Journal of Colloid and Interface Science 280 (2004) 283–288

www.elsevier.com/locate/jcis

Adsorption of methylene blue on Cu(II)-exchanged montmorillonite

Yu-Long Ma a,b,∗ , Zi-Rong Xu b , Tong Guo b , Ping You a
a College of Chemistry and Chemical Engineering, Ningxia University, Yinchuan 750021, People’s Republic of China
b Institute of Feed Science, Zhejiang University, Hangzhou 310029, People’s Republic of China

Received 4 December 2003; accepted 4 August 2004

Available online 21 September 2004

Abstract
Cu2+ -exchanged montmorillonite (CEM) was prepared using the method of ion exchange reaction. The goal was to determine the adsorp-
tion of methylene blue (MB) onto montmorillonite and CEM. The results showed that treatment with Cu2+ slightly reduced the adsorption
of montmorillonite. Temperature, pH value, and ionic strength in the medium had greater or smaller effects on interaction between MB and
the clay. The data obtained from MB adsorption onto the tested adsorbents followed the Langmuir and BET equations. Also, the adsorption
processes were endothermic and spontaneous in nature.
 2004 Elsevier Inc. All rights reserved.

Keywords: Montmorillonite; Copper; Methylene blue; Adsorption

1. Introduction replace antibiotics. Ion-type inorganic antibacterial mater-

Montmorillonite, an aluminosilicate clay, has a 2:1 layer
structure [1]. Each layer has a small net negative charge
because of isomorphous substitution of ions in the frame-
work. This charge imbalance is offset by interlayer hydrated
cations. These interlaminar cations can be exchanged with
other metal cations such as Ag+ , Cu2+ , or Zn2+ . It is well
known that montmorillonite has a great adsorption capac-
ity, which is attributed to its large specific surface area
(SSA) and high cation exchange capacity (CEC) [2–4]. Dale
et al. [5] reported that montmorillonite could adsorb organic
substances either on its external surfaces or within its in-
terlaminar spaces, by interaction with or substitution of the
exchange cations present in these spaces.
Overuse of antibiotics in humans and animals has led to
the rapid evolution of pathogenic bacteria that are resistant to
multiple drugs. These bacteria have been harmful to the pub-
lic heath and environment. To decrease antibiotic use and
overcome antibiotic resistance, it is very important and ur-
gent to prepare and develop safe antibacterial agents that can
ial (IIAM), which is prepared by an ion-exchange reaction
between clay and Ag+ or Cu2+ ions [6–9], has antibacter-
ial activity [10,11]. The mechanism of IIAM action against
bacteria may be attributed to adsorption by the clay and an-
tibacterial activity of loaded metal ions [12]. Therefore, it is
worthwhile to evaluate the adsorptive capacity of IIAM.
Methylene blue (MB) is widely used as a probe molecule
for investigating clay–organic systems [13–15]. The adsorp-
tion of MB on clay minerals is likely to be dominated by ion
exchange processes. In the present work, the adsorptive ca-
pacities of montmorillonite and Cu2+ -exchanged montmo-
rillonite (CEM) are assessed comparatively using MB mole-
cules, and the effects of pH, ionic strength in the medium,
and temperature of adsorption on the interaction between the
clay and MB are investigated.
2. Experimental
2.1. Materials

