Reactivity of metal oxides: Thermal and photochemical dissolution of MO and MFe2O4 (M=Ni, Co, Zn)

Article abstract of DOI:10.1016/j.jssc.2008.05.032

Dissolution rates of NiO, CoO, ZnO, α-Fe₂O₃ and the corresponding ferrites in 0.1 mol dm^-3 oxalic acid at pH 3.5 were measured at 70 °C. The dissolution of simple oxides proceeds through the formation of surface metal oxalate complexes, followed by the transfer of surface complexes (rate-determining step). At constant pH, oxalate concentration and temperature, the trend in the first-order rate constant for the transfer of the surface complexes (k_Me; Me=Ni, Co, Zn, Fe) parallels that of water exchange in the dissolved metal ions (k_-w). Thus, the most important factor determining the rates of dissolution of metal oxides is the lability of Me-O bonds, which is in turn defined by the electronic structure of the metal ion and its charge/radius ratio. UV (384 nm) irradiation does not increase significantly the dissolution rates of NiO, CoO and ZnO, whereas hematite is highly sensitive to UV light. For ferrites, the reactivity order is ZnFe₂O₄>CoFe₂O₄?NiFe₂O₄. Dissolution is congruent, with rates intermediate between those of the constituent oxides, Fe₂O₃ and MO (M=Co, Ni, Zn), reflecting the behavior of very thin leached layers with little Zn and Co, but appreciable amounts of Ni. The more robust Ni²⁺ labilizes less the corresponding ferrite. The correlation between log k_M and log k_-w is somewhat blurred and displaced to lower k_M values. Fe(II), either photogenerated or added as salt, enhances the rate of Fe(III) phase transfer. A simple reaction mechanism is used to interpret the data.

Full text of DOI:10.1016/j.jssc.2008.05.032

ARTICLE IN PRESS
Journal of Solid State Chemistry 181 (2008) 2350– 2358
Contents lists available at ScienceDirect
Journal of Solid State Chemistry
journal homepage: www.elsevier.com/locate/jssc
Reactivity of metal oxides: Thermal and photochemical dissolution of MO and
MFe2O4 (M ¼ Ni, Co, Zn)
Ã
Luis A. Garcıa Rodenas a,b, Miguel A. Blesa a,c,d, Pedro J. Morando a,b,c,
´
a Gerencia Qu´ımica, Comisio´n Nacional de Energ´ıa Ato´mica, Centro Ato´mico Constituyentes, Avenida General Paz 1499, 1650 San Mart´ın, Provincia de Buenos Aires, Argentina
b Instituto Jorge Sabato, Universidad Nacional de General San Mart´ın, Argentina
c Consejo Nacional de Investigaciones Cient´ıficas y Te´cnicas (CONICET), Argentina
d Centro de Estudios Ambientales, Escuela de Posgrado, Universidad Nacional de General San Mart´ın, Argentina
a r t i c l e i n f o
a b s t r a c t
Dissolution rates of NiO, CoO, ZnO, a-Fe2O3 and the corresponding ferrites in 0.1 mol dmꢀ3 oxalic acid at
pH 3.5 were measured at 70 1C. The dissolution of simple oxides proceeds through the formation of
surface metal oxalate complexes, followed by the transfer of surface complexes (rate-determining step).
At constant pH, oxalate concentration and temperature, the trend in the first-order rate constant for the
transfer of the surface complexes (kMe; Me ¼ Ni, Co, Zn, Fe) parallels that of water exchange in the
dissolved metal ions (kꢀw). Thus, the most important factor determining the rates of dissolution of
metal oxides is the lability of Me–O bonds, which is in turn defined by the electronic structure of the
metal ion and its charge/radius ratio. UV (384 nm) irradiation does not increase significantly the
dissolution rates of NiO, CoO and ZnO, whereas hematite is highly sensitive to UV light. For ferrites,
the reactivity order is ZnFe2O44CoFe2O4bNiFe2O4. Dissolution is congruent, with rates intermediate
between those of the constituent oxides, Fe2O3 and MO (M ¼ Co, Ni, Zn), reflecting the behavior of very
thin leached layers with little Zn and Co, but appreciable amounts of Ni. The more robust Ni2+ labilizes
Article history:
Received 17 January 2008
Received in revised form
15 May 2008
Accepted 23 May 2008
Available online 29 May 2008
Keywords:
Co, Ni and Zn oxides
Co, Ni and Zn ferrites
Hematite
Me–O bond lability
Dissolution kinetics
Photodissolution
Water exchange
less the corresponding ferrite. The correlation between log kM and log k
is somewhat blurred and
ꢀw
displaced to lower kM values. Fe(II), either photogenerated or added as salt, enhances the rate of Fe(III)
phase transfer. A simple reaction mechanism is used to interpret the data.
& 2008 Elsevier Inc. All rights reserved.
1. Introduction
behavior; heterogeneous acid–base, complexation and electron
transfer reactions are involved, and all of them bear important
similarities with analogous homogeneous phenomena [1].
In spite of the many similarities between dissolved and surface
metal ion chemistry, heterogeneous systems are more difficult to
characterize because of the influence of the solid history. Trends in
reactivity can be established by comparing the behavior of a given
oxide in contact with a series of related reagents. Such studies, for
instance, have established parallelisms between the stability of
the surface and dissolved metal complexes [2,3]. A parallelism has
also been established between the ligand-assisted mechanisms of
metal detachment from a dissolving mineral surface and the
mechanisms of ligand exchange around the corresponding metal
ion in solution [4]. Similar ideas have been suggested to apply to
the acid dissolution of NiO [5]. Here we show that, with adequate
precautions, trends can also be established in the dissolution
behavior of a series of related oxides in the presence of the same
dissolution reagent. Often, the history of the solid produces
quantitative changes in the observed kinetic behavior (rate
constants), but the qualitative behavior (mechanism) remains
unchanged.
Dissolution of metal oxides in acidic aqueous solutions is,
in principle, one of the simplest groups of heterogeneous
inorganic reactions. Dissolution always involves protonation of
oxo groups of the solid, rupture of Me–O bonds and phase transfer
of the metal [1]. Often, the anions of the acid also play an
important role, through surface complexation of metal ions,
enhancement of proton co-adsorption, and labilization of vicinal
Me–O bonds.
This group of reactions has been studied for very many years.
Most of the studies focus on one particular aspect of the process:
empirical rate laws to use in hydrometallurgy, influence of applied
potential in the case of oxide electrodes, and morphological
evolution of the metal oxide particles. Although a general picture
has not been completely developed, it is now widely accepted that
the inorganic chemistry of the interface governs the observed
Ã
Corresponding author at: Gerencia Quı´mica, Comisio´ n Nacional de Energı´a
Ato´ mica, Centro Ato´mico Constituyentes, Avenida General Paz 1499, 1650 San
Martı´n, Provincia de Buenos Aires, Argentina. Fax: +541167727886.
E-mail address: morando@cnea.gov.ar (P.J. Morando).
