Journal of Colloid and Interface Science 222, 12–19 (2000)
doi:10.1006/jcis.1999.6595, available online at http://www.idealibrary.com on
Adsorption and Interactions of Methyl Green with Montmorillonite
and Sepiolite
G. Rytwo∗, †,1 S. Nir,∗ M. Crespin,‡ and L. Margulies∗,2
∗ The Seagram Center for Soil and Water Sciences, The Faculty of Agriculture, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem,
Rehovot 76100, Israel; † Tel Hai Academic College, Upper Galilee 12210, Israel; and ‡ Centre de Recherche sur la Matiere Divisée,
CNRS and Université d’Orléans, 45071 Orléans Cedex 2, France
Received December 31, 1998; accepted October 19, 1999
The divalent organic cation, methyl green (MG), undergoes a
slow transformation (6 h) to a monovalent cation, carbinol
(MGOH+ ) upon dilution of its solution (10 mM), or in a buffer at
neutral pH. Adsorption isotherms of MG on montmorillonite were
determined by two procedures, both of which yield a final pH of
suspensions between 7 to 7.4. When the amounts of MG in suspen-
sion were lower than the cation-exchange capacity (CEC) of the
clay (0.8 molc /kg clay), no measurable amount of MG remained in
solution. The maximal amounts of MGOH+ adsorbed were larger
than those of MG2+ , being 1.15 and 0.75 mol MG/kg clay, respec-
tively, corresponding to 140% of the CEC in the first case. On a
charge basis the adsorption of added MG2+ amounts to 185% of the
CEC, which raises the possibility that a certain fraction of MG2+
transformed into the monovalent form during the incubation pe-
riod, since other divalent organic cations previously studied only
adsorbed up to the CEC (paraquat), or slightly above it (diquat).
Adsorption of MG on sepiolite (CEC = 0.15 molc /kg) further em-
phasizes the two patterns of its adsorption. The maximal adsorbed
amounts of MG2+ and MGOH+ were 0.09 and 0.30 mol/kg clay,
respectively. X-ray diffraction measurements gave lower values for
the basal spacings for montmorillonite-MG+ than for MGOH+ , sug-
gesting that MG2+ binds two clay platelets together, as in the case
of other divalent cations. A competition for adsorption between
MG and the monovalent organic cation, acriflavin (AF), gave lower
adsorbed amounts of AF when competing with MG+ , which is in-
terpreted to be due to the smaller basal spacing in this case, which
partially inhibits the entry of AF molecules into the interlamme-
lar space. Spectra of montmorillonite-MG particles in the visible
range exhibited significant differences between clay-MG and clay-
carbinol. °C 2000 Academic Press
Key Words: adsorption; clay minerals; montmorillonite; sepiolite;
methyl green; organic cations.
1 To whom correspondence should be addressed at Seagram Center for Soil
and Water Sciences, Faculty of Agriculture, Food and Environmental Quality
Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100,
Israel. Fax: +972-6-6900985. E-mail: rytwo@telhai.ac.il.
2 Deceased.
0021-9797/00 $35.00
Copyright ° C 2000 by Academic Press
All rights of reproduction in any form reserved.
INTRODUCTION
Methyl green (MG)3 is a basic triphenylmethane-type di-
cationic dye. It is used for staining solutions in medicine and bi-
ology (1). As it absorbs light in various parts of the UV-visible
spectrum, it has been used to deactivate excited molecules of
pesticides (“donor”) in order to stabilize them against photo-
decomposition. The feasibility of this method was first demon-
strated in the photostabilization of bioresmethrin (BR), a potent
and safe pyrethroid insecticide (2, 3). It was shown that BR
alone lost its insecticidal activity on Tribolium castaneum bee-
tles during one day of exposure to sunlight, but BR adsorbed on
montmorillonite together with MG at loads of 0.2 mol kg−1 clay
each, was still considerably active after 3 days of irradiation. In-
creasing the amount of MG adsorbed did not enhance the effect,
leading to the conclusion that the mechanism of the photosta-
bilization was an energy transfer rather than a screening effect.
It was also shown that the stabilization is not a result of light
scattering, since mixing thoroughly MG and BR with the clay
without adsorbing them did not cause a considerable decrease
in the photolability of BR. Strong interactions between the or-
ganic molecules of MG and BR and between them and clay were
shown by FTIR spectroscopy (3) and 13 C solid state MAS-NMR
spectroscopy (4), indicating that short distance intermolecular
interactions exist at the clay mineral surface.
Adsorption of the divalent cation methyl green to monoionic
Na-montmorillonite was studied by Margulies and Rozen (5),
who suggested that the adsorption of MG takes place through a
cation-exchange mechanism. At low MG loads each adsorbed
molecule replaced two Na ions, while at higher loads the ex-
change followed a 1 : 1 ratio. Adsorption of nonionic organic
3 Abbreviations used: AF, 3,6-diamino-10-methylacridinium chloride (acri-
flavin); BR, bioresmethrin; CEC, cation-exchange capacity; CV, N -[4-[bis[4-
(dimethylamino) - phenyl]methylene] - 2, 5-cyclohexadien-1-ylidine]-N -methyl
methanaminium chloride (crystal violet); DQ, 1,10 -ethylene-2,20 -dipyridylium
dibromide (diquat); MB, 3,7-bis(dimethylamino)phenazathionium chloride
(methylene blue); MG, 4-[[4-(dimethylamino) phenyl][4-(dimethylimino)-
2, 5-cyclohexadien - 1 - ylidine]methyl]-N -ethyl-N ,N -dimethylbenzenaminium
dichloride (methyl green); PQ, 1,10 -dimethyl-4,40 -bipyridinium (paraquat);
SSA, surface-specific area; UV, ultraviolet; XRD, X-ray diffraction.
