Colloids and Surfaces, 51 (1990) 207-217 207

Elsevier Science Publishers B.V., Amsterdam

Preparation of crystalline zirconia powders.

Characterization of the bulk phase and interfacial
behaviour

S. Ardizzone a, G. B a s s i a a n d G. L i b o r i o b
aDepartment of Physical Chemistry and Electrochemistry, University of Milan, Via Golgi 19,
20133 Milan (Italy)
bDepartment of Earth Sciences, University of Milan, Via Botticelli 23, 20133 Milan (Italy)
(Received 17 April 1990; accepted 25 June 1990)

Abstract

Calcinations, in oxygen, of ZrC1, and Zr (OH) 4, prepared hydrothermally, have been performed
over a wide temperature range (200-850°C). The product of the reaction was (except for the
850 °C sample) a phase mixture composed, at lower temperatures, of the stable monoclinic and
the metastable cubic and tetragonal forms and, at higher temperatures, of the monoclinic and
tetragonal polymorphs. Subjecting the powders to a severe mechanical grinding resulted, at any
temperature, in the transformation of the phase mixture into a pure monoclinic phase. Some of
the controversial aspects both of occurrence and of "stabilization" at low temperatures of Zr02
polymorphs are discussed. Results concerning the dependence on the temperature of preparation
and the features of the pure monoclinic phases (crystallites sizes, surface area, porosity) are pre-
sented. The paper also reports surface charge versus pH curves obtained on suspensions of the
powders in KN03. The effects induced by the temperature of preparation on the surface electri-
fication behaviour are discussed.

INTRODUCTION

The colloidal properties of zirconia may play a significant role in determin-

ing the final performance of a large class of ceramic materials, since concen-
trated suspensions of the oxide are used as precursors of high temperature
finished products [1]. Besides this, several technological appliances involve
the interfacial and surface properties of zirconia. For instance, the solid be-
haves as a catalyst in dehydration, hydrogenation and in hydrogen exchange
reactions [2 ], and may act as a support for metal catalysts [3]. Further, ZrO2
is present as a corrosion product of metallic zirconium or of zirconium based
alloys in boilers and in nuclear power reactors; consequently, the chemistry at
the ZrO2/fluid interface may be relevant to both the operation of plants and
to the control of activity dissemination [4,5]. These different technological

0166-6622/90/$03.50 © 1990 --- Elsevier Science Publishers B.V.

208

implications are not accompanied by a parallel evolution in the fundamental

research, since results on surface and interfacial behaviour of Zr02 are not
numerous and refer, mainly, to amorphous or non-characterized samples [4-
131.
Recently in this laboratory research on the interfacial characterization of
monoclinic zirconia has been performed using high purity commercial powders
obtained at high temperatures [14 ]. The hydrophobic conditions of the sur-
face, due to the temperature of preparation of the samples, introduced, how-
ever, serious difficulties in the determination of the surface parameters; fur-
ther, in the case of samples fully equilibrated in solution, the parameters of
zero charge (p.z.c. and i.e.p.) became coincident with the literature values re-
ported for bulk Zr (OH) 4. This effect was attributed to the presence of a surface
hydrous layer of Zr (OH)4 which formed upon rehydration of the thermally
dehydrated surface and shielded the inner crystalline oxide core. It is worth
noting that the only literature work dealing with the interfacial features of a
crystalline baddelyite [4 ] reports for the zero charge parameters a value of 6.5
which is typical of Zr ( OH ) 4. In a subsequent paper [5 ], the authors themselves
suggest the possible formation of hydrous layers at the surface of the crystal-
line baddelyite.
On the basis of the above, the present work was aimed primarily at obtaining
zirconia as a pure monoclinic oxide, at the lowest possible temperature, in or-
der to be able to characterize the intrinsic crystalline oxide surface not biased
by rehydration phenomena. The second objective was to study the effects of
increasing the temperature of preparation both on the bulk and on the inter-
facial features of the oxide.

EXPERIMENTAL

All reagents were of reagent-grade purity and used as received without fur-
ther purification. Solutions were prepared with doubly distilled water passed
through a Milli-Q apparatus.

