Kinetics and mechanism of the interaction of nitric oxide with pentacyanoferrate(II). Formation and dissociation of [Fe(CN)5NO]3-

Article abstract of DOI:10.1021/ic0342189

The interaction of NO with [Fe(CN)₅H₂O]^3- (generated by aquation of the corresponding ammine complex) to produce [Fe(CN)₅NO]^3- was studied by UV-vis spectrophotometry. The reaction product is the well characterized nitrosyl complex, described as a low-spin Fe(II) bound to the NO radical. The experiments were performed in the pH range 4-10, at different concentrations of NO, temperatures and pressures. The rate law was first-order in each of the reactants, with the specific complex-formation rate constant, k_f = 250 ± 10 M^-1 s^-1 (25.4 °C, I = 0.1 M, pH 7.0), ΔH_f? = 70 ± 1 kJ mol^-1, ΔS_f? = +34 ± 4 J K^-1 mol^-1, and ΔV_f? = +17.4 ± 0.3 cm³ mol^-1. These values support a dissociative mechanism, with rate-controlling dissociation of coordinated water, and subsequent fast coordination of NO. The complex-formation process depends on pH, indicating that the initial product [Fe(CN)₅NO]^3- is unstable, with a faster decomposition rate at lower pH. The decomposition process is associated with release of cyanide, further reaction of NO with [Fe(CN)₄NO]^2-, and formation of nitroprusside and other unknown products. The decomposition can be prevented by addition of free cyanide to the solutions, enabling a study of the dissociation process of NO from [Fe(CN)₅NO]^3-. Cyanide also acts as a scavenger for the [Fe(CN)₅]^3- intermediate, giving [Fe(CN)₆]^4- as a final product. From the first-order behavior, the dissociation rate constant was obtained as k_d = (1.58 ± 0.06) × 10^-5 s^-1 at 25.0 °C, I = 0.1 M, and pH 10.2. Activation parameters were found to be ΔH_d? = 106.4 ± 0.8 kJ mol^-1, ΔS_d? = +20 ± 2 J K^-1 mol^-1, and ΔV_d? = +7.1 ± 0.2 cm³ mol^-1, which are all in line with a dissociative mechanism. The low value of k_d as compared to values for the release of other ligands L from [Fe^II(CN)₅L]^n- suggests a moderate to strong σ-π interaction of NO with the iron(II) center. It is concluded that the release of NO from nitroprusside in biological media does not originate from [Fe(CN)₅NO]^3- produced on reduction of nitroprusside but probably proceeds through the release of cyanide and further reactions of the [Fe(CN)₄NO]^2- ion.

Full text of DOI:10.1021/ic0342189

Inorg. Chem. 2003, 42, 4179−4189
Kinetics and Mechanism of the Interaction of Nitric Oxide with
Pentacyanoferrate(II). Formation and Dissociation of [Fe(CN)5NO]3-
,‡
,†
Federico Roncaroli,†,‡ Jose´ A. Olabe,* and Rudi van Eldik*
Institute for Inorganic Chemistry, UniVersity of Erlangen-Nu¨rnberg,
Egerlandstrasse 1, 91058 Erlangen, Germany, and Department of Inorganic, Analytical and
Physical Chemistry, INQUIMAE, Faculty of Exact and Natural Sciences, UniVersity of Buenos
Aires, C1428EHA Buenos Aires, Argentina
Received February 26, 2003
The interaction of NO with [Fe(CN)5H O]3- (generated by aquation of the corresponding ammine complex) to
2
produce [Fe(CN)5NO]3- was studied by UV−vis spectrophotometry. The reaction product is the well characterized
nitrosyl complex, described as a low-spin Fe(II) bound to the NO radical. The experiments were performed in the
pH range 4−10, at different concentrations of NO, temperatures and pressures. The rate law was first-order in
each of the reactants, with the specific complex-formation rate constant, k ) 250 ± 10 M-1 s-1 (25.4 °C, I ) 0.1
f
M, pH 7.0), ∆Hq ) 70 ± 1 kJ mol-1, ∆Sq ) +34 ± 4 J K-1 mol-1, and ∆Vq ) +17.4 ± 0.3 cm mol-1. These
3
f
f
f
values support a dissociative mechanism, with rate-controlling dissociation of coordinated water, and subsequent
fast coordination of NO. The complex-formation process depends on pH, indicating that the initial product
[Fe(CN)5NO]3- is unstable, with a faster decomposition rate at lower pH. The decomposition process is associated
with release of cyanide, further reaction of NO with [Fe(CN)4NO]2-, and formation of nitroprusside and other unknown
products. The decomposition can be prevented by addition of free cyanide to the solutions, enabling a study of the
dissociation process of NO from [Fe(CN)5NO]3-. Cyanide also acts as a scavenger for the [Fe(CN)5]3- intermediate,
giving [Fe(CN)6]4- as a final product. From the first-order behavior, the dissociation rate constant was obtained as
kd ) (1.58 ± 0.06) × 10-5 s-1 at 25.0 °C, I ) 0.1 M, and pH 10.2. Activation parameters were found to be ∆H q
d
) 106.4 ± 0.8 kJ mol-1, ∆S q ) +20 ± 2 J K-1 mol-1, and ∆V q ) +7.1 ± 0.2 cm mol-1, which are all in line
3
d
d
with a dissociative mechanism. The low value of kd as compared to values for the release of other ligands L from
II
[Fe (CN)5L]n- suggests a moderate to strong σ−π interaction of NO with the iron(II) center. It is concluded that the
release of NO from nitroprusside in biological media does not originate from [Fe(CN)5NO]3- produced on reduction
of nitroprusside but probably proceeds through the release of cyanide and further reactions of the [Fe(CN)4NO]2-
ion.
Introduction
messenger in neuronal communication, and as a macrophage
defense weapon.2 The widely recognized clinical use of
[Fe(CN)5NO]2- (nitroprusside, NP) as a potent hypotensive
agent spans four decades, but the attribution of its pharma-
cological effect to the release of NO to the biological medium
is more recent. It is assumed that the release is preceded by
the reduction of coordinated NO+ in NP, effected by
biologically relevant thiolates.3 This raises a question on
some unraveled kinetic properties of the [Fe(CN)5NO]3- ion.
The chemistry of nitric oxide (NO) complexes has long
been of interest to transition metal chemists and has been
comprehensively reviewed.1 It has seen a great revival since
the discovery of NO as an essential biological molecule, with
multiple physiological roles as a vascular regulator, as a
* To whom correspondence should be addressed. E-mail: vaneldik@
chemie.uni-erlangen.de (R.v.E.); olabe@qi.fcen.uba.ar (J.A.O.).
† University of Erlangen-Nu¨rnberg.
‡ University of Buenos Aires.
(1) (a) Richter-Addo, G. B.; Legzdins, P. Metal Nitrosyls; Oxford
University Press: New York, 1992. (b) Special issue devoted to nitric
oxide chemistry: Richter-Addo, G. B., Legzdins, P., Burstyn, J. N.,
Guest Eds., Chem. ReV. 2002, 102, 861-1270. (c) Mingos, D. M. P.;
Sherman, D. J. AdV. Inorg. Chem. 1989, 34, 293-377.
(2) (a) Feelish, M., Stamler, J. S., Eds. Methods in Nitric Oxide Research;
Wiley: Chichester, England, 1996 and references therein. (b) Nitric
Oxide, Principles and Actions; Lancaster, J., Jr., Ed.; Academic
Press: New York, 1996. (c) Butler, A. R.; Williams, D. L. H. Chem.
Soc. ReV. 1993, 22, 233-241.
10.1021/ic0342189 CCC: $25.00 © 2003 American Chemical Society
Published on Web 06/03/2003
Inorganic Chemistry, Vol. 42, No. 13, 2003 4179

Roncaroli et al.
In general, the mechanistic details on the coordination of
NO to transition metal centers, as well as its further
dissociation reactions, have been insufficiently disclosed.4
This is probably related to the difficulties in the handling of
NO solutions free of oxidizing impurities. More recent work,
particularly focused on the reactions of NO with porphyrin-
based molecules,4,5 has given a great impulse to these studies.