12
 MG-CLAY INTERACTIONS
compounds, such as benzyl benzoate and benzophenone, on
MG-montmorillonite particles was also studied, leading to the
conclusion that strong interaction exists between the organic
compound and the organo-clay at low and medium loads, be-
coming weak when the loads of MG are large. They concluded
that the fraction of the clay surface occupied by MG molecules
adsorbed and their spatial organization affect the adsorption of
added nonionic organic molecules. This fact was considered
an important factor in determining the degree of photostabi-
lization of pesticides by clay-MG. In that work the amount of
MG adsorbed was reported to reach almost twice the cation-
exchange capacity (CEC), unlike results with other divalent or-
ganic cations, such as diquat (DQ) or paraquat (PQ), where the
adsorbed amounts approximately reached the CEC.
In the current work we address the issue of MG adsorption
to montmorillonite and sepiolite by elucidating the existence of
pH-dependent transformations of MG between mono- and di-
valent states, which may result in two patterns of adsorption. Our
studies include adsorption experiments, competitive-adsorption
experiments between MG and the monovalent organic cation
acriflavin (AF), UV-visible measurements, and X-ray diffraction
(XRD) measurements of basal spacing of MG- montmorillonite
particles.
As in previous studies the determination of the adsorbed
amounts of MG was deduced from the absorption of its solu-
tions. We observed that the molar extinction coefficient of MG
changed with time after dilutions. This required the development
of a kinetic analysis of that process.
We will show that a consideration of the pH-dependent trans-
formation of methyl green from a divalent to a monovalent cation
can consistently explain peculiarities of its absorption spec-
tra as well as its pattern of adsorption to montmorillonite and
sepiolite.
MATERIALS AND METHODS
Materials
The clay mineral used was Wyoming Na-montmorillonite
SWy-1, purchased from the Source Clays Repository of The
Clay Minerals Society (Columbia, MO). The cation-exchange
capacity of this clay was reported by van Olphen and Fripiat
(6) as 0.764 molc kg−1 . The specific surface area (SSA),
7.56 × 105 m2 kg−1 , was measured as described by Carter
et al. (7). The initial exchangeable cation suite was as in Rytwo
et al. (8) for “SWy-1 crude” (i.e., 0.41 molc Ca kg−1 , 0.14 molc
Mg kg−1 , and 0.26 molc Na kg−1 ).
Yunclillos sepiolite (<200 mesh) was provided by TOLSA
S.A. (Madrid, Spain) with a 99% content of pure mineral. The
chemical composition of this clay is described by Ruiz-Hitzky
and Casal (9). The surface area (N2 , BET) is 3.40 × 105 m2 kg−1
and the CEC is near 0.15 molc kg−1 . The clay was essentially a
MG2+ homoionic.
Methyl green and acriflavin were purchased as di- and
monochloride salts from Fluka Chemica (Fluka Chemie AG,
Buchs, Switzerland). An analysis report presented by the man-
13
FIG. 1. Molecular structures of methyl green (MG), carbinol, and acriflavin
(AF).
ufacturer indicates that MG contained large amounts of ZnCl2
(0.67 mol/mol MG), which was considered in calculating MG
concentrations in solutions. Since adsorption affinity of mono-
valent and divalent organic dyes is several orders of magnitude
higher than that of inorganic cations (8), we ignored the influ-
ence of Zn on the adsorption of MG species. Analytical NaOH
was obtained from Merck (Darmstadt, Germany). The molecu-
lar structures of the organic cations used are shown in Fig. 1. All
materials were used without further treatment or purification.
Kinetic Analysis of MG Spectra
Light absorption by MG is severely influenced by the range of
pH used during the experiments. Similar effects were obtained
while diluting a concentrated MG solution. To measure the influ-
ence of dilution on the absorption of MG solutions as a function
of time, a 10 mM stock solution of MG was prepared. The final
pH of this solution was pHMG 3.92.
At t = 0 two new solutions were prepared: Solution A (sol.A)
was made by diluting 0.5 ml of the stock MG solution to a
total volume of 100 ml, by adding deionized water. The pH was
pHA 5.68. Solution B (sol.B) was prepared in the same way: 0.5
ml of stock MG solution was diluted to a total of 100 ml, but in
this case a buffer solution for pH 7.0 (B.D.H. Chemicals Ltd.,
Poole, England) was used, giving a final acidity pHB 7.04. The
concentration of MG in both solutions for the kinetic analysis
was 5 ×10−5 M.
Spectra of the solutions were monitored at various times, using
an HP 8452A diode array UV-Vis spectrophotometer (Hewlett-
Packard Company, Palo Alto, CA). After 170 min, 50 ml of
sol.B was used to prepare solution C (sol.C), in which some
drops of highly concentrated HCl were added to lower the pH to
pHC 5.79. Since the increase in volume was insignificant (less
than 100 µl), the concentration of MG remained almost un-
changed and was close to 5 ×10−5 M. Changes of the spectrum
of sol.C as a function of time were measured at various intervals.
Regression calculations were made on the absorption of some
of the peaks of the spectra, as described under Results and
Discussion.
14 RYTWO ET AL.
Adsorption Isotherms
Adsorption isotherms of MG on montmorillonite were made
under two different initial conditions: (i) the isotherm was mea-
sured with MG solution “as is”, with an initial pH of 4.8 (“acidic
MG”, MG2+ ), and (ii) the second isotherm was measured with