* Corresponding author. Fax: +86-0571-86091820. The clay samples used in this study were Na-montmoril-
E-mail address: yulongma796@sohu.com (Y.-L. Ma). lonite (NaM) and Ca-montmorillonite (CaM). They were
0021-9797/$ – see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2004.08.044
284 Y.-L. Ma et al. / Journal of Colloid and Interface Science 280 (2004) 283–288
collected from Xinghe and Chifeng, respectively, in the In-
ner Mongolia autonomous region of China. After collection,
samples were ground and washed in deionized water sev-
eral times at a ratio of 10 g clay:100 ml water. The mix-
tures were stirred with the addition of more water for 3 h
and then kept standing overnight. After the clear liquid was
decanted from the top, water was added, stirred, and al-
lowed to settle. The upper part of the suspensions were col-
lected, centrifuged, kept in an air oven at around 60 ◦ C for
slow evaporation to dryness, and then ground and sieved
through a 300-mesh sieve. X-ray diffraction and infrared
spectroscopy measurements showed that the purified sam-
ples were pure NaM and pure CaM [16]. The CEC values
of NaM and CaM were 1.47 and 1.05 mmol g−1 , respec-
tively, which was determined by leaching with 1 M ammo-
nium acetate at pH 7, washing with 90% ethanol, displacing
the NH+ 4 with 1 M NaCl, and measuring the amount dis-
placed with an autoanalyzer [17]. The SSA values were 90.4
and 84.8 m2 g−1 , respectively, measured on NOVA Version
3.70n by N2 adsorption at 77 K and application of the BET
equation [18].
Samples of both NaM and CaM were rehydrated with
deionized water at a ratio of 10 g clay:100 ml water to
which Cu2+ (CuSO4 ·5H2 O, analytical grade) was added in
an amount of ≈1.5 times the CEC of the clay. The result-
ing mixtures were agitated for 24 h. The NaM and CaM
were then separated by centrifugation and washed under
agitation with 100 ml of deionized water. The washed ma-
terials were dried at 60 ◦ C and then ground and sieved
through a 300-mesh sieve. The supernatants were diluted
properly and copper concentrations were measured using
an atomic absorption spectrophotometer (AA-6501, Japan).
The concentrations of copper in Cu2+ -exchanged NaM
(NaM–Cu) and Cu2+ -exchanged CaM (CaM–Cu) were 0.86
and 0.54 mmol g−1 , respectively, calculated as the difference
between the copper concentrations in a control sample pre-
pared without montmorillonite and the sample supernatant
concentration.
2.2. Adsorption studies
Adsorption of MB (analytical grade, obtained from Bei-
jing Chemical Reagents Company, China) was carried out
in a batch process by varying adsorptive concentration, pH
value, and salt concentration of the medium and temper-
ature. A 100-mg sample was mixed with a 100-ml MB
solution of known concentration. The mixture, in a 250-ml
conical flask, was shaken in the water bath of a ther-
mostat at a particular temperature for a period of time.
The pH value was adjusted by adding a few drops of di-
lute NaOH or HCl before shaking. The mixture was al-
lowed to settle and centrifuged. The MB concentration in
the supernatant was determined with a spectrophotometer
(Model 751, Shanghai). Each experiment was performed in
triplicate.
3. Results and discussion
3.1. Adsorption equilibrium and adsorbed amount
The amounts of MB adsorbed on montmorillonite with
time are shown in Fig. 1. The adsorption processes of MB
molecules occurring on each of the tested adsorbents reached
equilibrium after 2 h. This result is in agreement with the
investigation of Sethuraman and Raymahashay [19]. How-
ever, Zhao and Wang [20] reported that the equilibrium in
adsorption of MB on bentonite, which consists mainly of
montmorillonite, was reached within 3 h. Some workers de-
scribed the adsorption process of MB on montmorillonite
particles as divided into two parts [15,21]: rapid adsorption
on the external surface of the clay by ion exchange and ag-
gregation, followed by deaggregation of the MB dimer and
the monomer migrating to the interlamellar position of the
clay. Therefore, the rate of adsorption interaction between
MB and montmorillonite decreases with time [19].
In the present study, the saturated amounts of MB ad-
sorbed onto NaM and CaM were 2.52 and 2.19 mmol g−1
clay, which correspond to 171 and 209% of the CEC, re-
spectively. This result is relatively high as compared to that
of the investigation from Rytwo et al. [22], who showed that
the amount of MB adsorbed on montmorillonite was 150%
of the CEC of the clay when the amounts of MB added were
between 0.95 and 1.65 mmol dye g−1 clay. The discrepancy
may result from the different amounts of MB added, which
were between 1.25 to 12.5 mmol dye g−1 clay in this study,
because an increase in dye concentration can promote dye
aggregation on the clay.
The adsorption capacity decreased in the order NaM >
NaM–Cu > CaM > CaM–Cu. The adsorption data were fit-
ted with Langmuir adsorption rate equation [20],