In the case of classical coordination compounds, several
families of nitrosyl complexes of the type {X5M(NO)}
(mainly group 8 metals, particularly Ru, and X ) cyanides,6
ammines,7 polypyridines,8 etc.) have been well characterized
chemically and spectroscopically, including the correspond-
ing aqua complexes {X5M(H2O)}. However, kinetic studies
on the formation and dissociation reactions are generally
absent, with the exception of very recent publications dealing
with iron-nitrosyl complexes containing aqua,9 cyano,10 and
different chelating ligands.11
Transition metal-nitrosyl complexes span variable ge-
ometries, coordination numbers, and electronic properties,
due to the differences in electronic configurations of the metal
centers and covalent M-NO interactions.12,13 In the En-
emark-Feltham electron-counting formalism,13 the com-
plexes are described as {MNO}n (regardless of the spectator
ligands), where n stands for the number of electrons
associated with the metal d and π*NO orbitals. There are no
assumptions on the actual degree of electron density on M
and the NO group, thus avoiding extreme ways of describing
the nitrosyl oxidation states as NO+ or NO-, on the basis of
the NO coordination mode (linear or bent, respectively),
which sometimes leads to unusual metal oxidation state
assignments. However, the delocalized nature of the {MNO}
fragment is frequently relaxed, by describing NO as a
noninnocent ligand which can act as a diamagnetic, strongly
π*-accepting NO+, as the equally diamagnetic NO- (iso-
electronic with O2), or as the paramagnetic, neutral NO. This
limiting idealized behavior is used, for example, for the
diamagnetic {MNO}6 complexes with Werner-type spectator
ligands, usually described as low-spin MIINO+ species, as
is the case with NP. In the case of porphyrin {MNO}6
systems, however, this description should hardly be general-
ized, because MIIINO structures (or others) could also be
inferred according to different spectroscopic measurements
and their interpretation. Similar ambiguities have also been
analyzed for the {MNO}7 systems, which are most relevant
to our presently reported work. In addition to the systems
with total spin S ) 1/2 (which are usually described as low-
spin FeII with the bound NO-radical (S ) 1/2)), the complexes
3
with total spin S ) /2 afford different possible bonding
descriptions, with iron in the +I, +II, or +III oxidation states
and NO in the +I, 0, or -I states, respectively.12-14
The coordination chemistry of NO in aqueous solutions
deals specifically with the radical species as the only possible
reactant, because both NO+ and NO- (or its acid-base
related “nitroxyl”, HNO) are transient species which react
very rapidly before coordinating to the metal, generating
nitrite or N2O, respectively.1,2 Depending on the oxidation
state of the counterpart aqua ion (usually II or III with group
8 metals), NO may be redox-active during the coordination
process, involving some degree of electron transfer to the
metal. For example, a recent study on the coordination of
NO to [FeIII(CN)5H2O]2- showed that prior reduction to
Fe(II) was needed with NO+ formation, rapid conversion to
nitrite, and final coordination and proton-assisted dehydration
of the latter species, leading to NP as an unique product.10
In the present study, we address the coordination reaction
of NO to the [FeII(CN)5H2O]3- ion. The kinetics and
mechanisms for the coordination of a large series of L ligands
to this aqua complex have been comprehensively studied.6,15
The present inclusion of NO appears to be quite attractive,
given the scarce available kinetic information on its coor-
dination ability. Great significance is assigned to such a
biologically relevant issue as is the mechanism of controlled
NO release from reduced NP.
(3) (a) Clarke, M. J.; Gaul, J. B. Struct. Bonding (Berlin) 1993, 81, 147-
181. (b) Butler, A. R.; Glidewell, C. Chem. Soc. ReV. 1987, 16, 361-
380.
(4) (a) Ford, P. C.; Lorkovic, I. M. Chem. ReV. 2002, 102, 993-1018.
(b) Wolak, M.; van Eldik, R. Coord. Chem. ReV. 2002, 230, 263-
282.
(5) Hoshino, M.; Laverman, L.; Ford, P. C. Coord. Chem. ReV. 1999,
187, 75-102.
(6) Baraldo, L. M.; Forlano, P.; Parise, A. R.; Slep, L. D.; Olabe, J. A.
Coord. Chem. ReV. 2001, 219-221, 881.
(7) Tfouni, E.; Krieger, M.; McGarvey, B. R.; Franco, D. W. Coord. Chem.
ReV. 2003, 236, 57-69.
(8) (a) Callahan, R. W.; Meyer, T. J. Inorg. Chem. 1977, 16, 574-581.
(b) Togano, T.; Kuroda, H.; Nagao, N.; Maekawa, Y.; Nishimura, H.;
Howell, F. S.; Mukaida, M. Inorg. Chim. Acta 1992, 196, 57-63.
(9) Wanat, A.; Schneppensieper, T.; Stochel, G.; van Eldik, R.; Bill, E.;
Wieghardt, K. Inorg. Chem. 2002, 41, 4-10.
(10) Roncaroli, F.; Olabe, J. A.; van Eldik, R. Inorg. Chem. 2002, 42,
5417-5425. In this reference, we argued that NO was acting as an
outer-sphere electron transfer reagent toward the [FeIII(CN)5H2O]2-
ion. A colleague commented that our kinetic data could hardly support
such a behavior, if the available redox potential data for the NO+/NO
couple and basic mechanistic requirements for the claimed outer-sphere
path were considered. He suggested another possibility, implying an
initial association of NO with coordinated cyanide in the precursor
complex, leading to the stabilization of the produced NO+ in the rate-
determining step, and followed by the further chemistry as described
by us. This could be accepted as an inner-sphere electron transfer
pathway involving NO, without changing our main argument in favor
of prior reduction of Fe(III) in the absence of a change in the first
coordination sphere, followed by the coordination of nitrite and
conversion to NO+ as found in NP.
(11) (a) Schneppensieper, T.; Finkler, S.; Czap, A.; van Eldik, R.; Heus,
M.; Nieuwenhuizen, P.; Wreesmann, C.; Abma, W. Eur. J. Inorg.
Chem. 2001, 491. (b) Schneppensieper, T.; Wanat, A.; Stochel, G.;
Goldstein, S.; Meyerstein, D.; van Eldik, R. Eur. J. Inorg. Chem. 2001,
2317. (c) Schneppensieper, T.; Wanat, A.; Stochel, G.; van Eldik, R.
Inorg. Chem. 2002, 41, 2565.
Experimental Section
Materials. NO was purchased from Air Liquide and purified
from higher nitrogen oxides by passing through an Ascarite II
(Aldrich) column. Sodium nitroprusside dihydrate (Fluka), sodium
dithionite (Merck), pyrazine (Aldrich), sodium cyanide (Merck),
mercaptosuccinic acid (Aldrich), sodium nitrite (Merck), sodium
perchlorate, and other chemicals used for the buffer systems were
used without any further purification. Na3[Fe(CN)5NH3]‚3H2O was
prepared and purified as described in the literature starting from
(12) Westcott, B. L.; Enemark, J. L. In Inorganic Electronic Structure and
Spectroscopy, Volume II: Applications and Case Studies; Solomon,
E. I., Lever, A. B. P., Eds. Wiley: New York, 1999.
(13) Enemark, J. H.; Feltham, R. D. Coord. Chem. ReV. 1974, 13, 339-
(14) Scheidt, W. R.; Ellison, M. K. Acc. Chem. Res. 1999, 32, 359-359.
(15) Macartney, D. H. ReV. Inorg. Chem. 1988, 9, 101.
406.
4180 Inorganic Chemistry, Vol. 42, No. 13, 2003

Interaction of NO with Pentacyanoferrate(II)
sodium nitroprusside.16 Its purity was checked by elemental analysis
and by its reaction with pyrazine forming the [Fe(CN)5pz]3- ion.17
Methods. All the solutions were prepared using distilled and
purified water (Milli-Q system) and deoxygenated by bubbling N2
in Schlenk tubes. They were handled and mixed using gastight
syringes. NO-saturated solutions were prepared by slowly bubbling
NO through previously deoxygenated buffer solutions. The maxi-
mum concentration of NO in aqueous solution at room temperature
is already known from our earlier work (1.8 × 10-3 M) and was
checked with an ISO-NOP electrode interfaced to an ISO-NO Mark
II nitric oxide sensor from World Precision Instruments, calibrated
as described previously.11a These solutions were diluted with buffer
solutions to reach the final desired NO concentrations. Solutions
of [Fe(CN)5L]3- (L ) NH3, H2O) were prepared by dissolving the
necessary amount of Na3[Fe(CN)5NH3]‚3H2O in deoxygenated and
thermostated ((0.1 °C) buffer solutions. At pH < 8 and within
2-3 min, aquation of [Fe(CN)5NH3]3- generates the [Fe(CN)5H2O]3-
complex.18 These solutions were generally used within 30 min after
preparation. Solutions containing NP were stored in the dark and
handled under diffuse light. Acetate buffers in the pH rang 4-6,
phosphate and Tris buffers in the pH range 6-8, and borate and
CAPS (3-[cyclohexylamino]-1-propanesulfonic acid) buffers in the
pH range 8-10 were used. The buffer concentrations were 0.01
M, unless otherwise stated. pH measurements were done with a
Mettler Delta 340 pH meter.
