Release of NO from reduced nitroprusside ion. Iron-dinitrosyl formation and NO-disproportionation reactions

Article abstract of DOI:10.1021/ic050070c

The kinetics and mechanism of the thermal decomposition of the one-electron reduction product of [Fe(CN)₅NO]^2- (nitroprusside ion, NP) have been studied by using UV-vis, IR, and EPR spectroscopy and mass-spectrometric and electrochemical techniques in the pH range of 4-10. The reduction product contains an equilibrium mixture of [Fe(CN)₄NO] ^2- and [Fe(CN)₅NO]^3- ions. The first predominates at pH <8 and is formed by the rapid release of trans-cyanide from [Fe(CN)₅NO]^3-, which, in turn, is the main component at pH >9-10. Both nitrosyl complexes decay by first-order processes with rate constants around 10^-5 s^-1 (pH 6-10) related to the dissociation of NO. The decomposition is enhanced at pH 4 by 2 orders of magnitude with protons (and also metal ions) favoring the release of cyanides from the [Fe(CN)₄NO]^2- ions and the ensuing rapid delivery of NO. At pH 7, an EPR-silent intermediate I₁ is detected (VNO, 1695 and 1740 cm^-1) and assigned to the trans-[Fe^II(CN) ₄(NO)₂]^2- ion, an {Fe-(NO)₂} ⁸ species. At pH 6-8, I₁ induces a disproportionation process with formation of N₂O and the regeneration of nitroprusside in a 1:2 molar ratio. At lower pHs, I₁ leads, competitively, to a second paramagnetic (S = 1/2) dinitrosyl intermediate I₂, [Fe(CN)₂(NO)₂]^1-, a new member of a series of four-coordinate {Fe(L)₂(NO)₂} complexes (L = thiolates, imidazole, etc.), described as {Fe(NO)₂}⁹. Other decomposition products are hexacyanoferrate(_II) or free cyanide, depending on the pH, and precipitates of the Prussian-Blue type. This study throws light on the conditions favoring rapid release of NO, to promote vasodilatory effects upon NP injection, and describes new processes related to dinitrosyl formation and NO disproportionation, which are also relevant to the diverse biological processes associated with NO and N₂O processing.

Full text of DOI:10.1021/ic050070c

Inorg. Chem. 2005, 44, 2781
−2790
Release of NO from Reduced Nitroprusside Ion. Iron-Dinitrosyl
Formation and NO-Disproportionation Reactions
Federico Roncaroli,†,‡ Rudi van Eldik,*,‡ and Jose´ A. Olabe*,†
Department of Inorganic, Analytical and Physical Chemistry, INQUIMAE,
Faculty of Exact and Natural Sciences, UniVersity of Buenos Aires,
C1428EHA Buenos Aires, Argentina, and Institute for Inorganic Chemistry,
UniVersity of Erlangen-Nu¨rnberg, Egerlandstrasse 1, 91058 Erlangen, Germany
Received January 17, 2005
-
The kinetics and mechanism of the thermal decomposition of the one-electron reduction product of [Fe(CN)5NO]2
(nitroprusside ion, NP) have been studied by using UV
and electrochemical techniques in the pH range of 4 10. The reduction product contains an equilibrium mixture of
[Fe(CN)4NO]2 and [Fe(CN)5NO]3 ions. The first predominates at pH <8 and is formed by the rapid release of
trans-cyanide from [Fe(CN)5NO]3-, which, in turn, is the main component at pH >9
10. Both nitrosyl complexes
10) related to the dissociation of NO.
