Kineticomechanistic Study of the Redox pH Cycling Processes Occurring on a Robust Water-Soluble Cyanido-Bridged Mixed-Valence {CoIII/FeII}2 Square

Article abstract of DOI:10.1021/acs.inorgchem.8b01147

A kineticomechanistic study of reversible electron-transfer processes undergone by the water-soluble, cyanido-bridged mixed-valence [{Co^III{(Me)₂(μ-ET)cyclen}}₂{(μ-NC)₂Fe^II(CN)₄}₂]^2- square has been carried out. The oxidation reaction consists of a two-step process with the participation of a solvent-assisted outer-sphere complex, as a result of the establishment of hydrogen bonds that involve the oxo groups of the oxidant (peroxodisulfate) and the terminal cyanido ligands of the tetrametallic square. The formally endergonic reduction reaction of the fully oxidized ([{Co^III{(Me)₂(μ-ET)cyclen}}₂{(μ-NC)₂Fe^III(CN)₄}₂]) core by water, producing hydrogen peroxide from water even at low pH values, is also a two-step process. Each one of these processes requires a set of two preequilibria involving the association of OH^- and its subsequent deprotonation by a further OH^- anion. The structure of the square compound in its fully protonated form has also been determined by X-ray diffraction and shows the existence of strong hydrogen-bonding interactions, in agreement with the rather high basicity of the terminal cyanido ligands. Likewise, density functional theory calculations on the tetrametallic complex showed zones with negative electrostatic potential around the Fe^II centers of the square that would account for the establishment of the hydrogen bonds found in the solid state. Spectroelectrochemistry experiments demonstrated the singular stability of the {Co^III/Fe^II}₂^2- complex, as well as that of their partially, {Co₂^III/Fe^IIIFe^II}^-, and fully, {Co^III/Fe^III}₂, oxidized counterparts because no hysteresis was observed in these measurements.

Full text of DOI:10.1021/acs.inorgchem.8b01147

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Kineticomechanistic Study of the Redox pH Cycling Processes
Occurring on a Robust Water-Soluble Cyanido-Bridged Mixed-
Valence {CoIII/FeII}2 Square
†
‡
§
†
Laura Alcazar, Paul V. Bernhardt, Montserrat Ferrer,† Merce Font-Bardia, Albert Gallen,
́
̀
Jesus Jover, Manuel Martínez,*,† Jack Peters,† and Timothy J. Zerk‡
†
́
†
̀
̀
́
̀
̀
Departament de Química Inorganica i Organica, Seccio de Química Inorganica, Universitat de Barcelona, Martí i Franques 1−11,
E-08028 Barcelona, Spain
‡School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane 4072, Queensland, Australia
§
́
̀
́
Unitat de Difraccio de RX, Centres Científics i Tecnologics de la Universitat de Barcelona, Universitat de Barcelona, Sole i Sabarís
1−3, E-08028 Barcelona, Spain
S
* Supporting Information
ABSTRACT: A kineticomechanistic study of reversible electron-
transfer processes undergone by the water-soluble, cyanido-bridged
mixed-valence [{CoI I I {(Me)2 (μ-ET)cyclen}}2 {(μ-
NC)2FeII(CN)4}2]2− square has been carried out. The oxidation
reaction consists of a two-step process with the participation of a
solvent-assisted outer-sphere complex, as a result of the establish-
ment of hydrogen bonds that involve the oxo groups of the oxidant
(peroxodisulfate) and the terminal cyanido ligands of the
tetrametallic square. The formally endergonic reduction reaction of
the fully oxidized ([{CoIII{(Me)2(μ-ET)cyclen}}2{(μ-
NC)2FeIII(CN)4}2]) core by water, producing hydrogen peroxide
from water even at low pH values, is also a two-step process. Each
one of these processes requires a set of two preequilibria involving
the association of OH− and its subsequent deprotonation by a further OH− anion. The structure of the square compound in its
fully protonated form has also been determined by X-ray diffraction and shows the existence of strong hydrogen-bonding
interactions, in agreement with the rather high basicity of the terminal cyanido ligands. Likewise, density functional theory
calculations on the tetrametallic complex showed zones with negative electrostatic potential around the FeII centers of the
square that would account for the establishment of the hydrogen bonds found in the solid state. Spectroelectrochemistry
experiments demonstrated the singular stability of the {CoIII/FeII}2 complex, as well as that of their partially, {Co2
/
2−
III
FeIIIFeII}−, and fully, {CoIII/FeIII}2, oxidized counterparts because no hysteresis was observed in these measurements.
Although the number of examples containing redox-active
bridging ligands (based on units such as tetrathiafulvalene or
perylene bisimide) or blocking ligands (bearing moieties such
as ferrocenyl) is much larger than that including redox-active
metal corners, some macrocycles containing CuII/CuIII, CoII/
CoIII, CoIII/CoIV, FeII/FeIII, RuII/RuIII, or MoII/MoIII as redox-
active pairs have been reported.6−10 Surprisingly, these reports
lack any reference to discrete cyanido-bridged (Prussian blue)
species, although these compounds usually display redox
activity because of the presence of multiple Fe and/or Co
oxidation states.11 Nevertheless, an increasing number of
interesting examples of Prussian blue molecular polymetallic
assemblies have been reported because of the fact that they
usually exhibit both optical and magnetic bistability, which
arises from a reversible metal-to-metal charge transfer between
INTRODUCTION
■
Self-assembly of macrocyclic metal complexes is a powerful
method for the construction of well-defined architectures with
interesting structure-specific properties such as guest recog-
nition, optical and/or magnetic properties, or multiredox
behavior among many others.1−3 Given the fact that a large
variety of metallamacrocycles are built with redox-active
transition-metal fragments as corners and/or redox-active
ligands as linkers, it is rather surprising that the number of
studies that specifically aim at taking advantage of that inherent
property is still rather limited. Nevertheless, in recent years,
there has been a growing interest in these redox-active species
because of their potentially useful properties (optical,
magnetic, and catalytic activity, molecular recognition, etc.).
Published comprehensive reports4,5 provide examples of
electroactive metallic assemblies whose redox activity is
localized in the organic ligands and/or metallic corners.
Received: April 27, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Mechanistic Process Involved in the Preparation of the Cyanido-Bridged [{CoIII{(Me)2(μ-ET)cyclen}}2{(μ-
NC)2FeII(CN)4}2]2− Square Complex of This Study
Figure 1. X-ray crystal structure of (H3O)6[{CoIII{(Me)2(μ-ET)cyclen}}2{(μ-NC)2FeII(CN)4}2]·(ClO4)4·4H2O. (a) Square unit of the mixed-
valence complex. The H atoms, solvent, and counterions are omitted for clarity. Selected bond lengths (Å) and angles (deg): Co1−N10 1.909(3),
Co1−N5 1.918(3), Co1−N2 1.925(3), Co1−N3 1.935(3), Co1−N1 1.982(3), Co1−N4 1.989(3), Fe1−C17 1.880(4), Fe1−C14 1.885(4), Fe1−
C15 1.896(4), Fe1−C18 1.902(4), Fe1−C13 1.910(4), Fe1−C16 1.921(4); N10−Co1−N5 88.81(13), C18−Fe1−C13 90.49(15), C18−N10−
Co1 173.1(3), N10−C18−Fe1 173.5(3), C13−N5−Co1 171.7(3), N5−C13−Fe1 176.0(3). (b) Part of the extensive 3D network found in the
structure, showing the variety of hydrogen-bond lengths.
the CoIII (or FeIII) and FeII centers. This family of compounds
across the full aqueous pH range, in deep contrast with that of
other reported redox-active metallomacrocycles) opens up the
possibility of performing a thorough study of its participation
in chemical processes, such as reversible oxidation, proto-
nation, and OH− association.