UV-vis spectra were recorded on a Varian Cary 5G spectro-
photometer. IR spectra of the products of the reaction of
[Fe(CN)5H2O]3- with NO were measured on an ATI Mattson
Infinity FTIR 9495-120-12500 spectrometer, with a Perkin-Elmer
liquid cell (CaF2 windows, 0.2 mm spacer). Solutions were prepared
by dissolving 10 mg of Na3[Fe(CN)5NH3]‚3H2O in 100 mL of
buffered solutions at pH 7 and 10. NO was bubbled through for 45
min, and the solutions were kept in the dark until the next day.
Finally, N2 was bubbled to eliminate excess of NO, and the volume
was reduced to ca. 5 mL. During the concentration process, a small
amount of a blue solid formed which was removed by filtration
before measuring the IR spectrum. The mass spectrometric mea-
surements were done using a GC-MS instrument, equipped with a
Gerstel CIS3 injection module, a capillary GC Perkin-Elmer
Autosystem XL, and an attached Quadrupole-MS Perkin-Elmer
TurboMass. Databases were from commercial providers (NIST,
NBS). The EPR spectra of 3 × 10-4 M solutions of [Fe(CN)5NO]3-
(pH 10, prepared as described in a following paragraph) in frozen
water (140 K) were recorded in the X-band of a Bruker ESP 300E
spectrometer. The spectra were recorded at 9.47 GHz, 0.635 mW
power, 100 kHz modulation frequency, and 9.434 G modulation
amplitude.
absorbance at 526 nm was measured. In this way, the conversion
into NP as a function of the amount of added NO could be studied.
In order to determine the conversion into NP at different pH, NO
was bubbled for 5 min through 40 mL of the 10-4 M [Fe(CN)5L]3-
(L ) NH3, H2O) solution, followed by mixing with the mercap-
tosuccinic acid solution.
Sodium dithionite was always used as a reductant for NP to
generate the [Fe(CN)5NO]3- complex (λmax ) 345 and 440 nm; ꢀ
) 3500 and 550 M-1 cm-1, respectively),20 or the [Fe(CN)4NO]2-
complex (λmax ) 615, 430, and 350 nm; ꢀ ) 380, 100, and 300
M-1 cm-1, respectively).20 In all cases, NP was about two times in
excess to avoid multielectron reduction. The amount of dithionite
was determined by titration of a solution of NP. After the addition
of more than 1 equiv of dithionite, the band at 345 nm decreases,
and a new band at 440 nm appears.
Kinetic experiments on the complex-formation reaction with NO
at ambient pressure were performed on a Durrum (Dionex) stopped-
flow instrument, equipped with a 2 cm optical path flow cell and
coupled to an online data acquisition system. Equal volumes of
both reactant solutions were mixed. Kinetic traces were analyzed
by the Olis Kinfit program. At least five kinetic runs were performed
at each condition, in many cases up to 12. In general, two or three
independent experiments were performed under the same experi-
mental conditions. The final observed rate constant, kobs, was taken
as the average of measurements that differed less than 15%. In the
preliminary experiments, 10-4 M solutions of [Fe(CN)5H2O]3- were
mixed with an equal volume of 1.8 × 10-3 M solutions of NO. No
electrolytes for ionic strength control were added. The temperature
was kept at 25.4 °C. At pH e 7, the complex and NO solutions
contained the same buffer (0.01 M), whereas at pH g 8 the complex
solution was in 0.002 M acetic acid/acetate buffer at pH 5 and the
NO solution was in 0.04 M boric acid/borate buffer at the final
required pH value.
The influence of nitrite ions on the observed rate constant was
studied by using a saturated NO solution with known amounts of
NaNO2, viz. a final concentration of (0.2-3.0) × 10-2 M and pH
7.0. Some complementary experiments which involved changing
the [Fe(CN)5H2O]3- complex concentration or the buffer concentra-
tion were also done at pH 7.0.
The dependence of the complex-formation reaction on the
concentration of NO, temperature and pressure was studied at pH
7.0 and I ) 0.1 M (NaClO4). For the NO concentration dependence
study, the [Fe(CN)5H2O]3- concentration was in the range (1.0-
5.0) × 10-5 M, and the NO concentration was (0.23-0.90) × 10-3
M in order to have at least a 17-fold excess of NO at 25.4 ( 0.1
°C. A plot of kobs versus [NO] showed a linear dependence, and
the slope enabled us to obtain the second-order rate constant, kf.
For the temperature dependence study, 5 × 10-5 M [Fe(CN)5H2O]3-
and 0.90 × 10-3 M NO were used, with the temperature varying
in the range 15.0-44.6 °C. From the obtained kobs values and the
NO concentration, kf was calculated for each temperature. Finally,
an Eyring plot was used to determine the thermal activation
parameters. For the pressure dependence study, a homemade high-
pressure stopped-flow instrument was used, which has been
described elsewhere.21 The 6 × 10-5 M [Fe(CN)5H2O]3- and 9 ×
10-4 M NO solutions were used at 25.6 ( 0.1 °C. The pressure
was varied from 10 to 130 MPa. Time for temperature and pressure
equilibration had to be allowed for after pressurizing the system.
Experiments in which aliquots of NO solutions were added to a
buffered solution of Na3[Fe(CN)5NH3]‚3H2O were performed in a
1 cm quartz cuvette directly attached to a round flask with sideway
gas connections. In order to quantify the conversion into NP, an
aliquot of the mother solution (ca. 10-4 M, at the desired pH) was
mixed with the NO solution and then with a mercaptosuccinic acid
solution, which led to the formation of the well characterized, stable
red adduct between NP and mercaptosuccinic acid.19 The mixture
was transferred to the cell under a flow of nitrogen, and the
(16) Kenney, D. J.; Flynn, T. P.; Gallini, J. B. J. Inorg. Nucl. Chem. 1961,
20, 75.
(17) (a) Toma, E. H.; Malin, J. M. Inorg. Chem. 1973, 12, 1039. (b) Toma,
H. E.; Malin, J. M. Inorg. Chem. 1973, 12, 2080.
(18) (a) Toma, H. E.; Malin, J. M. Inorg. Chem. 1974, 13, 1772-1774.
(b) Toma, H. E. Inorg. Chim. Acta 1975, 15, 205-211.
(19) Szacilowski, K.; Stochel, G.; Stasicka, Z.; Kisch, H. New J. Chem.
1997, 21, 893-902.
(20) Cheney, R. P.; Simic, M. G.; Hoffman, M. Z.; Taub, I. A.; Asmus, K.
D. Inorg. Chem. 1977, 16, 2187-2192.
(21) (a) van Eldik, R.; Palmer, D. A.; Schmidt, R.; Kelm, H. Inorg. Chim.
Acta 1981, 50, 131. (b) van Eldik, R.; Gaede, W.; Wieland, S.; Kraft,
J.; Spitzer, M.; Palmer, D. A. ReV. Sci. Instrum. 1993, 64, 1355.
Inorganic Chemistry, Vol. 42, No. 13, 2003 4181

Roncaroli et al.
known to undergo a rapid aquation reaction,18a according to
reaction 1:
This time increased with increasing pressure and varied between
20 and 60 min. Fortunately, the [Fe(CN)5H2O]3- solution appeared
to be more stable at higher pressures, where equilibration times
were longer than under ambient conditions. When the kinetic traces
became reproducible, they were averaged to obtain the kobs values.
A plot of ln kobs versus P enabled the determination of the activation
volume. A standard least-squares method (statistical function of
Microsoft Excel) was employed to estimate the error limits of rate
constants and activation parameters throughout this study.
[Fe(CN)5NH3]3- + H2O h
[Fe(CN)5H2O]3- + NH3 k1, k-1 (1)
Data reported for reaction 1 are the following: k1 ) 1.75
× 10-2 s-1, k-1 ) 452 M-1 s-1, and K1 ) 3.9 × 10-5 M at
25 °C.18a,24,25 We, therefore, expect a rapid (ca. 40 s half-
life), almost complete aquation in a few minutes after
dissolution of the ammine complex, provided the work is
carried out in the pH range 4-8, where NH3 is trapped as
For studies of the reverse dissociation reaction, ca. 3 × 10-4
M
solutions of [Fe(CN)5NO]3- were prepared in two ways: either
through reduction of NP with sodium dithionite, or through the
reaction of a 1.5 × 10-3 M buffered solution of [Fe(CN)5H2O]3-
(pH 5, 0.01 M), with an equal volume of a 1.8 × 10-3 M solution
of NO (pH 10.0, 0.1 M). In the first case, an aqueous solution of
NP was mixed with a solution of ca. 0.5 equiv of sodium dithionite
in 0.1 M buffer at pH 10.0. The final complex concentration of
[Fe(CN)5NO]3- was calculated using the extinction coefficient
reported in the literature (see a preceding description).20 Pyrazine
or sodium cyanide, as well as sodium perchlorate, was added to
the final solutions where required. Spectral changes were recorded
on a Varian Cary 5G spectrophotometer in the range 270-800 nm
and were analyzed with the Specfit program.22 Preliminary experi-
ments were performed at 25.0 and 45.1 °C.
