Você está na página 1de 17

Adsorption of Hydrolyzable Metal Ions at the

Oxide-Water Interface
III. A Thermodynamic Model of Adsorption
ROBERT

O. J A M E S AND T H O M A S W. H E A L Y

Department of Physical Chemistry, University of Melbourne, Parkville 3052 Victoria, Australia


Received March 18, 1971; accepted August 6, 1971
A detailed quantitative model for the adsorption of hydrolyzable metal ions at the
oxide-water in terrace is presented in terms of the competition between the free energy
changes faavorble to adsorption, i.e., coulombic and chemical energy changes, and the
unfavorable change in solvation energy. Because of tile quadratic dependence of
solvation energy changes on the charge of the ion, this term decreases the adsorption
of highly charged species. As hydrolysis occurs and the ionic charge is lowered, the
coulombic and chemical energy contributions dominate the adsorption energy change
and adsorption is abruptly enhanced.
If the oxide has a dielectric constant similar to that of the solvent then solvation
energy changes are minimal and do not prevent adsorption of unhydrolyzed species.
INTRODUCTION

indicates that aqueous Cr ( I I I ) exists in the


same environment in the pores of silica gel
as in bulk solution. Nortia and Laitinen
(4) have measured the electronic spectra of
transition metals adsorbed in ion exchange
resins and reached similar conclusions. Additional electronic spectra of Ni ( I I ) and Co ( I I )
adsorbed on silica gel have been measured b y
H a t h a w a y and Lewis (5, 6) and there is
little perturbation in the normal aqueous environment of the ions;
d. adsorption is sensitive to the nature of
the substrate in that silver halides, crystalline silica, silicates, and latex particles behave in a similar manner but adsorption on
titania is quite different in character; and
e. adsorbed species to not reverse the
charge of silver halides, silicas, silicates, etc.,
until surface precipitation occurs; unhydrolyzed metal ions do reverse the charge on
TiO2 and a shift in the P Z C of TiO2 is observed.
These essentially negative findings cannot
be resolved b y examining data on adsorption

The literature on adsorption of hydrolyzable metal ions contains m a n y suggestions


that adsorption and hydrolysis are "related."
This conclusion can also be made from the
results of Papers I (1) and I I (2) of the
present series; but, in addition, it was found
that :
a. adsorption cannot depend in any teneral way on the appearance of a specific ionic
species such as the first or other hydrolysis
product;
b. adsorption cannot be due in any general
way to the appearance of specific polynuclear complexes;
c. the adsorbed species, whatever they
m a y be, are separated from the surface b y at
least one layer of water molecules so t h a t
direct chemical bonding is precluded; consequently, the metal ions are not required to
lose their p r i m a r y hydration sheaths. Evidence for the retention of p r i m a r y hydration
sheaths also comes horn E P R line broadening studies b y Cornet and Burwell (3) which
Copyright zc@1972 b y Academic Press, Inc.

Journal of Colloid and Interface Ncience, Vol. 40, No. 1, J u l y 1972

65

66

JAMES AND HEALY

of metal ions in terms of the traditional adsorption isotherms which we can conveniently classify into Gouy-Chapman, Stern,
and Stern-Grahame types. To illustrate this
further, the adsorption data for Co (II) at
10-4 M fixed equilibrium concentration on
SiO2 are plotted;in Fig. 1 as a function of
~b0, the total double layer potential The
conversion from p H to ~b0 is made via the
Nernst equation, i.e.,

~/o = (leT~e)In (a~+/a+ )


[1]
= 59.2 (pH~zc - p H ) at 25C,
where 0 is in millivolts; a~+ and a~+
0 are the
activities of potential determining ions at a
given p H and the p H of the point-of-zero
charge (PZC), respectively.
Curve 1 of Fig. 1 is the calculated isotherm assuming a Gouy-Chapman model for
Co (II) adsorption from a 2:2 electrolyte,
i.e.,

{ 6.0S X

10--11/Z}

[2]

1/2

(Coo~+) {exp (ze~o/2kT)


I

-10
-~ 2.10

]I.

t~
>-

PCO2+

= 2rCco~+ exp {-ze(~0 4- ~ ) / k T } ,

I~CO2+

Here Fco2+ is the adsorption density (mole


em-2), z is the charge on the ion, and C is
the equilibrium concentration (mole liter-i).
The isotherm in this form is very limited
in application except at low potentials and
concentrations. The Gouy-Chapman isotherm can validly be applied by using the
diffuse double layer potential rather than
~b0 and such an isotherm is given by curve 2
of Fig. 1 where zeta-potentials, obtained
from streaming potential measurements (7)
are used to approximate the diffuse layer
potential.
Neither of these Gouy-Chapman isotherms fits the experimental Co (II)-SiO2 adsorption data.
The Stern and Stern-Grahame isotherms
to describe adsorption in a compact rather
than diffuse double layer can be simplified
(8, 9) to:

.lc

s 1~10
~-

-11

+118

(n)/i02

pzc Si02
2
0

///

./

4 pH 6
-118 mV -236

2
8
-354

10
-472

FIG. 1. A c o m p a r i s o n of e x p e r i m e n t a l d a t a ( O )

on Co(II) adsorption on SiO2 at ]0-4 M Co(II)


with various standard theoretical isotherms, viz:
(1) Gouy-Chapman (G); (2) Gouy-Chapman G');
(3) Stern (/0); (4) Stern (/~).
Journal of Colloid and Interface ~cience, Vol. 40, No. 1, July I972

[3]

where is the specific adsorption (i.e., chemisorption) potential of an ion charge z and
radius r. A Stern isotherm, generated by
using ~V0, r = rio, = 0.78 A, and = 0,
is shown as curve 3 of Fig. 1. Again there is
no agreement between theory and experiment.
This Stern isotherm with ~0 as the adsorption plane potential is made more realistic
by using a potential ~bx at some distance out
into the double layer. Curve 4 is therefore
derived from Eq. (3) with = 0, r = x and
~b0 = ~ . The ~alues of ~x for the present
purposes were taken at x = 3.54 A along a
simple Gouy-Chapman potential profile at
10-3 M ionic strength for 1:1 supporting
electrolyte.
Using the zeta-potential in the Stern equation produces an isotherm that falls below
curve 2 of Fig. 1.
If we now use a Stern-Grahame model
with a finite negative value of inserted
into Eq. [3], curves 3 and 4 are displaced
symmetrically to the left (i.e., to lower p H )
in Fig. 1 and further away from the data.

