Mechanistic studies on the reactions of cyanide with a water-soluble Fe(III) porphyrin and their effect on the binding of NO

Article abstract of DOI:10.1021/ic1023345

The reaction of the water-soluble Fe^III(TMPS) porphyrin with CN^- in basic solution leads to the stepwise formation of Fe ^III(TMPS)(CN)(H₂O) and Fe^III(TMPS)- (CN)2. The kinetics of the reaction of CN^- with Fe^III(TMPS)- (CN)(H₂O) was studied as a function of temperature and pressure. The positive value of the activation volume for the formation of Fe ^III(TMPS)(CN)2 is consistent with the operation of a dissociatively activated mechanism and confirms the six-coordinate nature of the monocyano complex. A good agreement between the rate constants at pH 8 and 9 for the formation of the dicyano complex implies the presence of water in the axial position trans to coordinated cyanide in the monocyano complex and eliminates the existence of Fe^III(TMPS)(CN)(OH) under the selected reaction conditions. Both Fe^III(TMPS)- (CN)(H₂O) and Fe ^III(TMPS)(CN)2 bind nitric oxide (NO) to form the same nitrosyl complex, namely, FeII(TMPS)(CN)- (NO⁺). Kinetic studies indicate that nitrosylation of Fe^III(TMPS)(CN)2 follows a limiting dissociative mechanism that is supported by the independence of the observed rate constant on [NO] at an appropriately high excess of NO, and the positive values of both the activation parameters ΔS and ΔV found for the reaction under such conditions. The relatively small first-order rate constant for NO binding, namely, (1.54 (0.01) × 10^-2 s^-1, correlates with the rate constant for CN^- release from the Fe^III(TMPS)(CN)2 complex, namely, (1.3 ± 0.2) × 10^-2 s^-1 at 20 °C, and supports the proposed nitrosylation mechanism.

Full text of DOI:10.1021/ic1023345

ARTICLE
pubs.acs.org/IC
Mechanistic Studies on the Reactions of Cyanide with a Water-Soluble
Fe(III) Porphyrin and Their Effect on the Binding of NO
Maria Oszajca,†,‡ Alicja Franke,† Mazgorzata Brindell,‡ Graz_yna Stochel,*,‡ and Rudi van Eldik*,†
†Inorganic Chemistry, Department of Chemistry and Pharmacy, University of Erlangen-N€urnberg, Egerlandstrasse 1, 91058 Erlangen,
Germany
‡Faculty of Chemistry, Jagiellonian University, Ingardena 3, 30-060 Krakꢀow, Poland
ABSTRACT: The reaction of the water-soluble FeIII(TMPS)
porphyrin with CNꢀ in basic solution leads to the stepwise
formation of FeIII(TMPS)(CN)(H2O) and FeIII(TMPS)-
(CN)2. The kinetics of the reaction of CNꢀ with FeIII(TMPS)-
(CN)(H2O) was studied as a function of temperature and
pressure. The positive value of the activation volume for the
formation of FeIII(TMPS)(CN)2 is consistent with the opera-
tion of a dissociatively activated mechanism and confirms the
six-coordinate nature of the monocyano complex. A good
agreement between the rate constants at pH 8 and 9 for the
formation of the dicyano complex implies the presence of water in the axial position trans to coordinated cyanide in the monocyano
complex and eliminates the existence of FeIII(TMPS)(CN)(OH) under the selected reaction conditions. Both FeIII(TMPS)-
(CN)(H2O) and FeIII(TMPS)(CN)2 bind nitric oxide (NO) to form the same nitrosyl complex, namely, FeII(TMPS)(CN)-
(NOþ). Kinetic studies indicate that nitrosylation of FeIII(TMPS)(CN)2 follows a limiting dissociative mechanism that is supported
by the independence of the observed rate constant on [NO] at an appropriately high excess of NO, and the positive values of both
the activation parameters ΔSq and ΔVq found for the reaction under such conditions. The relatively small first-order rate constant
for NO binding, namely, (1.54 ( 0.01) ꢁ 10ꢀ2 sꢀ1, correlates with the rate constant for CNꢀ release from the FeIII(TMPS)(CN)2
complex, namely, (1.3 ( 0.2) ꢁ 10ꢀ2 sꢀ1 at 20 °C, and supports the proposed nitrosylation mechanism.
’ INTRODUCTION
mechanism of NO interaction with iron porphyrins is still far
from complete and requires more detailed studies.
Nitric oxide (NO) was proclaimed the molecule of the year by
Science in 1992 since the important biological significance of this
messenger molecule was confirmed by many extensive studies on
the role of NO in mammalian biology.1 Since the physiological
and pathophysiological role of NO is often related to its
coordination to metal centers, in particular iron, the interac-
tion of hemoproteins and different modified, water-soluble
iron porphyrin complexes with NO has been intensively
studied over the past years.2ꢀ6 The dynamics of the formation
and stability of nitrosyl porphyrin complexes depends on
various factors, for instance, the nature and overall charge
on the porphyrin, oxidation state of the iron center, spin
reorganization, lability of the metal center, and the nature of
the axial ligands.7 In addition to the character of the prosthetic
group, numerous other factors connected with the protein
structure have a large influence on the reversible binding of
NO to the iron center. Thus, a comparison of the results ob-
tained from studies on the nitrosylation of iron porphyrins
with those obtained from studies on the binding of NO to
hemoproteins enabled the elucidation of the role of the
protein structure on the properties of the heme coordination
center and the regulation of NO binding. Because of the many
factors that govern NO coordination to iron hemes and NO
dissociation from nitrosyl complexes, the elucidation of the
Systematic studies on the interaction of NO with [FeIII(TMPS)]3ꢀ
,
[meso-tetrakis(2,4,6-trimethyl-3-sulfonatophenyl)porphinato]-
iron(III), a water-soluble synthetic model iron porphyrin, indicated
that this complex is a reasonable model for the ferriheme protein,
particularly met-myoglobin.2 The mechanism of reversible
binding of NO to FeIII(TMPS) was shown to depend strongly
on the pH of the solution.6 At low pH where FeIII(TMPS) exists
as a six-coordinate diaqua-ligated species [FeIII(TMPS)-
(H2O)2]3ꢀ, coordination of NO occurs via a dissociatively
activated mechanism and is controlled by the lability of the
metal center.2 An opposite, associatively activated mechanism
was observed for the binding of NO to the five-coordinate
monohydroxo species [FeIII(TMPS)(OH)]4ꢀ at high pH.
On the basis of detailed mechanistic studies, it was proposed
that formation of FeII(TMPS)(OH)(NOþ) is controlled by
FeIII-NO bond formation and spin reorganization connected
with this process.6
Mechanistic studies on the binding of NO to FeIII(TMPS)-
(OH)(MeIm) in an ionic liquid revealed that coordination
of methylimidazole (MeIm) to the monohydroxo porphyrin species
Received: November 22, 2010
Published: March 23, 2011
r
2011 American Chemical Society
3413
dx.doi.org/10.1021/ic1023345 Inorg. Chem. 2011, 50, 3413–3424
|

Inorganic Chemistry
ARTICLE
led to a changeover in mechanism for the formation of FeII(TMPS)-
(OH)(NOþ) from an associatively activated mode for FeIII-
cyanide was purchased from Aldrich Chemical Co., Inc. and used as
source for cyanide throughout the study.
(TMPS)(OH) to
a
dissociatively activated mode for
Solution Preparation. Samples of FeIII(TMPS)(OH) were pre-
pared at pH 8 and 9 using TAPS (>99%, ROTH) and CHES (>99%,
Sigma) buffer solutions (0.05 M), respectively. The ionic strength of the
solutions was adjusted with NaNO3 to 0.15 M. Deionized water was
used for the preparation of all solutions. Since experiments with NO
required an inert atmosphere, all solutions were deoxygenated by argon
and handled using gastight syringes. Appropriate concentrations of NO
(0.2ꢀ1.8 mM) were prepared by mixing a saturated solution of NO
(1.8 mM) with a deoxygenated buffer solution under inert atmosphere.
Cyanide derivatives of FeIII(TMPS) were obtained by addition of an
appropriate excess of NaCN.
FeIII(TMPS)(OH)(MeIm).8 In a preliminary study, we also att-
empted to generate the mono(methylimidazole) FeIII(TMPS)
derivative in aqueous solution. However, formation of only the
bis(methylimidazole) complex FeIII(TMPS)(MeIm)2 was ob-
served over a wide pH range (pH = 5ꢀ13). The binding of NO
to the bis(methylimidazole) porphyrin complex revealed only
formation of FeII(TMPS)(OH)(NOþ) from FeIII(TMPS)-
(OH) existing in equilibrium with FeIII(TMPS)(MeIm)2. Sub-
sequent reductive nitrosylation, which occurs much faster in the
presence of MeIm, rendered it impossible to monitor the binding
of NO to FeIII(TMPS)(MeIm)2.