In this paper we present
a study of the thermal and
photochemical dissolution of two oxides with rock-salt structure,
0022-4596/$ - see front matter & 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2008.05.032

ARTICLE IN PRESS
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
2351
CoO and NiO (bunsenite), and one with wurtzite structure,
2. Materials and methods
ZnO. All of them are highly soluble in mineral acids, but the
rates of dissolution differ appreciably, the typical penetration
rates at pH
2.1. Synthesis and characterization of the solid
1
ranging from 10ꢀ5 to 10ꢀ10 cm sꢀ1 [1]. The
chosen dissolution reagent was oxalic acid, which is known to
be especially aggressive towards metal oxides, due to its
combined properties of being a strong acid, a strong chelating
agent and, under appropriate conditions, a good reductant [6].
Because of these properties, oxalic acid is an important
constituent of many formulations used for chemical cleaning
of metal surfaces. We also report data on a-Fe2O3. Although
hematite has been much studied [7–9], the measurements
were necessary because one of our objectives is to describe
the dissolution of mixed oxides, namely ferrites of stoichiometry
MFe2O4 (spinels with M ¼ Co, Ni and Zn). Co and Ni ferrites
are important constituents of the scales grown onto stainless
steels. Furthermore, spinel ferrites constitute an important
group of materials with many uses, and they are often synthesized
from aqueous media by soft preparative procedures. Hence, the
reverse dissolution reaction is important not only because of the
corrosion processes the materials may undergo in wet atmo-
spheres, but also in order to understand the nature of the interface
ferrite/water in synthetic procedures. As described below, the
nature of the surface in contact with aqueous solutions may differ
appreciably from bulk composition. We found previously [10]
that, in general, the rate of dissolution is intermediate, higher than
that of the more robust constituent metal oxide, and lower than
that of the more labile one. Frequently, after incongruent
dissolution of a small fraction of the solid, the stoichiometry of
dissolution becomes congruent, and both metal ions are trans-
ferred at the same rate.
Understanding of the factors that govern the dissolution of
mixed metal oxides is limited by the need to distinguish
between ‘intrinsic’ (structural chemistry) and ‘extrinsic’ factors
(particle size, morphology and history). We have attempted
to minimize extrinsic factors by controlling the preparative
procedures, in order to focus on the response of each constituent
to the medium conditions. Chemical parameters may affect
the rates of dissolution of the component oxides in opposite
ways. For example, Fe(III) oxides’ dissolution rates are enhanced
by the presence of reductants, whereas all M(II) oxides
involved in the studied ferrites dissolve faster in the presence of
oxidants. Thus, in this paper the influence of O2 and Fe2+ is
studied.
Nearly stoichiometric spherical particles of NiO, ZnO and
ferrites of narrow size distribution were prepared using
a
´
modified procedure based on Tamura and Matijevic’s method
[18]. For NiO, ZnO and ferrites, KOH (0.08 mol) was added to a
cylindrical reaction vessel containing 750 cm3 of double-distilled
water, and in the case of ferrites KNO3 (0.16 mol) was also added.
N2 was bubbled through the solution for 2 h. For pure oxides,
50 cm3 of a solution containing 0.04 mol of M(NO3)2 (M ¼ Ni and
Zn) was poured into the vessel. For mixed oxides, the solution
contains FeSO4 and M(NO3)2 (M ¼ Ni, Co, Zn); the total amount of
metal ions was kept constant at 0.04 mol, with a molar ratio
[M]:[Fe] ¼ 0.5. The reaction vessel containing M(OH)2 (pure
oxides) or the coprecipitated Fe(OH)2 and M(OH)2 (mixed oxides)
was immersed in a bath at 90 1C, and the solids were aged for 4 h
with continuous stirring and N2 purge.
The solids were filtered through Millipore membranes
(0.45 mm pore size) and heated for 1 h at 700 1C and then cooled
to room temperature in an air atmosphere.
Attempts to prepare CoO always yielded a solid containing
variable amounts of Co3O4. Therefore, a commercial sample was
used (Merck, 99.99% purity).
Hematite was prepared by transformation of ferrihydrite [19].
In all, 40 g Fe(NO3)3 ꢁ 9H2O was dissolved in 500 cm3 twice
distilled water preheated at 90 1C and ferrihydrite was precipi-
tated with 300 cm3 1 M KOH preheated at 90 1C. Then 50 cm3 1 M
NaHCO3 was added, preheated to 90 1C and the suspension was
kept (pH ¼ 8–9) in a closed polyethylene flask at 90 1C for 48 h.
Metal content was measured by ICP optical emission spectro-
metry in a Perkin-Elmer Optima 3100 ICO-OES spectrometer AS 90
autosampler with conical vials or by atomic absorption spectro-
metry in a Perkin-Elmer A Analyst 100 apparatus. Iron content in
the case of ferrites was measured spectrophotometrically using
mercaptoacetic acid [20] in a Shimadzu UV-210A spectrophot-
ometer. Particle morphology and size were characterized by
scanning electron microscopy (SEM) using a Phillips SEM 500
apparatus, and near-surface composition was determined by EDS.
Specific surface area (BET method) measurements were obtained
in a Micromeritics AccuSorb 2100 E equipment and diffraction
patterns were recorded in
a Siemens D5000 diffractometer
equipped with a graphite monochromator, and CuKa radiation.
The diffractograms of NiO [21], CoO [22], ZnO [23] and Fe2O3 [24]
showed sharp peaks at the expected spacings.
The existence of a small fraction of hematite in cobalt and
nickel ferrites was demonstrated by X-ray diffraction and EDS.
The effect of light on the rate of dissolution of metal oxides has
been explored by several groups [11–15]. Again, similarities are
found between the photochemistry of dissolved and surface metal
complexes; this interaction with light may or may not produce
dissolution rate enhancement.
Hematite is present as
a thin surface layer; the hematite
Photochemical dissolution starts always through the absorp-
tion of a photon by the solid. With UV light (l ¼ 384 nm), the
main primary photophysical event is the generation of an
electron–hole pair. Direct absorption by surface chromophores
(oxalato complexes) is also possible, but the effective absorption
cross-section is comparatively small. The semiconducting proper-
ties of the oxides are therefore relevant. Both CoO and NiO are p-
type [16] wide band semiconductors, whereas ZnO [1] and a-
Fe2O3 are n-type [17]. Our final goal is to establish relationships
between the kinetic properties of the dissolved ions and the rates
of dissolution.
diffraction peaks disappear when ca. 12% of the solid is dissolved
with oxalic acid and the diffractograms show only the sharp peaks
of NiFe2O4 [25], CoFe2O4 [26] and ZnFe2O4 [27]. All the experi-
ments were carried out with these pickled samples.
Table 1 summarizes the bulk composition (average of three
determinations) and the main characteristics of the synthesized
oxides.
SEM pictures of ferrites show spherical particles with average
diameters in the order of 0.1, 0.2, and 0.3 mm for M ¼ Zn, Ni, and
Co ferrites, respectively.