IONS AT OXIDE-WATER

The use of q~ in the Stern equation with


0 equal to the zeta-potential is unrealistic;
however, such a calculation again does not
reproduce the form, position or shape of the
experimental results.
The property that hydrolyzable metal ions
show of increased adsorption, when the bulk
solution conditions are suitable for hydrolysis, has led several workers to postulate
that one or other of the hydrolysis complexes,
e.g., MOIt+ is ~pecificalIy adsorbed at the
solid-liquid interface. If adsorption isotherms were calculated via Grahame's isotherm Eq. [3] by using parameters of the
hydrolyzed metal ion only, e.g., MOH + togerber with a specific adsorption potential
for that ion, 4Mon+, then it would be possible to fit a curve to the data. However, this
does not explain why the unhydrolyzed
species, e.g., M 2 does not adsorb to a comparable extent in the Stern layer at the interface.
The intention of this paper is to present
a model for the adsorption of hydrolyzable
metal ions, in terms of simple electrostatic
ion-solid and ion-solvent interactions, which
will provide a general understanding of the
observed phenomena.
OUTLINE

OF THE

MODEL

The Gouy-Chapman and Stern isotherms


involve only a simple coulombic term and
thus the work to bring an ion charge ze up
to a plane where the potential is ~ is ze~.
Grahame (8, 9) included a specific adsorption potential so that ze~ is replaced by
ze(~ + 4), where ~ is a negative quantity
and is more appropriately termed a superequivalent adsorption potential (10).
The chemical description of ~ has been in
dispute; Grahame's suggestion (8) that 0
represents a covalent "surface" bond has
been opposed by Bockris et al. (10) Andersen
and Boekris (11), Levine et al. (12) and
Barlow and McDonald (13).
In the model proposed by Andersen and
Bockris (11) for superequivalent adsorption
of electrolytes at the mercury-water inter-

INTERFACE.

III

67

face, the following interactions and effects


were considered:
i. metal (eleetrode)-ion interactions;
it. water-electrode interactions;
iii. ion-primary water interactions,
iv. Born energy contributions; and
v. changing dieIeetric of solvent;
where both (iv) and (v) account for changes
in the secondary solvation sheath of the
adsorbing ion. The approach of Levine et al.
(12) and Barlow and McDonald (13) has
been to define 6 in electrostatic terms as a
result of the adsorption of real or discrete
ions.
It is important to emphasize that the
Grahame, the Levine et al., the Bockris et al.,
and the Barlow-McDonald models all result
in superequivalent adsorption where 4,, a free
eneNy, is negative. If a ~ term is to be used
for metal ion adsorption on oxides and if it is
to be used for the free ion and its lower
charged hydrolysis products, then 4 must
oppose adsorption rather than aid it. Simple
electrostatic adsorption will always be favorable for cation adsorption on negative surfaces until the ion is hydrolyzed completely
to anionic complexes; and yet metal ions,
as shown in Fig. 1 and Ref. (1, Figs. 1-12),
characteristically do not adsorb until a critical pH is obtained. The most likely interaction that could prevent adsorption of cations
on negative surfaces is a solvation term.
Changes in solvation energy are expressed
as changes in secondary or primary hydration depending on whether the outer or inner
co-ordination spheres respectively are considered. If we regard the first layer of water
molecules on the solid as electrically saturated and at a very low interfaeial dielectric constant and if the solid is an insulator,
i.e., also of low dielectric constant, then
work must be done to remove part of the
secondary hydration layer of a cation and
replace it by interfacial water of very low
dielectric constant. As already discussed the
primary hydration sphere of the cation is
not perturbed by the field at the surface.
In the present model, therefore, the work
Journag of Colloid and Interface Science, Vol. 40, No. 1, J u l y 1972

68

JAMES AND ttEALY

of cation adsorption is separated into a simple


eoulombic term (zeC/~) and a secondary solvation energy term. Because we are dealing
with real discrete ions and real surfaces, ~ ,
which is the potential due to a smeared out
charge, will be an underestimate of the real
potential (14). Rather than at this stage
using say, a Levine et al. (12) model to
estimate the true potential, the simple ze~
term will be "corrected" where necessary
with a "chemical" free energy term.
The adsorption of metal ions at the solid
oxide-liquid interface is therefore treated in
terms of eompeting energy changes as the
ion approaches the interface. Namely, the
changes in electrostatic free energy, together
with the possibility of short-range attractive
forces as opposed by the changes in the secondary solvation energy as parts of the solvation sheath, are rearranged or replaced.
THEORY
As a variety of soluble complex metal ions
will be formed as solution conditions change,
all possible species must be considered as
potential adsorbates. The free hydrated
metal ion M (H20)~ + is in equilibrium with
all of its hydrolysis products, M (H20)6-~"
(OH) (~-~)+ and ligand complexes,
-

M (H2t))~_~L~

H 2 0 ( *K~> 1\;1 (OH)a(~ -1)+ -]-

a~x(,~-~)+

[5]

aMn+ a x -

All these complexes will in turn be distributed between the solution and the interface, depending on the total adsorption
energy for each type of ion and the number
of sites available in the compact double
layer.
The matrix of competing equilibria involved in such a process is depicted schematicMly in Fig. 2, where X - represents a typical
ligand other than H20 or OH-, and might be
for example, the chloride ion, C1-.
Hence, we may write for each of the species
i of Fig. 2, the change in free energy of adsorption under standard conditions, 1 AGad~
0 ,
as the combination of the change in cou0
lombic energy, AGo,I~ the change in secondary solvation energy AGolv;, and a specific
as yet undefined adsorption energy contribution, AGohem~,i.e.,
aG0ae~ = AGcoul~
o
0
0
+ zXG~ol~
+ AGch~r~i.
[6]
The free energy of adsorption may then be
used in the Grahame equation to calculate
the adsorption density of species i, of bulk
concentration Ci, i.e.,
l~i = 2rhyd C i

H +, [4]

where
*K1 = aM(om(~-l)+.ai~+

[4]

aMn+

and
5~+

K-

(n--y)+

The concentration of these species depends


on the solution conditions and the respective
stability constants. Using the terminology
of Sillen and Martell (15), we consider, for
example, the hydrolysis and complexation
reactions:

~{[n+
~q +

where

Ks MX(~_I)+,

Journal of Colloid and Interface Science, Vol. 40, No. 1, July 1972

exp

/,

[7]

provided that the density is less than monolayer coverages. If both Ci and AGd~ are
large then the complete Stern expression
must be used to evaluate the adsorption
density. In Eq. [7], rhya is the hydrated radius
of the ion.
In order to use this model and before Eq.
[7] can be solved, each of the above listed
1 T h e detailed a r g u m e n t s concerning the s t a n d ard s t a t e are presented elsewhere (16). I n summary, the s t a n d a r d s t a t e for the p r e s e n t s y s t e m
is 1 a r m pressure, 298K, Galvani p o t e n t i a l of the
liquid phase equal to zero, the Volta or outer
p o t e n t i a l equal to zero (i.e., the PZC), infinite
dilution of ions in the liquid phase a n d the dielectric c o n s t a n t of the solid equal to t h a t of the
liquid phase a t infinite dilution.