MG solution in which the pH was adjusted to 7.4 with a few
drops of concentrated NaOH (“neutral MG”, carbinol+ ). Both
solutions were left overnight at 25◦ C in order to reach equi-
librium. Aliquots of aqueous 1 ×10−2 M MG and carbinol+
solutions were added dropwise under continuous stirring to
5 mL of a 0.005 kg montmorillonite kg−1 suspension in
20-mL polyethylene bottles. The final volume was brought to
15 mL with deionized water. The bottles were sealed and kept
at 25◦ C under continuous agitation. The final pH measured was
between 7.0 and 7.4 for montmorillonite samples and 6.8 and
7.3 for sepiolite samples, in both types of isotherms. Narine and
Guy (10) have reported that the adsorption of various organic
cations was essentially independent of pH in the range of 4.5
to 8.5. After 1, 3, and 7 days, 3 mL of suspension was taken
from each bottle and filtered through S&S FP030/2 disposable
filters (Schleicher & Schuell, Dassel, Germany), with 0.45-µm-
pore-diameter cellulose-acetate membranes. The concentration
of MG in each of the filtrates was determined by measuring the
absorption at 248 nm using an HP 8452A diode array UV-Vis
spectrophotometer. To avoid analytical problems observed in the
measurements of MG concentrations, the samples were diluted
with a buffer solution (pH 7.0) and left to stabilize for 24 h at
room temperature, under darkness. Under these conditions the
molar extinction coefficient of MG (which corresponds to the
carbinol+ molecules) was determined as ε = 13600 M−1 cm−1 .
The adsorbed amounts of MG2+ and carbinol+ on montmo-
rillonite had been confirmed for several points of the isotherm
using CHNSO measurements, with the same results obtained
with deviations of less than 5%. All experiments were carried
out in triplicate.
Competitive Adsorption
Three cases of competitive adsorption between the mono-
valent dye AF and the two species of MG were studied: (a)
equal amounts of charges of AF and MG2+ ; (b) equal amounts
of charges of AF and the presumed carbinol+ ; and (c) AF com-
peting with carbinol+ , at a 2 : 1 molar ratio of AF : carbinol+ ,
that would correspond to equal amounts of charges if carbinol+
molecules were divalent.
For these experiments separate aqueous 2 ×10−2 molc L−1
solutions of MG2+ , carbinol+ , and AF (Fig. 1) were prepared.
Solutions containing the relevant volumes of one of the MG
species and AF were prepared to yield a final concentration of
0.01 molc L−1 of each organic cation for experiments a and
b, and 0.01 molc L−1 AF with 0.005 molc L−1 carbinol+ for
experiment c. Aliquots of these solutions were added to mont-
morillonite suspensions as described above. Filtration and spec-
troscopic measurements were performed at various times.
The spectra were mathematically decomposed into individ-
ual components, as in Rytwo et al. (11). The spectrum of each
cation was unaffected by the presence of the other, indicating
the absence of significant interaction between the two organic
cations in solution.
Concentrations of AF were determined by measuring the ab-
sorption at 450 nm (ε = 37200 M−1 cm−1 ), using the HP 8452A
spectrophotometer. Limit of detection for AF was 2.7 ×10−7 M,
corresponding to an optical density of 0.01 in 10-mm cuvettes.
UV-Visible Absorption Spectra
Parts of the samples prepared for the adsorption isotherm were
diluted as needed, and the spectra of the unfiltered suspensions
of dye-montmorillonite complexes were measured. Since the
main influence of the clay in the range between 450 and 700 nm
is manifested by a diagonal baseline due to particle scattering,
this can be corrected as described by Russell (12).
X-Ray Diffraction
In samples prepared for the adsorption isotherm the solids
were separated from the filtrates. The filters were washed back-
ward with 2 ml of water. The organo-clay was diluted with 10 ml
of water and sonicated. The measurement of the basal spac-
ing of MG2+ - and carbinol+ -montmorillonite complexes was
performed on samples prepared from these solutions, by slow
sedimentation and evaporation on glass plates. Parts of these
solutions were filtered again through identical filters as in the
adsorption measurements, and the MG concentration in these fil-
trates was determined by measuring the absorption as described
above. The organic-cation concentration of most filtrates was
below the limit of detection, indicating almost complete ad-
sorption. The XRD measurements employed a Siemens D-500
diffractometer set up by reflection, using the λ Cu which was
filtrated by a Ni foil.
RESULTS AND DISCUSSION
Kinetic Analysis of Spectra
As described under Materials and Methods, we prepared two
MG solutions, with the same concentration (0.05 mM) but with
different pH: Solution A with pHA 5.68 and Solution B with
pH 7.04. Spectra were measured after various times. A few spec-
tra are presented in Figs. 2a (sol.A) and 2b (sol.B).
After 170 min, 50 ml of sol.B was used to prepare sol.C,
in which a few drops of very concentrated HCl were added to
change the pH to pHC 5.84. Changes of the spectrum of sol.C
as a function of time after addition of HCl are shown in Fig. 2c.
The optical densities of Solutions A and B decreased with
time at 314, 422, and 632 nm and increased at 248 nm. The
effect was more pronounced in Solution B. The time needed to
reach equilibrium was about 6 h. The present results show that
the spectral changes were completely reversible upon increasing
the acidity of the solutions, as in Fig. 2c (Solution C).
 MG-CLAY INTERACTIONS 15