In the case of ferrites, the different response to light
of both constituent oxides makes difficult the prediction
of the effect of irradiation on dissolution kinetics; the results
of the present study permit to integrate the mechanism of
photochemical dissolution with the mechanism of thermal
dissolution.
2.2. Kinetic experiments
Thermal kinetic experiments were performed in the dark
under N2 or O2 atmosphere, in a magnetically stirred vessel
immersed in a thermostat at 70.070.1 1C. The oxide (ca. 20 mg)

ARTICLE IN PRESS
2352
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
Table 1
Characteristics of the synthesized oxides
Oxide
Semiconduction type
Composition
Ni1.04
Surface area (m2 gꢀ1
)
Average particle diameter (mm)
NiO
p
p
n
n
p
n
n
O
45.1
0.79
4.12
8.72
10.3
3.9
0.02
1.2
0.3
0.1
CoO
Co0.98O
ZnO
Zn1.02O
Fe2O3
NiFe2O4
CoFe2O4
ZnFe2O4
Fe2.03O3
Ni1.15Fe1.90O4
Co1.00Fe2.00O4
Zn0.88Fe2.08O4
0.2
0.3
0.1
31.2
was poured over 100 cm3 0.1 M oxalic acid solution; ionic strength
was fixed at 0.5 M with NaClO4 and pH was set at 3.5 with NaOH.
At adequate time intervals, 1 cm3 samples were taken and filtered
through a 0.45-mm cellulose acetate membrane. Metal ions in the
filtrate were measured as described before. In experiments
designed to probe the influence of ferrous salts, adequate amounts
of Fe(NH4)2(SO4)2 were added to the initial solution.
1
0.8
0.6
0.4
0.2
0
Photodissolution experiments were carried out in a reaction
vessel illuminated from above with a 1000-W xenon lamp
through an interference filter centered at 384 nm. Incident light
intensity per unit solution volume was determined by chemical
actinometry with trisoxalatoferrate(III), Io ¼ 9.4 ꢂ10ꢀ6 einstein
s
ꢀ1 dmꢀ3
.
3. Results
0
2
4
6
3.1. Thermal dissolution
10-3 t / s
In large excess oxalic acid, the stoichiometry of the thermal
dissolution for the MO oxides is given by Eq. (1), whereas for
hematite Eq. (2) applies. Metal oxide solubility is in principle high
enough to guarantee total dissolution at pH 3.5. However, M(II)
oxalates are sparingly soluble, and they may precipitate. This is
indeed the case of Co(II) that reaches a maximum dissolution
fraction (f ¼ [M]/[M]T, where [M]T is the expected concentration
for total dissolution) of 0.4 in the dark and 0.6 under light:
Fig. 1. Thermal dissolution profiles for NiO (’), CoO (m), ZnO (K) and Fe2O3 (E)
in 0.1 M oxalic acid at 70 1C and pH 3.5 (I ¼ 0.5 M NaClO4).
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
MOðsÞ þ 2HC2Oꢀ4 ! MðC2O4Þ22ꢀ þ H2O
(1)
Fe2O3ðsÞ þ 6HC2Oꢀ4 ! 2FeðC2O4Þ33ꢀ þ 3H2O
(2)
The stoichiometry of dissolution for the ferrites is shown in
Eq. (3). The thermal redox decomposition of 1.7 ꢂ10ꢀ4 mol dmꢀ3
trisoxalatoferrate(III) yields less than 5% Fe(II) at 70 1C
and therefore does not affect appreciably the dissolution
stoichiometry:
MFe2O4ðsÞ þ ð6 þ nÞHC2Oꢀ4 þ ð2 ꢀ nÞHþ
! MðC2O4Þnð2nꢀ2Þꢀ þ 2FeðC2O4Þ33ꢀ þ 4H2O
(3)
0
5
10
10-3 t / s
15
20
Fig. 1 shows the time evolution of the dissolution fractions for
NiO, ZnO, a-Fe2O3 and CoO as a function of time. The maximum
dissolution fraction of CoO is smaller than NiO and ZnO; probably
coating of the particles by CoC2O4 produces passivity.
J
Fig. 2. Upper curve: time evolution of the Co (K) and Fe ( ) thermal dissolution
fractions (f ¼ [Me]/[Me]T) for CoFe2O4. Lower line: linear contracting volume
J
expression (Eq. [4]) for Co (K) and Fe ( ) dissolution. Experimental conditions:
0.1 M oxalic acid at 70 1C and pH 3.5 (I ¼ 0.5 M NaClO4).
Fig.
2
shows the dissolved fraction fMe/time profiles
(fMe ¼ [Me]/[Me]T, with Me ¼ M or Fe; [Me]T is the expected
concentration for total dissolution) for both metals in cobalt
ferrite. Dissolution is congruent (fFe ¼ fCo) within the time span of
our experiments and proceeds at a constant specific dissolution
rate k (mol mꢀ2 sꢀ1). Nickel and zinc ferrites follow the same
congruent behavior (not shown).
In all pure and mixed oxides the thermal profiles are
deceleratory and can be fitted by a simple contracting volume
rate law (Eq. (4), where k0 is the specific dissolution rate constant
(sꢀ1)) up to appreciable degrees of conversion. This law implies
constant reactivity per unit area, approximation valid for rather
large particles with appreciable kinetic robustness. Fitting to this
kinetic law is limited essentially by the formation of insoluble Co
or Ni oxalates that arrests dissolution (see Fig. 1). The formation of
insoluble oxalates is not important in the case of ferrites. In pure
oxides of general formula MxO, the specific dissolution rate k
(mol mꢀ2 sꢀ1) relates to k0 through Eq. (5), where S0 is the initial

ARTICLE IN PRESS
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
2353
value of the specific surface area S, Mw is the oxide molecular
weight and x is the metal stoichiometric coefficient in the oxide
MxO (x deviates from unity as indicated in Table 1). This regime
implies that the penetration rate p (cm sꢀ1) usually defined as
p ¼ ꢀdr/dt (where r is the particle radius) is also constant and
proportional to k and k0 [1]:
The shapes of the curves for Ni and Co ferrites can be described
by Langmuir–Hinshelwood Eq. (7), where Rmax ¼ kMFeðmIIÞax=kM
is the maximum acceleration factor and KL the binding constant
of Fe(II) on the surface (mediated by oxalate). Table 3 shows the
values of kFeðIIÞ
M max, KL and Rmax for nickel and cobalt ferrites. For zinc
ferrite, if Eq. (7) applies, Rmaxffi1 and/or KLffi0. The order of
reactivity at high ferrous concentrations ( kFeðIIÞ
M max) is CoFe2O4bZn-
0
1=3
1 ꢀ ð1 ꢀ fÞ
¼ k t
(4)
Fe2O44NiFe2O4. The maximum acceleration factors are similar for
nickel and cobalt ferrites, but in the second case higher
concentrations of Fe2+ are required to achieve this factor because
KL is lower:
k ¼ 3xk0=S0Mw
(5)
RmaxKL½FeðIIފ
In ferrites of general formula M1ꢀyFe2+xO4 (y ¼ (3/2)x), the
R ¼
(7)
specific dissolution rate k relates to k0 through Eq. (6). The specific
1 þ KL½FeðIIފ
rates of dissolution of each individual ion Fe and
respectively, kFe ¼ {(2+x)k} and kM ¼ {(1–y)k}:
k ¼ 3k0=S0Mw
M
are,
(6)
Fig.