−vis, IR, and EPR spectroscopy and mass-spectrometric
−
-
-
−
5
1
decay by first-order processes with rate constants around 10- s- (pH 6
−
The decomposition is enhanced at pH 4 by 2 orders of magnitude with protons (and also metal ions) favoring the
-
release of cyanides from the [Fe(CN)4NO]2 ions and the ensuing rapid delivery of NO. At pH 7, an EPR-silent
-
intermediate I1 is detected (
ν
NO, 1695 and 1740 cm-1) and assigned to the trans-[FeII(CN)4(NO)2]2 ion, an
{
Fe-
8
(NO)2
}
species. At pH 6
−
8, I1 induces a disproportionation process with formation of N2O and the regeneration
1/2)
complexes
9. Other decomposition products are hexacyanoferrate(II)
of nitroprusside in a 1:2 molar ratio. At lower pHs, I1 leads, competitively, to a second paramagnetic (S
dinitrosyl intermediate I2, [Fe(CN)2(NO)2]1-, a new member of a series of four-coordinate
Fe(L)2(NO)2
(L thiolates, imidazole, etc.), described as Fe(NO)2
)
{
}
)
{
}
or free cyanide, depending on the pH, and precipitates of the Prussian-Blue type. This study throws light on the
conditions favoring rapid release of NO, to promote vasodilatory effects upon NP injection, and describes new
processes related to dinitrosyl formation and NO disproportionation, which are also relevant to the diverse biological
processes associated with NO and N2O processing.
Introduction
although a photoredox process readily delivers NO upon
irradiation with UV-vis light.4 On the other hand, NP shows
electrophilic reactivity toward different nucleophiles
(OH-, amines, thiolates, etc.)5,6 and can be reduced chemi-
cally or electrochemically forming the paramagnetic one-
electron reduction product, [FeII(CN)5NO]3-, ill-defined two-
electron reduction products presumably related to the nitroxyl
anion (NO-), or nitrogen hydrides (NH2OH, NH3) through
multielectron processes.6,7
Among the metallonitrosyls, sodium pentacyanonitro-
sylferrate (Na2[Fe(CN)5NO]‚2H2O, nitroprusside, NP) is
widely used in physiological and medical experimentation
as a “gold-standard” NO donor drug. NP has a number of
clinical uses, the most common of which is hypotensive
anaesthesia during surgery.1 This {FeNO}6 species, in the
Enemark-Feltham formalism,2 is best described as an Fe-
(II) center (d6, low-spin) containing a NO+ (nitrosonium)
ligand.3 As such, it is extremely inert for thermal dissociation,
(2) Enemark, J. H.; Feltham, R. D. Coord. Chem. ReV. 1974, 13, 339-406.
(3) (a) Swinehart, J. H. Coord. Chem. ReV. 1967, 2, 385-402.
(b) Westcott, B. L.; Enemark, J. H. In Inorganic Electronic Structure
and Spectroscopy; Solomon, E. I., Lever, A. B. P., Eds.; Wiley-
Interscience: New York, 1999; Vol. 2, Chapter 7, pp 403-450.
(4) Wolfe, S. K.; Swinehart, J. H. Inorg. Chem. 1975, 14, 1049-1053.
(5) Bottomley, F. In Reactions of Coordinated Ligands; Braterman, P.
S., Ed.; Plenum Press: New York, 1989; Vol. 2, pp 115-222.
(6) Olabe, J. A. AdV. Inorg. Chem. 2004, 55, 61-126.
* 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 Buenos Aires.
‡ University of Erlangen-Nu¨rnberg.
(1) (a) Butler, A. R.; Megson, I. L. Chem. ReV. 2002, 102, 1155-1165.
(b) Wang, P. G.; Xian, M.; Tang, X.; Wu, X.; Wen, Z.; Cai, T.;
Janczuk, A. J. Chem. ReV. 2002, 102, 1091-1134. (c) Clarke, M. J.;
Gaul, J. B. Struct. Bonding (Berlin) 1993, 81, 147-181. (d) Butler,
A. R.; Glidewell, C. Chem. Soc. ReV. 1987, 16, 361-380.
(7) Masek, J.; Maslova, E. Collect. Czech. Chem. Comm. 1974, 39, 2141-
2160.
10.1021/ic050070c CCC: $30.25
Published on Web 03/16/2005
© 2005 American Chemical Society
Inorganic Chemistry, Vol. 44, No. 8, 2005 2781

Roncaroli et al.