Measurements. Ambient pressure stopped-flow measurements
were performed on an SX 18.MV Applied Photophysics apparatus
equipped with a thermostat ((0.1 °C). In a typical experiment, buffer
solutions (deoxygenated if required) of the porphyrin complex were
rapidly mixed with appropriate concentrated substrate solutions (NO or
NaCN) in the ratio 1:1. All reactions with NO were initiated using pre-
equilibrated solutions of FeIII(TMPS)(OH) with cyanide. All kinetic
measurements were performed under pseudo first-order conditions, that
is, by using at least a 10-fold excess of NO or NaCN (representing the
total cyanide concentration in the solution) over the porphyrin concentra-
tion. Reported rate constants are mean values of at least 10 kinetic runs. Time-
resolved spectra were recorded with the use of the stopped-flow apparatus
described above, equipped with a J&M TIDAS diode-array detector. High
pressure stopped-flow experiments were performed in the pressure range of
10 to 130 MPa on a custom built apparatus.15 OLIS KINFIT software
(Bogart, GA, 1989) was used for the analysis of kinetic traces. Ambient
pressure UVꢀvis spectra were recorded on a Shimadzu UVꢀ2101PC
spectrophotometer equipped with a thermostatted cell compartment
(CPS-260). UVꢀvis spectra under high pressure conditions (up to 150
MPa) were recorded in a quartz pill-box cell on the spectrophotometer
described above including a custom built high pressure cell.16
It follows from the above discussion that the number and
electronic nature of the axial ligands in synthetic model ferric
porphyrins can regulate the dynamics of NO binding as well as
the stability and reactivity of the nitrosyl adducts. Therefore, to
investigate the influence of the nature of the axial ligand on the
reactivity of iron(III) porphyrins with NO, we selected the
cyanide ligand which behaves as a strong σ donor and π acceptor.
The cyanide ion has been extensively used in the study of the
properties of iron porphyrins, hemes, and hemoproteins,9ꢀ12 as
well as in many modeling studies on simple iron complexes.13
Although both ferric and ferrous iron porphyrins can interact
with cyanide, the affinity of cyanide for the iron(III) heme center
is much higher. The binding of cyanide to the ferric heme center
leads to the formation of low-spin dicyano FeIII(Por)(CN)2 or
mixed-ligand FeIII(Por)(CN)(L) complexes.9,10 Cyanide for
example binds to the heme a3 oxygen site of cytochrome c
oxidase and strongly inhibits the catalytic activity of the enzyme,
and as a result studies in search for methods to treat cyanide
poisoning have received significant attention.14
In the present study we chose the water-soluble porphyrin
FeIII(TMPS) and by addition of appropriate amounts of sodium
cyanide under selected pH conditions generated the six-coordi-
nate FeIII(TMPS)(CN)(H2O) and FeIII(TMPS)(CN)2 species.
Detailed kinetic studies on the reactions of the cyanide deriva-
tives of FeIII(TMPS) with NO were performed. To obtain a
better understanding of the mechanism of NO coordination to
cyano porphyrin complexes, we supplemented our study with
kinetic measurements on the formation of the mono- and dicyano
FeIII(TMPS) complexes. All measurements were performed
at two selected pH values (namely, 8 and 9), for which the
FeIII(TMPS) complex mainly exists as FeIII(TMPS)(OH) since
the pKa of FeIII(TMPS)(H2O)2 is 6.9. Comparison of the kinetic
data from two different pH values helped us to identify the nature
of the monocyano species which participated in the observed
reactions. Furthermore, the selected pH conditions enabled us to
follow the formation of the monocyano porphyrin complex and
the subsequent formation of the dicyano complex.
’ RESULTS AND DISCUSSION
UVꢀvis Spectral Data for the Reaction of FeIII(TMPS)(OH)
with Cyanide. The UVꢀvis spectrum of FeIII(TMPS)(OH)
exhibits a Soret band maximum at 416 nm. After addition of an
excess of sodium cyanide to the FeIII(TMPS)(OH) complex, two
subsequent reactions were observed. The first fast reaction
leadingto formation of the monocyano species revealed an
absorbance increase in the Soret band at 416 nm and the
appearance of a new Q-band at 550 nm (Figure 1a). The second
reaction connected with conversion of the monocyano to the
dicyano complex occurred on a much slower time scale and
resulted in the spectral changes presented in Figure 1b. The
decrease in the Soret band at 416 nm was accompanied by a
simultaneous appearance of a new Soret band at 427 nm and an
increase in absorbance at 600 nm.
Kinetic and Thermodynamic Studies on the Reaction of
CNꢀ with FeIII(TMPS)(OH). Mixing equally concentrated solu-
tions of FeIII(TMPS)(OH) and sodium cyanide in the volume
ratio 1:1 at pH 8 or 9, and 20 °C, resulted in the formation of a
small fraction of both monocyano and dicyano species. The
spectral changes and stopped-flow kinetic measurements indi-
cated that binding of cyanide to [FeIII(TMPS)(OH)] occurs in
two steps.12,17ꢀ20 Under the selected pH conditions the follow-
ing reactions are to be considered:
’ EXPERIMENTAL SECTION
Materials. All chemicals used in this study were of analytical reagent
grade. NO gas was purchased from Praxair Deutschland GmbH & Co.
KG, Bopfingen (3.0) and bubbled through concentrated KOH solution
to remove higher nitrogen oxides (N2O3, NO2), then passed through an
Ascrite II column (NaOH on silica gel, Sigma Aldrich) and a phosphorus
pentoxide column. The iron(III) porphyrin, [meso-tetrakis(2,4,6-tri-
methyl-3-sulfonatophenyl)porphinato]-iron(III) hydroxide (tetrasodium
salt), was purchased from Frontier Scientific, Inc. Utah, U.S.A. Sodium
HCN a Hþ þ CNꢀ
KaðHCNÞ
ð1Þ
3414
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
Figure 1. (a) Spectral changes observed during formation of the monocyano porphyrin complex after addition of NaCN to FeIII(TMPS)(OH).
Experimental conditions: [FeIII(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, [NaCN] = 1.5 ꢁ 10ꢀ5 M, pH = 8, [TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3). (b)
Spectral changes observed during formation of the dicyano porphyrin complex after addition of NaCN to FeIII(TMPS)(OH). Experimental conditions:
[FeIII(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, [NaCN] = 5 ꢁ 10ꢀ4 M, pH = 8, [TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3).
k1
FeIIIðTMPSÞðOHÞ þ CNꢀ a FeIIIðTMPSÞðCNÞðOHÞ
k
ꢀ1
K1 ¼ k1=k
ð2Þ
ꢀ1
FeIIIðTMPSÞðCNÞðOHÞ þ Hþ a FeIIIðTMPSÞðCNÞðH2OÞ
Kb
ð3Þ
k2
FeIIIðTMPSÞðCNÞðH2OÞ þ CNꢀ a FeIIIðTMPSÞðCNÞ2 þ H2O
k
ꢀ2
K2 ¼ k2=k
ð4Þ
ꢀ2
The coordination of cyanide to FeIII(TMPS)(OH) in the first
step leads to the formation of the six-coordinate species FeIII-
(TMPS)(CN)(OH) (reaction 2). The strong labilizing effect of
CNꢀ is expected to increase the pKa value of coordinated H2O,
on which basis we suggest that under the studied pH conditions,
rapid protonation of OHꢀ follows the first cyanide binding
(reaction 3). Such a conclusion is in line with literature data
which show that the presence of an anionic ligand like OHꢀ or
CNꢀ in the trans position strongly increases the pKa value of
coordinated water up to ≈11ꢀ12 in comparison with that for the
diaqua-ligated form.20,21 In the next step binding of a second
cyanide ion leads to the formation of the six-coordinate FeIII-
(TMPS)(CN)2 species. The stopped-flow kinetic measurements
for cyanide binding to FeIII(TMPS)(OH) at pH 8 and 20 °C
were carried out under pseudo-first-order conditions, and re-
vealed a linear dependence of kobs(n) on [CNꢀ] (representing the
concentration of free cyanide at a particular pH) (eq 5) for
reactions 2 and 4 (Figure 2). It is important to note that Figure 2
refers to the free cyanide concentration (eq 6) on the abscissa,
and the pseudo-first-order kinetic conditions are fulfilled by the
presence of an appropriate large pool of nondissociated HCN in
solution (pKa(HCN) = 9.2 at 20 °C).