With these facts in mind, we are here reporting a redox-
reversible kineticomechanistic study of this tetrametallic
{CoIII/FeII}2 unit as a function of pH, oxidant concentration,
temperature, and pressure. In addition, a study of the charge
distribution of the fully reduced species and the one- and two-
electron oxidized forms by density functional theory (DFT)
calculations has been carried out in order to rationalize our
experimental observations. The results indicate an extreme
robustness of the square unit across the aqueous pH range, as
well as a significant contribution from specific outer-sphere
includes pentanuclear trigonal-bipyramidal{FeII /FeIII3} and
2
{FeII /CoIII3} units,12,13 octanuclear cubic {FeII/FeIII}4 and
{FeII2/CoIII}4 assemblies,14,15 homonuclear {FeII/FeIII}2
squares,16,17 and a significant number of heterotetranuclear
square complexes {FeII/CoIII}2.11,18−25 Recently, we have
reported the preparation of the diamagnetic cyanido-bridged
[{CoIII{(Me)2(μ-ET)cyclen}}2{(μ-NC)2FeII(CN)4}2]2−
square complex26 via a mechanistically designed, self-assembly
reaction (Scheme 1). The process involves a rate-limiting
outer-sphere redox reaction, followed by a fast substitution/
inner-sphere redox reaction, sequence that had been studied by
us for some time.27−34 Although this compound did not show
any evidence of complete intramolecular electron-transfer
phenomena, its striking stability in aqueous solution (stable
B
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. (a) Spectroelectrochemistry of a 2.0 × 10−3 M solution of {CoIII/FeII}22− in H2O (pH 7 and I = 0.1 M LiClO4). A platinum flag working
electrode, an Ag/AgCl/saturated NaCl reference electrode, and a platinum wire counter electrode were used. Cell path length = 1.0 mm. (b)
ReactLab-calculated spectra of the fully oxidized {CoIII/FeIII}2, fully reduced {CoIII/FeII}22−, and intermediate {CoIII2/FeIIIFeII}− complexes
II
II
III(1)
III(2)
= 0.7322 V vs NHE].
[aqueous solution, pH 7, I = 0.1 M LiClO4, EFe /Fe
= 0.628 V, and EFe /Fe
recognition of water molecules, which actuate in a way that
cannot be attributed to mere electrostatic effects. These
features allow consideration of the complexes studied as having
a versatile multiredox chemistry with implications in electro-
chromism.
Moreover, a complex 3D network is formed as a result of the
establishment of extensive hydrogen-bonding interactions that
involve the O atoms of the H2O molecules, H3O+ cations, and
−
ClO4 anions and the N atoms of all terminal cyanido ligands
(Figure 1b). Interestingly, three of these interactions are
singularly strong [O2W···N6 = 2.525(5) Å, O6W···N7 =
2.554(5) Å, and O4W···N9 = 2.573(5) Å] in comparison with
those reported for analogous square {CoIIIFeII}2 com-
pounds,20,24,37 in agreement with the basic character exhibited
by the terminal CN groups of the compound.
RESULTS AND DISCUSSION
■
Crystal Structure of (H3O)6[{CoIII{(Me)2(μ-ET)-
cyclen}}2{(μ-NC)2FeII(CN)4}2]·(ClO4)4·4H2O. Even though
the central core of the title compound has already been
inferred via several techniques,26 its crystal structure was
unknown. Nevertheless, knowledge of the protonation and
redox properties of the compound has enabled us to design a
way to produce it in a form that could be crystallized.
Caramel-colored crystals of the title complex were obtained
by slow concentration of a 1 M HClO4 solution of the parent
compound Na2[{CoII I {(Me)2 (μ-ET)cyclen}}2 {(μ-
NC)2FeII(CN)4}2] at 4 °C. The structure contains centrosym-
metric cyanido-bridged {Fe2(μ-CN)4Co2} units bearing a
nearly planar, square-shaped geometry, in which the Co and
Fe ions are bridged by cyanide ions. The anion structure,
together with selected bond distances and bond angles, is
shown in Figure 1a.
Spectroelectrochemistry: Redox Stability. We previ-
ously measured the cyclic voltammetry of the title complex in
an aqueous solution and showed that it exhibits two discrete
FeIII/II redox couples.26 Both FeIII/II redox couples are
reversible, which indicates that the complex is stable in each
of the fully oxidized FeIIIFeIII, fully reduced FeIIFeII, and
intermediate FeIIIFeII forms, while the CoIII/II redox couples are
irreversible under the same conditions. Herein we measured
the spectroelectrochemistry of the complex at pH 7, poising
the potential at increasingly positive potentials (oxidation) and
then returning incrementally to the starting potential
(reduction). At each poised potential, the observed spectrum
was stable for a minimum of 1 h, and no hysteresis was
observed in the spectra of the oxidation and reduction sweeps
(Figure S2), which testifies to the stability of the complex.
Modeling the data from Figure 2a using ReactLab Redox36
revealed the spectrum of the intermediate FeIIIFeII complex
(Figure 2b, blue curve). The separation of the two FeIII/II
couples (ΔE > 100 mV) is much larger than that expected for
two noninteracting Fe centers (2RT/F ln 2 = 35.6 mV),38
which shows that the charge increase at one oxidized Fe center
influences the redox potential of the other Fe center, thus
raising its redox potential.
The distances between adjacent Fe and Co centers through
the cyanide linkers are very similar [4.936(2) and 4.956(2) Å],
and the angles Fe1−Co1−Fe1 and Co1−Fe1−Co1 are
88.47(1) and 91.53(1)°, respectively, thus displaying a regular
square macrocyclic core, in good agreement with other
reported examples of squares containing {CoIII−NC−FeII}
edges.18,20,35 The coordination bond lengths about the Co ion
are in the range of 1.909(3)−1.989(3) Å, characteristic of low-
spin (LS) CoIII ions.12 The Fe−C distances are within
1.880(4) and 1.921(4) Å, confirming the LS character of the
Fe ions present in the compound and its divalent oxidation
state.23 The total compound charge, as well as the absence of a
ligand-to-metal charge-transfer (LMCT; FeIII−CN) band near
400 nm (characteristic of the ferricyanide chromophore) in the
electronic spectrum, is in agreement with the FeII centers in the
complex.
DFT Calculations. In view of the feasibility of pursuing the
kineticomechanistic studies of the reversible redox processes
III
indicated in Figure 2 and given the fact that the {Co
/
FeII}2 core is extremely robust, DFT calculations have been
conducted in order to investigate the possible differences
among the {CoIII/FeII}22−/{CoIII2/FeIIIFeII}−/{CoIII/FeIII}2
triad that could explain the reactivity observed.
2−
In addition to the discrete tetranuclear metal units
[{CoIII{(Me)2(μ-ET)cyclen}}2{(μ-NC)2FeII(CN)4}2]2−, per-
chlorate anions, hydronium cations, and lattice H2O molecules
are also found in the asymmetric unit. The anionic squares are
stacked, forming columns along the b axis (Figure S1) with an
angle of 57.4° between the planes of consecutive macrocycles.
These structures were optimized with DFT calculations
using the crystal structure of the {Co III/FeII}2 core as the
2−
starting point (the detailed procedure is described in the
Experimental Section). From the calculations, the oxidation
process does not entail major changes in the structure; the
C
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
central square unit of the {CoIIIFeII}2 complex remains
practically the same, with Fe···Co, Fe···Fe, and Co···Co
distances of 4.99, 6.93, and 7.20 Å, respectively. These values
are very similar to those found in the crystal structure, where
the average Fe···Co separation is 4.95 Å, while the Fe···Fe and
Co···Co distances are 6.90 and 7.09 Å. In the same line, the
computed Fe···Co···Fe and Co···Fe···Co angles are close to
those in the crystal structure: 87.8 versus 88.5° and 92.2 versus
91.5°, respectively. In all cases, as in the determined crystal
structure, the four metals remain in the same plane and the
octahedral Fe(CN)6 units show an eclipsed arrangement, with
twist angles between the iron hexacyanido units always below
light-blue surface for {CoIII/FeII}22−, orange-to-blue surface for
{CoIII2/FeIIIFeII}−, and green-to-dark-blue surface for {CoIII/
2−
FeIII}2 in Figure 4. The MEP map of {CoIII/FeII}2 clearly
2−
reflects the bonding found in the crystal structure, where the
square tetrametallic core interacts with H3O+ through
hydrogen bonding of the terminal, negatively charged, cyanido
ligands bound to the Fe centers. The different FeII and FeIII
atoms in {CoIII2/FeIIIFeII}− can be clearly distinguished with
the MEP map; the former is located in a zone with negative
electrostatic potential (orange, Figure 4b, top), while the latter
belongs to a space where the surface charge is more postive
(yellow, Figure 4b, bottom).