+
NH4 and cannot react as a nucleophile (note that the pKa
+
for NH4 is 9.24).26 The reactant and product complexes
absorb at 396 nm (ꢀ ) 450 M-1 cm-1)24 and 440 nm (ꢀ )
640 M-1 cm-1),18b respectively, allowing for an easy
discrimination between these species. Even up to pH 10,
reaction 1 is expected to be driven almost completely to the
right for sufficiently diluted solutions of the complex. In this
way, at the (1-5) × 10-5 M concentration level, the cyano-
bridged dimerization reaction of the aqua complex can be
prevented.27 No complications related to the protonation of
coordinated cyanide (the pKa values for bound HNC ligands
are between 0 and 3)17,28 or to the deprotonation of
[Fe(CN)5H2O]3- 27b are expected to occur in the selected pH
range 4-10. Finally, the [Fe(CN)5H2O]3- ion is known to
decompose thermally by release of cyanide (k ) ca. 10-4
s-1 at 25 °C);29 this unavoidable process can be controlled
by using freshly prepared solutions in all experiments.
The reaction under study can be described by eq 2:
The dependence of the dissociation rate constant, kd, on tem-
perature and pressure was studied at pH 10.2 (0.05 M buffer), I )
0.1 M (NaClO4), [CN-] ) 3 × 10-2 M, and a complex concentra-
tion of ca. 3 × 10-4 M (obtained from the reaction between
[Fe(CN)5H2O]3- and NO as already described). Some experiments
were carried out by varying the complex concentration between 1
× 10-4 and 3 × 10-3 M. The selected temperature range was 25.0-
56.0 °C. The kinetic traces were analyzed as described, through
the decrease in the 345 and 440 nm bands of [Fe(CN)5NO]3-
.
Duplicate experiments were done at 25.0, 34.9, 50.5, and 45.1 °C.
For the pressure dependence study, a 1.5 cm optical path length
quartz pillbox cell was used. It was first filled with argon and then
with the final [Fe(CN)5NO]3- solution. A high pressure optical cell
unit described elsewhere23 was used, and the spectral changes that
accompany the back reaction were recorded at 347 nm on a
Shimadzu UV-2101 PC spectrophotometer. Kinetic traces were
analyzed by the Olis Kinfit program. A single-exponential model
was used, and kinetic traces could be fitted for approximately 5
half-lives of the reaction. The remaining conditions were the same
as for the temperature dependence study. Some experiments were
performed using 0.05 M CAPS buffer instead of 0.05 M borate
buffer, without the addition of NaClO4. The temperature was 45.1
( 0.1 °C. Two experiments under each condition were done in the
range 5-150 MPa.
[FeII(CN)5H2O]3- + NO h
[Fe(CN)5NO]3- + H2O k2, k-2 (2)
The product of reaction 2 has been previously obtained
upon reduction of NP with different chemical reducing agents
(viz. dithionite, tetrahydroborate, and sodium in liquid
ammonia),30-32 as well as electrochemically33 and by pulse
radiolysis techniques.20 It has been isolated as a sodium salt32
(no crystal structure is available) and has been characterized
by different spectroscopic techniques (UV-vis,20 IR,32 and
EPR30). Recent DFT calculations on [Fe(CN)5NO]3-, together
with the Ru and Os analogues, have confirmed the original
EPR assignment30 and have supported the electronic structure
The reaction between [Fe(CN)4NO]2- and NO was studied by
mixing buffered solutions of approximately 1.8 × 10-4
M
(24) Toma, H. E.; Batista, A. A.; Gray, H. B. J. Am. Chem. Soc. 1982,
104, 7509-7515.
[Fe(CN)4NO]2- with a saturated solution of NO at pH 5.0 or 7.0
(25) Maciejowska, I.; van Eldik, R.; Stochel, G.; Stasicka, Z. Inorg. Chem.
1997, 36, 5409-5412.
and I ) 0.1 M (NaClO4). Kinetic traces were recorded at 615 nm.
(26) Fasman, G. D. Handbook of Biochemistry and Molecular Biology,
Physical and Chemical Data; CRC Press: Cleveland, OH, 1976; Vol.
1.
Results and Discussion
(27) (a) Davies, G.; Garafalo, A. R. Inorg. Chem. 1976, 15, 1101-1106.
(b) Davies, G.; Garafalo, A. R. Inorg. Chem. 1980, 19, 3543-3544.
(28) Malin, J. M.; Koch, R. C. Inorg. Chem. 1978, 17, 752-754.
(29) Zerga, H. O.; Olabe, J. A. Inorg. Chem. 1983, 22, 4156-4158.
(30) van Voorst, J. D. W.; Hemmerich, P. J. Chem Phys. 1966, 45, 3914-
3918.
(31) Bertolino, J. R.; Della Ve´dova, C. O.; Sala, O. Polyhedron, 1989, 8,
361-365.
(32) Nast, R.; Schmidt, J. Angew. Chem., Int. Ed. Engl. 1969, 8, 383.
(33) Masek, J.; Maslova, E. Collect. Czech. Chem. Commun. 1974, 39,
2141-2160.
Since Na3[Fe(CN)5NH3]‚3H2O is used as starting material,
it is important to describe the appropriate conditions for the
generation and handling of the reactants. This complex is
(22) (a) Binstead, R. A.; Zuberbu¨hler, A. D. Specfit; Spectrum Software
Associates: Chapel Hill, NC, 1993-1999. (b) Zuberbu¨hler, A. D. Anal.
Chem. 1990, 62, 2220.
(23) Spitzer, M.; Ga¨rtig, F.; van Eldik, R. ReV. Sci. Instrum. 1988, 59,
2092.
4182 Inorganic Chemistry, Vol. 42, No. 13, 2003

Interaction of NO with Pentacyanoferrate(II)
as corresponding to a {FeNO}7 system, with an angular
FeNO bond and a spin distribution at the singly occupied
molecular orbital (SOMO) mainly localized at the NO ligand
(ca. 62%), with around 29% at iron.34 On this basis, the
electronic structure of [Fe(CN)5NO]3- can be described as
a low-spin {FeIINO} species, also supported by the NO
stretching frequency at 1608 cm-1, which is quite down-
shifted compared to the one in NP, viz. 1940 cm-1, formally
described as {FeIINO+}.35 It is in agreement with a reduction
mainly localized at the NO ligand, which decreases signifi-
cantly the N-O bond order.
The [Fe(CN)5NO]3- ion decomposes by releasing cyanide
according to reaction 3:
[Fe(CN)5NO]3- h [Fe(CN)4NO]2- + CN- k3, k-3 (3)
Equilibrium and kinetic parameters have been determined
and are k3 ) 2.7 × 102 s-1, k-3 ) 4 × 106 M-1 s-1, and K3
) 6.8 × 10-5 M.20 These values indicate that variable
amounts of the penta- and the tetra-cyano species will be
present in aqueous solution depending on the pH. A dominant
contribution of [Fe(CN)5NO]3- appears in the mixtures
prepared at pH values greater than 8, as demonstrated by
the UV-vis spectra. However, some [Fe(CN)4NO]2- ion is
always present, and its relative amount increases on decreas-
ing the pH since the proton-assisted elimination of cyanide
displaces equilibrium to the right. Reaction 3 also predicts
that, in the presence of added cyanide, the solutions obtained
by reduction of NP will contain predominantly the penta-
cyano complex (see a following paragraph for studies on
the dissociation of NO).
The Complex-Formation Reaction (Forward Step in
Reaction 2). Figure 1 describes preliminary experiments in
which spectra were obtained upon successive addition of 1
mL aliquots of an NO-saturated solution to an aqueous
solution of Na3[Fe(CN)5NH3]‚3H2O at pH 10. Approximately
5 min elapsed between the addition of subsequent aliquots,
and the complete experiment was performed within 30 min.
The intensity of the 440 nm band of [Fe(CN)5H2O]3- (formed
via fast aquation in reaction 1)18a decreases upon addition
of NO, while new bands at 347 and 620 nm appear. Although
0.95 equiv of NO was consumed, the band at 347 nm reached
only 20% of the intensity expected for a 100% conversion
of [Fe(CN)5H2O]3- into [Fe(CN)5NO]3- (reaction 2). Upon
further addition of NO, all of the already-mentioned bands
disappeared, and only a shoulder at ca. 330-350 nm
remained. NO (1.9 equiv) was consumed in the total process.