The vasodilatory action of NP can be observed on the
minute time scale after injection. It can be inferred that free
NO must be generated in solution in order to activate soluble
guanylate cyclase (sGC) to trigger the vasodilation process.1
Activation involves a very fast (ca. 108 M-1 s-1) coordination
of at least one NO ligand at the iron center of the enzyme.8
The mechanism by which the NO+ ligand in NP is reduced
and released into bodily fluids has not been clearly defined
in the literature.1,9 It can reasonably be presumed that
thiolates or other biologically relevant reductants may be
operative at the beginning, but the chemistry of the reduction
product(s) is still unclear, particularly those aspects dealing
with the rates of dissociation of NO from iron.9 It is known
that the NO ligand generally promotes ligand labilization at
the trans position,2,3 and this is the case with reaction 1 for
[Fe(CN)5NO]3-
NO vs cyanides from reduced NP is particularly intriguing
as is the ensuing fate of NO in the solution. The possibility
of further reactivity, other than its eventual coordination to
heme proteins (e.g., sGC), should also be considered.
In the studies of coordination of NO to the [FeII(CN)5H2O]3-
ion at pH 7, we found that the initial product [Fe(CN)5NO]3-
was unstable under an excess of NO.13 A prelim-
inary investigation revealed that [Fe(CN)4NO]2-, not
[Fe(CN)5NO]3-, was the active species promoting decom-
position and that NP was formed as one of the products.
This result suggested that an NO-disproportionation process
could be involved. We, herein, present the results of a
stoichiometric and kinetic study in the pH range of 4-10
centered on the spontaneous decay of [FeII(CN)5NO]3- and/
or [Fe(CN)4NO]2- using UV-vis, IR, EPR, electrochemical,
and mass-spectrometric techniques. The study involves
avoiding light exposure1,4 and maintaining anaerobic condi-
tions. The latter is important because [FeII(CN)5NO]3- can
be rapidly transformed back to NP in aerobic media.14 Our
studies can be placed in the context of the still underdevel-
oped chemistry of bound NO in transition metal complexes;15
we address specifically the nitrosyl transfer problem from a
M-NO donor (viz., NP) to another metal acceptor,16 and
we include the possible intermediacy of dinitrosyl com-
pounds.
k1
[Fe(CN)5NO]3-
{
} [Fe(CN)4NO]2- + CN-
(1)
k-1
The equilibrium in reaction 1 is rapidly established in
solution with k1 ) 2.7 × 102 s-1 and k-1 ) 4 × 106 M-1 s-1
at 25.0 °C.10 The predominance of [Fe(CN)5NO]3- or
[Fe(CN)4NO]2- is strongly pH-dependent, and the relative
concentration of the latter species increases with decreasing
pH. Both anionic species have been obtained as solid salts
and characterized spectroscopically, including structural
results for [Fe(CN)4NO]2-.10,11 Although it has long been
recognized that [Fe(CN)5NO]3- is inert for the release of
NO,12 the NO-dissociation rate constant has only recently
been measured (k-NO ) 1.6 × 10-5 s-1, 25.0 °C).13 Con-
sequently, the rapid delivery of NO in the biological media
has been frequently ascribed to the labile [Fe(CN)4NO]2-
ion with the additional proposal that the remaining cyanides
are subsequently released.12 No kinetic evidence has been
provided for these processes. The timing of the release of
Experimental Section
General Procedures. All chemicals were of analytical grade and
were used without further purification. Eighty-five percent sodium
dithionite (Na2S2O4) was from Acros. Acetate, bis-tris or phosphate,
and borate buffers were used with NaCl to adjust the ionic strength.
The pH measurements in the range of 4-10 were done at room
temperature with a 744 Metrohm pH meter. The preparation of the
solutions was done in Schlenk tubes. The amount of the reducing
agent, usually sodium dithionite (Na2S2O4) or sodium tetrahydro-
borate (NaBH4), necessary for the reaction was added as a solid to
N2- or Ar-saturated solutions of NP containing the appropriate buffer
and ionic strength. At least a 2-fold excess of NP over the reducing
agent was always employed. The solutions were always protected
from light and were transferred using gastight syringes and a
vacuum/Ar or N2 line. NO was purchased from Air Liquide and
purified from higher nitrogen oxides by passing them through an
ascarite II column. The concentration of NO in the saturated
solutions prepared under NO bubbling was 1.8 mM.