0.5°. The consecutive oxidation of the {CoIIIFeII}2 complex
2−
The {CoIII/FeIII}2 complex reacts with hydroxide to produce
2−
hydrogen peroxide (H2O2) and the reduced {CoIII/FeII}2
produces open-shell species with one unpaired electron in the
oxidized Fe centers. For the half-oxidized {CoIII2/FeIIIFeII}−
species, the unpaired electron is localized on the new FeIII
center (Figure 3a), while for the fully oxidized {CoIII/FeIII}2
species (see below). The experimental data indicate that this is
an outer-sphere redox process. Nevertheless, the small size of
the OH− anion could suggest an encapsulation of the anion in
the central cavity of the tetranuclear metallic core prior to
electron transfer. This hypothesis has been tested by
optimizing a {CoIII/FeIII}2OH− species in which the hydroxide
anion is placed in the geometrical center of the cavity. After
optimization, the hydroxide group is observed to migrate to
the external part of the complex, where it can interact with the
dangling methyl groups of the cyclen ligand. Because the metal
centers surrounding the cavity do not exert any stabilizing
effect on the incoming OH− group, a mechanism where the
hydroxo groups are inside the cavity would not be expected.
Kinetics of the Oxidation Reaction of the {CoIII/FeII}2
2−
Figure 3. Electron spin densities computed for complexes (a) {CoIII
/
2
FeIIIFeII}− and (b) {CoIII/FeIII}2. Color code: C, gray; N, blue; Fe,
brown; Co, purple. H atoms have been omitted for clarity.
Complex. The oxidation reaction of simpler but related
(CoIII/FeII)− mixed-valence dinuclear complexes, sodium μ-
cyanido-1κN:2κC-(2-methyltetraazacycloamine)-1κ5N-penta-
cyanido-2κ5C-[1-cobalt(III)-1-ferrate(II)], with peroxodisul-
fate has been already studied, and the behavior has been
explained as an initial one-electron oxidation of FeII to FeIII
complex, the species can be either a triplet or an open-shell
singlet; in practice, both possibilities have practically the same
energy and display very similar structures. All of the results
appearing throughout this manuscript refer to the open-shell
singlet {CoIII/FeIII}2 complex. As in the {CoIII2/FeIIFeIII}−
species, the electron spin density remains localized on the FeIII
centers (Figure 3b).
The molecular electrostatic potential (MEP) maps show the
formal charge distribution of the total electron density of the
complexes, i.e., the regions with charge depletion and
accumulation that can be used to explain the reactivity and
outer sphere interactions (Figure 4). As expected, the
positively charged areas are always located around the Co
centers, while the Fe surroundings are more negative in all
species. The maximum and minimum surface charge values
vary according to the total charge of the complexes, i.e., red-to-
producing SO4 and SO4•−, followed by a fast (non-rate-
2−
•−
determining) second one-electron process involving the SO4
radical.30,34,39,40 In view of the spectroelectrochemical data
shown above, as well as the redox nonequivalence of the two
Fe centers,26 the oxidation of the new {CoIII/FeII}2 mixed-
2−
valence square was pursued from a kineticomechanistic
perspective at varying concentrations of peroxodisulfate, pH,
temperature, and pressure. As collected in Table S1, in all cases
the time-resolved spectra agree with the operation of a two-
step consecutive process in the acidity range between pH 8.5
and 1 M HClO4. Furthermore, the Specfit- or ReactLab-
Kinetics calculated spectra41,42 of the intermediate compound
agree with that of the single-electron-oxidized {CoIII
/
2
FeIIIFeII}− mixed-valence compound determined from the
spectroelectrochemical measurements (see Figures 2 and S3).
Figure 5 collects an example of the trends observed for the two
pseudo-first-order rate constants derived (kobs1 and kobs2) with
the concentration of the oxidant.
By a comparison with our previous data on the (CoIII/FeII)
dinuclear complexes, the reaction can be explained by the
sequence shown in Scheme 2, which produces a rate law such
as that indicated in eq 1 for both rate-determining reaction
steps.43−45 Nevertheless, it is clear that for the observation of
the second step in the reaction sequence (ox2a and ox2b), the
•−
SO4 radical generated by the first oxidation step (ox1a and
2−
Figure 4. MEP maps on the 0.002 isosurfaces of (a) {CoIIIFeII}2
,
(b) {CoIII2/FeIIIFeII}−, and (c) {CoIIIFeIII}2. Red and blue colors
represent negative and positive values of the electrostatic potential
(VE).
ox1b) has to react preferentially with an excess of the reducing
2−
{CoIII/FeII}2 to further produce the {CoIII2/FeIIIFeII}−
intermediate (oxf1) . Any other option would make the
D
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
From the limiting value obtained from fitting the data to eq
1, the values of kox and KOS at different temperatures and pH
can be derived for both reaction steps (ox1 and ox2, Scheme
2). By measurement of the temperature dependence of kox and
fitting the data to the Eyring equation (Figure 6a), the
corresponding thermal activation parameters can be calculated
for both electron-transfer processes. The pressure dependence
2−
of the values of kobs at concentrations of S2O8 at which the
plot has practically reached saturation (kox ≈ kobs at [S2O82−] =
0.05−0.07 M) was also studied in order to determine the
activation volumes of the two steps. Figure 6b shows some of
the ln kox versus P plots at the pH where a definite neat
oxidation is observed (see the next section),46 and Table 1
collects a summary of the equilibrium kinetic and activation
data determined.
Figure 5. Plot of the values of the first-order rate constants, kobs, for
2−
2−
the two-step S2O8 oxidation of the {CoIII/FeII}2 mixed-valence
square at different temperatures and acidities.
Scheme 2
Kinetics of the {CoIII/FeIII}2 Complex Reduction. From
our previous studies carried out on the simpler (CoIII/FeII)−
mixed-valence dinuclear complexes, a pH-dependent reversible
redox process has been demonstrated.30,40 Consequently, the
reduction process of the new {CoIII/FeIII}2 square (obtained
by peroxodisulfate oxidation in the acidic conditions of {CoIII/
26
FeII}2
)
in alkaline solutions to produce H2O2 was also
2−
studied from a kineticomechanistic perspective. As indicated in
the Experimental Section, the presence of H2O2 in the final
reaction mixture was confirmed by standard peroxide test kits;
no reaction of the parent {CoIII/FeII}22− square with H2O2 has
been observed with excesses of up to 100-fold. Interestingly,
for this new tetranuclear {CoIII/FeIII}2 compound, the
reduction process reverting to the mixed-valence {CoIII/
FeII}2 square complex, with the formation of H2O2, is
2−
observed even at acidic pH values (the reaction is observed
from pH 3 upward). As found for the previously described
oxidation reaction by peroxodisulfate, the reduction process is
clearly a consecutive two-step sequence from the time-resolved
fitting of the full spectra by the standard software indicated in
the Experimental Section.42,43
Figure 7 summarizes the trends observed for the values of
kobs1 and kobs2 with the concentration of OH−. For the
reduction of the simpler (CoIII/FeIII)− dinuclear complexes,
two equilibrium processes involving the OH− reductant
precede the rate-determining formation of peroxide; these
are the formation of an OH− adduct, followed by its
deprotonation by another OH−. Thus, an [OH−]2 dependence
of the kobs rate constants is observed.40 In the present case, this
is not so, and a double titration plot is observed with pKeq
values within the 3−4 and 11−12 ranges. Table S2 collects all
reaction sequence show a single rate-determining step
behavior.
koxKOS[S2O82−
]
kobs
=
1 + KOS[S2O82−
]
(1)
From the curvature shown in the plots, it is clear that both
kobs rate constants show the typical limiting behavior obtained
for outer-sphere redox processes, where the formation of the
precursor outer-sphere complex (KOS) is dominant.44 Although
this effect had not previously been observed for oxidations by
S2O82−, the reaction of the dinuclear (CoIII/FeII)− species with
[CoIII(oxalate)3]3− effectively showed this behavior. Important
interactions in the outer-sphere precursor complex had been
established as responsible of this fact.29,40
Figure 6. (a) Eyring plot for variation of the values of kox for the two steps observed with the temperature at pH 4.0. (b) ln kox versus P plot for the
reaction carried out at pH 2.0 and 25 °C.
