca2? ion concen-

trations in solution before and after the adsorption process

were determined by use of a Varian AA280 FS atomic

absorption spectrophotometer.

Acknowledgments This work was supported by the Research Fund

of the Technical University of Istanbul.

References

1. Stuzhin PA, Ercolani C (2003) In: Kadish KM, Smith KM,

Guilard R (eds) The porphyrin handbook, vol 15. Academic

Press, New York

2. Leznoff CC, Lever ABP (1989–1996) In: Phthalocyanines:

properties and applications, vols 1–4. VCH, Weinheim

3. Arslanoglu Y, Sevim AM, Hamuryudan E, Gul A (2006) Dyes

Pigment 68:129

4. McKeown NB (1998) Phthalocyanine materials: synthesis struc-

ture and function. Cambridge University Press, Cambridge

5. Ogawa M, Kuroda K (1995) Chem Rev 95:399

6. Takagi K, Shichi T (2000) In: Ramamurthy V, Schanze KS (eds)

Solid state and surface photochemistry, vol 5. Marcel Dekker,

New York

7. Drain CM, Hupp JT, Suslick KS, Wasielewski MR, Chen X

(2002) J Porphyr Phthalocyanines 6:243

8. Lindsay-Smith JR (1994) In: Sheldon RA (ed) Metalloporphyrins

in catalytic oxidations, chap 11. Marcel Dekker, New York

9. Wohrle D, Suvorova O, Gerdes R, Bartels O, Lapok L, Baziakina

N, Makarov S, Slodek A (2004) J Porphyr Phthalocyanines

8:1020

10. Barloy L, Battioni P, Mansuy D (1990) J Chem Soc Chem

Commun 19:1365

11. Bedioui F (1995) Coord Chem Rev 144:39

12. Sakoda K, Kominami K (1993) Chem Phys Lett 216:270

13. Sasai R, Sugiyama D, Takahashi S, Tong Z, Shichi T, Itoh H,

Takagi K (2003) J Photochem Photobiol A 155:223

14. Takagi S, Shimada T, Yui T, Inoue H (2001) Chem Lett 2:128

15. Takagi S, Shimada T, Eguchi M, Yui T, Yoshida H, Tryk DA,

Inoue H (2002) Langmuir 18:2265

N

N

N

N

N

N

NN

Co

S

N

CH3

CH3

SN

CH3

CH3

S

N

CH3

CH3

S

NCH3

CH3

S

N

CH3

CH3

S N

CH3

CH3

S

NCH3H3C

S

N

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

I

I

I

I

II

I

I

Fig. 10 Structure of octakis(2-trimethylammoniumethylsulfanyl)por-

phyrazinatocobalt octaiodide (QCoPz)

Preparation of novel heterogeneous catalysts 399

123

16. Van Damme H, Crespin M, Obrecht F, Cruz MI, Fripiat JJ (1978)

J Colloid Interface Sci 66:43

17. Elvers B, Hawkins S, Ravenscroft M, Schulz D (1989) Ullmann’s

encyclopedia of industrial chemistry, vol A13, 5th edn. VCH-

Verlag, Weinheim

18. Dannenberg S, Kroder M, Dilling W, Cypionka H (1992) Arch

Microbiol 158:93

19. Chanda M, Grinshpun A, O’Driscoll KF, Rempel GL (1984) J

Mol Catal 26:267

20. Gottschalk G, Knackmuss HJ (1993) Angew Chem Int Ed

32:1398

21. Meunier B, Sorokin A (1997) Acc Chem Res 30:470

22. Raghavacharya C (1997) Chem Eng World 32:53

23. Wakamatsu T, Fuerstenau DW (1973) AIME Trans 254:123

24. Erdinc A (2009) MSc Thesis, Bogazici University, Istanbul,

Turkey

25. Lagergren S (1898) Kungliga Svenska Vetenskapsakademiens,

Handlingar 24:1

26. Ho YS (2010) J Radioanal Nucl Chem 285:403

27. Gerente C, Lee VKC, Le Cloirec P, Mckay G (2007) Crit Rev

Environ Sci Technol 37:41

28. Dogan M, Alkan M, Turkyılmaz A, Ozdemir Y (2004) J Hazard

Mater B109:141

29. Gimbert F, Morin-Crini N, Renault F, Badot PM, Crini G (2008)

J Hazard Mater 157:34

30. Ozdemir AB, Turan M, Celik MS (2003) J Environ Eng 129:709

31. Langmuir I (1918) J Am Chem Soc 40:1361

32. Leechart P, Nakbanpote W, Thiravetyan P (2009) J Environ

Manag 90:912

33. Freundlich H (1906) Z Phys Chem 57:385

34. Al-Duri B (1996) In: McKay G (ed) Use of adsorbents for the

removal of pollutants from wastewaters, chap 7. CRC Press, New

York

35. Redlich O, Peterson DL (1959) J Phys Chem 63:1024

36. Temkin MJ, Pyzhev V (1940) Acta Physicochim 12:217

37. Fritz W, Schlunder EU (1974) Chem Eng Sci 29:1279

38. Sabah E, Saglam H, Kara M, Celik MS (1997) In: Proceedings of

fifth southern hemisphere meeting on mineral technology,

Argentina, p 277

39. Dhar HP, Conway BE, Joshi KM (1973) Electrochim Acta

18:789

40. Grchev T, Cvetkovska M, Stafilov T, Schultze JW (1991) Elec-

trochim Acta 36:1315

41. Frumkin AN, Damaskin BB (1964) In: Bockris JOM, Conway BE

(eds) Modern aspects of electrochemistry, vol 3. Butterworths,

London

42. Quan C, Khoe G, Bagster D (2001) Water Res 35:478

43. Sabah E (1998) PhD Thesis, Osmangazi University, Eskisehir,

Turkey

44. Kara M, Yuzer H, Sabah E, Celik MS (2003) Water Res 37:224

45. Krishna BS, Murty DSR, Prakash BSJ (2000) J Colloid Interface

Sci 229:230

46. Celik MS, Yoon RH (1991) Langmuir 7:79

47. Xin H, Nai-yun G, Qiao-li Z (2007) J Environ Sci 19:1287

48. Eren E, Afsin B (2008) J Hazard Mater 151:682

49. Hermosin MC, Martin P, Cornejo J (1993) Environ Sci Technol

27:2606

50. Celik MS, Somasundaran P (1988) J Colloid Interface Sci

122:163

51. Gulgonul I (2004) PhD Thesis, Istanbul Technical University,

Istanbul, Turkey

52. Takagi S, Eguchi M, Tryk DA, Inoue H (2006) J Photochem

Photobiol C 7:104

53. Ukrainczyk L, Chibwe M, Pinnavaia TJ, Boyd SA (1995) Envi-

ron Sci Technol 29:439

54. Erkan I (2008) MSc thesis, Karadeniz Technical University,

Trabzon, Turkey

55. Polat M, Gul A (2000) Dyes Pigment 45:195

56. Kavran G, Akkus H, Gul A, Erim FB (2002) J Sep Sci 25:514

57. Sabah E, Celik MS (2002) Sep Sci Technol 37:3081

58. Wang S, Li H, Xu L (2006) J Colloid Interface Sci 295:71

400 A. M. Sevim et al.

123