4 shows that the dissolved oxygen does not affect
significantly the rate of dissolution of cobalt ferrite, except
perhaps at high conversions, where dissolution is somewhat
arrested; on the other hand, nickel ferrite dissolution rate is
significantly enhanced. The profile for zinc ferrite dissolution, not
shown, is not affected by oxygen.
Oxygen quenches the effect of Fe(II) addition, because the
oxidation to Fe(III) is very fast. Sub-stoichiometric amounts of
oxygen simply change the concentration of Fe(II) (see Fig. 3).
Under excess oxygen, Fig. 4 applies.
Fig. 2 includes the fitting of the data according to the integrated
contracting volume rate law for both ions of CoFe2O4, Eq. (4).
Table 2 shows the kMe values for pure and mixed oxides; the
reactivity order in both cases is Zn4CobNi.
Addition of ferrous ions increases the rates of dissolution of
nickel and cobalt ferrites, without changing significantly that of
zinc ferrite. Fig. 3 shows the specific rate enhancement R, as a
function of [Fe2+]0. R is defined as kFMeðIIÞ=kM, where kFMeðIIÞ is the
specific rate at any given value of [Fe2+]o, and kM is the rate in the
absence of Fe2+ (from Table 2). At high [Fe2+]o, a constant
maximum rate is achieved ( kFMeðmIIÞax).
3.2. Photochemical dissolution
Fig. 5 shows the time evolution of the dissolution fractions
under illumination for the simple oxides. The results demonstrate
a negligible influence of light for Ni and Zn oxides (see Fig. 1). In
the CoO case, the dissolution fraction increases under light;
coating of the particles by CoC2O4 probably produces passivity as
it was observed in the dark dissolution. The photochemical
dissolution profile for a-Fe2O3 is sigmoidal, and the simple
contracting volume rate law does not apply.
Table 2
Thermal specific dissolution rates kMe (mol mꢀ2 sꢀ1) for the oxides in 0.1 M oxalic
acid at 70 1C and pH 3.5
Oxide
107 ꢂ kMe (mol mꢀ2 sꢀ1
)
The stoichiometry of hematite photodissolution is given by Eq.
(8), whereas that of MO dissolution does not change under light
(Eq. (1)):
Ni1.04
O
0.667
65.4
Co0.98O
Zn1.02O
98.4
Fe2.03O3
0.069
0.036a
1.10a
1.74a
Fe2O3ðsÞ þ HC2Oꢀ4 þ 5Hþ ! 2Fe2þ þ 2CO2 þ 3H2O
(8)
Ni1.15Fe1.90O4
Co1.00Fe2.00O4
Zn0.88Fe2.08O4
In homogeneous solution, photochemical decomposition of
trisoxalatoferrate(III) takes place with a quantum yield of 1.2
[28]. Under the experimental conditions used in this study, the
photonic flux suffices to reduce all Fe3+ in 10 s and the
stoichiometry of the photodissolution of ferrites is now given by
Eq. (9). In Eqs. (8) and (9) complexation of dissolved ions has been
ignored for simplicity:
a
Me ¼ M(II).
25
20
15
10
5
MFe2O4ðsÞ þ HC2Oꢀ4 þ 7Hþ
! M2þ þ 2Fe2þ þ 2CO2 þ 4H2O
(9)
Irradiation with lmax ¼ 384 nm (Io ¼ 9.4 ꢂ10ꢀ6 einstein sꢀ1
dmꢀ3) brings about kinetic changes that parallel those produced
by ferrous ion.
Table 4 shows the values of the photochemical dissolution rate
(
phkM) for oxides and the rate enhancement factors R0 ¼ phkM=kM,
where kM is the thermal dissolution rate (from Table 2). The
results demonstrate a negligible influence of light for Ni and Zn
oxides, and a little influence for Co oxides. For Ni and Co ferrites
there is an appreciable influence of light and the order of
reactivity is CoFe2O44ZnFe2O44NiFe2O4.
0
0
4
8
12
Even though the data in Table 4 are casted in terms of the
same kinetic law as found for thermal dissolution, cobalt ferrite
photodissolution rate at short times is lower than expected. Fig. 6
shows that a slightly sigmoidal profile is observed, similar to those
measured earlier in the dissolution of magnetite [29] and in the
105 [Fe2+]o/mol dm-3
Fig. 3. Dependence on [Fe2+
]
of the acceleration factor R ¼ kFMeðIIÞ=kM for the
o
specific dissolution rate of NiFe2O4 (’), CoFe2O4 (m) and ZnFe2O4 (K).
Experimental conditions: 0.1 M oxalic acid at 70 1C and pH 3.5 (I ¼ 0.5 M NaClO4).

ARTICLE IN PRESS
2354
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
Table 3
Langmuir–Hinshelwood parameters for the dissolution of ferrites in 0.1 M oxalic acid at 70 1C and pH 3.5 in the presence of ferrous salts and rate enhancement factor Rmax
107 ꢂ kFMeðmIIÞax (mol mꢀ2 sꢀ1
)
Rmax ¼ kFeðIIÞ
M max=kM
Ferrite
10ꢀ4 ꢂ KL ((mol Fe2+
)
ꢀ1 dm3)
Ni1.15Fe1.90O4
Co1.00Fe2.00O4
0.96
39.4
7.26
26.7
24.2
26.63
1
1
0.8
0.6
0.4
0.2
0.8
0.6
0.4
0.2
0
0
0
0
3
6
9
5
10
15
10-3 t / s
10-3 t / s
Fig. 5. Photochemical dissolution profiles for NiO (&), CoO (n), ZnO (o) and Fe2O3
(}) in 0.1 M oxalic acid at 70 1C and pH 3.5 (I ¼ 0.5 M NaClO4).
Table 4
Specific photodissolution rates phkM (mol mꢀ2 sꢀ1) for the oxides in 0.1 M oxalic
2
1
acid at 70 1C and pH 3.5 and rate enhancement factors R0 (lmax ¼ 384 nm,
Io ¼ 9.4 ꢂ10ꢀ6 einstein sꢀ1 dmꢀ3
)
R0 ¼ phkM=kM
107
ꢂ
phkM ðmol mꢀ2 sꢀ1
Þ
Oxide
Ni1.04
O
0.665
1.0
1.2
1.0
–
Co0.98O
77.6
Zn1.02O
98.6
–
Fe2.03O3
Ni1.15Fe1.90O4
Co1.00Fe2.00O4
Zn0.88Fe2.08O4
0.67
18.6
7.4
0.99
8.19
1.73
1
0.8
0.6
0.4
0.2
0
0
5
10
10-3 t / s
15
Fig. 4. Time evolution of the dissolved iron fraction in deaerated ( ) and oxygen-
ꢃ
J
saturated solution ( ): (A) CoFe2O4; (B) NiFe2O4. Experimental conditions: 0.1 M
oxalic acid at 70 1C and pH 3.5 (I ¼ 0.5 M NaClO4).
photochemical dissolution of Fe2O3 (Fig. 5). The difference is in
agreement with the change in the stoichiometry, and the
photochemical production of ferrous ion (see Section 4).