E
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Thermal and Pressure Activation Parameters for the Two Consecutive Reaction Steps Observed in the Oxidation of
a
the {CoIII/FeII}2 Mixed-Valence Square with S2O8 at Different pH Values
2−
2−
pH
K
OS1/M−1
298kox1/s−1
ΔHox1⧧/kJ mol−1
ΔSox1⧧/J K−1 mol−1
ΔVox1⧧/cm3 mol−1
8.5
4
1.5
0
70
100
90
150
OS2/M−1
1.3 × 10−3
5.1 × 10−4
3.8 × 10−4
0.62 (293 K)
298kox2/s−1
70
80
100 15
20
7
8
−90 20
−55 25
20 40
n.d.
n.d.
−38 4 (pH 2)
ca. 0
ΔVox2⧧/cm3 mol−1
5
−180 20
pH
K
ΔHox2⧧/kJ mol−1
ΔSox2⧧/J K−1 mol−1
8.5
4
1.5
0
30
55
40
40
1.8 × 10−4
90 10
−30 30
−26 20
18 40
n.d.
n.d.
2.2 × 10−4
90
6
2.0 × 10−4
1.2 × 10−3 (293 K)
100 12
80 10
−41 3 (pH 2)
−5.2 0.8
−20 50
a
The values of KOS (according to eq 1) are the average from the values at the different temperatures measured.
from the data. In this scheme, the values of kred1 and kred2
necessarily include the two equilibrium constants involving
hydroxide, which are in the above-mentioned range.
From the values obtained, it is clear that the validity of the
previous oxidation data collected at pH >3 has to be
considered with caution because a faster reduction of the
oxidized complex is occurring simultaneously in an approx-
imately catalytic reaction. Probably the observed process
corresponds de facto to an oxidation reaction but once the
buffer has been consumed.
In this regard, cyclic voltammetry experiments were
conducted at variable pH values in the range where a dramatic
increase on the reduction rate constants is observed (7−13;
Figure 7). The purpose of these measurements was to examine
the possible occurrence of an electrochemically catalyzed
oxidation of OH− to H2O2. Figure 8 shows that the fully
reduced complex is reversibly oxidized in two discrete steps to
the {CoIII/FeIIIFeII}− and {CoIII/FeIII} complexes at pH 7.
However, at higher pH values, the voltammetry is slightly
perturbed and the peak oxidative current (ipa) of the first wave
increases slightly above pH 10. This behavior is expected for an
electrochemical oxidation coupled to a catalytic chemical
reaction (Scheme S1) and confirms that hydroxide is only
oxidized by {CoIII/FeIII}2 at an appreciable rate above pH 10
under the nonbuffered experimental conditions used here. Two
clarifications should be noted in relation to Figure 8. First, the
nonspecific oxidation of hydroxide by the electrode is only
observed at very high potentials (>0.85 V vs NHE) and is not
responsible for the change in the peak anodic current (ipa).
Second, even though the change in ipa is small, it is definitive;
this is indicated by the stable baseline current during the initial
Figure 7. Logarithmic-scale plot of the fast and slow rate constants for
the two steps observed upon reduction of the {CoIII/FeIII}2 square
with OH− at different buffered pH values at 25 °C ([square] ≈ 4 ×
10−4 M; I = 1.0 M (NaClO4)).
of the values kobs obtained under the different conditions of the
study, and Scheme 3 collects the reaction sequence that results
Scheme 3
Figure 8. (a) pH-dependent voltammetry of a 7.3 × 10−4 M solution of {CoIII/FeII}22− in H2O (I = 0.1 M LiClO4. Scan rate = 10 mV s−1. Potential
initially swept in the positive direction. (b) pH dependence of the peak anodic current (ipa) at 0.662 V.
F
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
oxidation sweep (i.e., the current at 0.4 V vs NHE is the same
across the range of pH values during the initial sweep in the
positive direction).
and the nonprotonated and protonated nitrile units can explain
these facts (Schemes 4 and S2).51,52
As for the kinetic and activation parameters at pH > 2 (see
Table 1), there is a dramatic change going from the single (pH
1.5−2) to the doubly protonated mixed-valence square
complex (pH 0). The first step of the oxidation reaction
(kox1) is accelerated by a factor of 103, while the second step
rate constant undergoes only a 10-fold increase. These values
are clearly related to the thermal activation ΔH⧧ and ΔS⧧
values, which indicate a much higher degree of association in
the transition state and a lower enthalpy demand with
decreasing pH (both effects are more pronounced for the
first rate-determining step of the oxidation reaction).
Interestingly, the redox potential of the Fe centers becomes
more positive upon protonation,26,30 which would be expected
to produce higher enthalpy demands. Again, the proper
association of the oxidant and reductant via a solvent molecule
should be held responsible for the acceleration found (vide
supra); alternatively, direct interaction of the peroxodisulfate
anion with protonated nitrile groups could also explain the
changes (Scheme S3). As for the volumes of activation, these
show trends opposite to those of ΔS⧧, in agreement with the
operation of hydrogen-bonding interactions in the transition
state, and are mostly related to desolvation of the complexes, as
found for similar systems.48,49,53
2−
Mechanism of the {CoIII/FeII}2 Oxidation Process.
From the data collected, a key difference emerges between the
2−
oxidation kinetics of the dianionic {CoIII/FeII}2 square
complex and the parent monoanionc series of (CoIII/FeII)−
dinuclear species previously reported. While for the mixed-
valent monoanionic (CoIII/FeII)− dinuclear complexes the
buildup of outer-sphere precursor complexes with peroxodi-
sulfate (according to eq 1) is minimal, in the present case, the
equilibrium constants, KOS, indicated in Scheme 2 are large
enough for the observation of limiting kinetics in both
consecutive oxidation reaction rate-determining steps (Table
1 and Figure 5).44 In the previous studies with the (CoIII/
FeII)− dinuclear species, the only time this behavior was
observed was when [Co(oxalate)3]3− was used as the oxidant,
despite the unfavorable increase in charge repulsion.40 This
fact was attributed to a solvent-assisted nature of the outer-
sphere complex formed, which was also thoroughly described
for similar [Co(oxalate)3]3− reactivity.47 It is thus clear that in
the present case an outer-sphere solvent-assisted encounter
complex is formed with S2O82−, both with the dianionic {CoIII/
FeII}2 and monoanionic {CoIII2/FeIIIFeII}− species, in the
2−
first and second rate-determining reaction steps, respectively.
Furthermore, the fact that the ox2a and ox2b reaction steps are
observed in the oxidation process (Scheme 2), despite the
possible operation of reaction oxf2, indicates that the solvent-
assisted outer-sphere complex formed in ox1a involves solely
one of the oxo groups of peroxodisulfate, thus liberating the
Mechanism of the {CoIII/FeIII}2 Reduction Process. The
data collected and shown in Figure 7 clearly indicate that, as
found in previous studies,40,54 even though the reduction
potentials do not predict the formation of H2O2 (E° = +1.77
V), H2O2 is formed in a process that appears to be formally
endergonic. In any case, the reactivity of the iron hexacyanido/
H2O2 redox system is strongly dependent on the counterions,
pH, light, and reaction medium.55,56 Most of the processes lead
to decomposition of H2O2 or photochemical degradation of
the hexacyanido complexes to produce Prussian-blue-type
compounds. We have not observed this type of degradation on
our samples; furthermore, most of the FeII oxidations by H2O2
processes involve the coordination of a hydroperoxide anion to
a susbtitutionally active complex, which is not the case in the
system that we are dealing with here. There must be some sort
of stabilization of the peroxide in our systems, probably related
to adduct formation similar to that producing hydrolysis of the
cyanido ligands in other systems.
•−
resulting SO4 radical to the reaction medium. In this way, a
two-electron oxidation process is not operative, and the
liberated radical can react with both {CoIII/FeII}2 and
2−
{CoIII2/FeIIIFeII}−, ensuring the observation of the second step
(ox2a and ox2b, Scheme 2) of the full process (Scheme 4).