An IR spectrum of the final solution treated with an excess
of NO showed bands corresponding to the formation of NP
at 1936 and 2142 cm-1.35 The yield of NP was ca. 100% in
reference to the initial aqua complex concentration. The
product solution showed no reaction with pyrazine,17 con-
sistent with the extremely inert nature of the NO+ ligand
toward substitution in NP. The inset of Figure 1 shows kinetic
Figure 1. Spectral changes corresponding to successive addition of 1 mL
aliquots of NO-saturated solution (1.8 × 10-3 M) to 40 mL of a 3 × 10-4
M [Fe(CN)5NH3]3- solution (pH ) 10.0; 0.01 M borate buffer). The
continuous decrease of the 440 nm band parallels the initial increase and
subsequent decrease of the bands at 347 and 620 nm. The final pH was
unchanged. Inset: Kinetic traces at 440 and 350 nm upon mixing equal
volumes of an NO-saturated solution, (pH 10.0; 40 mM borate buffer) and
a 5 × 10-5 M [Fe(CN)5H2O]3- solution (pH 5; 0.02 M acetate buffer).
traces recorded at 440 and 350 nm for a related experiment
performed under excess of NO. It can be seen that a fast
process occurs in the first few seconds followed by a slow
decay of the bands. The kinetic traces are consistent with
reaction 2 and show that the rapid spectral change corre-
sponds to the formation of the nitrosyl complex. The
subsequent decay at 350 nm observed over longer times is
consistent with the low yields of the nitrosyl complex already
described and indicates that the [Fe(CN)5NO]3- complex
produced initially is not stable under the selected reaction
conditions.36
At pH 7, Figure 2 (obtained under similar conditions as
Figure 1) shows a similar but not identical picture. The band
at 440 nm also decays continuously upon addition of NO,
but no formation of an intermediate at ca. 350 nm is observed
as was observed in Figure 1. The final product absorbs at
ca. 320 nm accompanied by an increase in the baseline
around 550-800 nm with the clear appearance of two
isosbestic points. These spectral features were also observed
for similar experiments at pH 4-8. In the overall process,
(36) We found that both [Fe(CN)5NO]3- and [Fe(CN)4NO]2-, generated
as predominant species in solution at pH 10 and 6-7, respectively,
reacted with excess NO, showing a rapid decrease in their characteristic
absorption bands. This is supported by the fast cyanide release from
[Fe(CN)5NO]3- (see text) and also by the fast reaction of [Fe(CN)4NO]2-
with NO (kobs, 58 s-1, our measurements, see Experimental Section),
consistent with the reported fast decomposition of [Fe(CN)5NO]3- at
pH 7. However, when cyanide was added to the solution of
[Fe(CN)5NO]3- at pH 10, no decomposition was observed. Given
equilibrium 3, this indicates that NO must react with [Fe(CN)4NO]2-
,
but not with [Fe(CN)5NO]3-. The mode of decomposition is presently
under study. We report in the present study some preliminary results
that show the final formation of NP, with an increase in conversion
at increasing pH, together with the apparent formation of [Fe(CN)6]4-
under some conditions. If a disproportionation of bound NO is
operative (cf. ref 32), we are still unable to define the product of NO
reduction.
(34) Wanner, M.; Scheiring, T.; Kaim, W.; Slep, L. D.; Baraldo, L. M.;
Olabe, J. A.; Zalis, S.; Baerends, E. J. Inorg. Chem. 2001, 40, 5704-
5707.
(35) Paliani, G.; Poletti, A.; Santucci, A. J. Mol. Struct. 1971, 8, 63-74.
Inorganic Chemistry, Vol. 42, No. 13, 2003 4183

Roncaroli et al.
Table 1. Preliminary Results for Complex-Formation Reaction 2a
[Fe(CN)5H2O3-
(mM)
]
[NO2 ]
kobs
(s-1
-
b
pH
(mM)
)
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.01
0.02
0.06
0.10
4.0
5.0
6.0
7.0
8.0
9.0
10.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
0.22
0.24
0.22
0.23 (0.23)
0.24 (0.22)
0.24 (0.21)
0.22 (0.24)
0.22
0.23
0.35
0.40
0.41
2.0
5.9
9.4
15.6
19.2
30.1
0.54
0.26
0.22
0.24
0.35
a T ) 25.4 °C; [NO] ) 0.9 mM. b Values in parentheses were calculated
from the absorbance increase at 350 nm, the remaining from the absorbance
decrease at 440 nm.
Figure 2. Spectral changes corresponding to successive additions of 1
mL aliquots of NO-saturated solution, 1.8 × 10-3 M, to 40 mL of a 2.8 ×
10-4 M [Fe(CN)5H2O]3- solution (pH ) 7.0, 0.01 M Tris buffer). The final
pH was unchanged. Inset: Kinetic traces at 440 and 350 nm upon mixing
equal volumes of NO-saturated and 5 × 10-5 M [Fe(CN)5H2O]3- solutions
(same buffer system).
The possible influence of unavoidable nitrite impurities
always present in aqueous solutions of NO (which may reach
a level of around 10-4 M) inspired a series of experiments
on the dependence of kobs on the concentration of nitrite in
the test solutions (see Table 1). Increasing amounts of nitrite,
up to 30-fold as compared to [NO], were used, and it was
found that no significant effect was observed up to a
concentration of 6 × 10-3 M, with some increase in the
observed rate constant for higher concentrations. The latter
results are understandable, since both nitrite27 and NO
effectively bind to [Fe(CN)5H2O]3- in a competitive way,
and then the measured rate constant will be the sum of both
contributions. Since the second-order formation rate constant
for the coordination of nitrite is approximately an order of
magnitude smaller than for NO,27 the results for kobs in Table
1 can be understood. The error introduced by nitrite is at
the 1% level for NO solutions that contain the usual nitrite
impurity. This error is by far overcome by the normal
experimental error limits of ca. 10% in these measurements.
The percentage conversion of [Fe(CN)5H2O]3- into NP
in the presence of an excess of NO was studied as a function
of pH: % (pH) ) 35-45 (3-5.5); 60 (7); 75 (8); 95-100
(8.5-10). This suggests that, at the more acidic pH, the free
cyanide generated through reaction 3 becomes protonated
and is unable to recombine with the rapidly formed tetra-
cyano species in order to form NP.36
Figure 3 shows the dependence of kobs (s-1) on the
concentration of NO, from which the second-order rate
constant, k2 ) 250 ( 10 M-1 s-1 at 25.4 °C, I ) 0.1 M, and
pH 7.0, can be calculated from the slope of the line. As
shown in Table 3, this value is in close agreement with earlier
data for the rate constants obtained for the coordination of
neutral ligands to [Fe(CN)5H2O]3-, which cover a range from
200 to 450 M-1 s-1. In contrast, positively and negatively
charged ligands usually show larger or smaller rate constants,
respectively.15 The results in Figure 3, furthermore, indicate
that the rate constant for the reverse reaction, k-2, represented
0.95 equiv of NO were consumed. The IR spectrum of the
reaction product in the presence of an excess of NO also
showed the characteristic bands of NP, together with an
additional band at 2036 cm-1 that can be assigned to the
formation of [Fe(CN)6],4-37 consistent with the product
shoulder at 320 nm.38 The corresponding inset in Figure 2
shows a single kinetic process. The traces at 440 and 350
nm fit a single exponential function for at least three half-
lives, usually five half-lives, of the reaction.39 The observed
rate constants, kobs’s, obtained from kinetic traces at both
wavelengths were essentially the same. Note that the half-
lives for the absorbance decrease at 440 nm are practically
the same for the experiments at pH 10 and 7 (insets in Figures
1 and 2, respectively), although no intermediates are detected
in the latter case. This confirms that the decrease at 440 nm
effectively corresponds to the forward reaction in eq 2, with
the product decomposing faster at lower pH. Table 1
summarizes the data obtained for the preliminary experi-
ments. No pH dependence was observed for kobs in the range
4-7, with an average value of 0.23 s-1 at 25 °C. Even the
results in the pH range 8-10, using a different mixing
procedure in order to avoid the perturbations caused by
aquation of [Fe(CN)5NH3]3- (see Experimental Section),
revealed similar values for the rate constants.
(37) Jones, L. H. Inorg. Chem. 1963, 2, 777-780.
(38) Alexander, J. J.; Gray, H. B. J. Am. Chem. Soc. 1968, 90, 4260-
4271.
(39) The trace at 350 nm appears noisy in the stabilized region, because
the measurement was not performed at the appropriate wavelength,
namely 320 nm, where the final product absorbs. At that time, we
had in mind that an intermediate absorbing at 345 nm could be present.