For ferrites, at high ferrous ion concentrations, the effect of
light disappears, but the reactivity trend remains unchanged.
0
0
Fig. 7 shows the dependence on [Fe2+
]
o
of the relationship
2
4
6
10-3 t / s
R00 ¼ pHkMFeðIIÞ=kFMeðIIÞ, where kMFeðIIÞ and phkFMeðIIÞ are the thermal and
photochemical specific rate constants at any given value of [Fe2+]o,
respectively; at high concentrations the specific rates of thermal
Fig. 6. Time evolution of the photodissolved fraction for CoFe2O4 at lmax ¼ 384 nm
(Io ¼ 9.4 ꢂ10ꢀ6 einstein sꢀ1 dmꢀ3). Experimental conditions: 0.1 M oxalic acid at
70 1C and pH 3.5 (I ¼ 0.5 M NaClO4).
and photochemical dissolutions are equal, phkMFeðIIÞ ¼ kFMeðIIÞ
.

ARTICLE IN PRESS
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
2355
20
15
10
5
1
0.8
0.6
0.4
0.2
0
0
5
10
15
0
10-3 t / s
0
0.2
0.4
0.6
105 [Fe2+]o / mol dm-3
1
0.8
0.6
0.4
0.2
0
Fig. 7. Dependence on [Fe2+
]
o of the photoacceleration factor R00
¼
phkFMeðIIÞ=kFMeðIIÞ for
NiFe2O4 (’), CoFe2O4 (m) and ZnFe2O4 (K) at lmax ¼ 384 nm (Io ¼ 9.4 ꢂ10ꢀ6
einstein sꢀ1 dmꢀ3). Experimental conditions: 0.1 M oxalic acid at 70 1C and pH 3.5
(I ¼ 0.5 M NaClO4).
Oxygen inhibits the effect of light. In aerated media,
the dissolution of all ferrites proceeds at the same rate in the
dark and under light. Fig. 8 shows the coincidence of the
dissolution profiles for nickel and cobalt ferrites; for comparison,
the photochemical dissolution profile without oxygen is also
shown.
4. Discussion
0
2
4
6
8
10-3 t / s
4.1. Reactivity trends in the dark
Fig. 8. Time evolution of the dissolved iron fraction in the dark (m), under
Specific rates of dissolution at any given temperature depend
on solution composition. Solution pH, the nature and concentra-
tion of complexing agents, solution redox potential and ionic
strength are all important. The dependence of k on solution
variables has been dealt with repeatedly in the literature
([1,30,31] and references therein). In this paper, we are interested
in the reactivity trends of the series of oxides under a precise set
of experimental conditions. These conditions are 70 1C, 0.1 M
oxalic acid, pH 3.5 and I ¼ 0.5 M. The system does not contain
thermally active redox reagents, although oxalate is a good
reductant for Fe(III) under light. Internal redox reactions are
known for both dissolved FeIII(C2O4)33ꢀ and surface FeIII oxalato
complexes.
J
( ), and under light in deaerated
irradiation in oxygen-saturated solution
suspensions (n) at lmax ¼ 384 nm (Io ¼ 9.4 ꢂ10ꢀ6 einstein sꢀ1 dmꢀ3). (A) CoFe2O4;
(B) NiFe2O4. Experimental conditions: 0.1 M oxalic acid at 70 1C and pH 3.5
(I ¼ 0.5 M NaClO4).
actual nature of surface species, see Ref. [31]:
ꢄ MeOH þ H2Ox2 ꢄ Me ꢀ OxH þ H2O fast equilibrium; K1
(10)
ꢄ Me ꢀ OxH ! MenþðaqÞ rate-determining step; k1
ðV=SÞd½MenþŠ=dt ¼ qk ¼ k1fꢄ Me ꢀ OxHg
(11)
(12)
The trends in the specific dissolution rates discussed here do
not apply necessarily to other experimental conditions, because of
the diverse response of each system to these variables, in
particular redox potential. The conditions chosen in this work
are however representative of the so-called simple acid dissolu-
tion in the presence of complexing anions, or ligand-assisted acid
dissolution [31–33].
The general very simplified kinetic scheme for these reactions
is given in Eqs. (10) and (11), where Me ¼ Fe or M, Ox denotes
C2O24ꢀ and the symbol ꢄ identifies surface complexes. The first-
order rate constant k1 relates to the specific dissolution rate k
through Eq. (12), where V is the solution volume, {ꢄMe–OxH} is
the surface concentration (mol mꢀ2) and q is the stoichiometric
coefficient for Me in M1ꢀyFe2+xO4. For a detailed description of the
Eq. (10) describes adsorption as an electroneutral reaction,
although it is well known that other protolytic species of dissolved
oxalate and of surface complexes may be involved [34]. When
more than one surface complex may be transfered to solution, the
rate law (12) involves more than one term, but this complication
is not relevant for studies conducted at fixed solution composi-
tion. Eq. (11) describes the rate-determining step as the detach-
ment of the complexed metal ion from the solid and its transfer to
aqueous solution; no attempt was made to identify the speciation
of the dissolved metal or to define the participation of H+ in the
stoichiometry or in the kinetics. In general, it is agreed that in acid
dissolution the required number of protons equals the charge
borne by the metal ion [5]. The pH dependence of the rate is
embodied in k1. In oxalic acid excess, Men+(aq) is an oxalate

ARTICLE IN PRESS
2356
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
complex. Eqs. (10) and (11) predict that k increases with oxalate
concentration, until a maximum value is reached when total
coverage of surface sites is achieved (y ¼ 1), and {ꢄMe–OxH} ¼ Ns,
the total density of surface sites. At constant oxalate and pH, a
constant k value results, leading to the contracting geometry
regime.
For hematite, Eqs. (10) and (11) also apply in the dark. In
agreement with Siffert and Sulzberger [13], no Fe(II) is formed,
despite the relatively high thermodynamic driving force,
DG0 ¼ ꢀ0.87 eV at pH 3 for the equilibrium represented by Eq. (8).
In terms of Eq. (12), the differences in k1 for M and Fe are
compensated by differences in {ꢄMe–OxH}. The more labile
cation, with larger k1, is therefore depleted in the surface layer.
The leached layer is generated in the very early stages of
dissolution (‘instantaneously’ in the time scale of our experi-
ments), and must be very thin because no deviation from
congruency is observed even in the earliest measurements. Such
a deviation has been observed in other cases [35].
The reactivity order for mixed oxides was found to be
ZnFe2O44CoFe2O4bNiFe2O4 (see Table 2), with values of the
dissolution rates lying in between those of MO and Fe2O3.