Scheme 4
Interestingly, although the redox potential of the Fe center
increases with decreasing pH, the trend observed for the rate
constants measured for both reaction steps (Scheme 3) is
opposite. The previously studied similar reactions on (CoIII/
FeIII) dinuclear complexes40,54 indicated that the formation of
an outer-sphere OH− aggregate and its further deprotonation
by a second OH− group agree with the data obtained; a
[OH−]2 dependence was observed for the values of kobs. In the
present study, such a trend is not observed and, although in the
pH range between 5 and 10 the increases of both kred1 and kred2
are practically linear, the values increase dramatically upon
going to pH 11 to reach a final plateau. Increasing the acidity
to pH 3 produces an opposite immediate decrease of the
In this respect, from the data in Table 1, it is also clear that
the values of KOS decrease by ca. half upon going from the
{CoIII/FeII}2 to {CoIII2/FeIIIFeII}− complexes, opposite to
2−
that expected considering the electrostatics involved; clearly,
some other factor has to be dominant. The only possibility to
come to terms with this fact relates precisely again with the
existence of a solvent-assisted outer-sphere complexation,
where H2O is more tightly bound to more anionic
complexes.48−50 Nevertheless, when the fully protonated
H2{CoIII/FeII}2 species is considered (pH 0), the trend is
maintained with equivalent (if not larger) values of KOS, thus
implying that the outer-sphere solvent-assisted complexation
should occur on the nonprotonated nitrile units; even
simultaneous hydrogen bonding between a molecule of H2O
second-step rate constant and a buildup of the {CoIII
/
2
FeIIIFeII}− intermediate complex; a further increase in the
acidity makes both reduction processes too slow to be
observed.
From these data, it is clear that outer-sphere involvement of
OH− with the oxidized {CoIII/FeIII}2 square complex plays a
crucial role in its reduction to the mixed-valence parent species
G
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
and the formation of H2O2. In this respect, it is interesting to
note that the aggregation of hydroxide with the reduced
counterparts are remarkable, in accordance with the absence of
hysteresis observed in these measurements. DFT calculations
have also been conducted, and the results agree with the
{CoIII/FeII}2 has already been established by UV−vis
2−
spectroscopy.26 Furthermore, the pH values for which dramatic
changes in the rate constants are observed can be easily
associated with the deprotonation of a coordinated H2O
molecule (pKa values of around 3−5) or of a coordinated
hydroxide (pKa values of around 9−12).57 With all of these
facts in mind, we propose the reaction sequence shown in
Scheme 5 as responsible for the trends observed. From the
surprising stability and robustness of the {CoIIIFeII}2 square
2−
complex core in any of its oxidized forms. In this respect, MEP
maps support the establishment of hydrogen bonds between
H2O and the terminal cyanido ligands. In all, the remarkable
robustness displayed opens up the possibility of studying its
participation in a wide range of chemical processes in aqueous
solution.
In this respect, the kineticomechanistic study of the two-step
sequential oxidation reaction of the {CoIII/FeII}22− square with
peroxodisulfate features a striking difference with the
previously reported oxidation reaction of monoanionic
(CoIII/FeII)− dinuclear species. For the oxidation reaction of
this anionic tetranuclear square-shaped complex, the actuation
of an outer-sphere H2O-assisted encounter complex involving
an oxo group of the peroxodisulfate anion is evident. The
activation parameters are in agreement with the existence of
these hydrogen-bonding interactions in the transition state.
Reduction by H2O of the fully oxidized {CoIII/FeIII}2 square,
producing H2O2, occurs even at acidic pH values and,
according to the interpretation of the kineticomechanistic
results obtained, requires the existence of two equilibrium
processes prior to the rate-determining step. These involve the
existence of an OH− adduct, with either {Co2III/FeIIIFeII}− or
{CoIII/FeIII}, and its subsequent deprotonation by another
OH− before the rate-determining generation of H2O2.
Clearly, a fine-tuning of the building block redox potentials
via ligand design should expand the current study to further
chemically robust polynuclear mixed-valence compounds. This
robustness, with the goal being including that of the CoIII
centers upon reduction, should provide discrete polynuclear
molecules with very interesting electron-transfer properties.
Scheme 5
data in Figure 7, the values of both pKdep1 and pKdep2 are found
to be equivalent and within the 11−12 range; as expected,
because of the fact that the values of Kdep are directly
associated with the deprotonation of a {{CoIII/FeIII}2·OH−}
adduct in reactions red1a and red2b (Scheme 5). Furthermore,
compound {{CoIII/FeIII}2·H2O} is expected to be more acidic
than {{CoIII2/FeIIIFeII}−·H2O}, as is effectively shown by the
value of Kadd (pKadd1 being one unit lower than pKadd2; 2−3
versus 3−4). As a whole, the process is rather similar to the
one established for oxidation of the dinuclear (CoIII/FeII)
parent mixed-valence complexes. For the present complexes,
though, the higher acidity of the initial {CoIII/FeIII}2 oxidized
square allows for full dissociation of the initial adduct, even at
high pH. Furthermore, the difference between the acidities of
{{CoIII/FeIII}2·H2O} and {{CoIII2/FeIIIFeII}−·OH−} (or the
half-reduced intermediate species) allows for a full separation
of the effects of Kadd and Kdep for the reaction sequence
indicated in Scheme 5.
EXPERIMENTAL SECTION
■
2−
General Procedures. The {CoIII/FeII}2 square mixed-valence
compound was prepared following the procedures already described
in the literature;26 UV−vis, inductively coupled plasma mass
spectrometry, and cyclic voltammetry were used as characterization
techniques.
A solution of the above complex in 1.0 M HClO4, followed by
long-standing in a refrigerator, produced some aggregates of XRD-
quality crystals.
Buffer solutions were prepared using the standard procedures for
3.0 > pH < 11; out of that range, HClO4 or NaOH solutions were
used. In all cases, the concentration of the buffer was at least 10-fold
that of the reactants and the ionic strength was set at 1.0 with
NaClO4.58,59
Detection of H2O2 after reduction of the {CoIII/FeIII}2 oxidized
square, was carried out using Quantofix peroxide test sticks (Peroxid
25 from Sigma-Aldrich) at pH 7.3. Blank experiments in the presence
of free S2O82− or solutions of the initial {CoIII/FeII}22− square mixed-
valence compound showed no presence of H2O2.
DFT Calculations. All DFT calculations were carried out using
H2O as the solvent (PCM, see below) with the Gaussian09 (revision
D.01)60 electronic structure package with a 10−8 convergence
criterion for the density matrix elements, using the hybrid functional
B3LYP.61−63 The standard 6-31G* basis set64−66 was used for all H,
C, N, and O atoms, while the Stuttgart basis set (SDD),64,67 including
the associated effective core potential to describe the core electrons,
was employed for the Fe and Co atoms. Ultrafine integration grids
were used in all calculations to ensure a satisfactory convergence. In
all cases, the solvation energies were computed in H2O with the (IEF-
PCM) continuum dielectric solvation model68,69 using the SMD radii
and nonelectrostatic terms. The dispersion correction terms70 were
CONCLUSIONS
■
The structure of the heterotetranuclear cyanido-bridged
square-shaped [{CoIII {(Me)2 (μ-ET)cyclen}}2 {(μ-
NC)2FeII(CN)4}2]2− mixed-valence complex has been un-
ambiguously established by X-ray diffraction (XRD); the
structure shows extensive hydrogen-bonding interactions with
H2O molecules, H3O+, and the N atoms of the terminal
cyanido ligands. Spectroelectrochemical measurements indi-
cate that the stability of the complex, as well as those of their
partially ({Co2III/FeIIIFeII}−) and fully ({CoIII/FeIII}) oxidized
H
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
included in all of the calculations by using the D3 method of Grimme
et al.71 Vibrational analysis was performed for all of the computed
structures to ensure the nature of the stationary, which have zero
imaginary frequencies.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.8b01147.
■
S
X-ray Structure Analysis. A caramel prismlike specimen of
C36H74Cl4Co2Fe2N20O27, approximate dimensions 0.178 mm × 0.198
mm × 0.477 mm, was used for X-ray crystallographic analysis. The X-
ray intensity data were measured on a D8 Venture system equipped
with a multilayer monochromator and a Mo microfocus (λ = 0.71073
Å). The frames were integrated with the Bruker SAINT software
package72 using a narrow-frame algorithm. The integration of the data
using a monoclinic unit cell yielded a total of 126114 reflections to a
maximum θ angle of 30.58° (0.70 Å resolution), of which 9844 were
independent (average redundancy 12.811, completeness = 99.6%, Rint
= 5.80%, and Rsig = 2.83%) and 8040 (81.67%) were greater than
2σ(F2). The final cell constants of a = 28.454(6) Å, b = 10.644(2) Å, c
= 21.226(4) Å, α = 90.00(3)°, β = 91.14(3)°, γ = 90.00(3)°, and
volume = 6427(2) Å3 are based upon refinement of the XYZ centroids
of reflections above 20σ(I). Data were corrected for absorption effects
using the multiscan method (SADABS).73 The calculated minimum
and maximum transmission coefficients (based on the crystal size) are
0.6550 and 0.7461.