Figure 2. Dependence of kobs(n) on [CNꢀ] for the reaction of CNꢀ
with FeIII(TMPS)(OH) at pH 8: (a) binding of first CNꢀ to FeIII-
(TMPS)(OH), inset: typical kinetic trace at λ = 416 nm, [CNꢀ] = 5.9 ꢁ
10ꢀ6 M; (b) binding of second CNꢀ to FeIII(TMPS)(CN)(H2O),
inset: typical kinetic trace at λ = 450 nm, [CNꢀ] = 5.9 ꢁ 10ꢀ6 M.
Experimental conditions: [FeIII(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, pH = 8,
[TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3), T = 20 °C.
kobsðnÞ ¼ kn½CNꢀꢂ þ k
n ¼ 1, 2
ð5Þ
ꢀn
Since the ratio of the rate constants for reactions 2 and 4 at pH 8
is sufficiently large (k1/k2 ≈ 4 ꢁ 103), the observed rate
constants, kobs(1) and kobs(2) were determined independently
by fitting a single exponential function to the recorded kinetic
traces. Taking into account that formation of FeIII(TMPS)-
(CN)2 is preceded by protonation of FeIII(TMPS)(CN)(OH)
where
!
KaðHCNÞ
½CNꢀꢂ ¼
½NaCNꢂ
ð6Þ
½Hꢂþ þ KaðHCNÞ
3415
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
Table 1. Rate and Equilibrium Constants for the Reversible
Binding of Cyanide to FeIII(TMPS)(OH) and FeIII(TMPS)-
(CN)(H2O) at pH 8 and 9, and 20 °C
pH 9 is linear, it can be concluded that under the applied
conditions K1Kb[Hþ][CNꢀ] . 1 þ K1[CNꢀ], such that eq 8
can be simplified to the linear form expressed as in eq 5.
However, this simplification is only true at appropriately high
[CNꢀ] since at lower [CNꢀ] a curved dependence of kobs(2) on
[CNꢀ] will be observed. Therefore, the square dependence of
the observed rate constant on the cyanide concentration ex-
pressed by eq 8 accounts for our inability to determine kꢀ2 after
simplification of this equation to the linear expression 5.
pH = 8
pH = 9
k1 [Mꢀ1 sꢀ1
ꢀ1 [sꢀ1
k2 [Mꢀ1 sꢀ1
ꢀ2 [sꢀ1
]
(1.92 ( 0.04) ꢁ 107
35 ( 6
k
]
]
(4.9 ( 0.1) ꢁ 103
(1.3 ( 0.2) ꢁ 10ꢀ2
(1.54 ( 0.01) ꢁ 10ꢀ2
(5.5 ( 0.9) ꢁ 105
(3.8 ( 0.6) ꢁ 105
(3.2 ( 0.1) ꢁ 105
(4.8 ( 0.1) ꢁ 103
k
]
k00ꢀ2 [sꢀ1 a
]
(1.66 ( 0.01) ꢁ 10ꢀ2
2
k2K1Kb½Hþꢂ½CNꢀꢂ
K1 [Mꢀ1
K2 [Mꢀ1
]
pH9
obsð2Þ
k
¼
þ k
ð8Þ
ꢀ2
]
1 þ K1½CNꢀꢂ þ K1Kb½Hþꢂ½CNꢀꢂ
K002 [Mꢀ1 a
]
(3.2 ( 0.4) ꢁ 105
a k00ꢀ2 represents kꢀ2 obtained from the reaction of FeIII(TMPS)(CN)2
The rate constant for the release of cyanide from FeIII(TMPS)-
(CN)2 was also determined in the reaction of FeIII(TMPS)-
(CN)2 with NO (denoted as k00ꢀ2) at both pH values, which gave
more accurate values of kꢀ2 (see section on “Kinetic studies on
the reaction of NO with FeIII(TMPS)(CN)2”) (Table 1). The
general agreement observed for the data at pH 8 and 9 in Table 1
indicates that a correct model was used to fit these data to obtain
pH independent parameters. The independence of the rate
constants k2 and k00ꢀ2 on pH substantiates our assumption that
at both studied pH values the major species that undergoes a
reaction with CNꢀ is FeIII(TMPS)(CN)(H2O). If the second
axial ligand in the monocyano derivative was OHꢀ, a dependence
of the rate constant for binding of the second CNꢀ on pH would
be observed.
with NO; K002 represents the equilibrium constant calculated using k00
.
ꢀ2
The pKa value of FeIII(TMPS)(H2O)2 is 6.9 and thus at pH 8
only about 7% of the porphyrin complex exists in the diaqua-
ligated form.6 Unfortunately, the pKa value for the FeIII(TMPS)-
(CN)(H2O) intermediate is presently unknown. Determination
of the pKa value for the latter species was not possible since its
formation could not be observed at higher pH. We assume that
the pKa value for FeIII(TMPS)(CN)(H2O) is g11, in agreement
with the value of 12.2 reported for FeIII(TMPyP)(CN)(H2O).20
Therefore, it is reasonable to conclude that the binding of CNꢀ
to the five-coordinate FeIII(TMPS)(OH) species leads to a
considerable decrease in the pKb value, and thus rapid proton-
ation leads to the monoaquamonocyano complex. Furthermore,
it is well-known that coordination of two negatively charged
ligands like OHꢀ or CH3Oꢀ to the relatively electron-rich iron
center in FeIII(TMPS) is not observed even at a high nucleophile
concentration, unless it is CNꢀ. Therefore, the formation
of FeIII(TMPS)(CN)(OH) is highly unlikely in the studied
pH range.6,22
Comparison of the second order rate constant k1 = (1.92 (
0.04) ꢁ 107 Mꢀ1 sꢀ1 for reaction 2 determined at pH 8, with the
rate constants for the reaction of cyanide with met-myoglobin
and met-hemoglobin, namely, (2ꢀ5) ꢁ 102 Mꢀ1 sꢀ1 at pH ≈ 7
and 20 °C, and 3 ꢁ 102 Mꢀ1 sꢀ1 at pH 9.2 and 20 °C,
respectively,23ꢀ26 shows that the reaction with hemoproteins
occurs almost 5 orders of magnitude slower than with the
FeIII(TMPS)(OH) complex and seems not to depend on the
nature of the binding species, namely, HCN or CNꢀ. There are
three key factors which cannot be ignored when differences
between cyanide binding to model porphyrins and hemoproteins
are discussed: (i) the protein environment which regulates ligand
movement into the distal pocket, (ii) HCN deprotonation, and
(iii) dissociation of coordinated water. A detailed study of various
myoglobin mutants showed that in the case of HCN binding,
ligand diffusion into the protein is not the rate-determining
Figure 3. Dependence of kobs(2) on [CNꢀ] for the reaction of CNꢀ
with FeIII(TMPS)(CN)(H2O), reaction 4 at pH 9. Experimental
conditions: [FeIII(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, [CHES buffer] =
0.05 M, I = 0.15 M (NaNO3), 20 °C, λ = 455 nm.
to form FeIII(TMPS)(CN)(H2O), the observed rate constant
kobs(2) can be expressed as in eq 7.
k2Kb½CNꢀꢂ½Hþꢂ
pH8
obsð2Þ
k
¼
þ k
ð7Þ
ꢀ2
1 þ Kb½Hþꢂ
It is reasonable to assume that at pH 8, Kb[Hþ] . 1 and eq 7
reduces to eq 5. The insets of Figures 2a and 2b present typical
kinetic traces recorded for reactions 2 and 4 at pH 8 at 416 and
450 nm, respectively. The rate constants for cyanide binding (kn)
and release (kꢀn) calculated from the slope and intercept of these
plots, respectively, are summarized in Table 1.
Formation of the monocyano derivative at pH 9 was found to
be too fast to be measured by stopped-flow technique. Therefore,
kinetic measurements were performed only for reaction 4 under
such conditions. The observed rate constants were determined
by fitting a single exponential function to the kinetic traces. A plot
of the resulting kobs(2) values versus the cyanide concentration
was found to be linear over the selected [CNꢀ] range (Figure 3),
but the extrapolation of kobs(2) to [CNꢀ] = 0 did not give an
intercept that could be used to determine k
.
ꢀ2
On the basis that the formation of FeIII(TMPS)(CN)(OH)
occurs in a fast pre-equilibrium step, which is followed by fast
protonation to form FeIII(TMPS)(CN)(H2O) and the second
cyanide binding reaction, the observed rate constant can be
expressed as in eq 8. Since the dependence of kobs(2) on [CNꢀ] at
3416
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
step.25,26 The major factor that controls cyanide binding to the
protein heme center is deprotonation of HCN, which depends
on the polarity of the distal pocket and its accessibility to
solvent.25,26 In contrast to that, it has been reported that binding
ꢀ
of the anion N3 is mainly governed by its accessibility to the
heme site.25 It seems likely that also in the case of CNꢀ, ligand
movement into the distal pocket is the dominant barrier.