Although the phase transfer of M(II) and Fe(III) are independent
processes, the transfer of each ion exerts influence on the rate of
transfer of the other. The order of reactivity is the same as that for
the simple MO oxides, with lower rates defined in part (see below)
by the extent of depletion of the divalent cation in the leached
layer. The leached layers of the various ferrites may be viewed as a
disrupted layer of the less reactive oxide, with different contents
of M(II). Two factors determine the steady-state penetration rate:
(a) the degree of enrichment of the leached layer in the less
reactive ion, and (b) the disruption of surface layer produced by
the release of the more reactive ion.
Zinc and cobalt ferrites dissolve at rates that are approximately
66 and 59 times lower than the rate of dissolution of the pure
oxides ZnO and CoO, respectively, and 53 and 32 times higher,
respectively, than the rate of dissolution of hematite. The leached
layer may be viewed as very disrupted Fe2O3, causing the bulkier
Zn2+ a larger disruption. The possible influence of the different
structure of both leached layers on the effect of ferrous ion is
mentioned below. The rate of hematite dissolution is only 10
times lower than that of nickel oxide; hence, the leached layer in
nickel ferrite must contain appreciable amounts of both ions. The
rate of dissolution of nickel ferrite is similar to that of hematite,
indicating that the reactivity of surface Fe(III) is not largely
affected by Ni(II). These comparisons of the rates of dissolution of
pure and mixed oxides must of course be viewed with caution,
because of the possible influence of particle size, shape and
structure.
Table 5
Rate constants k
for water exchange [37]
ꢀw
Metal ion
k
(sꢀ1
)
ꢀw
Ni2+
Co2+
Zn2+
Fe3+
3 ꢂ104
2 ꢂ106
7 ꢂ107
2 ꢂ102
-4
-6
-8
ZnO
CoO
NiO
ZnFe2O4
Fe2O3
CoFe2O4
NiFe2O4
log k-w
-10
0
2
4
6
8
10
Fig. 9. Relationship between log kMe and log k
Fe2O3 oxides (’).
for MFe2O4 (K) and for MO and
ꢀw
ferrites the slope is 0.471 instead of 0.616. Thus, the dissolution of
ferrites is less sensitive to the lability of M (measured by kꢀw). The
presence of Fe(III) in the solid framework not only produces an
overall lowering of the line in Fig. 9. The deceleration factor
(already discussed) is larger for the more labile ions, thus
decreasing the slope.
Sellers and Williams [38] studied the kinetics of dissolution of
four different stoichiometric spinel ferrites MFe2O4 (M ¼ Fe(II), Ni,
Co, Mn) in oxalic acid media under more drastic conditions, 140 1C
and pH 2.5. The order of reactivity was found to be Fe3O44Mn-
Fe2O44CoFe2O44NiFe2O4. Including in our own series the value
for the initial rate of magnetite dissolution [29], the order
Fe3O44ZnFe2O44CoFe2O4bNiFe2O4 is obtained. The reactivity
trend is similar to ours, even though under the conditions used by
Sellers and Williams, the stoichiometry of the reaction is
reductive, Eq. (9), more comparable with our photochemical
dissolution (see below). The position of Fe in both trends merits
comment; it reflects the operation of a different mechanism. It has
been amply documented that Fe(II) salts in the presence of
carboxylate ligands accelerate the dissolution of Fe(III) oxides,
through a mechanism involving the formation of surface mixed
valence bridged dimers ꢄFeIII–L–FeII–L [29], which undergo
internal electron exchange and produce more labile surface ꢄFeII
species. Note that in the systems studied by Sellers the
concentration of Fe2+ increases as the dissolution proceeds, albeit
by different mechanisms for M ¼ Fe(II) and for M¼Fe(II).
In a series of metal oxides undergoing acid dissolution in the
presence of a complexant, it is expected that k reflects the lability
of the oxo bonds linking the surface metal complex to the solid
framework. In aqueous solution, a good measure of the lability of
Me–O bonds is the rate of water dissociation from the coordina-
tion sphere (kꢀw) that governs the rate of water exchange (Eq.
(13)). The values of kꢀw correlate with the rates of complexation in
solution (Eigen–Wilkins mechanism [36]):
The effect of oxygen confirms the above discussion. Iron(III)
oxides dissolve faster in reductive media, whilst nickel(II) oxide
dissolves faster in oxidizing media [1]. In agreement, dissolved
oxygen brings about the faster dissolution of nickel ferrite, and
does not affect appreciably the dissolution rates of the other two
ferrites (Fig. 4). Catalysis of nickel(II) oxides dissolution by oxygen
involves oxidation of surface metal centers, and enhances the
transfer rate of Ni(III); the last rapidly reduces back to Ni(II) and
the stoichiometry is not affected [1], except in the possible
parallel oxidation of oxalate by dissolved oxygen (undetectable
under our experimental conditions).
MeðOH2Þnmþ þ H2Oà ! MeðOH2Þmꢀ1ðÃOH2Þnþ þ H2O;
k
(13)
ꢀw
Table 5 shows the values of k
for Ni2+, Co2+, Zn2+ and Fe3+
ꢀw
[37], and Fig. 9 shows the plot of log kMe vs. log k
for pure and
ꢀw
mixed oxides. The data corresponding to the first fit to a linear
relationship, log kMe ¼ ꢀ9.636+0.616 log k
,
demonstrate that
ꢀw
the lability of the oxo bonds linking the surface oxalato complex
to the solid framework is determined by the intrinsic properties of
Me. Basically, the (charge/radius) ratio and the crystal field effects
determined by the d-electron configuration govern the rate. For

ARTICLE IN PRESS
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
2357
The effect of added Fe(II) on cobalt and nickel ferrites
dissolution is also in line with these ideas. The rate of Fe(III)
transfer is increased as discussed above, and the steady-state
dissolution condition is shifted to higher rates.
FeII. The sigmoidal profile indicates that light triggers dissolution but,
more important, changes the stoichiometry of dissolution. In oxalate
solution under light, all iron is rapidly converted to FeII, and a very
efficient external reductant is thus generated. Eqs. (14)–(17) describe
the involved chemistry.
In cobalt ferrite, the rate enhancement in the presence of Fe(II)
is large, and the rate of dissolution becomes 41% that of pure CoO.
Thus, the possibility that the composition of the leached layer is
drastically altered must be considered. It has been documented
previously that electron injection may be followed by electron
hopping within the solid cobalt ferrite, in adjacent FeIII–O–FeII
moieties, which propagates appreciably in the solid; thus, a thick
disrupted region may be generated [39,40]. The coordination
environment of FeIII in the inverse spinel is probably also
important in facilitating the labilization by electron transfer.