The structure was solved and refined using the Bruker SHELXTL
software package74 using the space group C12/c1, with Z = 4 for the
formula unit C36H74Cl4Co2Fe2N20O27. The final anisotropic full-
matrix least-squares refinement on F2 with 446 variables converged at
R1 = 6.42% for the observed data and wR2 = 7.10% for all data. The
goodness-of-fit was 1.121. The largest peak in the final difference
electron density synthesis was 1.580 e− Å−3, and the largest hole was
−1.945 e− Å−3 with an root-mean-square deviation of 0.132 e− Å−3.
On the basis of the final model, the calculated density was 1.644 g
cm−3 and F(000) 3280 e−. Table S3 collects the relevant data for the
diffraction.
Values of observed rate constants, crystal data structure
refinement, stacking of the anionic squares of the {CoIII/
FeII}22− complex, potential speciation plot, time-resolved
and Specfit-calculated spectra, and additional schemes
(PDF)
Accession Codes
CCDC 1827523 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
via www.ccdc.cam.ac.uk/data_request/cif, or by emailing
data_request@ccdc.cam.ac.uk, or by contacting The Cam-
bridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
Corresponding Author
*E-mail: manel.martinez@qi.ub.es.
ORCID
Paul V. Bernhardt: 0000-0001-6839-1763
Montserrat Ferrer: 0000-0002-4034-2596
■
́
Jesus Jover: 0000-0003-3383-4573
Manuel Martínez: 0000-0002-6289-4586
Notes
The authors declare no competing financial interest.
Electrochemistry. Electrochemistry experiments were carried out
with a BioLogic SP-150 instrument using a glassy carbon working
electrode, a Ag/AgCl (3 M KCl) reference electrode, and a platinum
wire counter electrode on 1 × 10−3 M solutions of the sample and
using 0.1 M NaClO4 as the supporting electrolyte, unless otherwise
stated.
ACKNOWLEDGMENTS
■
́
Financial support from the Spanish Ministerio de Economia y
Competitividad (CTQ2015-65707C2-1/FEDER) is acknowl-
edged.
Spectroelectrochemistry experiments were carried out with the
same instrument and the same standard setup in a CHI cell in a
HP8453 UV−vis spectrophotometer.
REFERENCES
■
(1) Vardhan, H.; Yusubov, M.; Verpoort, F. Self-assembled metal-
organic polyhedra: An overview of various applications. Coord. Chem.
Rev. 2016, 306, 171−194.
Kinetics. The kinetic profiles for the reactions at ambient pressure
were followed by UV−vis spectroscopy in the 700−300 nm range on
HP8452A or Cary 50 instruments equipped with thermostated
multicell transport. Runs with t1/2 < 10 s were performed using an
Applied Photophysics SX20 MV stopped-flow instrument with
photodiode-array detection (J&M TIDAS). For runs carried out at
elevated pressures with t1/2 > 10 s, an already-described pressurizing
cell system setup connected to a TIDAS J&M instrument was used.48
For experiments run at variable pressure with t1/2 < 10 s, a previously
described pressurized stopped-flow mixing unit connected with fiber
optics to a J&M TIDAS instrument was used.75
(2) Zarra, S.; Wood, D. M.; Roberts, D. A.; Nitschke, J. R. Molecular
containers in complex chemical systems. Chem. Soc. Rev. 2015, 44,
419−432.
(3) Chakrabarty, R.; Mukherjee, P. S.; Stang, P. J. Supramolecular
Coordination: Self-Assembly of Finite Two- and Three-Dimensional
Ensembles. Chem. Rev. 2011, 111, 6810−6918.
(4) Amijs, C. H. M.; van Klink, G. P. M.; van Koten, G.
Metallasupramolecular architectures, an overview of functional
properties and applications. Dalton Trans. 2006, 308−327.
(5) Croue, V.; Goeb, S.; Salle, M. Metal-driven self-assembly: the
case of redox-active discrete architectures. Chem. Commun. 2015, 51,
7275−7289.
The observed rate constants were derived from absorbance versus
time traces at the wavelengths where a maximum increase and/or
decrease of the absorbance were observed. Calculation of the
observed rate constants from absorbance versus time monitoring of
the reactions, studied under first-order concentration conditions, was
carried out using the Specfit or ReactLab software packages;41,42 the
general kinetic technique is that previously described.40,76 All postrun
fittings were carried out by the standard available commercial
programs.
For the oxidation reactions, solutions of the square compound were
prepared in the relevant buffers at I = 1.0 M (NaClO4) and mixed
with the relevant amounts of a peroxodisulfate-buffered (I = 1.0 M
NaClO4) stock solution and buffer at I = 1.0 M NaClO4. For the
reduction process, the oxidized square was dissolved in 0.001 M
HClO4, and this was added to the relevant buffered solutions for the
study (I = 1.0 M NaClO4).
(6) Beer, P. D.; Berry, N.; Drew, M. G. B.; Fox, O. D.; Padilla-Tosta,
M. E.; Patell, S. Self-assembled dithiocarbamate-copper(II) macro-
cycles for electrochemical anion recognition. Chem. Commun. 2001,
199−200.
(7) Fabrizi de Biani, F.; Corsini, M.; Zanello, P.; Yao, H.; Bluhm, M.
E.; Grimes, R. N. Electronic Properties of Mononuclear, Dinuclear,
and Polynuclear Cobaltacarboranes: Electrochemical and Spectroelec-
trochemical Studies. J. Am. Chem. Soc. 2004, 126, 11360−11369.
(8) Chifotides, H. T.; Giles, I. D.; Dunbar, K. R. Supramolecular
Architectures with π-Acidic 3,6-Bis(2-pyridyl)-1,2,4,5-tetrazine Cav-
ities: Role of Anion-π Interactions in the Remarkable Stability of
Fe(II) Metallacycles in Solution. J. Am. Chem. Soc. 2013, 135, 3039−
3055.
I
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(9) Chisholm, M. H.; Patmore, N. J.; Reed, C. R.; Singh, N. Oxalate
Bridged Triangles Incorporating Mo2 Units. Electronic Structure
heterobimetallic molecular square. Chem. Commun. 2010, 46, 8995−
8997.
4+
(25) Nihei, M.; Sekine, Y.; Suganami, N.; Oshio, H. Thermally Two-
stepped Spin Transitions Induced by Intramolecular Electron
Transfers in a Cyanide-bridged Molecular Square. Chem. Lett. 2010,
39, 978−979.
(26) Alcazar, L.; Aullon, G.; Ferrer, M.; Martínez, M. Redox-Assisted
Self-Assembly of a Water-Soluble Cyanido-Bridged Mixed Valence
{CoIII/FeII}2 Square. Chem. - Eur. J. 2016, 22, 15227−15230.
(27) Bernhardt, P. V.; Martínez, M. The First Structurally
Characterized Discrete Dinuclear μ-Cyano Hexacyanoferrate Com-
plex. Inorg. Chem. 1999, 38, 424−425.
(28) Bernhardt, P. V.; Macpherson, B. P.; Martínez, M. Discrete
Dinuclear Cyano-Bridged Complexes. Inorg. Chem. 2000, 39, 5203−
5208.
and Bonding. Inorg. Chem. 2010, 49, 7116−7122.
(10) Matsumoto, T.; Newton, G. N.; Shiga, T.; Hayami, S.; Matsui,
Y.; Okamoto, H.; Kumai, R.; Murakami, Y.; Oshio, H. Programmable
spin-state switching in a mixed-valence spin-crossover iron grid. Nat.
Commun. 2014, 5, 3865.
́
́
(11) Aguila, D.; Prado, Y.; Koumousi, E. S.; Mathoniere, C.; Clerac,
R. Switchable Fe/Co Prussian blue networks and molecular
analogues. Chem. Soc. Rev. 2016, 45, 203−224.
