The Nature of Laponite and Its Aqueous Dispersions

DUDLEY W. THOMPSON 1 AND JAMES T. BUTTERWORTH

School of Chemistry, University of Bristol, Bristol BS8 1TS, U.K.

Received May 15, 1991; accepted October 30, 1991

A study has been made of the physical properties and aqueous dispersion characteristics of the synthetic
magnesium silicate Laponite XLG. Results obtained from electron microscopy, electron diffraction,
ultracentrifuge analysis, and small angle X-ray scattering studies indicate that the primary particles of
this material are thin plate-like crystals of reasonably uniform size, having (plate) diameters of ca. 20
rim. In dilute, colloidally stable aqueous dispersions the average thickness of discrete crystals is within
the range 2-4 nm. Significant dissolution of this Laponite material occurs in aqueous solutions of pH
< 9 with solution concentrations of magnesium exceeding 10-3 tool dm -3 at pH ~< 7. In the region of
pH 7-8 precipitation of colloidal silica results in a marked increase in the cation-exchange capacity
(CEC) of Laponite dispersions. CEC values obtained for dispersions having pH values within the range
8-10 indicate that the Laponite particles develop no significant pH-dependent surface charge in aqueous
electrolyte (NaC1) solutions. The results of this investigation indicate that with regard to particle mor-
phology and high pH surface charge characteristics Laponite XLG constitutes a reasonable model material
for studies of the behavior of natural smectitie clay minerals. It is clear however that the use of Laponite
for such studies requires careful attention to dispersion preparation and "purification" methods and to
the choice and control of experimental pH ranges. © 1992AcademicPress,Inc.

INTRODUCTION can be prepared reproducibly, in large

Laponite materials are synthetic magnesium amounts, and free from contaminants com-
silicates manufactured by Laporte Industries, monly associated with natural clay samples,
Ltd., principally for use as fillers and thick- they appear to constitute useful model mate-
ening agents for aqueous preparations. The rials for studies of the surface and colloid
chemical composition and X-ray diffraction properties of smectite clays (4-8). However,
characteristics of these materials indicate that in addition to some degree of uncertainty as
they are layer-lattice silicates having structures to the exact structure and physical form of La-
similar to those of natural, trioctahedral ponite particles, other properties of these ma-
smectite clay minerals ( 1). Light and neutron terials, such as their apparent instability in so-
scattering data (2, 3) suggest that Laponite lutions of neutral or acid pH and their elec-
particle morphology is also similar to that of trokinetic and colloid stability properties in
smectite clays, and that in dilute, colloidally simple inorganic electrolyte solutions (5),
stable aqueous dispersions the particles exist would seem to differ significantly from those
as discrete plate-like crystals of diameter 20- of natural smectite clays. In view of this, an
30 nm and thickness within the range 1-4 rim. attempt has been made in the present work to
Results obtained from electron microscopy clarify the nature of Laponite materials and
have been claimed to support this particle to determine their suitability for model studies
morphology ( 1, 3, 4), although direct evidence of natural clay systems.
of this is by no means convincing.
MATERIALS AND METHODS
Owing to the fact that Laponite materials
The material investigated was Laponite
1 TO whom correspondence should be addressed. XLG, which is a gel-forming grade of Laponite
236
0021-9797/92 $5.00
Copyright© 1992by AcademicPress,Inc.
All rightsof reproductionin any formreserved. Journalof Colloidand InterfaceScience, Vol. 15I, No. I, June 1992
 LAPONITE DISPERSIONS 237

containing no stabilizing additives. This ma- pH Titration of Laponite Dispersions