Preparation of monoclinic Zr02

A first series of samples was prepared using ZrC14 as the precursor of the
oxide. The salt was calcinated in 02 at three different temperatures (500, 650
and 850 ° C ) for variable time lengths (4-8 h). A second series of samples was
prepared by calcinating, in an 02 atmosphere, at different increasing temper-
atures (200-800 ° C ), a hydrothermal intermediate. This product was prepared
by a modification of the procedure developed by Breeuwsma [ 15] to obtain the
precursor of hematite. Essentially the reaction is an alkaline precipitation of
Zr(OH)4 from an acidic solution of ZrC14 at the boiling point. At the end of
the precipitation, the heating was continued for 2 h and afterwards the product
 209

was washed with water to constant conductivity, dried at 60 ° C and stored in a

desiccator.
In order to obtain pure phase samples, several attempts have been made by
changing both the temperature and the time length of calcination. The final
procedure adopted consisted of heating the intermediate for 10 h at the given
temperature (200-800 °C) and afterwards grinding the powder severely for at
least 15 min, either manually, in an agate mortar, or by ball milling, having
previously divided the sample into small fractions. This treatment was fol-
lowed by 2 h calcination at the same temperature. Structural characterization
of the powders was performed by X-ray diffraction using filtered Cu radiation.

Specific surface area

Determinations were performed by B E T N2 adsorption using a Sorptomatic

(Carlo Erba) apparatus. Before the measurements dry samples were further
degassed for 2 h at 200 ° C.

Surface charge determination

Surface charge arising from the adsorption of H + / O H - was determined by

(potentiometric) acid-base titrations on aqueous suspensions of the powders.
Titrations were performed with the "fast" technique [16] in the presence of
KNO3. Before beginning the titration the oxide powder was equilibrated with
the base electrolyte for 18 h in an N2 atmosphere. Titrations were, in any case,
performed from acid (pH ~ 5) to alkaline pH ( p H - 9.5) and vice versa at the
same ionic strength. For the lower temperature samples (200 ° C ) some degree
of hysteresis between acid and alkaline runs (maximum ~pH-- 0.15) was ap-
parent only at the highest KNO3 concentration in the region of neutral pH
values. In the case of higher temperature samples some degree of hysteresis (of
the same order of magnitude of the 200 ° C sample) was apparent also at the
lower ionic strengths. Curves in Figs 3 and 5 refer to acid going titrations.
The pH was determined with a 0.001 pH unit resolution using an Amel 336
pH meter connected to separate glass and calomel electrodes. Concentrations
of the base electrolyte were varied by addition of the solid salt.

RESULTS AND DISCUSSION

Monoclinic zirconia preparation and characterization

The monoclinic phase is the only stable crystallographic form of the tetra-
valent zirconium oxide in the low temperature region, starting from the first
onset of crystallinity and up to the transition to the tetragonal phase
(-1150°C).
210

X-ray determinations performed on samples obtained by direct calcination

of ZrC14 proved, however, that a pure monoclinic phase could be obtained only
in the case of samples heated at 850°C; for lower temperatures, other zirconia
polymorphs (500 °C cubic and tetragonal; 650 °C tetragonal) formed together
with the monoclinic phase. A similar trend was shown by samples obtained
through calcinations of the hydrothermal intermediate. Also in this case a mix-
ture of cubic, tetragonal and monoclinic phases forms up to 400 ° C, while, for
higher temperatures, tetragonal and monoclinic phases occur together, no pure
monoclinic phase formed in the temperature range investigated.
The occurrence of metastable ZrO2 polymorphs, in the low temperature re-
gion, is described in the literature and Boganov et al. [ 17] showed that lattice
parameters of low-temperature cubic and tetragonal phases agree well, taking
into account temperature corrections, with the parameters of the high temper-
ature phases. Literature results concerning metastable ZrO2 polymorphs are,
however, often controversial both in relation to the origin of the stabilization,
at low temperatures, of the metastable forms and also in relation to the con-
ditions of occurrence of the different phases. The reactions studied in the lit-
erature regard mainly calcinations of alkoxides or hydroxides in variable tem-
perature ranges, in any case within the range of thermodynamic stability of
the monoclinic phase [18-29]. Some of the divergent aspects relative to the
phenomenological formation of the metastable phases can be mentioned. Only
the present results and those of few other authors [19,23 ] indicate the for-
mation of a third polymorph with the cubic structure at the lower end of the
temperature range, while only the tetragonal form is generally described to
occur, as a metastable phase, together with the monoclinic form in the low
temperature region. Some authors report on the possibility of obtaining one of
the metastable forms as a pure phase by modulating, at low temperature, the
experimental conditions of the reactions [23,25]; in the present and in other
literature cases, on the contrary, calcinations performed in oxygen always
yielded a phase mixture (except for the 850°C sample). The phase composi-
tion of the present samples, further, did not seem to be, at variance with other
results [23], significantly affected by the length of the heating at the same
temperature (10, 30 or 60 h). Performing the calcination of the hydroxide in
an oxygen free atmosphere is described to yield a black non-stoichiometric
oxide with stabilization of the tetragonal form [20].
In the present case calcinations performed in an N2 atmosphere at 200 °C
yielded a pure monoclinic phase in the absence of metastable polymorphs (this
procedure was not adopted as a general routine due to the possible presence of
bulk and surface Zr 3+ defects which might affect the interfacial characteriza-
tion of the samples).
The picture which emerges from the comparison of results from different
sources shows the presence of large differences in the description of occurrence
of the metastable phases. Likewise, debate is still open on the reasons for the
 211