IONS AT OXIDE-WATER INTERFACE. III


SOLID

69

M(OH) n

II Kso
n+

M(H20)6

.~.ads M(H20)
K,

tl q ,oH- % k,, x-

n-l+
n-l+
M(OH)(H20)5 K~d2s M(OH)(H20)5

IIK2,OH-

n-l+
M X(H20)5

n-l+
Mx(H20) 5

II k2~X-

M(O H)2(H20~4-2+ ..~ds M(OH):2[H20)4


n-2+

n-2+ ~--- MX2(H20)4


n -2+
MX2(H20)4

n-3+
n-3+
M(OH)3(H20)3 ad~ MOH)3(H20)3

n-3+
MX3(H20)3 ~

llk4,X-

tlK4,0 Hn-4+
M(OH)4(H20)2

~
ads
K5

n-3+
MX3(H20)3

n-4+
M(OH)4(H20)2

n-Z,+
MX4(H20)2
~

n-4+
MX4(H20)2

etc.
/lllll[lltlltllllllllll[llllllllllllllllllllILllll~lllllllllllllllll

SOLID

OXIDE

e.g., SiO2

Fro. 2. T h e m a t r i x of competing equilibria i n v o l v i n g aquo, hydroxide, and some o t h e r ligand (X)


complexes, free in solution, adsorbed at an interface, or i n v o l v e d in p r e c i p i t a t e formation.

energy contributions must be evaluated for


each of the species in terms of the experimental parameters such as pit, ionic strength
dielectric of the solution, ionic properties
such as ionic charge, crystal and hydrated
radii, and hydrolysis constants, and finally
substrate properties such as the pHrzc and
the dielectric constant of the solid.
The Change in Secondary Solvation Energy

The calculation of the change in the secondary solvation energy due to displacement of the secondary solvation sheath by
the interfacial solvent and the solid is rather
complex because the values assigned to certain variables in the expression depend on
the particular model chosen to represent the
interfaeial region. However, the general form
of the equation is independent of these variables and hence further analysis is worthwhile.
The free energy required to establish a
field in a continuous dielectric medium about
a spherically symmetric ion (11) is given

(joule per ion) by

G0 = ~lff.(.

X.DdV

[81

where/), the electric displacement vector, is


given by D = ze/4~p 2 and J(, the electric
field vector is given by X = ze/(4~re0epp2)
where ep is the dielectric of the medium distance p from the ion and eo is the electric
permittivity of free space which has a value
of 8.85 X 10-I2 coulomb v -~ m-I in SI
units. Thus if the ion was removed from the
medium of continuous dielectric eo.i and
and placed in another medium of dieIectrie
ep.s then the change in free energy is given
(11) now in polar coordinates as,

z2e2
AG- 2~) 2

ff sin O
=o

_o

=,coP 2

[9]

Born (17) used this approach to estimate


the contribution of secondary solvation
Journal of Colloid and Interface Science,

Vol. 40, No. I, July 1972

70

JAMES AND HEALY

energy to the total hydration energy of


aqueous ions. This was done b y estimating
the energy change when an ion was removed
from a vacuum, dielectric st = 1, and placing
it in water, dielectric Ef = 78.5. Since the
first layer of water molecules adjacent to the
ion is essentially electrically saturated the
lower limit of integration for the radial coordinate was taken as the hydrated ionic
radius, i.e., P = rlo~ -k 2 ~ ' w a t e r For greater
values of P the dielectric was assigned the
approximately constant value of 78.5.
Hence, on integration and substitution of
appropriate values we obtain the equation

z2e2N
AV0econdary

hydratlon

--

8~(r~o. + 2r~)~0

[~0]

0 Laidler et al. (18-21) have studied the hydration of ionic and dipolar species in detail
and have estimated hydration energies for
m a n y chemical species. Results indicate that
the Born equation is of reasonable accuracy
for estimating the secondary hydration energy but calculations of the absolute hydration free energy, i.e., including the first
of water molecules around the ion, must be
corrected for variation of the dielectric of
water as the field strength varies for the effective ionic radius in solution and the orientation and structure of water.
Using this Born equation to calculate the
secondary solvation energy for Co 2+ and
CoOH + b y substituting the values rion =
0.78 X 10-1 m and rmo = 1.38 X 10-1 m
and e~(mo) = 78.5 the values obtained are
0
zXGco~+
--- - 7 . 7 X 10~ joule mole -I and
~G 0coo~+ = 1.9 -t- 10~ joule mole -1. Reduction in charge greatly reduces the magnitude
of the secondary solvation energy.
When an ion moves to an interface, work
must be done to remove the secondary solvation sheath. The change in solvation energy
is not as great as in the above example because only part of the secondary solvation
sheath is replaced b y the solid and the interfacial adsorbed water molecules.
Journal of Colloid and Interface Science, Vol. 40, N o . 1, J u l y I972

In the region adjacent to charged interfaces electric fields of considerable magnitude exist and the magnitude of these fields
depends on the charge and potential of the
surface and the ionic strength of the medium. If these fields are large enough, that
is, greater than 10s v m -1, the water molecules adsorbed on the interface will undergo
partial or complete electrical saturation, in
which case the dielectric of these molecules
is reduced from the bulk value of 78.5 at
25C to values which may be as low as 6 if
the field strength exceeds 109 v m -1 The
variation of the dielectric of water as a function of distance from the interface for varying surface potentials is shown schematically
in Fig. 3. I t is necessary to choose some such
model of the interface to set the limits of
integration for use in Eq. [9] so as to compute the change in secondary solvation energy as the ion moves to the surface. As well
as requiring the integration limits, the value
of the dielectric ~ of the interracial water is
also required in terms of the field strength
and distance from the interface.
Several models which allow derivation of
expressions for Ep in terms of the electric
field strength, which in turn is a function of
the potential and the distance from the interface are reported in the literature (22-29,
11).
The equation used b y H u n t e r (29) for

.c:

-buik
- . ~, . _ .

80

(*)
60

2 401" 2 / ! !3

~3

i
i
I

20

3 I
.----4

I"-<
0

Distance,

I2
~

12

Distance,

FIG. 3. Actual continuous (A); a n d approxim a t e discontinuous (B) r e p r e s e n t a t i o n s of the


v a r i a t i o n of the dielectric c o n s t a n t as a f u n c t i o n
of distance for: (1) a surface of low electric field;
(2) a surface of high electric field; (3) a h y d r a t e d
ion, or v e r y high electric field.

IONS AT OXIDE-WATEI~ INTERFACE. III

71

each of these regions. Let these regions be:


i. the solid;
ii. the layer of adsorbed water at the
interface; and
iii. the water outside the adsorbed layer.
There will be a discontinuity in the dielectric
of the medium at each boundary of the three
regions as shown in Fig. 3B.
The interfaeial dielectric e~ has an upper
limit of 78.5 at low field strength and a lower
limit of 6 at high field strengths. Another
useful approximation is to assume that the
hydrated metal ion is surrounded by one
layer of electrically saturated water molecules and that outside this layer the water
molecules have bulk dielectric properties.
Hence, the effective radius of the hydrated
Ebulk -- 6
ion
used in the solvation energy calculations
+ 6.
[11]
e~ = (1 -1- B(d~/dx)~ 2)
is equal to (rio~ + 2rw). The discontinuities outlined above now define the limits of
The constant B has a value of 1.2 X 10-~7
integration
to use in E q [9] for calculating
m 2 v --2 in SI units. The electric field strength
(d~/dx)~ can be estimated from the Gouy- the change in solvation energy of the adsorbing ion.
Chapman model of the double layer. This
Because of this artificial division in the
approach of using a diffuse double layer
model there are two general locations of the
model is clearly not rigorous, although it
hydrated metal with respect to the interface
will lead to reasonable values for the variawhich give two different sets of integration
ables provided the concentration of eleclimits. The first is for the ease where the pritrolyte and the surface potential are not too
mary hydration sheath of the ion and the
high. A more appropriate Stern model for
water adsorbed on the solid do not overlap,
example (30) is not warranted at present but
as is shown in Fig. 4a. The second ease
will be required in future treatments.
covers the possibility that the primary hyAlthough it is possible to obtain an imdration sheath of the ion may include the
plicit continuous function for the variation
(see Fig. 3A) of the dielectric constant as a
(a)
function of the distance from an ion by the
Andersen and Bockris and Laidler et al. equations, the expression for the dielectric of
water as a function of the distance from an
interface is rather more complex. Also bex~
cause of qnantization of energy on the atomic
scale, this approach is no more rigorous than
X ~
ION
using a discontinuous model of the interface.
Instead of using a well-behaved continuous
FIG. 4. Representation of the solid-liquid interfunction for e~, it is possible to make an face showing two of the possible locations of an
equally good approximation to the system by adsorbed ion: (a) in the diffuse double layer; and
(b) near the inner Helmholtz plane. In the latter
dividing the interface into three regions and ease considerable rearrangement in the solvation
using a constant value of the dielectric in sheath of the adsorbed ion takes place.

low fields is convenient for approximate estimates of the interracial dielectric, but since
its use is limited to low fields only, it cannot
be used in the computation in the present
model. The equation of Saeher and Laidler
(28) which uses the value n2 for the high
field limit where n~ is the optical refractive
index, can be modified to obtain agreement
to the experimental value if n~ 2 is replaced
by 6. This equation then gives similar results
to that of Andersen and Boekris (11) with
the advantage that is is a simpler expression.
Thus, the expression chosen for the present purposes to calculate the dielectric of
the interracial water is

Journal of Colloid and Interface ~cience, Vol. 40, N o . 1, J u l y 1972

72

JAMES AND HEALY

adsorbed water on the solid, i.e., the primary hydration sheath and the adsorbed
water overlap as shown in Fig. 4b.
For ease 1 (Fig. 4a):
AG~I~ - 32~eo

~o Jo=o P=,o/co~O

sin0 ( 1

1)

ddodP
[12]

-t-~

=o o=o P=(r~+2rw)/o~0

p2

,)

eb:lk d~ dO dP

es~ia

z2e2 ( 1

1 )

32L;r o
ze

[13]

+ 32 (ro + 2r ) 0
joule ion -1.
For case 2 (Fig. 4b) :
--

--z2e2 L 27r L . . . . . . . Irin+2rw


32~r%o =o o=o

AG~l~

(re-t-2rw)/eos 0 sin

p~rion+2rw

p2

(1 1)

;i-rig

ebulk

d e dO d P

z2e2 f2] f ,2

f (r6-t-2rw)]Cos0 sin
i

LP=re/eOS 0

[141

p2

(1

Specific 'Chemical' Interactions

{~nt

=o J o=o

p2"

The present model has tacitly assumed


that the solid surface and the solution have
a smooth continuous charge distribution.
While this assumption was necessary to
evaluate the coulombic and solvation energy
charges, it does not include other energy contributions such as the attractive image and
dispersion forces described by Andersen and
Bockris (11). The adsorption of nonionic
species (36) is one instance where attrac-

Ebulk

Jr- ~

e~o~ia

z2e2 (

P=(r~+2r~)/cos e

d~ dO dP

eb:lk
1

1 ~ o \rlon q- 2rw
_

If the distance r~ is taken as rlon, then


Eq. [15] gives the change in secondary solvation free energy in moving from the bulk
solution to the inner Helmholtz plane (IHP)
at the solid-liquid interface.
It should be noted that the second term
Eq. [15] is effectively independent of the
solution parameters so that work has to be
done to place an ion at the I H P even at the
pH = pHpzc condition when the surface is
neutral and no electric field exists provided
esomd iS less than ~ (i.e., 78.5). ~
Using Eq. [15], the change in secondary
solvation energy for the adsorption of a free
aquo metal ion or any of its hydrolysis or
ligand complex products to some position in
the interface can be estimated from the following parameters: the pH, the pHpzc , the
dielectric of the solid, esolid, the ionic radii
and ionic charge, the ionic strength of the
solutions and the bulk dielectric of the solvent, in this ease, water. This leaves only the
location of the equilibrium position to be
determined. This could be obtained with
some difficulty by minimizing the total free
energy of adsorption of any ion. Since at
this stage we have not used a continuous
function for the variation of dielectric constant with distance, such a minimization procedure has not been included and we have
assumed that the equilibrium distance is the
IHP.

r~ )(1

2(rio. -t- 2r~)2

ze
-Jr- 327reo(r~ q- 2r~)

1)

-ei,.~ ed,k

[151

e~oUa

ei~

joule ion -1.


Journal of Oolloidand Interface Science~Vol. 40, No. 1, July 1972

2 It should be emphasized that for cation adsorption we have defined the IHP as the locus of
centers of hydrated cations containing a primary
hydration sphere only, whereas the OHP is the
locus of centers of fully hydrated (primary and
secondary) cations.

IONS AT OXIDE-WATER

tire noneoulombic interactions are shown to


exist.
If the surface consists of discrete sites,
several specific coordinate interactions may
be possible. These would include the attractive short-range dispersion forces and 'hydrogen bonding' in the interfacial region. It
is possible that the hydroxyl groups in the
coordination sphere of the hydrolysis products of the ion may act as bridging ligands
between the adsorbed metal ion and adjacent
surface groups. This would be an additional
attractive interaction to be included with
those outlined above. However, it is not
clear to the present authors why hydroxyl
ions in the interface should function any
more favorably in this respect than the coordinated water molecules and again it is
unlikely that there is any change in the first
layer of interfaeial water or the inner coordination sphere of the cation.
Stability constants for the reaction between soluble silica or the silicate anion and
metal ions are reported in the literature (15);
and ion exchange of metal ions on silica has
been studied under nondouble layer conditions of very high ionic strength. These
studies show that the metal-silicate bond is
slightly stronger than the metal-water bond,
consequently there may be a small but significant free energy change which favors the
formation of silica-metal bonds in aqueous
solution. The magnitude of the specific interaction for 20 metal ions with silica gel
has been measured by Dugger et al. (31)
under conditions when hydrolysis is negligible and typical values range from 0 keal
mole-~ bond -~ for K + ions, through - 2 . 0
kcal mole-~ bond -~ for Co 2+ to - 7 . 7 keal
mole-~ bond -1 for Fe ~+.
Although the suggested mechanism is not
well established, it must be recognized that
short-range attractive interactions of one
kind or another do exist, even for unhydrolyzed species. In Ref. (I), specific adsorption potentials of approximately -2 and
5 keal mole -I for SiO2 and Ti02, respectively, were obtained. The present approach
is to use a simple eoulombic term and a

INTERFACE.