terial was found to disperse readily in water Laponite dispersions (ca. 0.5 g dm -3 ) were
to form essentially optically clear dispersions
titrated with hydrochloric acid using the fol-
within a few hours. Shear-thinning gel char-
lowing batch-titration method.
acteristics were exhibited by dispersions having
To 10.0 cm 3 samples of dialyzed dispersion
solids concentrations > ca. 30 g dm -3. At-
were added 1.0 cm 3 of either water or sodium
tempts to "purify" dilute dispersions by di-
chloride solution (10-2-1.0 mol dm-3). Small
alysis against water resulted in significant re-
volumes (< 0.5 cm 3) of standard hydrochloric
ductions in solids concentrations and in some
acid solution were added to the dispersions
cases visible evidence of particle aggregation.
using an Agla micrometer syringe. The mix-
In view of this, a more detailed investigation
tures were allowed to equilibrate in sealed
of dispersion preparation characteristics was
tubes for a period of 3-5 days with continuous
carried out as follows.
agitation. Equilibrium pH values were then
A series of dispersions was prepared by dis-
recorded using a glass electrode combined with
persing different amounts of Laponite powder
a double junction, silver-silver chloride ref-
in water ( 1.0-20.0 g dm-3). The dispersions
erence electrode, the outer compartment of
were sealed from the atmosphere and agitated
which was filled with the supporting electrolyte
continuously for a period of 24 h. Turbidity
solution. Samples of the final dialysate solution
()~ = 300 nm) and ultracentrifuge studies in-
used in the original dispersion preparation
dicated that the time required for complete
were also titrated in a similar manner.
dispersal of the Laponite particles was in the
region of 10 h. Simple centrifugation and ul-
trafiltration methods were found to be unsat-
Analysis of Laponite Dispersion Media and
isfactory for the phase separation of colloidally
Cation-Exchange Capacity ( CEC)
stable dispersions. In view of this, analysis of
Measurements
the dispersion media was carried out using the
following dialysis method. Samples of Laponite dispersions ( 10.0 cm 3
A sample of each dispersion (25.0 cm 3) was of ca. 10 g dm-3), which had been extensively
sealed in "Visking" dialysis tubing and the dialyzed against solutions of sodium chloride
tube was immersed in water (25.0 cm3). The ( l0 4 __ 1 0 - 1 mol dm -3) at pH 10 (NaOH),
systems were sealed from the atmosphere and were sealed in dialysis tubes. The tubes were
allowed to equilibrate for a period of 7 days, immersed in solutions of sodium chloride
with continuous agitation. After this period of (15.0 cm 3) of the appropriate concentration
time the pH values of the dialysate solutions to which small volumes ( < 0.5 cm 3) of hy-
were recorded and samples of these solutions drochloric acid solution had been added. The
were analyzed for sodium, magnesium, and systems were sealed from the atmosphere and
silicate. Sodium analyses were carried out us- allowed to equilibrate for a period of 3-5 days
ing flame emission spectroscopy; magnesium with continuous agitation. Samples of the di-
and silicate were determined using colorimet- alysate solutions (10.0 cm 3) were then re-
ric methods based on formation of the Mg 2+- moved for analysis and replaced by ammo-
eriochrome black T complex (9) and of/3-sil- nium acetate (40 g dm 3) buffer solution at
icomolybdic acid (10). the same pH as the dialysate solution. These
On the basis of results obtained from these systems were allowed to equilibrate for a pe-
analyses, subsequent investigations were car- riod of 3 days with constant agitation. The
ried out using dispersions that had been pre- dialysate solutions were then analysed for so-
pared initially at a solids concentration of 10 dium and magnesium in order to determine
g dm -3 and dialyzed extensively against the total amount of cation displaced by am-
aqueous alkali (NaOH or NH4OH ) at pH 10. monium ions. In all cases, the pH values of
Journal of Colloid and Inter'face Science, Vol. 151, No. 1, June 1992
238 THOMPSON AND BUTTERWORTH

corresponding initial and final dialysates were

'~E 10-0
found to vary by no more than +0.1. -o
2.0
E
Electron Microscopy and Diffraction 9.O
z
o pH
Transmission electron microscopy and 1.5
< B.O
electron diffraction studies were carried out
using a Jeol 100 CX Temscan electron micro-
scope. iJ
o
7.0
z
Ultracentrifuge Analysis 2
0.5
Ultracentrifugation studies were carried out
using a Beckman Model E instrument. Dis- (
persions were centrifuged in 4 ° sector cells ( 12 o 0' ~ - - - ~£ I A I I
0 2.0 4.0 6.0 8.0 10.0
or 30 m m optical pathlength) having Filled- LAPONITE CONCENTRATION/g din-3
Epon center pieces. Particle sedimentation was
FIG. 1. Equilibrium concentrations of solution species
detected using schlieren optics, and all mea- obtained by dispersing Laponite XLG in water at room
surements were made at a temperature of 25 temperature. ([3) Na ÷, (D) Mg 2+, (©) SiO2 (aq), (@)
_____l°C. pH.