Stoichiometric and Kinetic Results. (a) Chemical Analysis
of Reactants and Ionic Products. NP (2-3.5 mg) was dissolved
in 10 mL of a 0.01 M buffer solution (I ) 0.1 M, NaCl). Na2S2O4
was added, and an aliquot was immediately transferred to a
spectrophotometric cell to record the UV-vis spectrum. The initial
concentration of the reduced NP species, either [Fe(CN)5NO]3- or
[Fe(CN)4NO]2-, was calculated considering the molar absorbance
values reported in the literature10 (these values were also checked
experimentally in this study). The concentrations usually ranged
between 0.2 and 0.5 mM. The samples were allowed to react for 2
(8) Ballou, D. P.; Zhao, Y.; Brandish, P. E.; Marletta, M. A. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 12097-12101.
(9) Butler, A. R.; Calsy-Harrison, A. M.; Glidewell, C.; Sorensen, P. E.
Polyhedron 1988, 7, 1197-1202.
(10) Cheney, R. P.; Simic, M. G.; Hoffman, M. Z.; Taub, I. A.; Asmus, K.
D. Inorg. Chem. 1977, 16, 2187-2192.
(11) (a) Nast, R.; Schmidt, J. Angew. Chem., Int. Ed. Engl. 1969, 8, 383.
(b) van Voorst, J. D. W.; Hemmerich, P. J. Chem. Phys. 1966, 45,
3914-3918. (c) Schmidt, J.; Ku¨hr, H.; Dorn, W. L.; Kopf, J. Inorg.
Nucl. Chem. Lett. 1974, 10, 55-61. (d) Glidewell, C.; Johnson, I. L.
Inorg. Chim. Acta 1987, 132, 145-147.
(12) (a) Butler, A. R.; Glidewell, C.; Johnson, I. L.; McIntosh, A. S. Inorg.
Chim. Acta 1987, 138, 159-162. (b) Shafer, P. R.; Wilcox, D. E.;
Kruszyna, H.; Kruszyna, R.; Smith, R. P. Toxicol. Appl. Pharmacol.
1989, 99, 1-10. (c) Wilcox, D. E.; Kruszyna, H.; Kruszyna, R.; Smith,
R. P. Chem. Res. Toxicol. 1990, 3, 71-76. (d) Bates, J. N.; Baker, M.
T.; Guerra, R., Jr.; Harrison, D. G. Biochem. Pharmacol. 1991, 42,
S157-S165. (e) Kruszyna, H.; Kruszyna, R.; Rochelle, L. G.; Smith,
R. P.; Wilcox, D. E. Biochem. Pharmacol. 1993, 46, 95-102.
(f) Rochelle, L. G.; Kruszyna, H.; Kruszyna, R.; Barchowsky, A.;
Wilcox, D. E.; Smith, R. P. Toxicol. Appl. Pharmacol. 1994, 128,
123-128.
(13) (a) Roncaroli, F.; Olabe, J. A.; van Eldik, R. Inorg. Chem. 2003, 42,
4179-4189. (b) Values of NO-dissociation rate constants are very
scarce in the literature. By comparison with the dissociation rates of
other L ligands in the [FeII(CN)5L]3- complexes, NO was shown to
be more labile than the strong π-acceptor NO+ (k-NO+ cannot be
measured from NP) but more inert than NH3, a σ-only ligand. We
infer that an intermediate σ-π binding of NO to Fe(II), of comparable
magnitude to the one for pyrazine or dimethyl sulfoxide, is established.
(14) Morando, P. J.; Borghi, E. B.; Schteingart, L. M.; Blesa, M. A. J.
Chem. Soc., Dalton Trans. 1981, 435-440.
(15) (a) Ford, P. C.; Lorkovic, I. M. Chem. ReV. 2002, 102, 993-1018.
(b) Ford, P. C.; Laverman, L. E.; Lorkovic, I. M. AdV. Inorg. Chem.
2003, 54, 203-257.
(16) Ueno, T.; Suzuki, Y.; Fujii, S.; Vanin, A. F.; Yoshimura, T. Biochem.
Pharmacol. 2002, 63, 485-493.