"stabilization", at low temperatures, of the metastable phases and specifically

of the tetragonal polymorph. Garvie's largely accepted assumption [22,26] that
the tetragonal polymorph is formed as a crystal size effect is questioned in the
papers by Murase and Kato [27] and by B lesa et al. [5 ], where no relationship
between crystallite size and crystal structure is reported; further, Mitsuhashi
and coworkers [25 ] proposed, for the formation of the metastable polymorphs,
an interpretation different from Garvie's size effect, based on the concept of a
martensitic transformation. The present occurrence of the phase mixtures could
fit very well in the picture proposed by Garvie as the very small dimensions of
the crystallites (8 nm, 200 oC-22 nm, 800 °C ) are, at any temperature, within
the range supposed to induce stabilization, of the metastable phases, by surface
energy factors [26]. The effect on the phase composition of a severe grinding
treatment does not find ready explanation within Garvie's hypothesis. Results
in Fig. la indicate that a prolonged mechanical treatment of the powders re-
sults, at any temperature, in the transformation of the phase mixture to a pure
monoclinic oxide. These findings, while in agreement with results by Mitsu-

M+T+C M+T
g_ Mechanical treatment

a.

M )

25'

/
d
20 /
/
/
/
/
•"Q 15 /
/
/

/
/
J
10 /

-0 . . . . . ~OS

I I I I
200 400 600 800
T[°C)
Fig. 1. (a) Phase composition of the powders obtained by calcination of the hydrothermal inter-
mediate. The phase mixtures obtained at 500, 600 and 700 ° C, in order to be transformed to pure
monoclinic zirconia, had to be heated, before the grinding, for 50 h (at variance with the 10 h
heating of the other samples. (b) Dependence of the crystailites sizes (by Scherrer's equation ) on
the firing temperature.
212

hashi et al. [25] who observed an identical phenomenon, contrast with the
hypothesis of Garvie who reported, as a support to this theory, the occurrence
of just the opposite phenomenon, i.e., the transformation, by grinding, of a
monoclinic into a tetragonal phase [28].
In the light of the divergent evidences it seems hardly probable that one
single effect should be the origin of the stabilization of the metastable phases;
the present authors are inclined to believe that the occurrence, at low temper-
atures, of high temperature phases should, in any case, be the result of a spe-
cific balance between surface and bulk components and between thermody-
namic and kinetic factors. Further, in the authors opinion, the close structural
relationships between the hydrous oxide and the cubic and tetragonal forms
might play a significant role in the formation of the metastable phases. Nish-
izawa et al. [ 29 ] suggest, in fact, that the transition from the hydrous oxide to
the tetragonal form occurs as a topotactic crystallization mediated by a cubic
polymorph.
Independently of the analysis of the factors which might induce either sta-
bilization or non-stabilization of the polymorphs, Fig. la contains in s e a very
important result. The adopted procedure has allowed, in fact, a pure mono-
clinic phase to be obtained, starting from low temperatures and in a wide tem-
perature range, fulfilling thereby one of the main aims of this work. Figure lb
reports the average sizes of the crystallites of the monoclinic samples obtained
at the different temperatures. The values have been calculated by means of
Scherrer's equation [30 ] on the (111) peak of monoclinic Zr02. From the fig-
ure it can be observed that, at the lower temperature extreme, the sizes of the
crystallites are almost invariant while, for higher temperatures, the dimen-
sions show the expected increase with temperature. This latter increase matches
the observed decrease of the specific surface areas of the powders. Figure 2a
reports, collectively, the values of specific surface areas both experimental,
from BET determinations, and calculated from the dimensions of the crystal-
lites; these latter values have been obtained using the following formula, S--
6/pd where p is ZrO2 specific weight ( ~ 5.8 g cm-3). Examining the figure it
can be observed that, except for the 200 ° C sample, the agreement between the
two sets of data is rather good both in trend and also in the absolute values.
This agreement implies that the effective dimensions of the ZrO2 particles are,
at any temperature, comparable with the sizes of the crystallites, and that the
present ZrO2 powders are mainly composed of single crystal particles with di-
mensions increasing with temperature.
The case of the 200 ° C sample is, instead, difficult to explain, the calculated
surface area being much lower than the experimental one. The difference can
not be attributed to the presence of a higher degree of porosity in the 200 ° C
sample. Figure 2b shows, in fact, that the specific pore volume of the lower
temperature sample is comparable with the porosity of higher temperature
samples, the degree of porosity showing a maximum at 400 ° C. Two possible
 213
J i