Density Measurements
Density determinations were carried out acteristics, which are reported later in this pa-
using an Anton Paar digital precision density per. At higher Laponite concentrations mag-
meter. Measurements made on Laponite dis- nesium release became undetectable, whereas
persions having concentrations within the the amounts of sodium and silicate furnished
range 0.8-8.0 g dm -3 gave a particle density to the solution were found to be proportional
of 2.2 + 0.1 × 103 kg m -3. to the concentration of dispersed Laponite.
These results strongly suggest the presence of
Small Angle X-ray Scattering (SAXS) soluble sodium and silicate materials in this
Laponite sample.
SAXS studies were carried out using a two- The dissolution characteristics of Laponite
dimensional X-ray diffractometer incorporat- precluded "purification" of its aqueous dis-
ing a two-dimensional X-ray imaging system persions by extensive dialysis against water or
( l 1). Measurements were made on dilute ( 10- electrolyte solutions of neutral pH. Dialysis
20 g d m -3) colloidally stable dispersions of against aqueous alkali at pH 10, however, was
Laponite using a perspex cell (1 m m path found to remove essentially all of the released
length) fitted with thin ( 30-40 #m) mica win- sodium, silicate, and magnesium from the so-
dows. lution phase with no detectable changes in the
state or form of the dispersed particles.
RESULTS AND DISCUSSION
Chemical analyses of the "as-received" La-
Results obtained from analysis of Laponite ponite sample and material, which had been
dispersion media (Fig. 1 and Table I) indicated extensively washed by dialysis against aqueous
significant dissolution of the magnesium sili- alkali at pH 10, were found to be consistent
cate in dispersions having solids concentra- with the presence in the supplied sample of a
tions < ca. 2 g dm -3. This effect almost cer- magnesium silicate having the composition
tainly arises from the relatively low pH value Nao.sSi8Mgs.4Lio.4Oz0(OH)4 together with
of these dispersions as was indicated by sub- small amounts of Na2SiO3 (ca. 2.4%), Na2CO3
sequent studies of Laponite dissolution char- (ca. 0.5%), and Na2SO4 (ca. 0.5%).

Journal of Colloid and Interface Science Vol. 151, No. 1, June 1992
 LAPONITE DISPERSIONS 239

TABLE I

A m o u n t s of Solution Species Released on Dispersing L a p o n i t e X L G in W a t e r at R o o m T e m p e r a t u r e

Solution species released (mol g-i (Laponite))

Laponite concentration
(g dm-3) Na Mg SiO2(aq.) pH

0.05 1.46 X 10 3 1.55 X 10 -3 3.10 × 10 3 7.27

0.52 5.94 X 10 -4 8.74 × 10 -5 6.48 X 10 -4 7.75
2.51 3.53 X 10 4 0 2.05 X 10 -4 9.30
5.06 2.68 × 10 -4 0 1.29 × 10 4 9.54
10.04 2.27 X 10 -4 0 8.91 X 10 -5 9.78

Electron Microscopy in the formation of apparently structureless

film-like patches of material. Around the pe-
Imaging individual Laponite particles by riphery of these patches, however, were de-
transmission electron microscopy proved to tected what appeared to be thin plate-like par-
be difficult. Attempts to prepare suitable sam- ticles of extremely low electron density having
ples by electrostatically depositing the particles an average plate diameter in the region of 20
from dispersions having pH values within the nm (Fig. 2). We believe these to be individual
range 8-10 using protein films (12) were un- Laponite crystals, however, owing to the dif-
successful. The evaporation of small drops of ficulty of unambiguously identifying these
dilute (10-3-10 -4 g dm -3) dispersions in particles as discrete crystals, as opposed to ag-
aqueous sodium hydroxide (pH 10) resulted gregates of much smaller particles formed

FIG. 2. T r a n s m i s s i o n electron m i c r o g r a p h of L a p o n i t e X L G .