III

73

secondary solvation term to predict the pH


of abrupt increase in adsorption and then to
use a AGehem term to obtain better fit in
terms of the amount adsorbed. This procedure does not significantly affect the prediction of the pH of adsorption.

The Adsorption Isotherm


To obtain the total adsorption density at
any given pH, total metal concentration and
ionic strength due to all species, the sum of
of the individual adsorption densities is computed. For example, the adsorption density
Fco/H) = Fco2+ q- Fco~om+ -~- Fco(om20

+ Fco~om~- q- Fco(omd-

[16]

Let the surface equilibria for each species be represented by the equation:

(1 - o) + M~(- 5 ~

0i,

[17]

where M~ is the equilibrium concentration


of metal species i, 0~ is the fractional coverage of surface sites by species i, 0 is the
fraction of surface sites covered by aquated
metal ions i, i.e., 0 = ~ i 6i and Ki is the
equilibrium constant given by
K~ = exp (--AG~d~jRT).
From Eq. [17]
K~ = 0~/M~(1 - 0),

[18]
[191

and since
O=

EOi,
i

then

0 =

~K~M~
1 + ~ i K~ Mi

[20]

K~ M~

[21]

and
9i =

1 + ~ i K~ M~
Thus tile total (fractional) coverage, O,
of the total number of surface sites S may
be calculated. The estimate of the number of
surface sites can be made in either of two
ways, by using the surface density of surface
groups or charges, or by calculating the
number of hydrated species to form a close
Journal of Colloidand Interface Science, Vo], 40, No. 1, July 1972

74

JAMES AND ItEALY

packed monolayer. Since we have already


obtained such "monolayer" data (1), the
latter estimate has been used.
Using this formulation the adsorption
density at constant equilibrium concentration at varying pH values may be calculated, or a similar procedure can calculate
the fraction of metal adsorbed from the total
amount of metal added to a system as a
function of pH.
This analysis tacitly assumes that the free
0 , is indepenenergy of adsorption, AG~d,
dent of the fractional coverage, 0. This
approximation is appropriate if the ions or
molecules are considered as hydrated, i.e.,
having a relatively large area per adsorption
site as compared with the area that would be
covered by the bare or free ion itself

Summary of Model
The final form of the model is given by the
following set of equations:
Oi =

K~M~
1 q- ~_,i K~Mi'

where
Ki = exp

/,

and
AGds~ = AGo.h q- AGolvl -}- AG0h~m~.

The energy terms in SI units as shown are


calculated from:
i.
AGo.,~ = ziF,5~ joule mole-1,
using

A~-

ii.

= (z~e2N~ (

2(rio. + 2r~)~

;~

~u:~

1)

+
\32~e0,] \ r i o n -1-

joule mole -1,


using
ebulk - - 6

1 + (12 ~ i6=I~)(d#dx)~2/ + ~
and
&b

dx

2K

RT

sin h

( ~ 4

\ 2RT /

vm

-~

'

iii. Values of AG0themare available from our


earlier adsorption studies (1) for Co (II) on
SiO: and Ti02. For F e ( I I I ) , C r ( I I I ) , and
C a ( I I ) , we have selected reasonable values
for zXGhem, which are similar in magnitude
to those obtained by Dugger et al. (31). So
as not to prejudice the role of any particular
hydrolysis product the value used or chosen
0
for AG,ho~ is the same for all species for a
given metal
The equilibrium concentrations of all the
metal ion species are calculated using the
hydrolysis and complex stability constants
of the particular metal ion and the p i t and
ligand concentration of the solution
C O M P A R I S O N OF T H E O I ~ Y W I T H
EXPERIMENT

2RT
zF
((e "F~/2"r q- 1) q- (e~e~nRr - - 1)e-""~
r -t- 1) - - (e ~ / 2 R r
1)e-"~/'

In \ ( e ~ 0 / 2 R

where

Co = 2.3RT/zF (pHrzc -- pH) volt


K = 0328 X 101(I) 1/2 m -1
x = (rio, q- 2r,)
m
Journal of Colloid and Interface Science,

zi = sign and number of charges on the adsorbing ion


z = 1 (for 1:1 background electrolyte).

Vol. 40, No 1, July 1972

The predictions of the present theory are


now compared with experimental data for
i. F e ( I I I ) , C r ( I I I ) , Co(II), and C a ( I I )
on SiO2 with each metal at approximately
10-4 M (Fig. 5).
ii. Co (II) on SiO2 at various concentrations (Fig. 6).
iii. Co (II) on TiOo at three concentrations (Fig. 7).

IONS AT OXIDE-WATER INTERFACE. I I I


The fixed parameters used in these eMculations are summarized in T a b l e I. The
radius of the water molecule ( r ~ ) ~ a s taken
as 1.38 A a n d the dielectric constant of
w'ater (eb~lk)is ~8.5.
~ h e p H of the abrupt rise in adsorption
is not significantly, influenced b y the choice
o

"
"
defined more precisely in future studies (16)
is necessary to counteract the obvious under-

--

theory

~o

Fe(m

(m)

"

" o
' ~

'

cain),

Co(~),

o
29 i

4 "

pH

'

's

'

~0 '

~2

FIG. 5. Comparison of the experimental adsorption data for Fe(III), Cr(III), Co(II), and Ca(II)
on SiC2 as a function of pH with the amount adsorbed calculated from the model.
100

ii o

( : ~

a /,

2"OxIO3 ( e )
expt.

o//+//

@ 3.Ox I() M ( d )

~_

80;
o~ [
~_ r
oC~ 60 i[
~
i
; ~o~
~ ~0

'

C~OI

theory

a2
a-m 2 0 i

""
+

3
)1

expt. '
o,~,o-~
~i.~o-%
OMxlO-3M
.MxlO
M

iheory
~,

75

J/"

,j

/
rb~/'"
"1
(c)
OO/ /b
/~lb
C
~c)
~[]
= ~'
]J-PP'
o//'l' / " I
o

o//

i
t

IO

12

Fig. 7. Comparison of the experimental adsorption data for Co(II) on TiC2 as a function of pt-I
at various added concentrations with the amount
adsorbed as predicted by the model.
estimate in ~ and a probable overestimate
of aG~o~. Without it the percentage adsorption (or adsorption density) is underestimated.
The operation of the solvation term can be
seen b y the way the F e ( I I I ) and C r ( I I I )
data are separated and b y the way Co ( I I )
and C a ( I I ) are also separated. However
when Co ( I I ) and Ca ( I I ) adsorption on TiO~
are compared via eleetrokinetie data (2, 32)
there is little separation. This is a result of
the high dielectrie constant of TiC2 which
leads to negligibly small changes in solvation energy.
The effect of increasing the concentration
of the metal ion is to displace the theoretical
and experimental isotherms of Fig. 6 and 7
to higher p H values. Although the theory
predicts a smaller effect t h a n is observed,
the shifts are in the right direction and of
similar (relative) magnitude.

~pptn,

DISCUSSION
0

6
pH

10

12

FIO. 6. Comparison of the experimental adsorption data for Co(II) on SiC2 as a function of pit

at various added concentrations with the amount


adsorbed as predicted by the model.