This is evidently not the case, as shown in Figure 2 in the main series
of recorded spectra. The overall kinetic trace in the inset, and most
importantly the rising part, is significant and shows the same kinetic
performance as observed for the one at 440 nm. This is evidence for
the initial formation and very fast decay of [Fe(CN)5NO]3- at pH 7
(see text and previous comment in ref 36).
4184 Inorganic Chemistry, Vol. 42, No. 13, 2003

Interaction of NO with Pentacyanoferrate(II)
Table 3. Complex-Formation Rate Constants and Activation
Parameters for Different [Fe(CN)5L]n- Complexesa
kf
∆Hfq
∆Sfq
∆Vfq
L
(M-1 s-1
)
(kJ mol-1
)
(J K-1 mol-1
)
(cm3 mol-1
)
ref
N-Mepz+ b,c
550b
2350c
310
250
240
380
315
365
365
452
242
30
70.3
+42
17
15
40
this work
41
17
42,43
17
COc
NOc
dmsob
pzd
63
70
+15
+34
+16.7
+20.9
+21
+29.3
+10
+67
+12.5
+42
+17.4
64.4
64.4
64.4
67.4
62
78
63.5
76.9
hisc
+17.0
+14.4
pyd
NH3
c
24,25
Imc
+15.5
+13.5
42,43
44
CN-b
a For the following reaction: [Fe(CN)5H2O]3- + L f [Fe(CN)5L]n-
+
H2O, T ) 25.0 °C, except for L ) NO, T ) 25.4 °C. b I ) 1 M. c I ) 0.1
M. d I ) 0.5 M.
Scheme 1
Figure 3. Dependence of kobs on the concentration of NO for the complex-
formation reaction: T ) 25.4 ( 0.1 °C, pH ) 7.0 (0.01 M tris), I ) 0.1 M
(NaClO4), [Fe(CN)5H2O3-] ) (1-5) × 10-5 M, [NO] ) (0.23-0.90) ×
10-3 M. For a given [NO], distinctive values of kobs represent the spread in
the measurements, at a constant complex concentration. From the slope, k2
) 250 ( 10 M-1 s-1
.
Table 2. Kinetic Results for Complex-Formation Reaction 2 at pH )
7.0 and I ) 0.1 Ma
for NO (see Table 3). The consistent rate constants found
for the complex-formation reactions with different ligands
of similar charge, as well as the almost constant, positive
values for the activation entropy and activation volume,
suggest the operation of a dissociative mechanism.45 Two
possible pathways can be considered, as outlined in Scheme
1 for the reaction with NO.
The first pathway is presented by the route described by
reactions 4 and 6, which involves a dissociative interchange
(Id) mechanism.45 Reaction 4 implies the rapid formation of
a precursor complex, followed by the slow ligand interchange
step (6) between the first and second coordination spheres
of the precursor complex. In this case, the derived rate law
is given by kobs ) k6Kpc[NO]/(1 + Kpc[NO]) + k-6, which
simplifies to kobs ) k6Kpc[NO] under our reaction conditions
(see Figure 3).15 An alternative pathway is given by reactions
5 and 7 and represents a limiting dissociative (D) mecha-
nism,45 in which the slow release of water in reaction 5 to
form a pentacoordinate intermediate is followed by the rapid
coordination of NO. In this case, the rate law has a similar
form: kobs ) (k5k7[NO] + k-7)/(k7[NO] + k-5[H2O]), which
simplifies to kobs ) k5k7[NO]/k-5[H2O] under our conditions.
A variation on the Id mechanism has been proposed, viz.
the Dpc mechanism, in which the rate-limiting loss of the
[Fe(CN)5H2O3-
(mM)
]
[NO]
(mM)
temp
(°C)
pressure
(MPa)
kobs
(s-1
)
0.05
0.04
0.02
0.01
1.1
0.05
0.05
0.05
0.05
0.05
0.05
0.06
0.06
0.06
0.06
0.90
0.68
0.45
0.23
0.1
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
25.4
25.4
25.4
25.4
25.4
29.9
35.3
40.1
44.6
20.9
15.0
25.6
25.6
25.6
25.6
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
10
0.23 ( 0.01
0.16 ( 0.01
0.11 ( 0.01
0.054 ( 0.005
0.29 ( 0.02
0.30 ( 0.01
0.54 ( 0.02
0.82 ( 0.03
1.29 ( 0.01
0.13 ( 0.01
0.077 ( 0.008
0.237 ( 0.004
0.182 ( 0.003
0.135 ( 0.003
0.103 ( 0.001
50
90
130
a k2 (25.4 °C) ) 250 ( 10 M-1 s-1, ∆H2 ) 70 ( 1 kJ mol-1, ∆S2
)
q
q
+34 ( 4 J K-1 mol-1, ∆V2 ) +17.4 ( 0.3 cm3 mol-1
.
q
by the intercept is indeed very small and cannot be obtained
in this way.
Table 2 summarizes the kinetic data obtained for the
complex-formation reaction at different temperatures and
pressures, from which the activation parameters were cal-
culated (see Figures S1 and S2, Supporting Information).
Table 3 shows that the value of the activation enthalpy, 70
( 1 kJ mol-1, is very similar to values obtained for a wide
range of neutral ligands. In general, there is little variation
in this parameter with values around 60-80 kJ mol-1. The
positive activation entropy, viz. +34 ( 4 J K-1 mol-1, is
also consistent with values reported for other entering ligands,
although in this case the variation is larger. Finally, the
activation volume for the reaction with NO was found to be
+17.4 ( 0.3 cm3 mol-1. Some examples from earlier reports
for the coordination of other neutral and negatively charged
ligands to [Fe(CN)5H2O]3- are consistent with that found
(40) Toma, H. E.; Moroi, N. M.; Iha, N. Y. M. An. Acad. Bras. Cienc.
1982, 54, 315-323.
(41) Toma, H. E.; Malin, J. M.; Giesbrecht, E. Inorg. Chem. 1973, 12,
2084.
(42) Toma, H. E.; Martins, J. M.; Giesbrecht, E. J. Chem. Soc., Dalton
Trans. 1981, 1610-1617.
(43) Stochel, G.; Chatlas, J.; Mart´ınez, P.; van Eldik, R. Inorg. Chem. 1992,
31, 5480-5483.
(44) Finston, M. I.; Drickamer, H. G. J. Phys. Chem. 1981, 104, 7509-
7515.
(45) Langford, C. H.; Gray, H. B. Ligand Substitution Processes; W. A.
Benjamin, Inc.: Reading, MA, 1965.
Inorganic Chemistry, Vol. 42, No. 13, 2003 4185

Roncaroli et al.
aqua ligand occurs within the precursor complex.15 It can
be seen that the proposed mechanisms agree with the second-
order behavior found in Figure 3. The slight but significant
dependence of k2 on the charge of the entering ligand (see
Table 3) is accounted for in the Id and Dpc mechanisms by
the different values of Kpc, which must increase with the
positive charge on the entering ligand. On the other hand,
the D mechanism can account for this fact through the
electrostatic effects that influence the value of k7.43 Other
arguments have been presented in favor of the limiting D
pathway; one of them is based on the principle of micro-
scopic reversibility, since strong evidence exists for the
operation of a limiting D mechanism for the reverse
dissociation process. Furthermore, particular emphasis has
been placed on the fact that the volumes of activation cover
a range from +14 to +18 cm3 mol-1 for a wide range of
neutral and negatively charged entering ligands (Table 3).43
The range of values is rather consistent with the limiting
value calculated for the dissociation of a water molecule from
an octahedral metal center, viz. +13.1 cm3 mol-1.46 The
presently reported value for complex-formation with NO
agrees with this proposal. For a series of high-spin FeII(L)
complexes, with L ) aqua and diverse polycarboxylate
ligands, the reversible binding of NO afforded small values
for the activation volumes, supporting an Id mechanism. A
comparison within the {FeNO}7 systems must be done with
caution, because in the already-mentioned complexes the
electronic structure is quite different from that in [Fe(CN)5-
NO]3-.4,9,11 Similarly, high formation rates and small activa-
tion volumes have been found for the formation of NO
complexes with some ferroporphyrins.4
Figure 4. Spectral changes recorded during the dissociation of NO from
the [Fe(CN)5NO]3- complex in the presence of free cyanide: [Fe(CN)5NO3-
]
) 3 × 10-4 M, pH 10.2 (0.05 M borate buffer), I ) 0.1M (NaClO4), [CN-]
) 0.03 M, T ) 50.4 °C, cycle time 312 s. Inset: Kinetic traces at 440 and
347 nm, kobs ) 5.14 × 10-4 s-1
.
behavior. Only a shoulder at ca. 320 nm remained, which
can be assigned to the presence of [Fe(CN)6]4- (see preceding
description).38 No bands in the range 500-800 nm were
detected.