Ca kt
log10 = , (1)
Ca − Ct 2.303
Fig. 1. Influence of time on adsorption of MB onto the clay. Initial
MB concentration 2.5 mmol g−1 clay; temperature 298.15 K. (2) NaM;
(1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
 Y.-L. Ma et al. / Journal of Colloid and Interface Science 280 (2004) 283–288
Table 1
Rate constants and R 2 of MB adsorption on montmorillonite
NaM NaM–Cu CaM CaM–Cu
k (s−1 × 104 ) 6.00 5.65 4.94 4.73
R2 0.989 0.998 0.995 0.996
Fig. 2. Adsorption of MB on montmorillonite as a function of the initial
MB concentration. Contact time 3 h; temperature 298.15 K. (2) NaM;
(1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
where C a (mmol g−1 ) is the equilibrium adsorbate amount
on the unit mass of the adsorbent, while Ct (mmol g−1 ) is ad-
sorbate amount at time t, t is the adsorption time, and k is the
adsorptive rate constant. The values of k can be determined
from the slope of the linear plot of log10 [Ca /(Ca − Ct )] vs t.
The k value and linear correlation coefficient (R 2 ) for each
adsorbent are given in Table 1. The k value of NaM was
higher than that of CaM, suggesting that the adsorption ca-
pacity of NaM might be better than that of CaM. This result
is consistent with the saturated amount of the dye adsorbed
(Fig. 2). The slight difference between NaM and CaM in ad-
sorption capacity may be attributed to both CEC and SSA,
which are very important parameters to illuminate the ad-
sorption capacity of the clay. High CEC and large SSA
suggest that NaM has a relatively strong ability for cation
exchange and adsorption as compared to the CaM.
The results of this study indicated that treatment with
Cu2+ decreased the adsorption capacity of NaM and CaM,
but the degree of reduction was slightly different. The
data determined from the CEM showed that copper con-
centration in NaM–Cu (0.86 mmol g−1 ) was higher than
that in CaM–Cu (0.54 mmol g−1 ), suggesting that, as com-
pared to the interlaminar Ca2+ , the interlaminar Na+ is
easily exchanged with Cu2+ ions. This also implies that
the exchangeability of interlaminar cations follows the or-
der Na+ > Ca2+ > Cu2+ . This means that the rate of the
exchange reaction between MB and interlayer cations de-
creases in the order Na+ > Ca2+ > Cu2+ ; i.e., the adsorbed
copper in layers has a negative effect on the cation ex-
change adsorption of MB on montmorillonite. In addition,
285
Fig. 3. BET isotherm plots for adsorption of MB on the clay. Contact time
3 h; initial MB concentration 2.5 mmol g−1 clay; temperature 298.15 K.
(2) NaM; (1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
the loaded copper on the edges of montmorillonite will
have the negative charge neutralized; hence, the electrostatic
attractive interaction between MB molecules and clay par-
ticles should be reduced. Therefore, the degree of reduction
in the k value and the amount of dye adsorbed on NaM–Cu
were relatively high as compared to those on the CaM–Cu
due to both of them loading different concentrations of cop-
per.
3.2. Adsorption isotherms
MB adsorption on the respective adsorbents was de-
scribed by the BET isotherm,
Cr b − 1 Cr 1
= + , (2)
Ca (C0 − Cr ) ab C0 ab
where C r (mmol L−1 ) is the equilibrium adsorbate con-
centration in the aqueous phase, C a (mmol g−1 ) is the
amount of MB adsorbed onto the unit mass of the adsorbent,
C0 (mmol L−1 ) is the initial MB concentration in aqueous
phase, and a and b are BET constants.
The Langmuir isotherm was also applied as follows:
Cr Cr 1
= + , (3)
Ca K1 K1 K2
where K1 and K2 are Langmuir constants, and C r and C a
are the same as indicated above.
BET and Langmuir isotherms are given in Figs. 3 and 4,
respectively. Adsorption isotherms of MB on NaM and CaM
are slightly different from those on NaM–Cu and CaM–Cu,
respectively, suggesting that Cu2+ loading may influence
the adsorptive properties of montmorillonite. But the val-
ues of R 2 (Table 2) indicate that the adsorption data for the
MB–clay systems, irrespective of montmorillonite or CEM,
follow the BET and Langmuir isotherms. Both adsorption
isotherms are usually used to describe the adsorptive process
occurring between a solid surface and gas molecules. The
286 Y.-L. Ma et al. / Journal of Colloid and Interface Science 280 (2004) 283–288
Table 2
Isotherm constants of MB adsorption on montmorillonite
BET
a (mmol g−1 ) b R2
NaM 0.945 367.7 0.998
NaM–Cu 0.918 576.3 0.999
CaM 0.662 236.0 0.999
CaM–Cu 0.636 245.9 0.997
Fig. 4. Langmuir isotherm plots for adsorption of MB on the clay. Con-
tact time 3 h; initial MB concentration 2.5 mmol g−1 clay; temperature
298.15 K. (2) NaM; (1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
BET constant a and Langmuir constant K1 are mainly re-
lated to the adsorptive ability of the adsorbent. An increase
in value of the adsorptive constant is connected with an in-
crease in the adsorption capacity of the adsorbent. In this