The acceleration observed in nickel ferrite is also high,
indicating not only an easy release of FeIII, but also an enhance-
ment of the transfer rate of Ni(II); this metal is transfered 1.6
times as fast as in pure NiO. As indicated above (see Fig. 3), the
acceleration produced by Fe(II) at high concentrations is similar
for nickel and cobalt ferrites. At lower concentrations, the pre-
equilibrium of bridged dimer formation limits the observed rate
enhancement. Fig. 3 shows that Fe(II) is more easily incorporated
in the leached layer of nickel ferrite (larger KL); thus, except at the
highest concentration, the acceleration brought about by Fe(II) is
larger in nickel ferrite.
(a) Fast complexation:
ꢄ FeIIIOH þ H2Ox2 ꢄ FeIII ꢀ OxH þ H2O
(b) Induction period (generation of dissolved Fe(II)):
(14)
(b1) Phase transfer of Fe(III) and photochemical reduction in
solution
hv
ꢄ FeIII ꢀ OxH ! FeðOxÞ33ꢀ ꢀ! FeðOxÞ2
(15)
2ꢀ
(b2) Photochemical reduction in the solid surface and phase
transfer of Fe(II)
ꢄ FeIII ꢀ OxH þ ecꢀb !ꢄ FeII ꢀ OxHꢀ ! Fe ðOxÞ2
(c) Redox dissolution:
(16)
II
2ꢀ
ꢄ FeIII ꢀ OxH þ Fe ðOxÞ22ꢀ2 ꢄ Fe ꢀ OxH ꢁ ꢁ ꢁ Fe ðOxÞ2
II
III
II
2ꢀ
!ꢄ FeII ꢀ OxHꢀ ꢁ ꢁ ꢁ Fe ðOxÞ2ꢀ ! ꢀFe ðOxÞ22ꢀ þ Fe ðOxÞ3
III
II
III
3ꢀ
(17)
Concerning the ferrites, the behavior reactivity trends under
light are largely similar to those in the presence of Fe(II). Thus, the
reactivity order is the same (Table 4), oxygen inhibits the effect of
light (Fig. 8), and light has no influence at high Fe(II) concentra-
tions (Fig. 7). Also, light has no effect on the dissolution of zinc
ferrite. All these observations signal that the main effect of light is
to produce Fe(II). The slightly sigmoidal kinetic profile shown in
In the case of zinc ferrite, no acceleration by Fe(II) is observed,
although it has been reported that the acid dissolution of zinc ferrite
is accelerated at low pH by high concentrations of Fe(II) [41,42]. It is
thus possible that the lack of effect of Fe(II) under our conditions is
due to very low values of KL, rather than due to low values of kMFeðmIIÞax
.
We cannot give a rationale for the order of KL values thus derived
(NiFe2O44CoFe2O4bZnFe2O4). The fact that zinc ferrite contains
only octahedral iron whereas the other two ferrites contain part of
the iron in the more labile tetrahedral environments may be
Fig.
6 supports this interpretation. The acceleration period
corresponds to the build-up of Fe(II) in solution. Inspection of
the values of R0 ¼ phkM=kM for nickel and cobalt ferrites (Table 4)
also agrees with this idea. The larger KL values for nickel ferrite
render low levels of photolytic Fe(II) adequate to produce a large
enhancement. The slightly sigmoidal profile in the photochemical
dissolution of cobalt ferrite (Fig. 6), not observed in nickel ferrite,
can also be explained in terms of the respective KL values.
Fe(II) can be generated directly in the oxide surface, if the solid
absorbs the radiation, or can be generated by photolysis of
dissolved trisoxalatoferrate(III) [48]. The prevalence of either
possibility depends on the light fraction absorbed by the solid and
by the dissolved species. In the very initial stages, light is
absorbed essentially by the solid, but Fig. 6 shows that there is
no clear evidence of acceleration of the initial rate. Once
appreciable values of dissolved iron fraction are reached, solution
photolysis dominates the process.
Photochemical dissolution of iron oxides is considered an
important mechanism of iron mobilization in aquatic media.
Direct irradiation in water does indeed generate electron–hole
pairs, the former being able to promote iron phase transfer, in the
form of Fe(II). This process however requires the presence of a
hole scavenger, oxalic acid in our case, natural organic matter in
aquatic environments. In the presence of hole scavenger, the
alternative of solution photolysis becomes possible, and it is not
possible to demonstrate unambiguously if direct oxide photolysis
is involved, or if solution photolysis alone suffices to account for
the observations. It has been proposed that photochemical
dissolution of metal oxides is defined by the availability of
minority carriers (electrons for p-semiconductors, and holes for n-
semiconductors) [49]. Because of the effects described in this
paper, it is difficult to test this hypothesis. No correlation exists
between the dissolution behavior of NiFe2O4 (p-type) and CoFe2O4
(n-type) and the nature of the minority carriers.
relevant, regarding changes in both KL and kMFeðmIIÞax
.
As a result of all these factors, the reactivity order at high Fe(II)
concentrations is CoFe2O4bZnFe2O44NiFe2O4.
4.2. Reactivity trends under light
As stated, dissolution of NiO and ZnO is enhanced by oxidants
[1]. However, the photogeneration of holes by light absorption
does not lead to higher rates (Table 4). In the case of NiO, a p-type
semiconductor, this result is expected: the concentration of holes
does not change appreciably under light. On the other hand, the
lack of enhancement in n-type ZnO reflects the high rate of
thermal dissolution; the photochemical pathway, limited by the
light intensity and the low quantum yields, does not add
significantly to the overall rate [43,44].
The increased extent of p-type CoO dissolution at long times
(Figs. 1 and 5) reflects changes in the precipitation of sparingly
soluble Co(C2O4). The initial photochemical and thermal rates do
not differ much, and the different maximum degree of dissolution
achieved in both cases must be determined by the different onset
of precipitation. The rate enhancement factor quoted in Table 4 is
rather imprecise, and not very different from unity.
Thus, light does not seem to affect appreciably the rates of
dissolution of all three M(II) oxides. On the other hand, the influence
on hematite dissolution is drastic, amply documented in the
literature [7,13,45,46]. Fe(III) oxides dissolve very fast in the presence
of reductants; this is specially true for magnetite dissolution in the
presence of Fe(II) [47]. Surface FeIII oxalato complexes are good
electron acceptors, and this holds true both for external chemical
reductants and for photogenerated conduction band electrons [7].
Thus, a reactive pathway opens through the generation of surface

ARTICLE IN PRESS
2358
L.A. Garc´ıa Rodenas et al. / Journal of Solid State Chemistry 181 (2008) 2350–2358
5. Conclusions
[8] S. Banwart, S. Davies, W. Stumm, Colloids Surf. 39 (1989) 303.
´
[9] Y. Zhang, N. Kallay, E. Matijevic, Langmuir 1 (1985) 201.
[10] C.A. Figueroa, E.E. Sileo, P.J. Morando, M.A. Blesa, J. Colloid Interface Sci. 225
(2000) 403.
[11] T.D. Waite, F.M.M. Morel, J. Colloid Interface Sci. 102 (1984) 121.
[12] T.D. Waite, F.M.M. Morel, Environ. Sci. Tech. 18 (1984) 680.