Furthermore, under the studied conditions, CNꢀ binding to
Fe(TMPS)(OH) does not require displacement of a solvent
molecule since the sixth coordination site is vacant in FeIII-
(TMPS)(OH). In met-myglobin or met-hemoglobin the sixth
coordination site is occupied and the dissociation of a water
molecule is required. Therefore, the strength of solvent coordi-
nation seems to be crucial. In the case of ligand coordination to
the iron center that is not preceded by deprotonation at
physiological pH, for example, N3ꢀ, it has been shown that the
absence of water in the sixth coordination position entails a
significant increase in the binding rate constant.25,27
The ratio of the rate constants kn/k gave the equilibrium
ꢀn
constants K1 = (5.5 ( 0.9) ꢁ 105 Mꢀ1 and K2 = (3.8 ( 0.6) ꢁ
105 Mꢀ1 for reactions 2 and 4, respectively at pH 8. Despite the
small value of k2 in comparison to k1, K2 is relatively high, which
is due to a much smaller rate constant for cyanide release from
the dicyano than from the monocyano complex. At pH 8, the
equilibrium constant K2 calculated using the kꢀ2 value obtained
from the intercept is within the error limits in good agreement
with the K002 value calculated by using k00ꢀ2 obtained from the
reaction with NO (see further Discussion).
Unfortunately, the determination of activation parameters for
reaction 2 was not possible because of the high rate of the
reaction and the limitations of the instrumentation to follow such
reactions as a function of temperature and pressure. Therefore,
kinetic measurements as a function of temperature and pressure
were performed only for the formation of FeIII(TMPS)(CN)2 at
pH 8 and 9. In all the calculations of reaction and activation
parameters, the concentration of free cyanide was used and
calculated by taking the dependence of Ka(HCN) on temperature
and pressure into account.28,29 In addition, the influence of
temperature and pressure on the pH of the employed buffers
was also considered.30ꢀ32 The rate constants for CNꢀ binding
(k2) in the temperature range from 15 to 40 °C enabled to
construct an Eyring plot (Figure 4a) and to calculate ΔHq2 and
ΔSq2 from the slope and intercept, respectively.
Figure 4. Temperature and pressure dependence for the binding of
CNꢀ to FeIII(TMPS)(CN)(H2O). (a) Eyring plot of ln(k2/T) vs 1/T
(b) plot of ln(k2) vs pressure, T = 25 °C. Experimental conditions:
[FeIII(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, [NaCN] = 8 ꢁ 10ꢀ4 M, pH = 8,
[TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3), λ = 450 nm. To calculate
k2, the concentration of free cyanide was used and changes in pKa(HCN)
with temperature and pressure, as well as changes in the pH of the buffers
with temperature were taken into account.
the porphyrin complex at pH 8 and 9, respectively. The relatively
high excess of total cyanide used and the clean isosbestic points
observed over the applied temperature and pressure ranges
confirmed that the observed spectral changes can be exclusively
attributed to reaction 4. The thermodynamic equilibrium con-
stants (Kth2) at 20 °C were found to be in good agreement at
both studied pHs, namely, 2.7 ꢁ 105 and 1.9 ꢁ 105 Mꢀ1, and
correlate very well with the kinetically determined values (see
Tables 1 and 2).
The influence of temperature on the equilibrium constant
(Kth2) was monitored over the temperature range from 10 to
40 °C. The increase in temperature caused a shift in equilib-
rium (4) toward the reactants (Figure 5a). The values of
ln(Kth2) obtained over the selected temperature range show a
linear dependence on 1/T (Figure 5a, inset) and enabled to
determine ΔH° and ΔS° from the slope and intercept,
respectively. High pressure experiments were performed at
30 °C over the pressure range from 5 to 150 MPa. The ob-
served spectral changes were completely reversible and in-
dicated that increasing pressure shifted equilibrium (4) to-
ward the product side (Figure 5b). A plot of ln(Kth2) versus
pressure showed a linear dependence (Figure 5b, inset), and
the value of the slope was used to calculate ΔV°. All calculated
thermodynamic parameters for the reaction of FeIII(TMPS)-
(CN)(H2O) with cyanide are included in Table 2.
In addition, the effect of pressure on the kinetics of the
reaction of FeIII(TMPS)(CN)(H2O) with cyanide was stu-
died over the pressure range 10ꢀ130 MPa. Plots of ln(k2)
versus pressure were found to be linear and enabled to deter-
mine the volume of activation (ΔVq ) (Figure 4b). Pressure
2
experiments were repeated at least two times, and the reported
ΔVq2 values were calculated as the mean value. At the applied
concentration of cyanide the contribution of k to kobs(2)
ꢀ2
can be neglected and then kobs(2) = k2[CNꢀ]. The activation
parameters for the binding of cyanide to FeIII(TMPS)(CN)-
(H2O), determined at pH 8 and 9, are summarized in Table 2.
The activation parameters for release of CNꢀ from FeIII-
(TMPS)(CN)2 were determined in a more direct way, namely,
in the reaction with NO (see section on “Kinetic studies on
the reaction of NO with FeIII(TMPS)(CN)2”) and are in-
cluded in Table 2.
To determine the standard reaction parameters for reaction 4,
UVꢀvis spectra were monitored as a function of temperature and
pressure in the presence of 30- and 15-fold excess of NaCN over
3417
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
Table 2. Summary of Reaction and Activation Parameters for the Reversible Binding of CNꢀ to FeIII(TMPS)(CN)(H2O) at pH 8
and 9a
pH = 8
pH = 9
b
b
ꢀ2
kinetic and thermodynamic parameters
ΔHq (kJ molꢀ1
ΔSq (J molꢀ1 Kꢀ1
ΔVq (cm3 molꢀ1
Kth2 (Mꢀ1) at 20 °C
CNꢀ binding ΔXq
CNꢀ release ΔXq
CNꢀ binding ΔXq
CNꢀ release ΔXq
2
ꢀ2
2
)
43.3 ( 0.3
ꢀ26 ( 1
þ10.4 ( 0.4
92 ( 1
37 ( 1
94 ( 3
)
þ36 ( 2
þ11.2 ( 0.3
ꢀ50 ( 3
þ13 ( 1
þ41 ( 9
þ15 ( 1
)
2.7 ꢁ 105
1.9 ꢁ 105
ꢀ61 ( 1
ΔH° (kJ molꢀ1
ΔS° (J molꢀ1 Kꢀ1
ΔV° (cm3 molꢀ1
)
ꢀ54.6 ( 0.2
ꢀ82 ( 1
ꢀ3.9 ( 0.4
)
ꢀ107 ( 5
)
ꢀ5.0 ( 0.2
a X represents H, S, V, respectively. b Activation parameters determined through the reaction of FeIII(TMPS)(CN)2 with NO.
Scheme 1
and release of the second cyanide ion, confirm the six-coordinate
nature of the Fe(TMPS)(CN)(H2O) complex and suggest a
dissociative reaction mechanism. Since coordination of the
second CNꢀ has to be preceded by dissociation of H2O, the
mechanism of the reaction can be expressed by Scheme 1.
According to a steady-state approximation on FeIII(TMPS)-
(CN), the observed first-order rate constant (kobs(2)) can be
expressed as shown in eq 9.33
k20 k}2½CNꢀꢂ þ k0ꢀ2kꢀ} 2½H2Oꢂ
kobsð2Þ
¼
ð9Þ
k0 ½H2Oꢂ þ k}½CNꢀꢂ
ꢀ2
2
Under the selected conditions, it can be assumed that k0ꢀ2[H2O]
. k00 [CNꢀ] as no curvature in the dependence of kobs(2) on
2
[CNꢀ] was observed. Consequently, rate law (9) can be simplified
to the expression shown in eq 10, where k2 = k02k00 /k0ꢀ2[H2O]
2
and kꢀ2 = k00
ꢀ2
k02k}2½CNꢀꢂ
kobsð2Þ
¼
þ k}ꢀ2
ð10Þ
k0 ½H2Oꢂ
ꢀ2
According to eq 10, the activation parameters for the “on” reaction
can be expressed as a composite of three terms:
Figure 5. (a) Temperature induced spectral changes for reaction 4,
inset: plot of ln(Kth2) vs 1/T. Experimental conditions: [FeIII
-
0
00
0
ΔHq ¼ ΔHq 2 þ ΔHq 2 ꢀ ΔHq
ð11Þ
(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, [NaCN] = 3 ꢁ 10ꢀ4 M, pH = 8,
[TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3), P = 0.1 MPa; (b)
Pressure induced spectral changes, inset: plot of ln(Kth2) vs pressure.