study we fitted BET and Langmuir isotherms with adsorp-
tion of MB on either montmorillonite or CEM in the aque-
ous phase, finding that the change in values of the BET
constant a and the Langmuir constant K1 were correlative
with the change in the amount of MB adsorbed on each
adsorbent. This implies that the BET constant a and the
Langmuir constant K1 can be used to describe the ability
of montmorillonite to adsorb MB. In addition, the values of
the isotherm constant a and K1 obtained from MB adsorp-
tion on NaM and NaM–Cu were higher than those on CaM
and CaM–Cu, respectively, further suggesting that adsorp-
tion of MB on NaM is relatively strong as compared to that
on CaM.
3.3. Effect of ionic strength
Ionic strength is the most important variable in dye–clay
interaction [23]. In the present study, ionic strength of the
medium exhibited some effect on adsorption of both mont-
morillonite and CEM (Fig. 5). The amounts of MB adsorbed
on the tested adsorbents were decreased with increased con-
centration of NaCl in medium. This result was similar to the
investigation of Narine and Guy [23], who studied the varia-
Langmuir
K1 (mmol g−1 ) K2 (L mmol−1 ) R2
1.210 687.7 0.999
1.129 1001.2 0.999
0.994 223.9 0.990
0.920 310.3 0.998
Fig. 5. Adsorption of MB on montmorillonite with NaCl concentration in
medium. Contact time 3 h; initial MB concentration 2.5 mmol g−1 clay;
temperature 298.15 K. (2) NaM; (1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
tion in adsorption of MB on bentonite as a function of ionic
strength. The reason for this is that, on the one hand, hy-
drated cations in the aqueous phase compete effectively for
the adsorption sites on the clay. On the other hand, “ionic
atmosphere” may be progressively formed around MB mole-
cules with increased NaCl concentration and results in the
reduction of MB adsorption on the tested clay.
However, the degree of reduction by ionic strength in
adsorption of MB on CEM was higher than that on mont-
morillonite in this study. Morton et al. [24] reported that the
increase in Na+ concentration resulted in displacement of
the copper from the interlayer sites and shift in copper ad-
sorption from the interlayer to the edge sites. An increase in
copper concentration at the edge sites may reduce the mag-
nitude of the negative surface potential and thereby affect
the electrostatic adsorption of MB molecules on montmoril-
lonite.
3.4. Effect of pH
The amounts of MB adsorbed on each adsorbent in the
acidic medium were slightly low as compared to those in the
alkaline medium (Fig. 6). In the acidic medium, the H+ ion
can competitively exclude the adsorption of MB by exchang-
ing with cations on the surface or in the interlayer region of
the clay. In the alkaline medium, Cu2+ exchanged had a poor
influence on adsorption of montmorillonite. This may be at-
tributed to the reaction of OH− and Cu2+ ions and thereby
 Y.-L. Ma et al. / Journal of Colloid and Interface Science 280 (2004) 283–288
Fig. 6. Adsorption of MB on montmorillonite with pH. Contact time 3 h; ini-
tial MB concentration 2.5 mmol g−1 clay; temperature 298.15 K. (2) NaM;
(1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
Fig. 7. Effect of temperature on adsorption of MB to montmorillonite.
Contact time 3 h; initial MB concentration 2.5 mmol g−1 clay. (2) NaM;
(1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
reduces the negative effect of Cu2+ ions on MB adsorption.
In addition, MB molecules in the alkaline medium may be
dehydrogenated, which reduces positive charge, and conse-
quently should result in MB aggregation on the surface of
the clay.
In the present study, the amounts of MB adsorbed on each
adsorbent showed a minimum at a certain pH of the medium.
We guess that this phenomenon might be related to the point
of zero charge (PZC) on the surface of the clay, because
chemical adsorption is lowest when the medium pH is equal
to the PZC of clay.
3.5. Effect of temperature
The amounts of MB adsorbed on each adsorbent in-
creased with increasing temperature of adsorption (Fig. 7).
287
Fig. 8. Plots of log10 Kd vs 1/T for adsorption of MB on montmorillonite.
Contact time 3 h; initial MB concentration 2.5 mmol g−1 clay. (2) NaM;
(1) NaM–Cu; (Q) CaM; (P) CaM–Cu.
This is because the diffusion of MB molecules in the aque-
ous phase is gradually faster with increased temperature of
adsorption and thereby promotes the exchange interaction
between MB molecules and cations on the clay. Moreover, a
similarly increasing degree in amount of MB adsorbed on
each adsorbent was observed with increased temperature,
suggesting that temperature has a similar influence on the
adsorption capacity of montmorillonite and CEM.
3.6. Thermodynamic parameters
The thermodynamic parameters of the adsorption process
are obtained from experiments at various temperatures using
the equations [14],