2782 Inorganic Chemistry, Vol. 44, No. 8, 2005

Release of NO from Reduced NP Ion
days in the Schlenk tubes. Then aliquots were extracted, and the
different ions present in the final solutions were analyzed using
spectrophotometric techniques. The latter procedures were not
performed under anaerobic conditions. A Hewlett-Packard 8453A
diode array spectrophotometer was employed for the UV-vis
measurements. The o-phenanthroline and SCN- reagents were used
(30 mg) was added after saturation with Ar, and then the solution
was immediately transferred to the IR liquid cell. The rest of the
solution was stored in a syringe which was kept in a large Schlenk
tube under Ar to prevent any reaction with air. From time to time,
aliquots of the solution were loaded in the cell and the IR spectrum
was recorded with a Thermo Nicolet Avatar 320 FT-IR spectrometer
and a Spectratech liquid cell containing two 32 × 3 mm CaF2 disks
and a 0.1 mm spacer. The spectral ranges were 1750-2800 and
2150-1300 cm-1 in the H2O and D2O solutions, respectively. The
reaction of NO with the [Fe(CN)4NO]2- complex was studied by
bubbling NO through the complex solution.
To perform the labeling experiments starting with
[Fe(CN)515NO]2-, we added 55 mg of Na3[Fe(CN)5NH3]‚3H2O23
to a deoxygenated solution of 12 mg of Na15NO2 (Aldrich) in
20 mL of H2O. The solution was allowed to react for 1 h, and then
HCl was added to reach pH 2-4. Four hours later, the solution
was evaporated to dryness. The residue was dissolved in a few
milliliters of D2O and evaporated again. The IR spectrum was
checked for comparison with reported values for NP labeled with
15NO.24 For the labeling experiments, this sample was dissolved in
5 mL of D2O together with the buffer system (pH 7). Finally,
Na2S2O4 was added as described for the other IR experiments.
(d) EPR Spectroscopy. Samples (0.3 mM in the reduced species,
0.01 M in buffer, and I ) 0.1 M) were prepared as described before
and kept at 25.5 °C during the experiment. From time to time,
aliquots were extracted and transferred under N2 to EPR quartz
tubes. Each sample was frozen immediately in liquid N2. After all
of the samples were collected, they were measured in the X-band
of a Bruker ESP 300E spectrometer. The spectra were recorded at
9.44 GHz, 0.635 mW power, 100 kHz modulation frequency, and
9.3434 G modulation amplitude. To study the reaction between NO
and the [Fe(CN)4NO]2- complex, we prepared the solutions as
already described and bubbled NO through them. They were, then,
transferred to the EPR tube and frozen. The g values were estimated
by using the magnetic field and the microwave frequency without
performing a calibration.
(e) Electrochemical Detection of NO. An amiNO electrode from
Innovative Instruments, Inc., was used. The electrode was attached
to a homemade potentiostat interfaced to a computer. For these
measurements, 10 mg of NP was dissolved in 10 mL of Ar-saturated
0.1 M buffer solutions. After allowing for equilibration of the
electrode output, we added 3 mg of Na2S2O4 and recorded the
current as a function of time. FeSO4, CuSO4, CoSO4, MnSO4, and
NiCl2 (9-10 mM at pH 7.0) were used to study the influence of
metal ions. KSCN (20 mM), o-phenanthroline, or disodium
ethylenediaminetetraacetate was used to study the influence of other
ligands.
(f) Mass-Spectrometric Determinations of N2O. A 0.5 M
buffered solution containing 100 mg of NP in 10 mL was transferred
to a mechanically stirred reactor (volume ca. 24 mL) attached to a
vacuum system. After evacuation, 30 mg of Na2S2O4 was added
from a side flask. The reactor was linked through a capillary to a
SenSym SX15A pressure sensor and to an Extrel Emba II mass
spectrometer. The temperature was controlled at 25 °C. The
evolution of the total pressure as a function of time was recorded
during the experiment. At the end, the composition of the gaseous
products was analyzed with the mass spectrometer.25
to determine the concentrations of [Fe(H2O)6]2+ and [Fe(H2O)6]3+
,
respectively.17 Nitrite was quantified using the Griess reaction.18
Free CN- or HCN was determined with an Aquaquant 14429 kit
from Merck. NP was determined through the reaction with
thiosuccinic acid,19 as described elsewhere.20 The final value was
corrected using the initial concentration of NP. By knowing the
amount that was reduced from the initial spectrum, we determined
the concentration of NP that was effectively obtained through the
reoxidation of [Fe(CN)5NO]3-. The concentration of [Fe(CN)6]4-
was determined from the reaction with [Fe(H2O)6]3+ producing
Prussian-Blue.21 The sample was mixed with an equal volume of
a 0.1 M HNO3 solution containing 1 mM Fe(NO3)3. Calibration
was achieved by preparing standard solutions of [Fe(CN)6]4- in
the range of 0-0.2 mM. Samples and standard solutions were
allowed to react for 2 h, and the absorbance at 705 nm was
measured. No precipitation of Prussian-Blue was observed under
these conditions, and the calibration curve showed a good linear
distribution.