Experimental conditions: [FeIII(TMPS)(OH)] = 5 ꢁ 10ꢀ6 M,
[NaCN] = 1.5 ꢁ 10ꢀ4 M, pH = 8, [TAPS buffer] = 0.05 M, I = 0.15 M
(NaNO3), T = 30 °C. For the calculation of K2th, the concentration of
free cyanide was used, and changes in pKa(HCN) with temperature and
pressure, as well as changes in the pH of the buffers with temperature were
taken into account.
2
ꢀ2
0
00
0
ΔSq ¼ ΔSq 2 þ ΔSq 2 ꢀ ΔSq
ð12Þ
ð13Þ
2
ꢀ2
0
00
0
ΔVq ¼ ΔVq 2 þ ΔVq 2 ꢀ ΔVq
2
ꢀ2
Therefore the values of ΔHq , ΔSq , and ΔVq for the forward
2
2
2
reaction will include compensating effects arising from the differ-
ent contributions. In terms of ΔVq , the first reaction step in
Mechanism for the Reversible Binding of CNꢀ to Fe-
2
(TMPS)(CN)(H2O). The positive values of ΔVq for the binding
Scheme 1 involves the dissociation of a water molecule for which
3418
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
Scheme 2
structed for the reversible binding of CNꢀ to FeIII(TMPS)(CN)-
(H2O), which nicely demonstrates the dissociative character of both
the forward and the back reactions.
A comparison of the reaction parameters calculated on the
basis of kinetic data with those obtained from thermodynamic
measurements shows only minor discrepancies. The negative
value of ΔH° demonstrates the exothermic character of reaction
4 that involves the reversible binding of cyanide to FeIII(TMPS)-
(CN)(H2O). The reaction entropy and volume values were found to
be negative. The overall reaction volume does not only result
from the displacement of H2O by CNꢀ, but is also due to a change in
electrostriction. Charge concentration which occurs on going
from [FeIII(TMPS)(CN)(H2O)]4ꢀ to [FeIII(TMPS)(CN)2]5ꢀ
(Scheme 1) results in a volume contraction that arises from an
increase in electrostriction. Furthermore, a relatively small over-
all volume collapse confirms the lack of any spin state change on
the iron center on going from the low-spin monocyano to the
low-spin dicyano complex. As can be expected, the binding of the
first cyanide ion to the high-spin, five-coordinate FeIII(TMPS)-
(OH) complex will involve a change in spin state to form the
low-spin FeIII(TMPS)(CN)(H2O) complex as found for many
different six-coordinate mixed-ligand monocyano iron(III) porphyrin
species.10,18
Figure 6. Volume profile for the reversible binding of CNꢀ to FeIII-
(TMPS)(CN)(H2O) according to reaction 4 at pH 8.
ΔVq 2 ꢀ ΔVq ꢀ2 is expected to be ∼ þ13 cm3 molꢀ1, estimated
for the release of a water molecule in going from a six- to a five-
coordinate complex.34 This value will be offset by the contribution
0
0
UVꢀvis Spectral Data for the Reaction of NO with FeIII-
(TMPS)(CN)(H2O) and FeIII(TMPS)(CN)2. As previously men-
tioned, on mixing cyanide with an aqueous solution of FeIII-
(TMPS)(OH) at pH 8 and 9 in the concentration ratio 1:1 (or
lower) leads to the formation of both FeIII(TMPS)(CN)(H2O)
and FeIII(TMPS)(CN)2. Since the equilibrium constants for
reactions 2 and 4 at pH 8 are almost equal (see Table 1), it was
not possible to obtain a high yield of the monocyano complex at
room temperature. Addition of NO to a deoxygenated solution
of FeIII(TMPS) containing a small excess of total cyanide, result-
ed in spectral changes with no clean isosbestic points that
indicated the occurrence of three simultaneous reactions. The
fastest reaction leads to the formation of FeII(TMPS)(OH)-
(NOþ) with a characteristic Soret band maximum at 430 nm,
Q-band at 545 nm and isosbestic points at 420, 447, and 527 nm
(Figure 7a). The concurrent slower reaction corresponds to a
bathochromic shift of the Soret (434 nm) and Q (553 nm) band
maxima, and isosbestic points at 430, 457, and 549 nm. The
spectral changes associated with this reaction indicated the
formation of FeII(TMPS)(CN)(NOþ) (Figure 7b). On a much
longer time scale, small absorbance changes (no shift in Soret and
Q-band, but only absorbance increase) associated with nitrosylation
of FeIII(TMPS)(CN)2, were observed. In the presence of a high
excess of cyanide, where the main species is FeIII(TMPS)(CN)2,
of ΔVq 2, expected to be negative for the coordination of cyanide.
00
In many cases the value of ΔSq usually reflects the trend in ΔVq;
however, in our case the determined value of ΔSq2 was found to be
negative, that is, opposite to the positive ΔVq2 value. We assume
that the obtained value of ΔSq2 is a result of the composite nature
of k2. In general, considerable uncertainties contributing to the
determination of ΔSq limit the usability of this parameter in
drawing mechanistic conclusions.
The activation parameters for the release of cyanide (Table 2)
are in good agreement at both studied pH values. In accordance
with the principle of microscopic reversibility, the “on” and “off”
reactions are expected to pass through the same transition state.
The positive values of ΔSqꢀ2 and ΔVqꢀ2 can be interpreted in
terms of a dissociative mechanism in which Fe-CN bond cleavage
and changes in electrostriction are responsible for the positive
values. The dissociation of CNꢀ from [FeIII(TMPS)(CN)2]5ꢀ
leads to the formation of the five-coordinate intermediate species
[FeIII(TMPS)(CN)]4ꢀ (see Scheme 1). Although the overall
charge does not change on going from products to transition state,
charge dilution does take place which can be expected to cause a
reduction in the electrostricted water.35 This in turn will result in a
positive contribution to both ΔVqꢀ2 and ΔSqꢀ2. On the basis of the
reported activation volumes, a volume profile (Figure 6) was con-
3419
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
Figure 8. Plot of kobs vs [NO] for the reaction of FeIII(TMPS)(CN)-
(H2O) with NO. Experimental conditions: [FeIII(TMPS)(OH)] = 1 ꢁ
10ꢀ5 M, [NaCN] = 2.5 ꢁ 10ꢀ4 M, pH = 8, [TAPS buffer] = 0.05 M,
I = 0.15 M (NaNO3), λ = 446 nm, T = 20 °C.
New bands appeared at 434 and 553 nm, corresponding to the
conversion of FeIII(TMPS)(CN)2 to FeII(TMPS)(CN)(NOþ)
(Figure 7c).
Kinetic Studies on the Reaction of NO with FeIII(TMPS)-
(CN)(H2O). The formation of FeII(TMPS)(CN)(NOþ) was
studied under pseudo-first-order conditions by following the
reaction of NO with FeIII(TMPS)(CN)(H2O) in the presence of
different concentrations of NaCN and at different temperatures
to find the best reaction conditions. Because of the presence of
three species in solution, namely, FeIII(TMPS)(OH), FeIII-
(TMPS)(CN)(H2O), and FeIII(TMPS)(CN)2, which can si-
multaneously undergo nitrosylation, the appropriate spectral
range had to be selected to monitor the binding of NO to the
monocyano complex. The stopped-flow kinetic experiments
were performed at 446 nm that corresponds to the isosbestic
point for the reaction of NO with FeIII(TMPS)(CN)2 (see
Figure 7c) and shows negligible absorbance changes associated
with the conversion of FeIII(TMPS)(OH) to FeII(TMPS)-
(OH)(NOþ) (see Figure 7a). Furthermore, NO binding to
FeIII(TMPS)(CN)(H2O) was followed in the presence of a
relatively large excess of NaCN (1:25) such that the con-
centration of FeIII(TMPS)(OH) present in solution prior to
addition of NO was practically zero. Therefore, at this wave-
length only absorbance changes associated with the conversion
of FeIII(TMPS)(CN)(H2O) to FeII(TMPS)(CN)(NOþ) were
observed. The essential advantage of this approach was the
possibility to record single exponential kinetic traces,36 although
they were associated with quite small absorbance changes at this
wavelength. Furthermore, it is important to note that since the ratio
of the observed rate constants for NO binding to FeIII(TMPS)-
(CN)(H2O) and FeIII(TMPS)(CN)2 varies from 25 to 267 over
the [NO] range from 0.1 to 0.6 mM, respectively, these reactions do
not interfere with each other under the selected reaction conditions
(see Figures 8 and 9). The data in Figure 8 show that the observed
rate constants increase linearly with [NO] at low NO concentration
and display a slight saturation at higher NO concentrations. The
latter is ascribed to complications arising from side reactions that
occur under these conditions.