(12) Berlinguette, C. P.; Dragulescu-Andrasi, A.; Sieber, A.; Gudel,
̈
H. U.; Achim, C.; Dunbar, K. R. A Charge-Transfer-Induced Spin
Transition in a Discrete Complex: The Role of Extrinsic Factors in
Stabilizing Three Electronic Isomeric Forms of a Cyanide-Bridged
Co/Fe Cluster. J. Am. Chem. Soc. 2005, 127, 6766−6779.
(13) Shatruk, M.; Dragulescu-Andrasi, A.; Chambers, K. E.; Stoian,
S. A.; Bominaar, E. L.; Achim, C.; Dunbar, K. R. Properties of
Prussian Blue Materials Manifested in Molecular Complexes:
Observation of Cyanide Linkage Isomerism and Spin-Crossover
Behavior in Pentanuclear Cyanide Clusters. J. Am. Chem. Soc. 2007,
129, 6104−6116.
́
(29) Bernhardt, P. V.; Bozoglian, F.; Macpherson, B. P.; Martínez,
M. Tuning the metal to metal charge transfer energy of cyano-bridged
dinuclear complexes. Dalton Trans. 2004, 2582−2587.
́
(30) Bernhardt, P. V.; Bozoglian, F.; Macpherson, B. P.; Martínez,
M. Molecular mixed-valence cyanide bridged CoIII-FeII complexes.
Coord. Chem. Rev. 2005, 249, 1902−1916.
́
́
(14) Nihei, M.; Ui, M.; Hoshino, N.; Oshio, H. Cyanide-Bridged
Iron(II,III) Cube with Multistepped Redox Behavior. Inorg. Chem.
2008, 47, 6106−6108.
(31) Bernhardt, P. V.; Bozoglian, F.; Gonzalez, G.; Martínez, M.;
Macpherson, B. P.; Sienra, B. Dinuclear Cyano-Bridged CoIII/FeII
Complexes as Precursors for Molecular Mixed Valence Complexes of
Higher Nuclearity. Inorg. Chem. 2006, 45, 74−82.
́
́
̀
(15) Li, D.; Clerac, R.; Roubeau, O.; Harte, E.; Mathoniere, C.; Le
Bris, R.; Holmes, S. M. Magnetic and Optical Bistability Driven by
Thermally and Photoinduced Intramolecular Electron Transfer in a
Molecular Cobalt-Iron Prussian Blue Analogue. J. Am. Chem. Soc.
2008, 130, 252−258.
(32) Bernhardt, P. V.; Martínez, M.; Rodríguez, C. Molecular CoIII/
FeII Cyano-Bridged Mixed-Valence Compounds with High Nuclear-
ities and Diversity of CoIII Coordination Environments: Preparative
and Mechanistic Aspects. Inorg. Chem. 2009, 48, 4787−4797.
(33) Bernhardt, P. V.; Martínez, M.; Rodríguez, C. Outer-Sphere
Redox Reactions leading to the Formation of Discrete CoIII/FeII
pyrazine-Bridged Mixed Valence Compounds. Eur. J. Inorg. Chem.
2010, 2010, 562−569.
(16) Mondal, A.; Li, Y.; Herson, P.; Seuleiman, M.; Boillot, M. L.;
Riviere, E.; Julve, M.; Rechignat, L.; Bousseksou, A.; Lescouezec, R.
Photomagnetic effect in a cyanide-bridged mixed-valence {FeII FeIII
}
2
2
molecular square. Chem. Commun. 2012, 48, 5653−5655.
(17) Zheng, C.; Xu, J.; Wang, F.; Tao, J.; Li, D. Spin crossover and
reversible single-crystal to single-crystal transformation behaviour in
(34) Bernhardt, P. V.; Martínez, M.; Rodríguez, C.; Vazquez, M.
Discrete RhIII/FeII and RhIII/FeII/CoIII Cyanide-Bridged Mixed
Valence Compounds. Inorg. Chem. 2011, 50, 1429−1440.
two cyanide-bridged mixed-valence {FeIII2FeII } clusters. Dalton
2
Trans. 2016, 45, 17254−17263.
(35) Oshio, H.; Onodera, H.; Tamada, O.; Mizutani, H.; Hikichi, T.;
Ito, T. Cyanide-Bridged Fe-Fe and Fe-Co Molecular Squares:
Structures and Electrochemistry of [FeII4(μ-CN)4(bpy)8](PF6)4·
4H2O, [FeII2CoII2(μ-CN)4(bpy)8](PF6)4·CHCl3·2CH3CN, and
(18) Nihei, M.; Yanai, Y.; Hsu, I. J.; Sekine, Y.; Oshio, H. A
Hydrogen-Bonded Cyanide-Bridged [Co2Fe2] Square Complex
Exhibiting a Three-Step Spin Transition. Angew. Chem., Int. Ed.
2017, 56, 591−594.
[FeII CoIII2(μ-CN)4(bpy)8](PF6)6·2CHCl3·4CH3NO2. Chem. - Eur.
2
(19) De, S.; Jimenez, J. R.; Li, Y.; Chamoreau, L. M.; Flambard, A.;
Journaux, Y.; Bousseksou, A.; Lescouezec, R. One synthesis: two
redox states. Temperature-oriented crystallization of a charge transfer
J. 2000, 6, 2523−2530.
(36) King, P.; Maeder, M. Reactlab Redox; J Plus Consulting Pty Ltd:
Western Australia, 2014.
(37) Zheng, C.; Xu, J.; Yang, Z.; Tao, J.; Li, D. Factors Impacting
Electron Transfer in Cyano-Bridged {Fe2Co2} Clusters. Inorg. Chem.
2015, 54, 9687−9689.
(38) Bard, A. J.; Faulkner, L. R. Electrochemical Methods:
Fundamentals and Applications; Wiley: Chicago, IL, 2001.
{Fe2Co2} square complex in a {FeIILSCoIII
}
diamagnetic or
LS
2
{FeIIILSCoIIHS}2 paramagnetic state. RSC Adv. 2016, 6, 17456−17459.
(20) Zhang, Y. Z.; Ferko, P.; Siretanu, D.; Ababei, R.; Rath, N. P.;
́
̀
Shaw, M. J.; Clerac, R.; Mathoniere, C.; Holmes, S. M.
Thermochromic and Photoresponsive Cyanometalate Fe/Co Squares:
Toward Control of the Electron Transfer Temperature. J. Am. Chem.
Soc. 2014, 136, 16854−16864.
́
(39) Bernhardt, P. V.; Bozoglian, F.; Macpherson, B. P.; Martínez,
́
M.; Gonzalez, G.; Sienra, B. Discrete cyanide bridged mixed-valence
(21) Cao, L.; Tao, J.; Gao, Q.; Liu, T.; Xia, Z.; Li, D. Selective on/off
switching at room temperature of a magnetic bistable {Fe2Co2}
complex with single crystal-to-single crystal transformation via
intramolecular electron transfer. Chem. Commun. 2014, 50, 1665−
1667.
Co/Fe complexes: outer sphere redox behaviour. Eur. J. Inorg. Chem.
2003, 2003, 2512−2518.
́
(40) Bernhardt, P. V.; Bozoglian, F.; Macpherson, B. P.; Martínez,
́
M.; Merbach, A. E.; Gonzalez, G.; Sienra, B. Oxidation of Mixed-
Valence CoIII/FeII Complexes Reversed at High pH: A Kinetico-
Mechanistic Study of Water Oxidation. Inorg. Chem. 2004, 43, 7187−
7195.
(22) Siretanu, D.; Li, D.; Buisson, L.; Bassani, D. M.; Holmes, S. M.;
̀
̀
Mathoniere, C.; Clerac, R. Controlling Thermally Induced Electron
Transfer in Cyano-Bridged Molecular Squares: From Solid State to
Solution. Chem. - Eur. J. 2011, 17, 11704−11708.
(41) Maeder, M.; King, P. ReactLab; Jplus Consulting Pty Ltd.: East
Fremantle, Australia, 2009.
(23) Nihei, M.; Sekine, Y.; Suganami, N.; Nakazawa, K.; Nakao, A.;
Nakao, H.; Murakami, Y.; Oshio, H. Controlled Intramolecular
Electron Transfers in Cyanide-Bridged Molecular Squares by
Chemical Modifications and External Stimuli. J. Am. Chem. Soc.
2011, 133, 3592−3600.