We based our analysis on Prof. Zollinger’s suggestion that

the optical densities at 632, 422, and 314 nm are exclusively
contributed by the divalent MG2+ . Thus, regression calculations
can be made on the absorption due to these peaks, with

OD(t) = B1e−B2t + B3, [2]

where OD(t) represents the optical density at time t, and B1,

B2, and B3 are parameters found by regression. The parameter
B3 represents the OD at infinite time, B2 determines the rate
of the reaction. Since at t = 0, OD(0) = B1 + B3, it follows that
B1 = OD0 − OD∞ , i.e.,

OD(t) = (OD0 − OD∞ )e−kt + OD∞ . [3]

An analysis of OD(t) results for Solutions B and C yielded

FIG. 2. Changes in the spectrum of three MG solutions with time. Expla-
nations in text.
Prof. H. Zollinger (ETH-Zentrum, Zurich, personal commu-
nication, 1994; (13)) suggested that the effects can be explained
by a relatively slow addition of OH− to the MG cation, forming
a carbinol compound that should behave as a monovalent cation
(Fig. 1), the reaction being
MG2+ + OH− ↔ carbinol+ . [1]
very good simulations and is summarized in Table 1. It fol-
lows that k values are very similar for all the peaks that de-
crease in sol.B and increase in sol.C (at 314, 422, and 632 nm),
kB = 0.0138 min−1 ± 1.7%, kC = 0.0273 min−1 ± 2.2%, even
though they are calculated in a completely unrelated way. The
k value for the peak at 248 nm differs from the others in both
solutions. This corroborates Prof. Zollinger’s suggestion that the
adsorption bands at 632, 422, and 314 nm are assigned to the
MG2+ species, while the peak at 248 nm includes contributions
from both MG2+ and carbinol+ .
From the results in Table 1 we calculated the equilibrium
coefficient of reaction [1], K eq = 6.31 × 107 M−1 using the data
of Solution B for λ = 632 nm, assuming that at t = 0 all the cation
is in its divalent form, and knowing that the buffer solution keeps
the pH constant. Using this value to evaluate pHC yields 5.89,
which amounts to a small difference between the measured and
the calculated value. We focused mainly on the 632-nm peak,
since it is the most influenced by the change in pH.
The standard free energy of the reaction can be calculated
using the equilibrium coefficient, thus 1G ø = −RT ln K eq =
−44.3 kJ mol−1 .