a
\ \0
\
180 \
\
\
\
%
E 120 \

0 \0
z~

60 0 ~ ~0

.~ 20]
/

1.0 \\O
~ 0''/
O \

i
2OO 400 600 800
T(°C)

Fig. 2. (a) Experimental ( B E T ) ( O ) and calculated ( A ) (from the dimensions of the crystal-
lites) specific surface areas as a function of the temperature of preparation. (b) Specific pore
volume of the powders versus temperature of calcination.

explanations for the disagreement between experimental and calculated sur-

face areas at 200°C can be suggested: (1) the estimate of the size, through
Scherrer's equation, is not sufficiently accurate, in the case of very small crys-
tallites ( ~ 5 nm, would be the size of the crystallites consistent with the ex-
perimental surface area); (2) the particles at 200 °C are not smooth and the
increase in surface area is related to a high degree of surface roughness. This
latter hypothesis has been suggested by examining the particles using scanning
electron microscopy. The 200°C sample appears, in fact, to be composed by
coarse aggregates of smaller particles which in turn appear to be composed of
smaller particles and so on, so that any distinguishable element presents a
raspberry-like surface. Of course scanning electron micrographs can give just
mere indications of what might occur at dimensions lower than 10 nm.

Interracial behaviour

Figure 3 reports charge versus pH curves obtained in the case of the 200 ° C
sample. The curves present the classical pattern and give rise to a sharp com-
mon intersection point (c.i.p.) at pH 8.0. The "indifferent" role of KNOa, as
the base electrolyte, was previously checked on crystalline baddelyite by cross
comparison of microelectrophoretic determinations and surface charge results
[ 14 ]. With this conclusion the c.i.p, of charge curves in Fig. 3 can be given the
214

-3 '
-2 ~ ' C' b t
~'~.E-1 ~ a
0
~,o
1

3
6
I
7
I
8
I
9[

pH

Fig. 3. Charge versus pH curves, of a 200 °C sample, determined in KNOa: (a) 10-3; (b) 10-~; (e)
10-1M.

meaning of a point of zero charge (p.z.c.). It is interesting to note that this

value compares reasonably well with the value of pHpzc 8.65 which can be ob-
tained using Parks-Yoon equation [6,31] by inserting the structural parame-
ters of the monoclinic phase. It can be observed that the present experimental
p.z.c, is significantly different from the p.z.c, reported both for Zr(OH)4 and
for rehydrated baddelyite samples [4,7,8,13,14]. Differences in the p.z.c, be-
tween hydrated and nonhydrated oxides have been often reported in the lit-
erature [6,31-35]. In the case of MnO2 [34] and RuO2 [35] the p.z.c, under-
goes a significant change in the acid direction in passing from "dry" crystalline
forms to hydrous oxides. Likewise, hydroxides have been reported to present
p.z.c, values more acidic than the corresponding "dry" oxides (CuO, Fe203,
A1203) [31 ]. This seems to be also the case of Zr~V/oxygen compounds: pH 6.5
is, in fact, consistently, reported to be the p.z.c. (or i.e.p.) of Zr(OH)4
[5,8,12,13] and of samples of rehydrated baddelyite [4]. In this latter case
ZrO2 particles are supposed to be covered by a layer of amorphous or gel-like
Zr (OH) 4 formed upon rehydration of the surface [5 ]. The p.z.c, at pH 8 of the
present samples can be considered, instead, to be representative of the intrinsic
surface of the monoclinic oxide not modified by surface rearrangements. The
temperature of preparation of the present samples is, in fact, never so high as
to determine severe dehydration of the surface with consequent formation,
upon rehydration, of external layers different from the inner crystalline core
[ 36 ]. Figure 4 reports the trend of temperature variation for the values of p.z.c.
of the samples prepared at increasing temperatures. The values appear to be
scattered around a straight line and not to show any definite trend of variation
 215

lO

0
8 . . . . . . -o . . . . . . - -- -- "-<)- - - 0 - -- - "(3 -- - ~ D -

i I I I
o 200 400 600 800
T (°C)