Figure 4 shows that the NO ligand is substituted by
cyanide in the overall reaction 8.
[Fe(CN)5NO]3- + CN- h [Fe(CN)6]4- + NO
(8)
The Ligand Dissociation Reaction (Reverse Step in
Reaction 2). The [Fe(CN)5NO]3- complex can be generated
in solution via the reaction of [Fe(CN)5H2O]3- with NO at
pH 10, or upon reduction of NP with dithionite. The EPR
spectra recorded for frozen solutions of [Fe(CN)5NO]3- were
found to be the same as previously reported by van Voorst
and Hemmerich (see Figure S3, Supporting Information).30
This result, together with the characteristic UV-vis spectral
signatures, confirms the identity of the product in reaction
2. Spectral changes that correspond to the decay of
[Fe(CN)5NO]3- at pH 10 in the absence of any other potential
ligand could not be fitted to a single-exponential function.
This was the case for [Fe(CN)5NO]3- generated by any of
the procedures mentioned already. Upon addition of pyrazine
to the system, a decrease of the band at 347 nm and an
increase in the absorbance at ca. 450 nm were observed,
which were assigned to the formation of the [Fe(CN)5pz]3-
complex (λmax ) 452 nm, ꢀ ) 5000 M-1 cm-1).17 However,
first-order kinetics were still not observed. In addition, there
was also a small but appreciable absorbance at 615 nm due
to the presence of [Fe(CN)4NO]2-, which also decayed in
the same process. Finally, Figure 4 shows the spectral
changes observed after addition of sodium cyanide to the
[Fe(CN)5NO]3- complex solution. Now the kinetic traces at
347 and 440 nm (see inset in Figure 4) obeyed a first-order
As a result of cyanide addition, reaction 8 now appears
as a clean substitution process, free of the presence of the
initial [Fe(CN)4NO]2- complex (compare reaction 3). NO
was detected qualitatively as a product in the thermal
decomposition of a 3 mM [Fe(CN)5NO]3- solution using
electrochemical and mass spectrometric detection techniques.
The data for the dissociation reaction at different temper-
atures and pressures are summarized in Table 4.47 The
average observed rate constant (k-2) was found to be (1.58
( 0.06) × 10-5 s-1 at 25.0 °C, I ) 0.1 M, and pH 10.2.
Table 5 summarizes dissociation rate constants and activation
parameters for the release of different ligands L from
pentacyanoferrate(II) fragments.
The activation parameters, ∆H-2q, ∆S-2q, and ∆V-2q, were
found to be 106.4 ( 0.8 kJ mol-1, +20 ( 2 J K-1 mol-1,
and +7.1 ( 0.2 cm3 mol-1, respectively. The temperature
and pressure dependence plots are shown in Figures S4 and
S5 (Supporting Information), respectively. As seen from the
summary of data in Table 5, the activation enthalpies and
entropies are consistent with those reported for the release
of other ligands from the pentacyanoferrate(II) fragment.48-53
(47) Although not included in Table 4, some experiments were performed
by changing the initial complex concentration, by going up 10-fold
and down 3-fold with respect to the data in Table 4. This was done in
search for any possible second-order pathway allowing for alternative
decomposition modes of [Fe(CN)5NO]3-, viz., a bimolecular dismu-
tation process. No significant difference in the kobs values was found.
(46) Swaddle, T. W. Inorg. Chem. 1983, 22, 2663-2665.
4186 Inorganic Chemistry, Vol. 42, No. 13, 2003

Interaction of NO with Pentacyanoferrate(II)
Table 4. Kinetic Results for the Dissociation of NO from
Under steady-state conditions applied to the dissociative
mechanism, it follows that kobs, the first-order rate constant
for the decay of the nitrosyl complex, will increase with
increasing concentration of the scavenger ligand, cyanide,
up to the point where a saturation rate constant is reached,15
i.e., where k-7 in Scheme 1, the rate constant for the release
of NO, becomes rate-limiting. Because of the complications
introduced by reaction 3, a detailed study over a wide range
of cyanide concentrations could not be performed. However,
within the cyanide concentration range 3-60 mM, a constant
value of kobs was obtained, indicating that saturation condi-
tions were met. It is useful to compare the value of k-2 with
those for the release of other ligands by the pentacyanofer-
rate(II) fragment (see Table 5). It can be seen that NO is
released significantly more slowly than several other ligands
affording different σ-π bonding interactions with iron. The
rate is comparable with S-bonded DMSO, but it is much
faster than the values obtained for the release of cyanide or
CO. This suggests that the NO ligand binds to the pentacy-
anoferrate(II) fragment through the onset of a moderate σ-π
interaction. The existence of some (minor) degree of π
interaction is supported by the intense band at 345 nm, also
reported for many [FeII(CN)5L]3- complexes with L )
N-heterocyclic ligands, and assigned to metal-to-ligand
charge transfer (MLCT) transitions.6,17 Although the value
of the rate constant is of the same order as that measured
for the dissociation of NO in some five-coordinate ferro-
heme complexes, significant variations in the range from 10-5
to 10 s-1 have been measured by changing the heme
structure.54 The release of NO from different FeII-NO
chelate complexes also depends on the spectator ligands, with
values in the range from 10 to 103 s-1.11 A range of values
from 5 to ca. 10-4 s-1 have also been obtained for the
dissociation of NO from a series or ruthenium-nitrosyl
complexes.7,55 Much has still to be learned about the factors
controlling the rate of NO dissociation in the different metal
systems.56
[Fe(CN)5NO]3- a
temp
(°C)
pressure
(MPa)
kobs × 104
(s-1
)
25.0
30.3
34.9
38.3
45.1
50.5
56.0
45.1
45.1
45.1
45.1
45.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
5
0.158 ( 0.006
0.34 ( 0.01
0.68 ( 0.01
1.10 ( 0.05
2.6 ( 0.3
5.3 ( 0.2
9.8 ( 0.3
2.55 ( 0.01
2.24 ( 0.01
1.99 ( 0.01
1.84 ( 0.03
1.70 ( 0.01b
50
100
150
150
a pH 10.2; I ) 0.1 M; [Fe(CN)5NO3-] ) ca. 3 × 10-4 M; [CN-] )
0.030 M. k-2 (25.0 °C) ) (1.58 ( 0.06) × 10-5 s-1, ∆H-2q ) 106.4 ( 0.8
kJ mol-1, ∆S-2q ) +20 ( 2 J K-1 mol-1, ∆V-2q ) +7.1 ( 0.2 cm3 mol-1
.
b CAPS buffer was used instead of boric acid/borate buffer. This value was
used instead of the previous one in order to prevent errors associated with
the dependence of the pKa value of the boric acid/borate buffer upon
pressure.
Table 5. Dissociation Rate Constants and Activation Parameters for
Different [Fe(CN)5L]n- Complexesa
q
q
q
kd
∆Hd
∆Sd
∆Vd
ligand
(s-1
)
(kJ mol-1
)
(J K-1 mol-1
)
(cm3 mol-1
)
ref
CObc
<10-8
ca. 4 × 10-7
1.58 × 10-5
7.5 × 10-5
2.8 × 10-4
4.2 × 10-4
5.3 × 10-4
7.7 × 10-4
1 × 10-3
40
48
CN- cd
NOc
106.4
110.0
115
110.5
105.4
114
+20
+46.0
+75
+7.1
this work
41
DMSOc
N-Mepz+ e
pze
+0.9
+13.0
17, 49
17, 49
42
+58.6
+46.0
+44
hisc
2-Mepze
4-CNpye
pye
+19.4
+20.6
50
50
17
105
+50
1.1 × 10-3
103.8
108
93
97
102
+46.0
+60
3,5-Me2pye 1.2 × 10-3
+20.5
+20.6
+20.1
+16.4
17,51
51
3-CNpye
2.2 × 10-3
1 × 10-2
+16
- c
NO2
NH3
+42
+68
52
1.75 × 10-2
18a,25
c
a T ) 25.0 °C. b Estimated number, measured using pz as a scavenger.
c I ) 0.1 M. d Extrapolated from data reported at higher temperatures. e I
) 0.5 M.
The reported activation volume is appreciably smaller for
NO than for the release of other ligands, which span a range
+13 to +24 cm3 mol-1, but is still indicative of a dissociative
mechanism. A more detailed discussion of the observed
volume effects in terms of the overall volume profile for
reaction 2 is presented in subsequent paragraphs.