The overall features of the experimental


data are well reproduced. To examine more
detailed predictions, each of the parameters
for the solid oxide, the solution, the solute,
and the interface must be discussed.
Journal of Colloid and Interface Science, Vol. 40, No. I, J u l y 1972

76

JAMES AND HEALY


TABLE I

F I X E D PARAMETERS IN COMPUTATION OF ADSORPTION ISOTHERMS OF M E T A L IONS ON OXIDES AT 2 5 C

Metal (M)

rion

AGchem (kcal mole-t)

Ionic
strength(M)

k2~

Solid parameters

Fe(III)
1.2 X 10-4

0.64 (A) - 8 . 5 Fe 3, FeOH 2~,


etc.

10 - 2 3 7 . ;

Cr(III)
2 )4 10-4

0.69

--7.0 Cr 3~, C r 0 H 24,


etc.

10-~/30.: 4.1 5.~

Co(II)
1.2 X
1.2 X
1.2 X
3 X
2 X

0.78

- 6 . 5 Co 2~, CoOH +,
etc.

10-6
10-~
10-4
10-4

4.(

lO_~114.~/

9.6 9.~

3 X lO-~l
1 10 -~
1 10 -2

0.99

--7.0 Ca s~, CaOH ~,


etc.

ii .! SiOs
75 m s liter -1
~solid ~ 4.3
pHp~ = 2.0
Ref. (2)

I lO-a /

10 -8

Ca(II)
1.4 X 10-~

2.~

10-s 5.4 2.6

SiO2
75 m 2 liter-'
esolid ~ 4.3
pH,zo = 2.0
Ref. (2)
SiO2

75 m "~liter -i
esolid ~ 4.3
pHp~o = 2.0
Eel. (2)

Si02
75 m 2 liter -i
esolld = 4 . 3

pHp~ = 2.0
Ref. (2)

Co(II)

0.78

-4.0

Co s+, CoOH ~,

14.c 9.6 9.2

etc,

1.1 X 10-s

10-3

1.1 X 10-4

10-8

1.1 X 10-3

10-s

TiO2 , 115 m 2
litre-1
pHp~ = 5.50
Ref. (2)
pItpz = 5.75
Ref. (2)
pHpzo = 6.25
l~ef. (2)
solid ~

Ebulk

78.5

Oxide Parameters
T h e isoelectrie p o i n t or t h e p H of t h e P Z C ,
is t h e p r o p e r t y of t h e oxide w h i c h g o v e r n s
the surface charge a n d the surface potential,
~b0. T h e P Z C is n o t a n a d j u s t a b l e p a r a m e t e r
a n d m u s t b e e x p e r i m e n t a l l y d e t e r m i n e d for
e a c h oxide, a l t h o u g h t h e e m p i r i c a l f o r m u l a
of P a r k s (33) m a y b e of u s e i n c a l c u l a t i n g
t h e p H e z c F r o m t h e c a l c u l a t e d sm'face p o t e n t i a l ~0 a n d u s i n g t h e ionic s t r e n g t h d u e
to b a c k g r o u n d e l e c t r o l y t e a n d a s u i t a b l e dist a n c e x i n t o t h e d o u b l e layer, t h e p o t e n t i a l
Journal of Colloid and Interface Science, VoI. 40, /qo. 1, July 1972

at that plane ~ may be calculated. The


lower the value of pHezo, then the more
negative is the potential providing that
and x are held constant. Hence as the pI-I is
increased, cation adsorption becomes more
favorable.
The dielectric of the solid is the property
which makes the major contribution to
changes in the secondary solvation energy as
the ion approaches the interface. The dielectric must also be experimentally determined; it is possible, however, that due to

IONS AT OXIDE-WATER INTERFACE. III


formation of gel-like amorphous layers at the
surface (34, 35) the dielectric in this region
may differ from the value measured for
single crystals or dry powders. If the electric
field in the interface is small, i.e., at pH
values near the pHpzc there is little dielectric saturation of the adsorbed water molecules and hence, the change in solvation
energy under these conditions is governed by
the dielectric of the solid and the charge on
the ion. For insulating oxides, i.e., low e~o~a,
then AG,0o~ is of considerable magnitude.
For semiconducting oxides for which E~o~a
has much higher values or for metals with
an infinitely large dielectric constant, the
change in solvation energy is small and the
coulombic and specific chemical interactions
dominate the free energy of adsorption and
hence, the shape of the adsorption isotherm.
These predictions compare well with the
measured adsorption isotherms for Co (II)
on silica (e~olldsio~ parallel to optical axis
4.27 and e~o~idslo~ normal to optical axis 4.34
and pHpzc about 2.0), and for Co(II) on
titania (e~oUaTio~ parallel to optical axis 170
and e~ol~aT~o2 normal to optical axis 86 and
pHpzc about 5.6), shown in Figs. 6 and 7.
The surface area used in the experiments was
75 and 115 m 2 liter -~ for the SiO2 and Ti02,
respectively.
The fractional adsorption of Co (II) on
the silica substrate is low at low pH values
(low potentials); the adsorption then increases abruptly in the pH range 6.0-7.0.
It is proposed that this delay in adsorption
is due to the large change in the solvation
energy contribution which opposes the cou]ombic and specific interactions. The adsorption does increase when the solvation energy
is lowered by increasing the importance of
the lower charged hydrolysis products as
the pH is increased. Again, as the pH is increased, the coulombic attraction of a given
ion is increased.
Thus the adsorption of metal ions on
oxides, silicates, and glasses which have
similar dielectric properties to silica occurs
when the solution conditions are suitable for

77

some hydrolysis of the particular ion. Hence,


the empirical relationship, between the fractional adsorption and the pH at which hydrolysis is important, i.e., pH >_ p'K1 may
be rationalized. However, when titania is
the substrate, significant adsorption occurs
in the pH range 5 to 6 when the surface has
almost zero charge; adsorption on Ti02 increases abruptly as soon as the surface
acquires a negative charge as the pH is increased above the PZC. In this case, the
change in the solvation energy is minimal
and may even be of slightly attractive nature
rather than repulsive as for silica. Specific
interactions of the metal ions with the TiO~
surface and the coulombic interaction dominate the adsorption isotherm.
Solution Parameters

The pH of the solution is the parameter


which determines the surface charge and
potential of the oxide; and this solution
property provides the mechanism through
which the coulombic interactions between
the ions and the surface may be controlled.
At constant ionic strength, the electric field
d@/dx increases with increase in potential.
This means that electrical saturation and
depression of the dielectric constant of the
interracial region also increases with the pH,
and hence, the change in the secondary
solvation energy increases with increasing
pH.
The ionic strength, which is normally
fixed by using a fixed amount of 1:1 background electrolyte in the adsorption system,
controls, together with the potential, the
charge density in the diffuse double layer.
If the ionic strength is increased, at constant pH, by adding more electrolyte, the
electrical double layer around the colloid
particle is compressed. The electric field,
d~,/dx, at a given point, is increased and the
electrostatic potential at the same point
is decreased. Increased field strength ultimately results in higher solvation energy
contributions to the free energy of the process, and the eoulombie energy is also deJournal of Colloid and Interface Science, VoL 40, No. 1, J u l y 1972

78

JAMES AND HEALY

creased by reduction in potential, which


together reduce the adsorption of a given ion
as the background electrolyte concentration
is raised.