Jm~rnal o/'Colloid and lnterjace Science Vol. 151, No. 1, June 1992
 240 THOMPSON AND BUTTERWORTH

during the evaporation process, additional ev-

idence was sought using a number of different 6-0

techniques.
m 5'0-
Electron Diffraction
Electron diffraction by thin areas of dried ~'~--Z..0 -

Laponite films gave rising patterns with re- o

flections closely corresponding to the inter- 3.0

planar spacings of smectite crystals irradiated

in a direction normal to the 0001 crystal plane. 2.0
Tilting the specimen with respect to the elec- 2.0 /-,.0 6.0 8.0 I0-0
DISPERSION CONCENTRAT]ON/g dm-3
tron beam resulted in eliptical patterns, a
characteristic of thin plate-like crystals ( 13 ). FIG. 3. The influence of dispersion concentration on
the sedimentation coefl%ient (S) for Laponite particles
dispersed in aqueous NaC1 ( 10-4 tool dm -a)/NaOH (pH
Ultracentrifuge Analysis 10).
Centrifugation of dispersions having solids
concentrations within the range 1-10 g dm -3
at an acceleration of 30,270 g (20,410 rpm) independent measure of particle mass made
resulted in the development of narrow peaks using the approach to equilibrium method of
in the schlieren image, indicating a reasonably Archibald (15) at an acceleration of 1,750 g
uniform particle size. Results obtained for (4,908 rpm) gave the value 1.6 × 10 -21 kg.
dispersions containing different concentra- These mass values almost certainly underes-
tions of added sodium chloride showed that timate the true particle mass to some extent
at salt concentrations > ca. 10 -3 mol dm -3, owing to the effects of particle charge on the
the value of the sedimentation coefficient in- sedimentation process (14). However, in con-
creased markedly with increasing salt concen- junction with the electron microscope estimate
tration. This effect was almost certainly due of the average particle (plate) diameter (ca.
to particle aggregation at the higher salt con- 20 nm), the particle mass values obtained from
centrations and precluded the use of "swamp- ultracentrifuge analyses are consistent with a
ing electrolyte" to minimize the effects of par- crystal thickness of 2-3 nm.
ticle charge on the sedimentation process. In
view of these results all subsequent ultracen- S A X S Studies
trifuge studies were carried out on dispersions
Small angle X-ray scattering data obtained
containing ~< 10 -4 mol dm -3 added sodium
from colloidally stable dispersions of Laponite
chloride.
were analyzed using the Guinier approxima-
The influence of Laponite concentration on
tion (16) that at low scattering vector (Q) the
the value of the sedimentation coefficient is
scattered intensity ( I ( Q ) ) for a dilute system
shown in Fig. 3. Extrapolation of these data
depends only on the radius of gyration (RG)
to zero particle concentration gave an infinite
of the particles, and is not sensitive to the de-
dilution sedimentation coefficient (So) of 6.1
tails of particle shape:
X 10 -12 S. The diffusion coefficient (D) deter-
mined using the "synthetic boundary" method In I ( Q ) -~ const
(14) was f o u n d t o be 2.1 × 10 -11 m2s -1 . Sub-
- ~1 t ' - 1 21,o
D 2
(QRG < 1). [1]
stituting these values for So and D into the
Svedberg Equation (14) gave a value for the A Guinier plot of In I ( Q ) against Q2 (Fig.
average particle mass of 2.2 × 10 -21 kg. An 4a) indicated that for the Laponite particles,

Journal of Colloid and Interface Science. Vol. 151, No. 1, June 1992
 LAPONITE DISPERSIONS 241

phology of Laponite particles, which has been

.82f ,,.
indicated by previous scattering studies of a
similar material, Laponite RD (2, 3). The use
3"17I~
of different techniques has led to some varia-
tion in estimated particle dimensions; how-
F| "... °°~.... ever, there would seem to be little doubt that
the primary particles of Laponite XLG are
o.,5 . reasonably uniform in size, having an average
(plate) diameter in the region of 20 nm and
a thickness of 2-4 nm in dilute, colloidally
_,J , , t--{1, stable aqueous dispersions.
0 0-03 0"07 0.10 0-13
Q2