Fig. 4. Dependence of the values of the p.z.c, on the temperature of preparation.

with varying the temperature of preparation. This invariance, occurring in a

rather broad temperature range, was not expected a priori, especially because,
in the case of other oxides, marked dependences of the p.z.c, values on the
temperature of preparation are reported [37,38]. In the literature, the p.z.c.
variation was mainly attributed to modifications of the oxide bulk and surface
composition occurring over the investigated temperature range. The present,
apparent absence of trend with the temperature might, consequently, be inter-
preted the other way round, as the result of an invariance of the oxide surface
and {possibly) bulk composition. As a matter of fact, ZrO2 is not an oxide
which is readily prone to assume a non-stoichiometric composition, Zr 3+ de-
fects being induced only by heating treatments in vacuo of the oxide precursors
[24].
The invariance of the p.z.c, does not imply, however, that the variation of
the temperature of preparation is not inducing any modification in the elec-
trification behaviour of the present ZrO2 powders. Examining charge curves
obtained at 200, 400, 600 and 800 ° C, respectively, a regular increase in charge
density can be observed. Figure 5 reports the direct comparison of results ob-
tained on samples prepared at 200 and at 800 ° C. From the figure a definite
increase in charge density, at the same zlpH, is apparent in passing from low
to high temperature. Concerning the absolute values, the density of charge in
the case of the 200 ° C sample is relatively low but comparable with the values
reported in the case of other oxides prepared at low temperatures [39]. Charge
densities, for the 800°C sample, in turn, compare reasonably well with the
density of charge reported by Regazzoni et al. [4 ] for crystalline baddelyite. In
comparison with the charge in the "master curves", drawn by Fokkink et al.
[40] on results of hematite and rutile, the present charge densities appear to
be, respectively, lower (200°C samples) and higher (800°C samples) at the
three ionic strengths. In the figure the presence of a different effect can be
seen: at higher temperatures the curves are shifted at higher charge but the
effect of variation in the ionic strength is much lower the curves being less
separated than at low temperature. This occurrence implies that, with increas-
ing the temperature of preparation, the mechanism of charge formation is pro-
216
-4 i u , i
• D'

/ t/
JH /
//J
-2
/5" c b

z./S /

¢0 0

z / /
/ /
/ t S
/ / /
/ / ¢
I /i I I
4 -2 -1 0 1 2

PH - PHp.z,c.

Fig. 5. Charge curves of a 200 (a, b and c) and of a 800 ° C sample (a', b' and c' ) as a function of
pH, normalized with respect to the p.z.c., in KN03: (a,a') 10-3; (b,b') 5-10-3; (c,c') 10 -2 M.

gressively more sustained by amphoteric reactions than by surface complexa-

tion with the base electrolyte [41].
Besides these trends, which are rather well defined, other less definite effects
became apparent with increasing the temperature of ZrO2 preparation. Hys-
teresis between acid and alkaline going titrations increased at all ionic strengths
and the definition of the p.z.c, became less accurate, the curve at the higher
ionic strength (0.1 M) being somewhat shifted towards more acid pH values.
It is reasonable to believe that the "wetting" of the surfaces may become pro-
gressively more difficult with increasing the firing temperature so that the gen-
eral picture of the acid-base behaviour may become less definite.
Work is in progress, however, in this laboratory to further investigate these
latter aspects, studying the effects induced by varying pH and time of precon-
ditioning and by heat treatments in hydrogen.

ACKNOWLEDGEMENTS

Financial support from the Ministry of Public Education Research Funds

(40 and 60% ) is gratefully acknowledged.