S
H
log10 Kd = − , (4)
2.303R 2.303RT

G =
H − T
S, (5)
Ca
Kd = , (6)
Cr
where Kd is the distribution coefficient for the adsorption,

S is the entropy change, R is the molar gas constant,
H
is the enthalpy change, T is the absolute temperature,
G
is the Gibbs energy change, and C a and C r are the same as
indicted above. The values of
H and
S were determined
from the slope and intercept of the linear plot of log10 Kd vs
1/T (Fig. 8). The results are presented in Table 3.
Thermodynamic parameters demonstrate a spontaneous
and favorable reaction process. As shown in Table 3, the val-
ues of
H and
S were positive, whereas
G values were
negative for the adsorption of MB on the surfaces of all the
tested adsorbents. These results are in agreement with the
investigation of Ghosh and Bhattacharyya [14], who used
kaolinite as the adsorbent. But Gemeay [25] obtained neg-
ative
H and
S values for adsorption of rhodamine-6G
on Na–montmorillonite. In the present study, treatment with
288 Y.-L. Ma et al. / Journal of Colloid and Interface Science 280 (2004) 283–288

Table 3
Thermodynamic parameters for adsorption of MB on montmorillonite

H (kJ mol−1 )
S (J mol−1 K−1 ) −
G (kJ mol−1 ) at temperature R2
277 K 288 K 298 K 310 K
NaM 2.92 26.00 4.29 4.57 4.83 5.14 0.984
NaM–Cu 2.31 23.49 4.20 4.46 4.69 4.98 0.998
CaM 2.77 24.45 4.01 4.28 4.52 4.81 0.965
CaM–Cu 2.26 22.38 3.94 4.19 4.41 4.68 0.971