The reaction is an overall second order reaction, being first

order in [MG2+ ] and [OH− ].
TABLE 1
Prof. H. Zollinger suggested that the adsorption bands at 632,
Kinetic Coefficients and Correlation Factors for Diluted Solutions
422, and 314 nm can be assigned to the MG2+ cation, while the
of MG
peak at 248 nm includes MG2+ and the product formed by the
increase of the pH. λ k
The break of the conjugated double-bond system in the center [nm] [min−1 ] OD0 OD∞ r2
of the MG2+ molecule reduces the spread of the large π orbital, Solution B 248 0.0199 0.4511 0.6796 0.981
implying a decrease in the delocalization of the electron, that 314 0.0141 0.5217 0.1700 0.996
results in an increase in the energy needed for excitation, causing 422 0.0135 0.3553 0.1227 0.997
the disappearance of the band at 632 nm and a considerable 632 0.0137 1.4474 0.1771 0.999
decrease in the other bands in the visible part of the spectrum, Solution C 248 0.0161 0.6100 0.4901 0.983
making the solution colorless. Such an effect was reported for the 314 0.0280 0.1996 0.3534 0.994
422 0.0272 0.1451 0.2517 0.995
similar dye malachite green (Prof. S. Yariv, Hebrew University
632 0.0265 0.3357 0.9192 0.998
of Jerusalem, personal communication, 1998).
16 RYTWO ET AL.

Competitive adsorption
Figure 4 presents the results of competitive adsorption be-
tween different species of MG and AF. In Fig. 4a equal amounts
of charges of MG2+ and AF were added, whereas Figs. 4b and

FIG. 3. Adsorption isotherms of divalent methyl green (MG2+ ) and

carbinol+ on montmorillonite after 7 days. Points represent the mean exper-
imental values. Error bars represent standard deviation of the measurements.

Adsorption isotherms
As was pointed out under Materials and Methods, we de-
termined separately the adsorption of MG2+ and carbinol+ by
adding them from solutions at pH 4.8 and 7.4, respectively. After
long incubation times the measured pH varied between 7.0 and
7.4 in both cases. Since the transformation of MG2+ to carbinol+
upon elevation of pH required several hours (Fig. 2), whereas
a great deal of the adsorption occurs within the first hour, we
assumed that when MG is added as MG2+ , its adsorption is
primarily as that of a divalent cation.
Figure 3 shows the adsorption isotherms for both cases. When
the total amounts of MG in the suspension were lower than the
CEC of the clay, no measurable amount of dye remained in the
supernatant for both MG2+ and carbinol+ . Similar results were
reported for several monovalent and divalent organic cations
(11, 14–16). The maximal amount of MG adsorbed when added
as MG2+ was 0.75 mol MG/kg clay, i.e., 185% the CEC of
the clay, assuming divalency of the cation. Similar adsorbed
amounts were reported (11) for the monovalent organic cations
AF, CV, and MB (150, 200, and 175% of the CEC, respectively),
whereas for the divalent organic cations DQ and PQ (16) the
maximal adsorbed amounts were lower, 120 and 100% of the
CEC, respectively. Hence, it is suggested that a certain fraction
of MG sorbs as monovalent cation even when it is added as
MG2+ . For carbinol+ , the adsorbed amounts of MG were larger,
reaching 1.15 mol MG/kg clay, i.e., 140% of the CEC, if we
consider the carbinol as a monovalent cation.
When the added amounts of MG were slightly higher than the
CEC, we observed significant differences between the adsorbed
FIG. 4. Competitive adsorption between MG species and AF after 7 days
amounts after 1 or 3 days (results not shown). At higher
(a). AF and divalent MG2+ were added at equal amounts of charges (b). AF and
added amounts the differences decreased. There were almost no carbinol+ (MGOH+ ) were added at equal molar concentrations (c). The molar
differences between the adsorbed amounts after 3 or 7 days, sug- concentrations of AF added were twice those of MGOH+ . Points are mean
gesting that equilibrium has been reached between 1 and 3 days. experimental values. Standard deviation in all cases was smaller than symbols.
 MG-CLAY INTERACTIONS 17