Solute Parameters

increasing the adsorption density. This concept of variation of free energy of adsorption
with average ionic charge is illustrated
schematically in Fig. 8 for Co (II) adsorptiou
on SiO2 at pit values 4, 6, and 8 in Fig.
8A, B, and C, respectively and for Co (II)
on TiO2 in Fig. 8D. The principle feature of
0
0
the SiO2 data is that AGcoal and AG~olv
together produce a total free energy of adsorption that passes from positive to negatire as the charge per cation is decreased.
Notice that at pH 4, where CoOH + is not
present, the predominant species Co 2+ cannot adsorb; at pH 8, AG~a~ is just positive
for Co 2+ adsorption but strongly negative
for C o O H + . The solution concentrations of
both species are similar so that CoOIt +
adsorption predominates. It is thus the statistically weighted mean charge per species
that controls the adsorption where the
averaging is controlled by Eqs. [18] and
[19].
The term AGhom has not been included
in Fig. 8. It simply shifts the AGhom curve
uniformly up and does not affect its shape.
The representation of the combination of
the free energy terms may be extended to
other solid-metal ion combinations. As
rutile, TiO~ has a comparatively high dielectric, approximately the same as water, the
change in solvation energy for the adsorption of an ion onto TiO2 is small compared
with the change for the adsorption onto
silica. Thus the coulombic and specific interactions are responsible for adsorption, as
shown in Fig. 8D where now AGol~ is in
fact slightly negative. The ease of metal ion
adsorption on metals will be of similar form
to that computed here for TiO2. The funda~
mental difference between metal ion adsorption on insulating oxides like SiO2 compared
with adsorption on metals or high dielectric
constant oxides like TiO2 resides in the
present model in the solvation energy term.

The ionic crystal radius of the metal ion,


r~o,, together with the diameter of the water
molecule, 2r~, determine the distance between the oxide surface and the inner Helmholtz plane. Thus, for large ions, the IHP
is located further into the solution and consequently the adsorption potential ~ decreases with increasing size of the hydrated
metal ion. The expression for the solvation
energy change includes both the ionic radius
and the reciprocal of the hydrated radius.
Thus the solvation energy term decreases
with increasing ionic dimensions. Generally,
if the charge of the species is fixed and the
ionic dimensions are increased the sum Of the
eoulombie and solvation energy changes becomes slightly more negative, i.e., more
favorable to the adsorption process.
The other important property of the metal
ion is its ability to hydrolyze. The reduction
of the average ionic charge z as hydrolysis
proceeds, reduces both the coulombic and
solvation energy changes. However, because
the eoulombie energy varies linearly with
charge z at constant potential ~ , and the
solvation energy change varies with z2, the
decrease of charge due to hydrolysis reduces
the magnitude of the solvation term much
more than the coulombic term. Thus, the
sum of the coulombie and solvation energy
changes may become negative, i.e., favorable
to adsorption for the lower charged species,
while for the higher charged unhydrolyzed
ions the sum of the eoulombic and solvation
energy terms is often positive or unfavorable
to adsorption. Hence adsorption may occur
in the inner Helmholtz plane even without
the inclusion of specific interaction terms,
AGch
0 ..... When the specific interactions, if
Interracial Parameters
any, are included in the summation of the
energy contributions, the total energy of adThe thickness of the interracial region

0
sorptlon
AG~a~
is made more negative, hence where the adsorbed water molecules may
Journal of Colloid and Interface Science, Vo]. 40, N o . 1, J u l y 1972

IONS

AT

OXIDE-WATER

INTERFACE.

III

79

ness of 2.76 X 1 0 - 1 m m a y be a n u n d e r e s t i m a t e a t high e l e c t r o l y t e c o n c e n t r a t i o n s


a n d a t high surface p o t e n t i a l s b u t u n d e r t h e
u s u a l conditions of t h e a d s o r p t i o n experiments, it is t h o u g h t to be a r e a s o n a b l e
estimate.

be p a r t i a l l y or c o m p l e t e l y e l e c t r i c a l l y satur a t e d in response to t h e existing electric field,


is a s s u m e d to be t h e d i a m e t e r of t h e w a t e r
molecule, 2.76 X 10 -1 m. T h i s is t h e s a m e
d i s t a n c e as u s e d b y B o r n in his c a v i t y app r o x i m a t i o n to a h y d r a t e d cation. T h e t h i c k i

Co (ll)/SiO 2

-40

Co(ll)/sio 2
4O

(A)

(b)

-30

tu

-30

A G c ~

o
-20

-20

E
9
m

-IO

-IO
d$.

O
>O
r
w
z

z"',o

IO
20

uJ

20

solv.
30
40

% : -I18 mY./ ~x : -llOmV.

30

pH 6.0
IO-3M e l e c t r o l y t e .

40

~0:-237~v, f :-183 ~v.

IONIC

CHARGE, z.

IONIC

CHARGE, z

Co(ll)/TiO 2

Col, l l ) / S i O 2
-4C

-40

(D)
='-" -3C

-'-"

-30

-20

-2C

-IO

-IC

i
v
0
uJ
z
m

>-

:o,v

IO
20

30
40

o
A G a d s . ~
o AGcul.
AGsolv.

~o
uJ

20

pH 8.0
IO-3M e l e c t r o / y t e

30

+ 0 : - 3 5 5 mY, ,Ibx: - 2 0 4 mY.

40

pH

7.0

}bo:-89

mV ,

~.,x:-84 inV.

[
IONIC

2
CHARGE, z.

IONIC

CHARGE, z.

FIG. 8. The total free energy of adsorption of the ionic Co (II) species obtained from the suinmation
of the changes in coulombic and the secondary solvation energy at the oxide-water interface. (A) For
SiO2 , pHvzc = 2.0, solution pH 4.0 and 10-3 M ionic strength. (]3) For SiO2, pH Pzc = 2.0, solution
ptI 6.0 and 10-'~ M ionic strength. (C) For SiO.~, ptIpzc = 2.0, solution p i t 8.0 and 10-3 M ionic strength.
(D) For TiO2, pH Pzc = 5.6, solution pH 7.0 and 10-~ M ionic strength.
Journal of Colloid and Interface Science, o i . 40, N'o. 1, J u l y 1972