We now turn to a discussion of the volume profile
presented for reaction 2 in Figure 5, on the basis of the
activation volumes determined for k2 and k-2 in Tables 2
and 4, respectively. The partial molar volume of the transition
state is clearly much larger than that of either the reactant
or product states, illustrating the dissociative nature of the
ligand substitution process. Several volume profiles for the
reversible binding of NO to different iron complexes have
been reported recently,9,11c,57,58 and a number of them support
These results all suggest that the release of NO can be
described by a dissociative mechanism. Scheme 1 can also
be used for this analysis by starting with reaction 7 (in a
reversed manner), followed by reaction 5 and the final rapid
coordination of cyanide in reaction 9.
(54) Bohle, D. S.; Hung, C. H. J. Am. Chem. Soc. 1995, 117, 9584.
(55) (a) Lang, D. R.; Davis, J. A.; Lopes, L. G. F.; Ferro, A. A.;
Vasconcellos, L. C. G.; Franco, D. W.; Tfouni, E.; Wieraszko, A.;
Clarke, M. J. Inorg. Chem. 2000, 39, 2294-2300. (b) Lopes, L. G.
F.; Wieraszko, A.; El-Sherif, Y.; Clarke, M. J. Inorg. Chim. Acta 2001,
312, 15-22.
[Fe(CN)5H2O]3- + CN- h [Fe(CN)6]4- + H2O (9)
(48) Legros, J. J. Chim. Phys. Phys.-Chim. Biol. 1964, 61, 909.
(49) Alsheri, S.; Burgess, J. Inorg. Chim. Acta 1991, 181, 153.
(50) Sullivan, T. R.; Stranks, D. R.; Burgess, J.; Haines, R. I. J. Chem.
Soc., Dalton Trans. 1977, 1460.
(51) Blandamer, M. J.; Burgess, J.; Morcom, K. W.; Sherry, R. Transition
Met. Chem. 1983, 8, 354.
(52) Stochel, G.; van Eldik, R.; Hejmo, E.; Stasicka, Z. Inorg. Chem. 1988,
27, 2767-2770.
(53) Bal Reddy, K.; van Eldik, R. Inorg. Chem. 1991, 30, 596-598.
(56) The release of a spectator ligand from {MNO}7 complexes (as in eq
3) is a manifestation of the strong trans-effect resulting from the partial
2
population of the dz orbital by the unpaired electron derived from the
nitrosyl ligand (see text for the spin distribution in [Fe(CN)5NO]3-).34
The prior release of a spectator ligand has also been observed during
the studies of NO release from several trans-[Ru(NH3)4(X)(NO)]n+
complexes, with X ) his, ImN, ImC, H2O, P(OEt)3, and N-heterocyclic
ligands.7,55 It has been proposed that the k-2 values depend on the
magnitude of the trans-effect exerted by X.7
Inorganic Chemistry, Vol. 42, No. 13, 2003 4187

Roncaroli et al.
A remarkable aspect of the volume profile reported in
Figure 5 is the fact that the volume of activation for the
formation of the transition state is significantly larger than
that predicted for the dissociation of a water molecule from
an octahedral hexaaqua complex, viz. +13 cm3 mol-1.46
A
very large and similar volume of activation was also reported
for the reaction between [Fe(CN)5H2O]3- and ammonia
studied before.25 In the present case, the volume increase
almost equals the partial molar volume of a bulk solvent
(water) molecule, which means that no major volume
collapse during the reorganization of the five-coordinate
[Fe(CN)5]3- intermediate occurs, as was predicted for the
pentaaqua intermediate.46 This may be related to the highly
solvated nature of the pentacyano fragment in a polar solvent
like water, which could prevent significant ligand reorganiza-
tion on going from a six- to a five-coordinate state. Fur-
thermore, the volume of activation for the release of NO in
the back reaction is rather surprisingly small (compared to
all the other volume of activation data reported in Table 5),
which may indicate that specific solvation effects around the
NO radical on its release from the coordination sphere could
partially compensate the intrinsic volume increase associated
with FeII-NO bond cleavage. Thus, the volume profile for
the rather unique reversible binding of NO to pentacyano-
ferrate(II) that involves only a minor charge transfer from
NO to the Fe(II) center34 reveals some interesting features
not observed in related volume profiles reported before and
calls for further investigations on similar or related systems.
Figure 5. Volume profile for overall reaction 2.
the operation of a dissociative mechanism.9,11c,57,58 A direct
comparison of the reactions of NO with [Fe(H2O)6]2+ and
[Fe(CN)5H2O]3- indicates that both the complex-formation
and release of NO reactions are accompanied by significantly
larger activation volumes for the latter system.9 This
underlines our claim that the reactions involving the penta-
cyano complex are more dissociative in character than those
involving the hexaaqua complex, for which a dissociative
interchange mechanism was suggested.9 The overall reaction
volume for reaction 2 in Figure 5, viz. +10.3 ( 0.5 cm3
mol-1, is also more positive than the value of +4.8 ( 0.6
cm3 mol-1 reported for the hexaaquairon(II) system.9 This
may be partially due to the formal change in oxidation state
when [Fe(H2O)6]2+ reacts with NO to produce [FeIII(H2O)5-
(NO-)]2+ as compared to the formation of [FeII(CN)5NO]3-
in the case of the pentacyano complex.9 Other examples
where the volume profiles clearly supported the operation
of a limiting dissociative mechanism are for the reaction of
NO with metmyoglobin58 and a series of iron(III) porphyrin
complexes.57 In the latter case, it was shown that the rate
and mechanism of the binding of NO are controlled by the
water exchange process on the porphyrin complexes.59
The limiting dissociative mechanism suggested for reaction
2 is also in line with that expected on the basis of the σ-donor
and π-acceptor properties of the cyano ligands that will cause
significant cis and trans labilization effects in terms of the
displacement of coordinated water or NO.25 It is important
that the volume profile reported in Figure 5 is the first one
for a reversible complex-formation reaction with NO that
does not involve a formal oxidation or reduction of the metal
center as is the case for all other volume profiles reported
so far.9,11c,57,58,60
Conclusions
In contrast to the behavior found for the interaction of
[FeIII(CN)5H2O]2- with NO in which the complex was first
reduced by NO prior to coordination, the interaction of NO
with [FeII(CN)5H2O]3- shows that NO behaves as a common
Lewis base. Since the complex-formation rate constant is
controlled by the dissociation of water, no specific informa-
tion on the coordination properties of NO can be obtained
from the kinetic experiments. Furthermore, no specific
influence of the unpaired electron on NO can be disclosed,
as compared to the coordination ability of diamagnetic
ligands. The rate law and activation parameters support the
operation of a limiting dissociative mechanism, as found for
complex-formation reactions with other nucleophiles. The
release of NO is also governed by a dissociative mechanism,
revealing that NO binds to pentacyanoferrate(II) through the
onset of moderate to strong σ-π interactions, as suggested
by the low value of k-2. By comparison with the release of
NO+ as a ligand in NP, NO dissociates much faster (the
dissociation of NO+ is presently undetectable in MIINO+
complexes). Although NO should be a stronger Lewis base
than NO+, it should also be a much weaker π-acceptor. In
view of the reported value for k-2, it is feasible that NO is
not released through the dissociation of NO from [Fe(CN)5-
NO]3- following biological reduction of NP, but rather
through decomposition routes involving the decay of the
tetracyanonitrosyl complex as suggested. This aspect is
presently under further investigation.
(57) Laverman, L. E.; Hoshino, M.; Ford, P. C. J. Am. Chem. Soc. 1997,
119, 12663-12664.
(58) Laverman, L. E.; Wanat, A.; Oszajca, J.; Stochel, G.; Ford, P. C.; van
Eldik, R. J. Am. Chem. Soc. 2001, 123, 285-293.
(59) Schneppensieper, T.; Zahl, A.; van Eldik, R. Angew. Chem., Int. Ed.
2001, 40, 1678-1680.
(60) It should be kept in mind that some charge transfer from NO to Fe-
(II) does occur and spin density calculations indicate that 35% of the
unpaired electron is located on iron, although we can still approach
this description as FeII-NO.34
4188 Inorganic Chemistry, Vol. 42, No. 13, 2003

Interaction of NO with Pentacyanoferrate(II)
Acknowledgment. The authors gratefully acknowledge
financial support from the Deutsche Forschungsgemeinschaft
and the Volkswagen Foundation, as well as from the
University of Buenos Aires and the Argentinian agencies
ANPCyT and Conicet. J.A.O. is a member of the scientific
staff, and F.R. is a fellow of Conicet. We appreciate the
assistance of Manfred Po¨hlein (University of Erlangen-
Nu¨rnberg) with the GC-MS measurements.
Supporting Information Available: Additional figures. This
material is available free of charge via the Internet at http://pubs.
acs.org.
IC0342189
Inorganic Chemistry, Vol. 42, No. 13, 2003 4189