4c present results for equal amounts of charges of the mono-

valent carbinol+ and AF and a ratio of AF : carbinol+ of 2 : 1,
respectively.
In all three cases, when the added amount of total charges
was lower than the CEC, complete sorption of both organic
cations was achieved. At added amounts larger than the CEC,
AF adsorption is preferred over that of the MG species.
In case a the amount of MG adsorbed decreased when the total
amount of dyes added exceeded the CEC by more than 50% (at
1.2 molc total dyes added/kg clay). At that level it appears that
a maximum adsorption of MG molecules is achieved. Rytwo
et al. (17), who measured competitive adsorption on motmoril-
lonite of the divalent herbicides PQ or DQ with AF, reported a
similar behavior for the competitive adsorption of DQ with AF.
However, when PQ competed for adsorption with AF, no de-
crease of its adsorption with total amount added was observed,
although it adsorbed less than when it was added alone to the
clay. The maximal amount of AF adsorbed was 0.9 molc /kg.
The total amount of organic cations adsorbed was 1.20 molc /kg
clay. There were no differences between the measured adsorbed
FIG. 5. UV-visible spectra of clay-MG particles. Amounts of organic
amounts after 3 or 7 days of incubation; we show results after cations adsorbed (in mol kg−1 clay) are given at the right. The “no clay” lines
7 days of incubation. denote the pure dye solutions. Dashed lines correspond to MG-clay, and full
In case c, where the ratio between the number of molecules lines correspond to carbinol-clay.
of AF and the MG species was the same as in case a
(Fig. 5). The “no clay” lines represent the dye solutions. It can be
(2 AF : 1 carbinol+ ), the amount of carbinol+ sorbed was very
similar to the amount of MG2+ adsorbed in case a, but there was seen that carbinol+ solution does not absorb in the visible range.
a larger amount of AF sorbed when competing with the mono- At low added concentrations, the spectrum is similar, showing a
valent carbinol+ . We assume that this can be ascribed to MG2+ peak batochromically shifted to 672 nm. Such a red shift is as-
molecules binding clay platelets together in case a. Such effect cribed to the surface acidity of the interlayer space of the clay and
can partially inhibit the entry of AF molecules to the interlam- its influence on the MG molecule. Similar results were reported
melar space (17). for other organic dyes, such as methylene blue (20) or crystal
In case b, where equal numbers of AF and carbinol+ violet (21). Cenens and Schoonheydt (20) explain such red shifts
molecules were added, there is an increase in the amount of in terms of the conformation of the molecule on different sur-
carbinol+ sorbed, whereas the amount of AF adsorbed remains face sites. The π ⇒ π ∗ transitions of aromatic molecules shift
almost unaffected by the total concentration of MG molecules to red if the molecule is made more planar. In other cases (20)
added and is similar to the amount of AF adsorbed at case c. large batochromic shifts were ascribed to formation of hydrogen
The outcome of the competition experiments is also expected bonds between the proton donor in the clay surface and the dye
to be dependent on the kinetics of the process. We have seen cation. Such Brønsted acidity is usually due to hydroxyl groups
numerous examples where the order of addition of monovalent in the edges of the platelets or increased dissociation of water
organic cations did affect the adsorbed amounts in favor of the molecules surrounding hydrated cations. We think that in this
cation added first (24 h earlier), even after 30 days of incubation case the first explanation should be more feasible. It should be
(14,18,19). noted that the carbinol+ which is colorless in solution, becomes
blueish when sorbed, due to the acidic influence that imposes
Similarly, a dramatic effect of the order of addition on the
the reorganization of the large π orbital. In other words, the
adsorbed amounts was observed in the competition between AF
acidic influence of the surface induces a change from carbinol+
and the divalent organic cations PQ and DQ (19). The adsorbed
to MG2+ . The change from carbinol+ to MG2+ is not com-
amounts of the divalent organic cations were significantly larger
plete, since the pattern of adsorption of MG as monovalent or-
when added first, and those of AF were smaller, in comparison
ganic cation is retained when it is added to the clay as carbinol+
with the case when the competing cations were added simulta-