[13] C. Siffert, B. Sulzberger, Langmuir 7 (1991) 1627.
[14] M.I. Litter, M.A. Blesa, J. Colloid Interface Sci. 125 (2) (1988) 679.
[15] M.I. Litter, M.A. Blesa, Can. J. Chem. 70 (1992) 2502.
[16] B. Arnison, R. Segall, R. St.C. Smart, P.S. Turner, J.Chem. Soc. Faraday Trans. 1
(77) (1981) 535.
[17] N. Sato, Electrochemistry at Metal and Semiconductor Electrodes, Elsevier,
1998.
The dissolution rates of pure oxides correlate with the rates of
water substitution in aqueous Men+. Also, light accelerates dissolu-
tion when electron trapping by the surface complexes originates a
new, efficient dissolution pathway. This is true even for an n-type
semiconductor as hematite. No observable effect is found for oxides
prone to oxidative dissolution. Trapping of electrons, intrinsically
less reactive than holes, is responsible for important photodissolu-
tion. However, even in this case, light only triggers the onset of a
thermal dissolution pathway involving dissolved Fe(II).
[18] H. Tamura, E. Matijevic´, J. Colloid Interface Sci. 90 (1981) 100.
[19] U. Schwertmann, R.M. Cornell, Iron Oxides in the Laboratory, VCH, Weinheim,
1991.
Nickel ferrite dissolution is accelerated by both reductants
(Fe2+
) and oxidants (O2), demonstrating that the dissolving
[20] D.L. Leussing, N. Newman, J. Am. Chem. Soc. 78 (1956) 552.
[21] Data Base from International Center for Diffraction Data (ICDD), PDF 22-1189.
[22] Data Base from International Center for Diffraction Data (ICDD), PDF 42-1300.
[23] Data Base from International Center for Diffraction Data (ICDD), PDF 36-1451.
[24] Data Base from International Center for Diffraction Data (ICDD), PDF 33-0664.
[25] Data Base from International Center for Diffraction Data (ICDD), PDF 10-0325.
[26] Data Base from International Center for Diffraction Data (ICDD), PDF 22-1086.
[27] Data Base from International Center for Diffraction Data (ICDD), PDF 22-1012.
[28] C.G. Hatchard, C.A. Parker, Proc. R. Soc. A 235 (1956) 518.
[29] M.A. Blesa, H.A. Marinovich, E.C. Baumgartner, A.J.G. Maroto, Inorg. Chem. 26
(1987) 3713.
[30] G. Furrer, W. Stumm, Chimia 37 (1983) 338.
[31] M.A. Blesa, A.D. Weisz, P.J. Morando, J.A. Salfity, G.E. Magaz, A.E. Regazzoni,
Coord. Chem. Rev. 196 (2000) 31.
[32] B. Zinder, G. Furrer, W. Stumm, Geochim. Cosmochim. Acta 50 (1986) 1861.
[33] E. Wieland, B. Wherli, W. Stumm, Geochim. Cosmochim. Acta 52 (1988) 1969.
[34] C.A. Figueroa, E.E. Sileo, P.J. Morando, M.A. Blesa, J. Colloid Interface Sci. 244
(2001) 353.
interface contains appreciable amounts of both Ni(II) and Fe(III).
Dissolution of zinc and cobalt ferrites on the other hand is only
accelerated by reductants, because the leached layer contains low
concentrations of the divalent cation.
There is a linear relationship between kMe and k
for simple
ꢀw
oxides, and, although somewhat blurred, for spinel ferrites. The
lability of the metal ion governs the rate of dissolution. Fe(III) arrests
more the reactivity of the more labile ions, thus producing a lower
slope and a worse linear correlation. Photochemical dissolution is
important in Co and Ni ferrites, whereas it is ineffective for CoO, NiO
and ZnO. More important than the generation of electron/hole pairs
in the oxide, the well-known photochemistry of Fe(III) in oxalic
media seems to account for the results.
[35] S. Jacobo, A.E. Regazzoni, M.A. Blesa, J. Mater. Sci. 37 (2002) 3053.
[36] M. Eigen, R.G. Wilkins, Adv. Chem. Ser. 49 (1965) 55.
[37] F.M.M. Morel, J.G. Hering, Principles and Applications of Aquatic Chemistry,
Wiley-Interscience, New York, 1993.
Acknowledgments
[38] R.M. Sellers, W.J. Williams, Faraday Discuss. Chem. Soc. 77 (1984) 265.
[39] M.A. Blesa, A.J.G. Maroto, P.J. Morando, J. Chem. Soc. Faraday Trans. I 82 (1986)
2345.
This work was carried out with Grants from ANPCyT (PICT
13534), CONICET (PIP 5978) and CNEA (Program P5), and was part
of the LAGR Ph.D. Thesis.
´
[40] E.E. Sileo, L. Garcıa Rodenas, C.O. Paiva-Santos, P.W. Stephens, P.J. Morando,
M.A. Blesa, J. Solid State Chem. 179 (2006) 2237.
[41] K. Nii, Y Hisamatsu, Trans. NRIM 8 (5) (1966) 27.
[42] F. Elgersema, G.F. Kamst, G.J. Witkamp, G.M. van Rosmalen, Hydrometallurgy
29 (1992) 173.
References
´
[43] J. Domenech, A. Prieto, J. Phys. Chem. 90 (1986) 1123.
[44] G. Rogers, G. Simkovich, K. Osseo-Asare, Hydrometallurgy 10 (1983) 313.
[45] T.D. Waite, A. Torikov, J.D. Smith, J. Colloid Interface Sci. 112 (1986) 412.
[46] N.N. Kalyazin, A.F. Nechaev, Russ. J. Phys. Chem. 58 (1984) 693.
[47] E.C. Baumgartner, M.A. Blesa, H.A. Marinovich, A.J.G. Maroto, Inorg. Chem. 22
(1983) 2224.
[48] M.I. Litter, E.C. Baumgartner, G.A. Urrutia, M.A. Blesa, Environ. Sci. Technol. 25
(11) (1991) 1907.
[49] V.A. Myamlin, Y.V. Pleskov, Electrochemistry of Semiconductors, Plenum
Press, New York, 1967.
[1] M.A. Blesa, P.J. Morando, A.E. Regazzoni, Chemical Dissolution of Metal
Oxides, CRC Press, Boca Raton, FL, 1994.
[2] D.A. Dzombak, F.M.M. Morel, Surface Complexation Modeling: Hydrous Ferric
Oxide, Wiley-Interscience, New York, 1990.
[3] A.D. Weisz, Ph.D. Thesis, Universidad de Buenos Aires, 2001.
[4] C. Ludwig, W.H. Casey, P.A. Rock, Nature 375 (1995) 44.
[5] C. Ludwig, W.H. Cassey, J. Colloid Interface Sci. 178 (1996) 176.
[6] D. Suter, S. Banwart, W. Stumm, Langmuir 7 (1991) 809.
[7] B. Wehrli, B. Sulzberger, W. Stumm, Chem. Geol. 78 (1989) 167.