(b) UV-Vis Spectroscopy. (i) Slow Processes. Samples
(ca. 0.2 mM in the reduced species, 0.01 M in buffer, and
I ) 0.1 M) were transferred to a cell with a 1 cm optical path,
under N2. The absorbance changes in the range of 300-800 nm
were recorded on a double-beam Varian Cary 5G spectrophotometer
equipped with a temperature control system. The results were
analyzed with the SPECFIT program22 using a two-exponential
model (A f B f C) for the reaction at pH 4-8 and a single-
exponential model (A f B) for the one at pH 9-10. The molar
absorbance values of the intermediates were estimated by calculating
the initial complex concentration through the molar absorbance
reported in the literature.10 Analyses of residuals from the SPECFIT
treatments are presented in Figure SI1. All of the reactions were
done at 25.5 ( 0.2 °C.
(ii) Stopped-Flow (SF) Measurements. NO-saturated solutions
were prepared as described elsewhere.20 The ionic strength was
0.1 M (NaCl), and the buffer concentration was 0.01 M (pH 5-7).
NP was reduced with Na2S2O4 in the buffer solutions, and the final
reduced complex was 10-fold substoichiometric with respect to NO.
The NO and complex solutions were mixed with a SX 18MV
Applied Photophysics SF instrument at 25.0 ( 0.1 °C. Kinetic traces
were recorded at 620, 420, and 330 nm and were fitted to a single
exponential over more than 3 half-lives. The observed rate constants
did not differ significantly at the different wavelengths.
(c) IR Spectroscopy. NP (100 mg) was dissolved in 10 mL of
a solution 0.5 M in buffer (without the addition of NaCl) in H2O
or D2O, depending on the spectral range under study. Na2S2O4
(17) Vogel, A. A Textbook of QuantitatiVe Inorganic Analysis Including
Elementary Instrumental Analysis, 4th ed.; Longman Inc.: New York,
1978.
(18) Schmidt, H. H. H. W.; Kelm, M. In Methods in Nitric Oxide Research;
Feelisch, M.; Stamler, J. S., Eds.; Wiley: Chichester, UK, 1996; Part
VII, Chapter 33.
(19) Szacilowski, K.; Stochel, G.; Stasicka, Z.; Kisch, H. New J. Chem.
1997, 21, 893-902.
(20) Roncaroli, F.; Olabe, J. A.; van Eldik, R. Inorg. Chem. 2002, 41,
5417-5425.
(23) Kenney, D. J.; Flynn, T. P.; Gallini, J. B. J. Inorg. Nucl. Chem. 1961,
20, 75.
(24) Chaco´n Villalba, M. E.; Varetti, E. L.; Aymonino, P. J. Vib. Spectrosc.
1997, 14, 275-286.
(25) Gutie´rrez, M. M.; Amorebieta, V. T.; Estiu´, G. L.; Olabe, J. A. J.
Am. Chem. Soc. 2002, 124, 10307-10319.
(21) Sharpe, A. G. The Chemistry of Cyano Complexes of the Transition
Metals; Academic Press: London, 1976.
(22) Binstead, R. A.; Zuberbulher, A. D. SPECFIT; Spectrum Software
Associates: Chapel Hill, NC, 1993-1999.
Inorganic Chemistry, Vol. 44, No. 8, 2005 2783

Roncaroli et al.