(Figs. 3 and 4) and as will be further demonstrated with the XRD
neously (16,17,19).
results (Fig. 6). At larger added amounts, the carbinol samples
show an additional peak at 582 nm, exhibiting metachromasy.
With similar molecules such as CV and ethyl violet (22, 23), a
UV-Visible Spectra
peak shifted hypsochromically by about 50 nm was ascribed to
UV-visible spectra of the organo-clay particles exhibit sig- an influence on the π orbital. Such influence might be caused
nificant differences between the MG-clay and the carbinol-clay by an adjacent clay platelet, indicating flocculation, or by the
18
FIG. 6. Basal spacing of the organo-clay particles. Dashed lines and trian-
gles: MG2+ -clay samples. Full lines and circles: carbinol+ -clay samples.
formation of dimers on the surface of the clay. The peak at
582 nm does not appear in the MG2+ -clay suspensions, at that
loading. Since the CEC of the clay is 0.8 molc /kg, at an added
amount of 0.6 mol/kg of a monovalent cation most of the charges
should be neutralized, leading to a surface potential close to zero
(24) and to flocculation and fast settling of the particles that
was clearly observed. For the same amount of a divalent cation
there is a surplus of positive charge that leads to redispersion,
thus the peak ascribed to flocculation is not evident, in accord
with the fact that the suspension remains stable. At the largest
added amounts, the spectra of MG-clay particles exhibit a peak
at 632 nm, similar to that of MG in solution. This can be ascribed
to molecules loosely adsorbed (11) or to residuals of MG in the
samples, after washing.
XRD Results
XRD measurements as a function of the amount of MG2+
and carbinol+ sorbed are shown in Fig. 6. The clay without
any organic dye sorbed shows a basal spacing typical of mont-
morillonite with a bilayer of water (25). At low adsorbed
amounts the MG-clay samples show a c-spacing slightly smaller
than the carbinol-clay ones. At higher loads, the c-spacing of the
carbinol-clay increases more steeply.
When comparing these results with those presented in previ-
ous studies on the adsorption of other divalent cations (16), a re-
duction in c-spacing was measured only for the MG2+ species at
low adsorbed concentrations. When DQ and PQ were adsorbed,
such a reduction indicated water exclusion from the interlayer
space.
It should be noted that due to the divalency of MG, a given
adsorbed amount of charges amounts to half of the number of
MG molecules adsorbed, compared to CV or carbinol. When
comparing results with equal numbers of molecules adsorbed,
RYTWO ET AL.
the basal spacings measured are lower for the divalent-clay com-
plexes than for the monovalent-clay samples. For example, for
0.8 mol kg−1 of MG2+ added, the basal spacing was 1.52 nm,
whereas for the same amount of carbinol+ the basal spacing was
1.78 nm.
At 0.6 mol kg−1 of carbinol+ the basal spacing is only 1.5 nm.
This confirms our assumption that the metachromasy observed
in the UV-visible measurements is due to interactions with an
adjacent clay platelet, since such a low c-spacing could not allow
the formation of two layers of dye between the clay platelets.
Since the molecular structure of MG is very similar to that of
the monovalent organic cation CV, it is interesting to compare the
basal spacings obtained with these two molecules. For 0.8 mol
kg−1 of CV, the basal spacing was 2.1 nm (11). The difference
between monovalent and divalent cations was ascribed in previ-
ous studies to the fact that a divalent organic cation may bind to
two opposed plates, reducing the basal spacing for the MG2+ -
clay in comparison to the similar monovalent CV- or carbinol+ -
clay complexes.
Adsorption on Sepiolite
Adsorption of MG2+ and carbinol+ by sepiolite further em-
phasized the differences between these species. Monovalent
dyes are adsorbed by sepiolite up to 4-fold of its CEC (26,
27), which equals 0.15 molc /kg clay, whereas adsorbed amounts
of divalent organic cations only reach the CEC (unpublished
results). Figure 7 shows adsorption isotherms of MG2+ and
carbinol+ on sepiolite. The maximal adsorbed amounts of MG2+
and carbinol+ on sepiolite were 0.09 and 0.30 mol/kg, respec-
tively. The very large difference between the MG species can be
interpreted by the difference in valency between them. Mono-
valent dyes can form charged complexes, which consist of two
dye molecules per one negatively charged surface site, and were
also found to adsorb to silanol neutral sites on sepiolite.