pH-Titration and Solution Characteristics of

Laponite Dispersions
Results obtained from strong acid titrations
36 of Laponite dispersions in aqueous mixtures
%
of sodium hydroxide and sodium chloride
q -3.8
showed that these dispersions exhibit a sub-
-/~.0 stantial pH-buffer capacity at pH values < ca.
7. In addition, titration curves obtained for
-4.2 dispersions at different ionic strengths were
-4.l I I I I i I i I found to exhibit a clearly defined common in-
0 0,02 0"04 0.06 0,08
Q2 tersection point in the region of pH 7.4 (Fig.
5 ). This latter effect, which was confirmed by
FIG. 4. SAXS data obtained for a colloidally stable La- salt titration studies (17), is typical of the be-
ponite dispersion (20 g dm-3). (a) Guinier plot (Eq. 1 ),
havior of hydrous oxide dispersions and may
(b) Flat plate approximation plot (Eq. [2]).
well arise from the influence of ionic strength
on the ionization and protonation of hydroxyl
groups located at the Laponite crystal edges.
Re -~ 4.1 nm. If the particles are assumed to
This would seem to be supported by electro-
be of a plate-like form, then at higher Q values kinetic data obtained by Perkins et al. (5) for
the scattering parameters can be approxi- a very similar material, Laponite CP. In the
mately related to the plate thickness ( T ) in
case of simple hydrous oxide systems common
the manner shown by intersection points of titration curves obtained
l n ( Q 2 - I ( Q ) ) -~ const - QZT2. [2] for dispersions at different ionic strengths de-
fine the point of zero salt effect pH on the
A plot of ln(Q 2. I(Q)) against Q2 (Fig. 4b) surface proton charge, which in the absence
indicated a plate thickness T - 3.9 nm. For a of specific adsorption effects can be identified
disk-shaped particle of diameter D the esti- with the point of zero proton charge pH of the
mates of whole particle radius of gyration and surface (18 ). Results obtained from analysis
particle thickness indicate a minimum particle of Laponite dispersion media, however, indi-
diameter in the region of 11 nm: cated that the interpretation of titration data
D -- (8(R 2 - T2/12)) '/2. [3] obtained for dispersions of this material is
somewhat more complicated than in the case
The results obtained from EM, ED, and of simple oxide systems.
SAXS studies confirm the plate-like mor- Detectable dissolution of Laponite was

Journal of Colloid and Inte~ace Science, Vol. 151,No. 1, June 1992

242 THOMPSON AND BUTTERWORTH

N a + exchange at the Laponite surface rather

8,5
I [.oc,]
tool dnT3
than from an effect of ionic strength on La-
ponite solubility. This would seem to be sup-
ported by the fact that at the higher ionic
--o- 10.-3
strengths, over the range of p H 8.5-7.3, the
--41]--- 10_
-4
ratio of the solution concentrations of mag-
8,0
nesium/silicate remained within the range
0.66 _+ 0.05 which is reasonably consistent
pH
with the value of 0.68 expected for congruent
dissolution of Laponite according to
7-5
Nao.sSi 8Mgs.4Li0.4020 (OH)4 + 12H +
+ 8H20 ~ 0.8Na + + 8Si(OH)4
+ 5.4Mg 2+ + 0.4Li +. [3]
7,0
These results indicate that pH-titration data
obtained for Laponite dispersions could be in-
I I I I I I I fluenced both by the specific adsorption of
2,0 4-0 6.0 8-0 10.0 12.0 14.0 magnesium, and by the surface chemistry of
1 0 3 X ( C A _ C B ) / m o l drn-3 precipitated silica. In view of this it is difficult
to draw firm conclusions regarding the proton
FIG. 5. The influence of electrolyte concentration
(NaCI) on the pH-titration curves for Laponite XLG dis-
persions (4.9 g dm-3) (CA and CB = concentrations of
9-0
added acid and base, respectively).
ENaCrl
mo[
8.o 1° 3 1° 1
found to occur at p H values < ca. 9 (Fig. 6) Mg ,'% •
with the solution concentration of magnesium 7-0
Si02(oq.J 0 •
increasing in an essentially exponential man-
ner with decreasing pH. The concentration of ~ 6'0
dissolved silica, however, was found to attain
an essentially constant value (ca. 2.3 × 10 -3 ~.~ 5.0
mol d m -3) at p H values < ca. 7.3. This latter g
F-
result strongly suggested that precipitation of ~ 4.0
amorphous silica occurred in the titration sys-
tems in the region o f p H 7.3 (19) and this was o 3,0
clearly confirmed by electron microscope ex-
amination of the dialysate solutions. For par-
£ 2-0
ticular values of solution pH, concentrations
of dissolved magnesium were found to in-
1'0
crease with increasing concentration of the
supporting electrolyte (Fig. 6). In view of the
0
fact that corresponding solution concentra- 7.0 8.0 9.0
tions of silicate were found to be essentially pH
independent of electrolyte concentration it FIG. 6. The influence of p H o n the solution concentra-
seems almost certain that this effect arises from tions of m a g n e s i u m and silicate for dispersions of Laponite
the influence of ionic strength on the Mg 2+- X L G in a q u e o u s NaCI.