(24) Mercurol, J.; Li, Y.; Pardo, E.; Risset, O.; Seuleiman, M.;
Rousseliere, H.; Lescouezec, R.; Julve, M. [FeIILSCoIIILS]2 [Leftright-
arrow] [FeIIILSCoIIHS]2 photoinduced conversion in a cyanide-bridged
(42) Binstead, R. A.; Zuberbuhler, A. D.; Jung, B. SPECFIT32
[3.0.34]; Spectrum Software Associates: Marlborough, MA, 2005.
(43) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms;
McGraw-Hill: New York, 1981.
(44) Lappin, A. G. Redox Mechanisms in Inorganic Chemistry; Ellis
Horwood: Chichester, U.K., 1994.
(45) Tobe, M. L.; Burgess, J. Inorganic Reaction Mechanisms;
Longman: New York, 1999.
J
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(46) Ducommun, Y.; Merbach, A. E. Inorganic High Pressure
Chemistry; Elsevier: Amsterdam, The Netherlands, 1986; pp 69−114.
(47) Geselowitz, D. A.; Hammershoi, A.; Taube, H. Stereoselective
electron transfer reactions of (ethylenediaminetetraacetato)cobaltate-
(III), (propylenediaminetetraacetato)cobaltate(III) and (1,2-
cyclohexanediaminetetraacetato)cobaltate(III) with tris-
(ethylenediamine)cobalt(II). Inorg. Chem. 1987, 26, 1842−1845.
(48) Martínez, M.; Pitarque, M. A.; van Eldik, R. Outer-sphere redox
reactions in sterically hindered pentaam(m)inecobalt(III) complexes.
A temperature and pressure dependence kinetic study. J. Chem. Soc.,
Dalton Trans. 1994, 3159−3163.
(66) Ditchfield, R.; Hehre, W. J.; Pople, J. A. Self-Consistent
Molecular-Orbital Methods. IX. An Extended Gaussian-Type Basis for
Molecular-Orbital Studies of Organic Molecules. J. Chem. Phys. 1971,
54, 724−728.
(67) Dunning, T. H.; Hay, P. J. Modern Theoretical Chemistry;
Plenum: New York, 1976; pp 1−28.
(68) Tannor, D. J.; Marten, B.; Murphy, R.; Friesner, R. A.; Sitkoff,
D.; Nicholls, A.; Honig, B.; Ringnalda, M.; Goddard, W. A. Accurate
First Principles Calculation of Molecular Charge Distributions and
Solvation Energies from Ab Initio Quantum Mechanics and
Continuum Dielectric Theory. J. Am. Chem. Soc. 1994, 116,
11875−11882.
(69) Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy, R. B.;
Ringnalda, M. N.; Sitkoff, D.; Honig, B. New Model for Calculation of
Solvation Free Energies: Correction of Self-Consistent Reaction Field
Continuum Dielectric Theory for Short-Range Hydrogen-Bonding
Effects. J. Phys. Chem. 1996, 100, 11775−11788.
(70) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal
Solvation Model Based on Solute Electron Density and on a
Continuum Model of the Solvent Defined by the Bulk Dielectric
Constant and Atomic Surface Tensions. J. Phys. Chem. B 2009, 113,
6378−6396.
(71) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and
accurate ab initio parametrization of density functional dispersion
correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010,
132, 154104.
(72) SAINT, revision 7.68A; Bruker AXS Inc., 2009.
(73) Sheldrick, G. M. SADABS, revision 2008/1. Bruker AXS Inc.,
2008.
(74) Sheldrick, G. M. SHELXL. Acta Crystallogr., Sect. A: Found.
Crystallogr. 2008, 64, 112−122.
(75) Zerk, T. J.; Martínez, M.; Bernhardt, P. V. A Kinetico-
Mechanistic Study on CuII Deactivators Employed in Atom Transfer
Radical PolymCerization. Inorg. Chem. 2016, 55, 9848−9857.
(76) Crespo, M.; Martínez, M.; Nabavizadeh, S. M.; Rashidi, M.
Kinetico-mechanistic studies on C-X (X = H, F, Cl, Br, I) bond
activation reactions on organoplatinum(II) complexes. Coord. Chem.
Rev. 2014, 279, 115−140.
(49) Martínez, M.; Pitarque, M. A.; van Eldik, R. Outer-sphere redox
reactions of [CoIII(NH3)5(HxPyOz)](m‑3)‑ complexes. A temperature-
and pressure-dependence kinetic study on the influence of the
phosphorus oxoanions. J. Chem. Soc., Dalton Trans. 1996, 2665−2671.
́
(50) Bernhardt, P. V.; Gonzalvez, M. A.; Martínez, M. Kinetico-
mechanistic Study on the Oxidation of Biologically Active Iron(II)
Bis(thiosemicarbazone) Complexes by Air. Importance of NH-O2
Interactions As Established by Activation Volumes. Inorg. Chem.
2017, 56, 14284−14290.
(51) Basha, M. T.; Bordini, J.; Richardson, D. E.; Martínez, M.;
Bernhardt, P. V. Kinetico-mechanistic Studies on Methemoglobin
Generation by Biologically Active Thiosemicarbazone Iron(III)
Complexes. J. Inorg. Biochem. 2016, 162, 326−333.
́
(52) Bernhardt, P. V.; Martínez, M.; Rodríguez, C.; Vazquez, M.
Biologically Active Thiosemicarbazone Fe Chelators and their
Reactions with Ferrioxamine B and Ferric EDTA; a Kinetic Study.
Dalton Trans. 2012, 41, 2122−2130.
(53) Garcia-Amoros, J.; Stopa, G.; Stochel, G.; van Eldik, R.;
Martinez, M.; Velasco, D. Activation volumes for cis-to-trans
isomerisation reactions of azophenols: a clear mechanistic indicator?
Phys. Chem. Chem. Phys. 2018, 20, 1286−1292.
(54) Basallote, M. G.; Bernhardt, P. V.; Calvet, T.; Castillo, C. E.;
Font-Bardía, M.; Martínez, M.; Rodríguez, C. Mechanistic aspects of
the chemistry of mononuclear CrIII complexes with pendant-arm
macrocyclic ligands and formation of discrete CrIII/FeII and CrIII/
FeII/CoIII cyano-bridged mixed valence compounds. Dalton Trans.
2009, 9567−9577.
(55) Eaton, D. R.; Pankratz, M. The reaction of the
hexacyanoferrate(III) ion with hydrogen peroxide. Can. J. Chem.
1985, 63, 793−797.
(56) Rabai, G.; Kustin, K.; Epstein, I. R. Systematic design of
chemical oscillators. 57. Light-sensitive oscillations in the hydrogen
peroxide oxidation of ferrocyanide. J. Am. Chem. Soc. 1989, 111,
8271−8273.
(57) Atkins, P.; Overton, T.; Rourke, J.; Weller, M.; Armstrong, F. A.
Inorganic Chemistry; Oxford University Press, 2006.
(58) Good, N. E.; Winget, G. D.; Winter, W.; Connolly, T. N.;
Izawa, S.; Singh, R. M. M. Hydrogen Ion Buffers for Biological
Research. Biochemistry 1966, 5, 467−477.
(59) Perrin, D. D. Buffers of Low Ionic Strength for Spectrophoto-
metric pK Determinations. Aust. J. Chem. 1963, 16, 572−578.
(60) Frisch, M. J.; et al. Gaussian09, revision CD.01; Gaussian Inc.:
Wallingford, CT, 2009.
(61) Becke, A. D. Density-functional thermochemistry. III. The role
of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652.
(62) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-
Salvetti correlation-energy formula into a functional of the electron
density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789.
(63) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Results obtained
with the correlation energy density functionals of Becke and Lee,
Yang and Parr. Chem. Phys. Lett. 1989, 157, 200−206.
(64) Hariharan, P. C.; Pople, J. A. The influence of polarization
functions on molecular orbital hydrogenation energies. Theoret.Ch-
im.Acta 1973, 28, 213−222.
(65) Hehre, W. J.; Ditchfield, R.; Pople, J. A. Self-Consistent
Molecular-Orbital Methods. XII. An Extended Gaussian-Type Basis
for Molecular-Orbital Studies of Organic Molecules. J. Chem. Phys.
1972, 56, 2257−2261.
K
DOI: 10.1021/acs.inorgchem.8b01147
Inorg. Chem. XXXX, XXX, XXX−XXX