FIG. 7. Adsorption isotherms MG2+ and carbinol+ on sepiolite after 7 days.
Points represent the mean experimental values. Error bars represent standard
deviation of the measurements.
 MG-CLAY INTERACTIONS
Concluding Remarks
Both the time-dependent changes in the MG solutions and
its characteristics of adsorption on the clays montmorillonite
and sepiolite are explained on the basis of pH-dependent trans-
formations of MG molecules between MG2+ and MGOH+
(carbinol+ ).
When MG was added from a water solution, and the initial pH
was slightly acidic, it behaved and adsorbed mostly as a divalent
cation. These results were obtained, despite the fact that the final
pH of the suspension was neutral. We suggest that this is due to
the slow kinetics of the transformation of MG2+ to carbinol+ ,
whereas most of the adsorption process occurs during the first
hour. We tested the above proposal by adding MG solutions in
which the pH was previously adjusted to neutral. In this case
methyl green adsorbed mostly as a monovalent cation, which
can be deduced from the larger amounts of molecules adsorbed
and the larger basal spacings.
It is interesting to note that the UV-visible results show that
even carbinol molecules, when added at low rates, are highly
influenced by the acidity of the interlayer space, and the π orbital
is reorganized, leading to a blue color of the organo-clay, even
though the initial carbinol+ solution is completely colorless.
ACKNOWLEDGMENTS
This research was supported in part by a Grant G-0405-95
from G.I.F., the German- Israeli Foundation for Scientific Re-
search and Development, by Grant 8803-1-96, Ministry of Sci-
ence and Arts, Israel, and by the Association Franco-Israelienne
pour la Recherche Scientifique et Technique (AFIRST) for
French-Israeli collaboration. We are grateful to Prof. Shmuel
Yariv for reading the manuscript and for his helpful comments
and suggestions.
REFERENCES
1. Green, F. J., “The Sigma-Aldrich Handbook of Stains, Dyes and Indicators.”
Aldrich Chem. Co., Milwaukee, WI, 1990.
19
2. Margulies, L., Rozen, H., and Cohen, E., Nature 315, 658 (1985).
3. Margulies, L., Cohen, E., and Rozen, H., Pest. Sci. 18, 79 (1987).
4. Botto, R. E., Neil, P., Ruzo, L. O., and Margulies, L., TAMU NMR Newslett.
332, 43 (1986).
5. Margulies, L., and Rozen, H., J. Mol. Struct. 141, 219 (1986).
6. van Olphen, H., and Fripiat, J. J. (Eds.), in “Data Handbook for Clay
Materials and Other Non-metallic Minerals,” p. 19. Pergamon, P. Oxford,
1979.
7. Carter, D. L., Heilman, M. D., and Gonzalez, C. L., Soil Sci. 100, 356
(1965).
8. Rytwo, G., Serban, C., Nir, S., and Margulies, L., Clays Clay Miner. 39,
551 (1991).
9. Ruiz-Hitzky, E., and Casal, B., J. Catal. 92, 291 (1985).
10. Narine, D. R., and Guy, R. D., Clays Clay Miner. 29, 205 (1981).
11. Rytwo, G., Nir, S., and Margulies, L., Soil Sci. Soc. Am. J. 59, 554
(1995).
12. Russell, J. D., in “The Infrared Spectra of Minerals” (V. C. Farmer, Ed.),
p.11, Mineralogical Soc., London, 1974.
13. Zollinger, H., “Color Chemistry,” 2nd ed., p. 316, VCH, New York, 1991.
14. Margulies, L., Rozen, H., and Nir, S., Clays Clay Miner. 36, 270
(1988).
15. Nir, S., Rytwo, G., Yermiyahu, U., and Margulies, L., Colloid Polym. Sci.
272, 619 (1994).
16. Rytwo, G., Nir, S., and Margulies, L., Soil. Sci. Soc. Am. J. 60, 601
(1996).
17. Rytwo, G., Nir, S., and Margulies, L., J. Colloid Interface Sci. 181, 551
(1996).
18. Rytwo, G., Nir, S., and Margulies, L., Clay Miner. 28, 139 (1993).
19. Rytwo, G., Interactions between organic cations and montmorillonite, ad-
sorption and model calculations. Ph.D. Thesis (in Hebrew), Hebrew Uni-
versity, Jerusalem (1994).
20. Cenens, J., and Schoonheydt, R. A., Clays Clay Miner. 36, 214 (1988).
21. Yariv, S., Intern. J. Trop. Agric. 6, 1 (1988).
22. Yariv, S., Ghosh, D. K., and Hepler, L. G., J. Chem. Soc. Faraday Trans.
87, 1201 (1991).
23. Dobrogowska, C., Hepler, L. G., Ghosh, D. K., and Yariv, S., J. Thermal
Anal. 37, 1347 (1991).
24. Schramm, L. L., Yariv, S., Ghosh, D. K., and Hepler, L. G., Can. J. Chem.
75, 1868 (1997).
25. MacEwan, D. M. C., and Wilson, M. J., in “Crystal Structures of Clay
Minerals and Their X-Ray Identification” (G.W. Brindley and G. Brown,
Eds.), p. 197. Mineralogical Soc., London, 1980.
26. Aznar, A. J., Casal, B., Ruiz-Hitzky, E., Lopez-Arbeloa, I., Lopez-Arbeloa,
F., Santaren J., and Alvarez, A., Clay Miner. 27, 101 (1992).
27. Rytwo, G., Nir, S., Margulies, L., Casal, B., Merino, J., Ruiz-Hitzky, E.,
and Serratosa, J. M., Clays Clay Miner. 46, 340 (1998).