Figure 7. Spectral changes accompanying the nitrosylation of the
FeIII(TMPS) complex in the presence of cyanide. (a) Reaction between
NO and FeIII(TMPS)(OH) leading to the formation of FeII(TMPS)-
(OH)(NOþ), time 0ꢀ0.15 s; (b) simultaneous slower reaction between
NO and FeIII(TMPS)(CN)(H2O) leading to the formation of FeII-
(TMPS)(CN)(NOþ), time 0.15ꢀ20 s. Experimental conditions (a)
and (b): [FeIII(TMPS)(OH)] = 1 ꢁ 10ꢀ5 M, [NaCN] = 2 ꢁ 10ꢀ5 M,
[NO] = 0.9 mM, pH = 8, [TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3),
T = 10 °C. (c) Reaction between NO and FeIII(TMPS)(CN)2 leading to
the formation of FeII(TMPS)(CN)(NOþ). Experimental conditions (c):
[FeIII(TMPS)(OH)] = 5 ꢁ 10ꢀ6 M, [NaCN] = 5 ꢁ 10ꢀ4 M, [NO] =
0.54 mM, pH = 8, [TAPS buffer] = 0.05 M, I=0.15M(NaNO3), T=20°C.
The second-order rate constant calculated from the data in the
lower NO concentration range by fitting a linear dependence to the
only one slow reaction was observed after addition of NO. The
observed spectral changes with clean isosbestic points (427, 446,
and 533 nm) showed a shift of the Soret band to longer wavelength.
data presented in Figure 8 has a value of (8.2 ( 0.3) ꢁ 103 Mꢀ1 sꢀ1
.
It is important to note that the observed rate constants depend
on the cyanide concentration present in solution, that is, with
3420
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
Scheme 3
Scheme 4
Figure 9. Dependence of kobs on the NO concentration. Experimental
conditions: [FeIII(TMPS)(OH)] = 5 ꢁ 10ꢀ6 M, [NaCN] = 5 ꢁ 10ꢀ4
M, pH = 8, [TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3), λ = 460 nm,
T = 20 °C.
followed by dissociation of CNꢀ. In the next step rapid NO
binding to the five-coordinate monohydroxo complex leads to
formation of FeII(TMPS)(OH)(NOþ).
increasing excess of [NaCN] kobs also increases. Accordingly,
determination of an accurate value of the second order rate
constant for NO binding to FeIII(TMPS)(CN)(H2O) was
impossible (it is in the order of 103 Mꢀ1 sꢀ1 at 20 °C for
concentration ratios of [FeIII(TMPS)(OH)] to [NaCN] from
1:1 to 1:25). This indicates that equilibrium 2 does contribute to
the observed rate constant. The presence of three different
complexes in solution that simultaneously, but at different
rate constants, bind NO, made it very difficult to follow the
nitrosylation of only the monocyano complex. Scheme 2 sum-
marizes the three reactions which lead to two different nitrosyl
complexes, as well as the concomitant equilibria between the
reactant species.
The data at 20 °C presented in Scheme 2 indicate that the
estimated rate constant for NO binding to FeIII(TMPS)(CN)-
(H2O) is lower than the rate constant for NO binding to
FeIII(TMPS)(OH). This can be ascribed to dissociation of the
aqua ligand which has to precede NO binding to FeIII(TMPS)-
(CN)(H2O), leading to the formation of FeII(TMPS)(CN)-
(NOþ). Furthermore, the rate constant for cyanide release from
the monocyano complex is relatively high. Therefore, it can be
expected that for the applied [NO] range the monohydroxo
nitrosyl complex, FeII(TMPS)(OH)(NOþ), is produced in a
relatively high quantity, even though only traces of FeIII(TMPS)-
(OH) are present in solution prior to the addition of NO.
This conclusion is supported by the fact that nitrosylation
of FeIII(TMPS)(OH) was observed even if the ratio of the
FeIII(TMPS)(OH) complex to cyanide concentration was very
high, namely, 1:25. The presence of such a large excess of cyanide
will lead to the existence of mainly the dicyano species and only a
small fraction of the monocyano species, which was confirmed
by spectra recorded after mixing FeIII(TMPS)(OH) with cyanide in
the concentration ratio 1:25.
Kinetic Studies on the Reaction of NO with FeIII(TMPS)-
(CN)2. Kinetic studies on the binding of NO to FeIII(TMPS)-
(CN)2 were performed in the presence of a high excess of NaCN
to eliminate other possible side reactions. The binding of NO to
FeIII(TMPS)(CN)2 is described by reaction 14.
kon
FeIIIðTMPSÞðCNÞ2 þ NO a FeIIðTMPSÞðCNÞðNOþÞ þ CNꢀ
koff
ð14Þ
Single exponential kinetic traces were recorded at 460 nm and
20 °C. The observed rate constants decreased with increasing
[NO] to reach a limiting value at high [NO] (Figure 9). Such a
correlation between kobs and [NO] can be attributed to the
operation of a limiting dissociative mechanism as depicted in
Scheme 4. By applying a steady-state approximation to the five-
coordinate intermediate FeIII(TMPS)(CN), the observed first-
order rate constant can be expressed as in eq 15.33
k}ꢀ2k3½NOꢂ þ k}2kꢀ3½CNꢀꢂ
kobs
¼
ð15Þ
k}2½CNꢀꢂ þ k3½NOꢂ
The characteristic decrease in kobs with increasing [NO], is a
consequence of the smaller value of k00ꢀ2 in comparison to k
.
ꢀ3
Nevertheless, at a high [NO] the observed rate constant shows
no dependence on the NO concentration, which can be ac-
counted for in terms of a limiting dissociative mechanism in
which the release of CNꢀ from FeIII(TMPS)(CN)2 is the rate-
determining step of the reaction. Therefore, eq 15 can be reduced
to kobs = kon = k00ꢀ2. Thus, the values of kon, namely, (1.54 (
0.01) ꢁ 10ꢀ2 and (1.66 ( 0.01) ꢁ 10ꢀ2 sꢀ1 at pH 8 and 9,
respectively, were calculated using NO concentrations in the
range from 0.27 to 0.9 mM. These results suggest that the first-
order rate constant for the nitrosylation of FeIII(TMPS)(CN)2
corresponds to the first-order rate constant for the dissociation of
CNꢀ, that is, kon = kꢀ2. Therefore, the reaction of NO with
Taking into account that the binding of NO has to be preceded
by the dissociation of one axial ligand, the mechanism of the
reaction can be presented by Scheme 3. Accordingly, the mono-
cyano species reacts with NO along two parallel pathways leading
to two different reaction products as described above.
The major reaction pathway proceeds through the dissociation
of water, followed by the binding of NO to the five-coordinate
monocyano complex to produce FeII(TMPS)(CN)(NOþ). The
second pathway involves deprotonation of coordinated water,
FeIII(TMPS)(CN)2 can be used to determine the value of k
ꢀ2
more accurately, since the determination of k from the
ꢀ2
intercept was coupled to large uncertainties (see Table 1). The
upper limit of kꢀ3 is e0.1 sꢀ1 (see Figure 9) and much slower
than the release of NO by FeII(TMPS)(OH)(NOþ) (kꢀ4),6
3421
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
activation parameters for the “on” reaction with NO are
synonymous with those for cyanide release from FeIII(TMPS)-
(CN)2. Therefore, the activation parameters ΔHqꢀ2 = ΔHq
,
on
ΔSqꢀ2 = ΔSqon, and ΔVqꢀ2 = ΔVqon summarized in Table 2,
were obtained from the nitrosylation of FeIII(TMPS)(CN)2 as
described above.
A similar limiting dissociative mechanism was proposed for the
nitrosylation of FeIII(TMPS)(OH)(MeIm) in the ionic liquid
[emim][NTf2], 1-ethyl-3-methylimidazolium bis-trifluorometh-
ylsulfonylamide, where the formation of the six-coordinate FeIII-
(TMPS)(OH)(MeIm) complex was connected with the pre-
sence of methylimidazole impurities at a micromolar level in the
ionic liquid.8 The kinetic data showed an independence of kobs on
[NO], suggesting that the release of methylimidazole is the rate-
determining step of the reaction rather than the binding of NO.