Journal of Colloid and Inter~bee Science, Vol. 151, No. 1, J une 1992
 LAPONITE DISPERSIONS 243

surface charge characteristics of Laponite methods of preparation and purification of

crystal surfaces on the basis of simple pH- aqueous dispersion, the choice and control of
titration data alone. Further information con- experimental pH ranges, and the pH-history
cerning this aspect of Laponite dispersions of experimental systems.
was, however, obtained from direct measure-
ments of cation-exchange capacity. Over the ACKNOWLEDGMENTS
range of pH 7.5-10 the CEC was found to We thank Mr. B. J. R. Mayes of Laporte Industries,
remain essentially constant, at a value of 0.65 Ltd., for the gift of the Laponite sample and also Dr.
+ 0.01 mol eq g-i but increased significantly R. M. Richardson and Dr. J. Jutson of the University of
with decreasing pH at lower pH values (Fig. Bristol for carrying out the SAXS measurements.
7). These results would seem to indicate that
Laponite crystals do not in fact develop any REFERENCES
significant pH-dependent surface charge in
1. van Ophen, H., and Fripiat, J. J. (Eds.), "Data Hand-
solutions of simple electrolytes such as sodium
book for Clay Minerals and Other Non-Metallic
chloride; the increased values of the CEC at Minerals." OECD and Clay Minerals Society, Per-
pH < 7.5 almost certainly arises from the hy- gamon, Oxford, 1979.
droxylated surface of precipitated silica. 2. Avery, R. G., and Ramsay, J. D. F., J. Colloid Interface
Sci. 109, 448 (1986).
3. Rosta, L., and Gunten, H. R., J. Colloid Interface Sci.
CONCLUSIONS
134, 397 (1990).
With regard to particle morphology and 4. Ramsay, J. D. F., J. Colloid Interface Sci. 109, 441
(1986).
high-pH surface charge characteristics it would 5. Perkins, R., Brace, R., and Matijevic, E., J. Colloid
seem that Laponite dispersions can constitute Interface Sci. 48, 417 (1974).
reasonable systems for model studies of the 6. Lockhart, N. C., J. Colloid Interface Sci. 74, 509
behavior of smectite clay dispersions. The re- (1980).
sults obtained in this investigation, however, 7. Fripiat, J., Cases, J., Francois, M., and Letellier, M.,
J. Colloid Interface Sci. 89, 378 (1982).
clearly indicate that when used for this purpose
8. Grauer, Z., Malter, A. B., Yariv, S., and Avnir, D.,
considerable care must be exercised in the Colloids Surf 25, 41 (1987).
9. Vogel, A. I., "Quantitative Inorganic Analysis," 3rd
Ed. Longmans, Green, New York, 1966.
1"8
10. Chalmers, R. A., and Sinclair, A. G., Anal Chim.
Acta 33, 384 (1965).
1-6 11. Swanton, S. W., "The Physico-Chenqical Character-
ization of Zirconium Polymers in Aqueous Solu-
tions." PhD thesis, University of Bristol, 1989.
1-4
12. Green, R. S. B., Murphy, P. J., Posner, A. M., and
Quirk, J. P., Clays Clay Miner. 31,230 (1983).
1-2 13. Rymer, T. B., "Electron diffraction." Methuen, Lon-
E don, 1970.
c_)
14. Svedberg, T., and Pederson, K. P., "The ultracentri-
LU 1'0 fuge." Oxford Univ. Press, London, 1940.
O
15. Archibald, W. J., J. Phys. Colloid Chem. 51, 1204
0"8 (1947).
o 16. Guinier, A., and Fournet, G., "Small angle scattering
cL O QD----oO O of X-rays." Wiley, New York, 1955.
O'6 O O 17. Davis, J. A., and Leckie, J. P., J. Colloid Interface
I I I Sci. 43, 409 (1973).
7.0 8"0 g.0 10.0
pH 18. Sposito, G., "The Surface Chemistry of Soils." Oxford
Univ. Press, New York, 1984.
FIG. 7. The influence of pH on the cation-exchange 19. Alexander, G. B., Heston, W. M., and Iler, R. K., J.
capacity of Laponite XLG. Phys. Chem. 58, 453 (1954).

Journalof Colloidand InterfaceScience. Vol.151,No. 1,June 1992