copper slightly decreased the absolute values of
H ,
S, Acknowledgments

and
G, suggesting that the loaded copper has a small in-
fluence on the adsorption of MB molecules on the mont-
morillonite particles. The positive value of
H indicates
that the adsorption process is endothermic in nature, but the
required activation enthalpy is not very large. The adsorp-
tion is accompanied by an increase in entropy, suggesting
that the adsorption process in the aqueous medium is dif-
ferent from that in the gas phase. Normally, adsorption of
gases leads to a decrease in entropy due to orderly arrange-
ment of the gas molecules on a solid surface. Therefore, the
adsorption of MB from solution on a nonuniform montmo-
rillonite surface is more complicated. The negative values
of
G from this study show that the adsorption of MB
on either montmorillonite or CEM is spontaneous in ther-
modynamics at body temperature (37 ◦ C), room tempera-
ture (25 ◦ C), and the standard temperature for refrigeration
(4 ◦ C). In addition, there is no large change in value of any
of the thermodynamic parameters for all the tested adsor-
bents, suggesting that thermodynamic processes involved in
the clay–MB interaction are more or less uniform in na-
ture.
4. Conclusion
The amount of MB adsorbed on the NaM was higher
than that on the CaM. Treatment with Cu2+ ion decreased
the adsorption capacities of NaM and CaM. The abilities of
all the tested adsorbents to adsorb MB were relatively low
in the acidic medium as compared to those in the alkaline
medium. The amount of the dye adsorbed on the adsorbent
showed a minimum at a certain pH of the medium. The
adsorbed amount of MB on the tested clay gradually de-
creased with the increase of ionic strength in medium or
with the decrease of adsorptive temperature. The data ob-
tained from adsorption of MB on the surfaces of all the
four adsorbents obeyed the Langmuir and BET isotherms.
The adsorption processes for MB on either montmoril-
lonite or CEM were endothermic and spontaneous in na-
ture.
Financial support was obtained from the Chinese Na-
tional “985” Program and from the Science Research Foun-
dation of Ningxia University. We thank Fei-An Ye and Xin-
Yan Han for skilled technical assistance.
References
[1] G. Borchardt, in: J.B. Dixon, S.B. Weed (Eds.), Minerals in Soil Envi-
ronments, Soil Science of America, Madison, WI, 1989, p. 675.
[2] A.J. Ramos, E. Hernandez, Anim. Feed Sci. Technol. 62 (1996) 263.
[3] H.P. He, J.G. Guo, X.D. Xie, J.L. Pen, Acta Mineral. Sin. 19 (1999)
231.
[4] X.R. Hu, G.L. Lu, L.S. Chen, J.M. Gu, Y. Zhang, Acta Pharm. Sin. 37
(2002) 718.
[5] J. Dale, M. Kowalska, D.L. Cocke, Chemosphere 22 (1991) 769.
[6] A. Oya, T. Banse, F. Ohashi, S. Otani, Appl. Clay Sci. 6 (1991) 135.
[7] Y. Inoue, Y. Kanzaki, J. Inorg. Biochem. 67 (1997) 377.
[8] F. Ohashi, A. Oya, L. Duclaux, F. Beguin, Appl. Clay Sci. 12 (1998)
435.
[9] Z.R. Xu, Y.L. Ma, C.H. Hu, M.S. Xia, T. Guo, H.L. Jin, Asian-Aust.
J. Anim. Sci. 16 (2003) 1673.
[10] M. Rivera-Garza, M.T. Olguín, I. García-Sosa, D. Alcántara, G. Rod-
ríguez-Fuentes, Micropor. Mesopor. Mater. 39 (2000) 431.
[11] G. Čík, H. Bujdáková, F. Šeršeň, Chemosphere 44 (2001) 313.
[12] Y. Ye, Y.H. Zhou, M.S. Xia, C.H. Hu, H.F. Ling, J. Inorg. Mater. 18
(2003) 569.
[13] C. Viseras, A. Lopez-Galindo, Appl. Clay Sci. 14 (1999) 69.
[14] D. Ghosh, K.G. Bhattacharyya, Appl. Clay Sci. 20 (2002) 295.
[15] M.G. Neumann, F. Gessner, C.C. Schmitt, R. Sartori, J. Colloid Inter-
face Sci. 255 (2002) 254.
[16] J.Q. Pan, in: Y.X. Yang, N.X. Zhang (Eds.), Clay Minerals of China,
Geological Press, Beijing, 1994, p. 116.
[17] B.K. Theng, G.S. Hayashi, M. Soma, H. Esyama, Clays Clay Miner. 45
(1997) 718.
[18] M. Stadler, P.W. Schindler, Clays Clay Miner. 41 (1993) 288.
[19] V.V. Sethuraman, B.C. Raymahashay, Environ. Sci. Technol. 9 (1975)
1139.
[20] B. Zhao, G.Q. Wang, Ion Exch. Adsorpt. 18 (2002) 156.
[21] F. Gessner, C.C. Schmitt, M.G. Neumann, Langmuir 10 (1994) 3749.
[22] S. Rytwo, S. Nir, L. Margulies, Soil Sci. Soc. Am. J. 59 (1995) 554.
[23] D.R. Narine, R.D. Guy, Clays Clay Miner. 29 (1981) 205.
[24] J.D. Morton, J.D. Semrau, K.F. Hayes, Geochem. Cosmochim. Acta
65 (2001) 2706.
[25] A.H. Gemeay, J. Colloid Interface Sci. 251 (2002) 235.