The first-order rate constant, kon = (7.0 ( 0.2) ꢁ 10ꢀ2 sꢀ1 found
for the nitrosylation of FeIII(TMPS)(OH)(MeIm) is compar-
able with that obtained for the reaction of NO with FeIII(TMPS)-
(CN)2. On the contrary, NO binding to the five-coordinate
FeIII(TMPS)(OH) complex in basic aqueous solution follows an
associative mechanism.6 The obtained results indicate that the
presence of a strong ligand in the sixth axial position strongly
affects NO coordination to the metal center and results in a
changeover in the nitrosylation mechanism. Earlier conclusions
drawn from studies on water-soluble iron(III) porphyrins and
ferriheme proteins confirm that NO binding could be mainly
governed by axial ligand lability.2,3(c) However, the influence of
an overall spin transition and related structural changes can play a
crucial role in the reactivity of NO toward ferriporphyrins.6 The
ability of NO to substitute CNꢀ in FeIII(TMPS)(CN)2 suggests
a very high affinity of this signaling molecule for the iron(III)
center. Colman and co-workers previously reported that NO can
substitute inhibitors of cytochrome c oxidase like CNꢀ and CO,
and so act as “antidote” for their poisoning.37 Taking advantage
of the fact that cyanide dissociation is the rate-determining step
in the nitrosylation of cyanide ligated myoglobin, the koff rate
constant for cyanide release from ferric myoglobin was found to
Figure 10. Temperature and pressure dependence for the binding of
NO to FeIII(TMPS)(CN)2. (a) Eyring plot of ln(kon/T) vs 1/T. (b) Plot
of ln(kon) vs pressure, T = 25 °C. Experimental conditions:
[FeIII(TMPS)(OH)] = 7.5 ꢁ 10ꢀ6 M, [NaCN] = 5.6 ꢁ 10ꢀ4 M,
[NO] = 0.9 mM, pH = 8, [TAPS buffer] = 0.05 M, I = 0.15 M (NaNO3),
λ = 455 nm.
which indicates that OHꢀ has a stronger labilizing effect than
CNꢀ in the present system.
For more detailed information on the binding mechanism of
NO to FeIII(TMPS)(CN)2, the temperature dependence of kobs
was measured over the range 15 to 40 °C under limiting [NO]
conditions. At both selected pH values, plots of ln(kon/T) cor-
related linearly with 1/T (Figure 10a). The values of the
activation enthalpy (ΔHonq) (pH 8: 92.0 ( 0.6 kJ molꢀ1; pH
9: 94 ( 3 kJ molꢀ1) and activation entropy (ΔSonq) (pH 8: þ36
be 4.3 ꢁ 10ꢀ4 s
.
ꢀ1 23 This value is more than a 100 times smaller
than the value found in this study for the release of cyanide from
FeIII(TMPS)(CN)2.
( 2 J molꢀ1
K
ꢀ1; pH 9: þ41 ( 9 J molꢀ1
K
ꢀ1) are in good
’ CONCLUSIONS
agreement within the error limits at both pH values. Further
insight into the reaction mechanism was obtained by applying
high pressure (10ꢀ130 MPa) kinetic measurements, which
indicated positive activation volumes, namely, þ11.2 ( 0.3
and þ15.3 ( 0.5 cm3 molꢀ1 at pH 8 and pH 9, respectively.
The linear dependence of ln(kon) on pressure used to determine
ΔVqon, is presented in Figure 10b.
The reported study revealed detailed insight into the mechan-
ism of the reactions of CNꢀ with the water-soluble ferric
porphyrin, FeIII(TMPS). The obtained kinetic and thermody-
namic data for CNꢀ coordination to the Fe(III) center provide a
better understanding of the mechanism of nitric oxide binding to
the cyano complexes FeIII(TMPS)(CN)(H2O) and FeIII-
(TMPS)(CN)2. The positive activation volume and indepen-
dence of the rate constant on pH for the binding of the second
cyanide ion to the monocyano species imply the six-coordinate
nature of the complex and presence of a water ligand in the axial
position trans to cyanide. Nitrosylation of both FeIII(TMPS)-
(CN)(H2O) and FeIII(TMPS)(CN)2 resulted in the formation
of the same species, namely, FeII(TMPS)(CN)(NOþ). Detailed
kinetic studies on the coordination of NO to FeIII(TMPS)(CN)2
indicated that nitrosylation occurs via a limiting dissociative
mechanism.
The positive values of the activation parameters ΔSq and
on
ΔVqon reported for the binding of NO to FeIII(TMPS)(CN)2 are
consistent with a limiting dissociative mechanism in which
FeIIIꢀCN bond cleavage is the rate-determining step. We were
not able to determine the activation parameters for the release of
NO from FeII(TMPS)(CN)(NOþ), but on the basis of micro-
scopic reversibility we suggest that the reverse reaction passes
through the same five-coordinate intermediate as shown in
Scheme 4. Therefore, it is reasonable to expect that the back
reaction follows the same limiting dissociative substitution
mechanism. As mentioned above, according to the limiting
dissociative mode of NO binding to FeIII(TMPS)(CN)2, the
The obtained results allow us to draw a conclusion related to
the effect of the axial ligands on the reactivity pattern of NO with
ferriporphyrins. The presence of a strong nucleophile in the sixth
3422
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
axial position results in a changeover of the mechanism of NO
binding to the iron(III) center in comparison with the mechan-
ism observed for nitrosylation of five-coordinate porphyrin
complexes. Furthermore, the fact that NO is able to substitute
the CNꢀ ligand in FeIII(TMPS)(CN)2, further supports that NO
can be a better ligand than cyanide also in iron(III) hemopro-
teins. Accordingly, the mechanism of CNꢀ substitution by NO
proposed in the present study can shed more light on the mecha-
nism of the recovery process in the treatment of cyanide inhi-
bition of cytochrome c oxidase.
Inorg. Chim. Acta 1998, 277, 242–246. (c) Balch, A. L.; Noll, B. C.; Safari,
N. Inorg. Chem. 1993, 32, 2901–2905. (d) Schappacher, M.; Fischer, J.;
Weiss, R. Inorg. Chem. 1989, 28, 389–390. (e) Hada, M. J. Am. Chem. Soc.
2004, 126, 486–487. (f) Ikezaki, A.; Nakamura, M. Inorg. Chem. 2002,
41, 2761–2768.
(12) Scheidt, W. R.; Haller, K. J.; Hatano, K. J. Am. Chem. Soc. 1980,
102, 3017–3021.
(13) (a) Stochel, G. Coord. Chem. Rev. 1992, 114, 269–295. (b)
Stasicka, Z.; Wasielewska, E. Coord. Chem. Rev. 1997, 159, 271–294. (c)
Oszajca, J.; Stochel, G.; Wasielewska, E.; Stasicka, Z.; Gryglewski, R. J.;
Jakubowski, A.; Cieslik, K. J. Inorg. Biochem. 1998, 69, 121–127.
(14) (a) Beasley, D. M.; Glass, W. I. Occup. Med. (Lond.) 1998,
48, 427–431. (b) Cumings, T. F. Occup. Med. (Lond.) 2004, 54, 82–85.
(c) Brenner, M.; Mahon, S. B.; Lee, J.; Kim, J.; Mukai, D.; Goodman, S.;
Kreuter, K. A.; Ahdout, R.; Mohammad, O.; Sharma, V. S.; Blackledge,
W.; Boss, G. R. J. Biomed. Opt. 2010, 15, 017001. (d) Broderick, K. E.;
Potluri, P.; Zhuang, S.; Scheffler, I. E.; Sharma, V. S.; Pilz, R. B.; Boss,
G. R. Exp. Biol. Med. (Maywood, NJ, U. S.) 2006, 231, 641–649. (e)
Bhattacharya, R.; Tulsawani, R. J. Environ. Biol. 2009, 30, 515–520.
(15) (a) van Eldik, R.; Palmer, D. A.; Schmidt, R.; Kelm, H. Inorg.
Chim. Acta 1981, 50, 131–135. (b) van Eldik, R.; Gaede, W.; Wieland, S.;
Kraft, J.; Spitzer, M.; Palmer, D. A. Rev. Sci. Instrum. 1993, 64, 1355–1357.
(16) Spitzer, M.; Garting, F.; van Eldik, R. Rev. Sci. Instrum. 1988,
59, 2092–2093.
(17) Silver, J.; Taies, J. A. Inorg. Chim. Acta 1989, 159, 231–235.
(18) Wang, J.-T.; Yeh, H. J. C.; Johnson, D. F. J. Am. Chem. Soc. 1978,
100, 2400–2405.
(19) Goff, H.; Morgan, L. O. Inorg. Chem. 1976, 15, 2069–2076.
(20) Arifuku, F.; Ujimoto, K.; Kurihara, H. Bull. Chem. Soc. Jpn. 1986,
59, 149–154.