80

JAMES AND HEALY

The final parameter to determine is the


location of adsorbed metal ions in the electrical layer. There are at least two choices
possible, either in the diffuse double layer
or in the compact double layer. Adsorption
in the diffuse double layer is described b y the
Gouy-Chapman theory in which the ions
are distributed in a continuous fashion and
the adsorption density is calculated b y integrating over the whole diffuse double layer.
This approach is satisfactory for low surface
potentials and surface charge, i.e., at low
p H values and low salt concentrations, because most of the ions are sufficiently remote
from the interface that the solvation energy
changes are small, and only the eoulombic
energy, as calculated using experimentally
measured diffuse double layer potentials,
makes a significant contribution to the free
energy of adsorption. For adsorption into
the compact double layer, i.e., at the inner
Helmholtz plane, the distance between the
ion and the interface is fixed at the hydrated
ionic radius, (rio, + 2rw). The adsorption
density in the compact double layer is calculated b y assuming all adsorbed ions are
located at the inner Helmholtz plane.
In the above discussion of oxide, solution,
solute, and interfacial parameters, all but
one quantity has been fixed and is thereafter
not adjustable. For example rio~ , rw, e~olid ,
ebulk, *K1, etc., K and 0 have been fixed
by experiment or set at literature values.
0
The one adjustable parameter is AGc~,m,
the specific adsorption free energy of the
potential adsorbate ions. For TiO2 this was
evaluated b y electrokinetic experiments and
the adsorption isotherms. For SiO2, it was
not possible to evaluate AGh~ experimentally because extrapolation of adsorption data to the P Z C of SiO2 was not reliable. We were therefore forced to select a
reasonable value for AG~em for Si02 adsorption and to adjust to give best fit of the
magnitude of each metal adsorbed. Furthermore, for each metal we used the same
AGh,m for each hydrolysis product including the free ion. In this way the value of
Journal of Colloid and Interface Science, Vol. 40, No. 1, July 1972

AGhem does not preferentially promote any


one species more than another.
SUMMARY
A general mathematical model has been described for the adsorption of hydrolyzable metal
ions at the solid-liquid interface, in terms of
changes in the eoulombic, solvation, and specific
chemical energy interactions as the ion approaches
the interface for each of the possible species, including all hydrolysis products.
The physical representation of the model is
that the ion-solvent interactions present a barrier
to close approach of highly charged ions to the
interface between a low dielectric constant solid
and water. When the ionic charge is lowered by
hydrolysis or ligand complex formation, the ionsolvent interaction is decreased thus lowering the
energy barrier. The ions may then approach closer
to the interface which results in greater coulombic
and short-range interaction energies, which are
more favorable to adsorption.
ACKNOWLEDGMENTS
One of us (R. O. J.) acknowledges the award of a
Commonwealth Postgraduate Research Award.
Support from the Australian Mineral Industries
Research Association is gratefully acknowledged.
REFERENCES
1. JAMES, R.
Interface
2. JAMlgS, R.
Interface

O., AND I-IEALY, T. W., J . Colloid


Sci. 40, 42 (1972).
0., AND HEALY, T. W., J. Colloid
Sci. 40, 65 (1972).

3. CORNET, D., AND BIIRWELL, l~. L., JR., Or.


Amer. Chem. Soe. 90, 2489 (1968).
4. NORTIA, T., AND LAITINEN, S., Suom. Kemistilehti 41, 136 (1968).
5. HATHAWAY,B. J., AND LEWIS, C. E., J. Chem.
Soc., London 1969, 1176-1183.
6. HATIt&}VAY,B. J., AND LEWIS, C. E., or. Chem.
Soe., London 1969, 1183-1188.
7. WIESE, G. R., 5AMES, R. O., AND HEALY, W. W.,

General Discussions Faraday Society September 1971.


8. GaA~AM~, D. C., Chem. Rev. 41,441 (1947).
9. GR~JaAME, D. C., or. Chem. Phys. 23, 1166
(1955).
10. BOCKRIS, J. O'M., DEVANATtIAN, M. A. V.,
AND MULLER, K., Proc. Roy. Soc., Ser. A 274,

55 (1963).
11. ANDERSEN, T. N., AND BOCKRIS, J. O'M.,
Electrochim. Acta 9,347-371 (1964).
12. LEVINX, S., MINGINS, J., AND BELL, G. M.,
J. Electroanal. Chem. Interfacial Electrochem. 13,280--329 (1967).

IONS

AT

OXIDE-WATEI~

13. BARLOW, C. A., AND MAcDoNALD, J. R.,


Advan. Eleelrochem. Eleetrochem. Eng. 6,

1 (1967).
14. LEVINE, S., BELL, G. M., AND CALVERT, D.~
Can. 3-. Chem. 40,518 (1962).
15. SILLEN,L. G., AND MARTELL,A. E., "Stability

Constants of Metal-Ion Complexes." Spec.


Publ. 17. Chem. Soe., London (1964).
16. JAMES, R. O., AND HEALY, W. W., unpublished
data.
17. BOgN, N., Z. Phys. 1, 45 (1920).
18. LAIDLER, K. J., AND MUIRHLAD-GOuLD,J. S.,
Trans. Faraday Soe. 63, 953 (1967).
19. LAIDLER, I~. J., AND PEGIS, C., Proc. Roy. Soc.
Ser. A 241, 80 (1957).
20. MUIRIIEAD-GOULD,J. S., AND LAIDLER, K. J.,
Trans. Faraday Soc. 63,944 (1967).
21. MUIRHEAD-GOULD,J. S., AND LAIDLER, K. J.,
Trans. Faraday Soc. 63,958 (1967).

22. DEBYE, P., "Polar Molecules." Chem. Catalogue Co., New York, (1926); Dover, New
York (1964).
23. ONSAGEa, L., 3-. Amer. Chem. Soc. 58, 1486
(1936).
24. KIRKWOOD,J., 3-. Chem. Phys. 7,911 (1939).
25. BOOTH, F., J. Chem. Phys. 19,391, 1327, 1615
(1951).

INTERFACE.

III

81

26. GRAHAME, D. C., J. Chem. Phys. 18,903 (1950).


27. GRAHAME, D. C., 3-. Chem. Phys. 21, 1054
(1953).
28. SACHER, E., AND LAIDLER, I~. J., "Modern
Aspects of Electrochemistry" (J. O. Bockris
and B. E. Conway, eds.), Vol. 3. Butterworths, London (1964).
29. Hunter, I~. J., 3. Colloid Interface Sci. 22,231
(1966).
O. F., AND DE BRUYN, P. L., Y.
Colloid Sci. 19,302 (1964).
DUGGEa, D. L., STANTON, J. H., IRB, B. N.,
MCCONNELL, B L., CUMMINGS,W. W., AND
MAATMAN, I~. W., J. Phys. Chem. 68, 757
(1964).
PURCELL, G., AND SUN, S. C., Trans. AIME
226, 6 (1963).
PARKS, G. A., Chem. Rev. 65,177 (1965).
HEALY, T. W., HERRING, A. P., AND FUERSTENAU, D. W., 3-. Colloid Interface Sei. 21,
435 (1966).
NAIE, N. I~., AND THOEP, J. M., Trans. Faraday Soc. 61,962 (1965) ; 61,975 (1965).
GILES, C. H., MAcEWAN,T. H., NAKHWA,S. N.
AND SMITH, D., J. Chem. Soc., London 1960,
3973.

30. DEVEREUX,

31.

32.

33.
34.

35.
36.

Journal of Colloid and Interface Science, Vol. 40, No. 1, July 1972

Você também pode gostar