REFERENCES

1 A. Garg and E. Matijevid, J. Colloid Interface Sci., 128 (1988) 243.

2 Y. Nakano, T. Iizuka, H. Hattori and K. Tanabe, J. Catal., 57 (1978) 1.
3 M. Bensitel, O. Saur, J.C. Lavalley and G. Mabilou, Mater. Chem. Phys., 17 (1987) 249.
4 A. Regazzoni, M. Blesa and A. Maroto, J. Colloid Interface Sci., 19 (1983) 560.
 217

5 M. Blesa, A. Maroto, S.I. Passaggio, N.E. Figliolia and G. Rigotti, J. Mater. Sci., 20 {1985)
4601.
6 G. Parks, Chem. Rev., 65 (1965) 177.
7 P. Tewari and W. Lee, J. Colloid Interface Sci., 52 (1975) 77.
8 K.C. Ray and S. Khan, Indian J. Chem., 13 (1975) 577.
9 S.H. Ahmed, Can. J. Chem., 44 (1966) 1663.
10 C.B. Amphlett, L.A. McDonald and M.J. Redman, J. Inorg. Nucl. Chem., 6 (1958) 236.
11 A.J. Shor, K.A. Kraus, W.T. Smith Jr and J.S. Johnson Jr, J. Phys. Chem., 72 (1968) 2200.
12 S. Mattson and A.J. Pugh, Soil Sci., 38 (1934) 229.
13 F.S. Mandel and H.G. Spencer, J. Colloid Interface Sci., 77 (1980) 577.
14 S. Ardizzone, G. Chidichimo, A. Golemme and M. Radaelli, Croat. Chem. Acta, in press.
15 A. Breeuwsma, Ph.D. Thesis, Agricultural University, Wageningen, 1973.
16 A. Breeuwsma and J. Lyklema, Discuss. Faraday Soc., 52 (1971) 324.
17 A.G. Boganov, V.S. Rudenko and L.P. Makarov, Dokl. Akad. Nauk SSSR, 160 {1965) 1065.
18 G. Katz, J. Am. Ceram. Soc.,54 (1971)531.
19 R. Ruh and J. Rockett, J. Am. Ceram. Soc., 53 (1970) 360.
20 J.'Livage, K. Doi and C. Mazieres, J. Am. Ceram. Soc., 51 (1968) 349.
21 G.L. Clark and D.H. Reynolds, Ind. Eng. Chem., 29 (1937) 711.
22 R.C. Garvie, J. Phys. Chem., 69 (1965) 1238.
23 K.S. Mazdiyasni, C.T. Lynch and J.S. Smith, J. Am. Ceram. Soc., 49 (1966) 286.
24 M.J. Torralvo, M.A. Alario and J. Soria, J. Catal., 86 (1984) 473.
25 T. Mitsuhashi, M. Ichihara and U. Tatsuke, J. Am. Ceram. Soc., 57 (1974) 97.
26 R.C. Garvie, J. Phys. Chem., 82 (1978) 218.
27 Y. Murase and B. Kato. J. Am. Ceram. Soc., 66 (1983) 196.
28 J.A. Bailey, D. Lewis, Z.M. Librant and L.J. Porter, Trans. J. Brit. Ceram. Soc., 71 (1972)
25.
29 H. Nishizawa, N. Jamasaki, K. Matsuoka and H. Mitsushio, J. Am. Ceram. Soc., 65 (1982)
343.
30 H.P. Klug and L.E. Alexander, X-Ray Diffraction Procedures, Wiley, New York, 1962.
31 R.H. Yoon, T. Salmon and G. Donnay, J. Colloid Interface Sci., 70 (1979) 117.
32 J.V. Schuylenborgh and A.M.H. S~inger, Rec. Trav. Chim. Pay Bas, 68 (1949) 999.
33 W. Smit and H.N. Stein, J. Colloid Interface Sci., 60 (1977) 299.
34 T.W. Healy, A.P. Hearring and D.W. Ftirstenau, J. Colloid Interface Sci., 21 (1966) 435.
35 S. Ardizzone, A. Daghetti, L. Franceschi and S. Trasatti, Colloids Surfaces, 35 (1989) 85.
36 N.H.G. Penners, L.K. Koopal and J. Lyklema, Colloids Surfaces, 21 (1986) 457.
37 P. Siviglia, A. Daghetti and S. Trasatti, Colloids Surfaces, 7 (1983) 15.
38 C. Pirovano and S. Trasatti, J. Electroanal. Chem., 180 (1984) 171.
39 S. Ardizzone, J. Electroanal. Chem., 239 (1988) 419.
40 L.G.J. Fokkink, A. de Keizer and J. Lyklema, J. Colloid Interface Sci., 127 (1989) 116.
41 S. Ardizzone and G. Bassi, Mater. Chem. Phys., 25 (1990) 417.