(21) (a) Kobayashi, N. Inorg. Chem. 1985, 24, 3324–3330. (b)
Kobayashi, N.; Koshiyama, N.; Osa, T.; Kuwana, T. Inorg. Chem.
1983, 22, 3608–3614.
(22) Wolak, M.; van Eldik, R. Chem.—Eur. J. 2007, 13, 4873–4883.
(23) Ascenzi, P.; di Masi, A.; Gullotta, F.; Mattu, M.; Ciaccio, C.;
Coletta, M. Biochem. Biophys. Res. Commun. 2010, 393, 196ꢀ200 and
references therein.
’ AUTHOR INFORMATION
Corresponding Authors
*E-mail: vaneldik@chemie.uni-erlangen.de (R.v.E.), stochel@
chemia.uj.edu.pl (G.S.).
’ ACKNOWLEDGMENT
This work was supported by the International Ph.D. Study
Program at the Faculty of Chemistry, Jagiellonian University,
within the Foundation for Polish Science MPD Program cofi-
nanced by the European Regional Development Fund. The
research was carried out with equipment purchased with financial
support from the European Regional Development Fund within
the framework of the Polish Innovation Economy Operational
Program (contract no. POIG.02.01.00-12-023/08). A.F. and R.v.
E. gratefully acknowledge continued financial support from the
Deutsche Forschungsgemeinschaft.
’ REFERENCES
(1) (a) Koshland, D. E. Science 1992, 258, 1861–1865. (b) Culotta,
E.; Koshland, D. E. Science 1992, 258, 1862–1865.
(2) (a) Ford, P. C.; Lorkovic, I. M. Chem. Rev. 2002, 102, 993ꢀ118
and references therein. (b) Laverman, L. E.; Ford, P. C. J. Am. Chem. Soc.
2001, 123, 11614ꢀ11622 and references therein. (c) Ford, P. C.;
Laverman, L. E.; Lorkovic, I. M. Adv. Inorg. Chem. 2003, 54, 203–257
and references therein.
(3) (a) Franke, A.; Stochel, G.; Jung, C.; van Eldik, R. J. Am. Chem.
Soc. 2004, 126, 4181ꢀ4191 and references therein. (b) Theodoridis, A.;
van Eldik, R. J. Mol. Catal. A: Chem. 2004, 224, 197–205. (c) Hoshino,
M.; Ozawa, K.; Seki, H.; Ford, P. C. J. Am. Chem. Soc. 1993,
115, 9568–9575. (d) Hoshino, M.; Kogure, M. J. Phys. Chem. 1989,
93, 5478–5484. (e) Wanat, A.; Wolak, M.; Orzel, L.; Brindell, M.;
van Eldik, R.; Stochel, G. Coord. Chem. Rev. 2002, 229, 37–49.
(4) Laverman, L. E.; Wanat, A.; Oszajca, J.; Stochel, G.; Ford, P. C.;
van Eldik, R. J. Am. Chem. Soc. 2001, 123, 285–293.
(5) (a) Jee, J.-E.; Eigler, S.; Jux, N.; Zahl, A.; van Eldik, R. Inorg.
Chem. 2007, 46, 3336–3352. (b) Jee, J.-E.; Wolak, M.; Balbinot, D.; Jux,
N.; Zahl, A.; van Eldik, R. Inorg. Chem. 2006, 45, 1326–1337. (c) Jee,
J.-E.; Eigler, S.; Hampel, F.; Jux, N.; Wolak, M.; Zahl, A.; Stochel, G.; van
Eldik, R. Inorg. Chem. 2005, 44, 7717–7731.
(24) Sybert, D. W.; Mofat, K.; Gibson, Q. H. J. Biol. Chem. 1976,
251, 45–52.
(25) Brancaccio, A.; Cutruzzolz, F.; Allocatelli, C. T.; Brunori, M.;
Smerdon, S. J.; Wilkinson, A. J.; Dou, Y.; Keenan, D.; Ikeda-Saito, M.;
Brantley, R. E.; Olson, J. S. J. Biol. Chem. 1994, 35, 13843–13853.
(26) Dou, Y.; Olson, J. S.; Wilkinson, A. J.; Ikeda-Saito, M. Biochem-
istry 1996, 35, 7107–7113.
(27) (a) Giacometti, G. M.; Da Ros, A.; Antonini, A.; Brunori, M.
Biochemistry 1975, 14, 1584–1585. (b) Giacometti, G. M.; Ascenzi, P.;
Bolognesi, M.; Brunori, M. J. Mol. Biol. 1981, 146, 363–374. (c)
Bolognesi, M.; Coda, A.; Frigerio, F.; Gatti, G.; Ascenzi, P.; Brunori,
M. J. Mol. Biol. 1991, 213, 621–625. (d) Mattevi, A.; Gatti, G.; Coda, A.;
Rizzi, M.; Ascenzi, P.; Brunori, M.; Bolognesi, M. J. Mol. Recognit. 1991,
4, 1–6.
(28) Marques, H. M.; Brown, K. L.; Jacobsen, D. W. J. Biol. Chem.
1988, 263, 12378–12383.
(29) Lambier, A. M.; Heremans, K.; Dunford, H. B. Biophys. Chem.
1983, 18, 195–201.
(30) Kitamura, Y.; Ioth, T. J. Solution Chem. 1987, 16, 715–725.
(31) Goldberg, R. N.; Kishore, N.; Lennen, R. J. Phys. Chem. Ref.
Data 2002, 31, 231–370.
(6) Wolak, M.; van Eldik, R. J. Am. Chem. Soc. 2005, 127
13312–13315.
(7) Franke, A.; Roncaroli, F.; van Eldik, R. Eur. J. Inorg. Chem. 2007,
6, 773ꢀ798 and references therein.
(32) By using the concentration of free cyanide to calculate the reaction
and activation parameters for reaction 4 makes them independent of the
ionization of HCN (ΔHHCN = 45 kJ molꢀ1, ΔSHCN = ꢀ22 J molꢀ1 Kꢀ1
and ΔVHCN = ꢀ6.5 cm3 molꢀ1). Since [CNꢀ] depends on [Hþ] (eq 6),
changes in pH with temperature and pressure were taken into account.
According to the thermodynamic data for the ionization of the buffers,
namely, TAPS (ΔH = 40 kJ molꢀ1, ΔS = ꢀ26 J molꢀ1 Kꢀ1 and ΔV =
0.5 cm3 molꢀ1) and CHES (ΔH = 39 kJ molꢀ1, ΔS = ꢀ47 J molꢀ1 Kꢀ1
and ΔV = 1 cm3 molꢀ1), the [Hþ] calculated at particular temperatures
were used to calculate [CNꢀ], whereas changes in pH with pressure
(8) Schmeisser, M.; van Eldik, R. Inorg. Chem. 2009, 48, 7466–7475.
(9) (a) Ikezaki, A.; Ikeue, T.; Nakamura, M. Inorg. Chim. Acta 2002,
335, 91–99. (b) Nakamura, M.; Ikeue, T.; Ikezaki, A.; Ohgo, Y.; Fujii, H.
Inorg. Chem. 1999, 38, 3857–3862. (c) Nakamura, M.; Ikeue, T.; Fujii,
H.; Yoshimura, T. J. Am. Chem. Soc. 1997, 119, 6284–6291.
(10) Li, J.; Noll, B. C.; Schulz, C. E.; Scheidt, W. E. Inorg. Chem.
2007, 46, 2286–2298.
(11) (a) Patra, R.; Rath, S. P. Inorg. Chem. Commun. 2009,
12, 515–519. (b) Bartczak, T. J.; Wolowiec, S.; Latos-Grazynski, L.
3423
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424

Inorganic Chemistry
ARTICLE
were negligible since the values of ΔV for the ionization of the buffers are
very small.
(33) Malin, J. M.; Toma, H. E.; Giesbrecht, E. J. Chem. Educ. 1977,
54, 385–386.
(34) Swaddle, T. W. Adv. Inorg. Bioinorg. Mechanisms; Academic
Press Inc.: New York, 1983; Vol. 2, p 95.
(35) van Eldik, R.; Asano, T.; le Noble, W. J. Chem. Rev. 1989,
89, 549–688.
(36) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms, 2nd
ed.; McGraw-Hill, Inc.: New York, 1995; 3.6 and 4.1, pp 62ꢀ64 and 75.
(37) Colman, J. P.; Dey, A.; Decreau, R. A.; Yang, Y.; Hosseini, A.;
Solomon, E. I.; Eberspacher, T. A. Proc. Natl. Acad. Sci. U.S.A. 2008,
105, 9892–9896.
3424
dx.doi.org/10.1021/ic1023345 |Inorg. Chem. 2011, 50, 3413–3424