Table 1. IR and UV-Vis Spectroscopic Data for Reactants,
Intermediates, and Products
stretching
frequencies
λ
max (nm)
compound
(cm-1
)
(ꢀ, M-1 cm-1
)
[Fe(CN)5NO]2-
[Fe(CN)4NO]2-
[Fe(CN)5NO]3-
[Fe(CN)6]4-
2142, 193626 498 (8), 394 (25), 330 (40)27
2115, 180026 615 (380), 430 (100), 350 (300)10
2088, 164826 430 (550), 345 (3500)10
203826
422 (5), 322 (300)28
[Fe(CN)4(OH)NO]2-
[Fe(CN)4(NO)2]2- (I1)
[Fe(CN)2(NO)2]- (I2)
N2O
191010
1695, 1737 333 (600), 425 (200), 600 (120)
1810, 1737 330 (650), 400 (300), 500(150)
223029
Results
A summary of the compounds that appear as main
reactants, products, or proposed reaction intermediates along
with the relevant IR and UV-vis spectroscopic data is
presented in Table 1.10,26-28
(a) pH 9-10. Predominance of [Fe(CN)5NO]3- at the
Beginning of Decomposition. The [Fe(CN)5NO]3- ion is
the most predominant initial species in this pH range after
the fast reduction of NP and the establishment of equilibrium
1. Figure 1 shows the successive UV-vis spectra reveal-
ing a decreasing intensity of the absorption bands of
[Fe(CN)5NO]3- (Table 1). The single-exponential decay
is valid over 3 half-lives affording a value of kobs ) 4 ×
10-5 s-1 at pH 9. Note that some [Fe(CN)4NO]2- can be
observed through the weak absorption at ca. 620 nm.
Figure 2 shows the successive spectra for the IR measure-
ments in D2O. An intensity decrease can be observed for
the stretching vibrations of the cyano and NO groups in
[Fe(CN)5NO]3- at 2088 and 1648 cm-1, respectively.26
Product formation is associated with the onset of a new band
at 2038 cm-1, typical of the [Fe(CN)6]4- ion.21,26 The
concentration of this ion was also determined in the solutions
(ca. 20-30% with respect to initially reduced species),
although a strictly quantitative evaluation was disturbed by
the formation of precipitates. Again, minor quantities of
[Fe(CN)4NO]2- were detected through the bands at 2115 and
1800 cm-1.26 A weak shoulder at 1910 cm-1 can be assigned
to [Fe(CN)4(OH)(NO)]2-.10 NO was detected electrochemi-
cally during the first 15 min and was absent after 10 h. N2O
was also observed in the IR spectra (not shown) after ca.
10 h of reaction through its characteristic band at
2230 cm-1.29 Nitrite was also detected over the whole pH
range of 4-10( yield < 5%). Finally, Figure 2 allows us to
detect some gradual increase in the characteristic absorptions
of NP at 2142 and 1936 cm-1.26 The quantitative determi-
nation at the end of reaction showed, at pH 10, around a
10% yield of NP with respect to the initial [Fe(CN)5NO]3-
concentration.
Figure 1. UV-vis spectral changes observed during the decomposition
of a ca. 0.2 mM reduced nitroprusside solution (reductant: Na2S2O4) at
pH 9.0 (I ) 0.1 M, T ) 25.5 °C, 0.01 M borate buffer) with a cycle time
of 2400 s. Inset: kinetic traces at 348 and 440 nm fitted to a single
exponential by SPECFIT, kobs ) (4.0 ( 0.3) × 10-5 s-1
.
Figure 2. IR spectral changes recorded during the decomposition of a
solution ca. 17 mM in reduced nitroprusside (reductant: Na2S2O4) at
pH 10 (T ) 25 °C, 0.5 M borate buffer) in D2O: kobs ≈ 4 × 10-5 s-1
(decay of νNO). Peak numbers correspond to the following (as detailed in
Table 1): 1, [Fe(CN)5NO]2-; 2, [Fe(CN)4NO]2-; 3, [Fe(CN)5NO]3-; 4,
[Fe(CN)6]4-; and 5, [Fe(CN)4(OH)NO]2-
.
successive UV-vis spectra with significant quantities of both
complexes at the beginning. The inset includes the absor-
bance traces at two selected wavelengths fitted to a two-
exponential model. A monotonic decay at 615 nm along with
an initial increase and subsequent decrease of an intermediate
I1 with maximum at 336 nm can be observed. The output
leads to values of kfirst ) 3-6 × 10-5 s-1 for the first reaction
step and of ksecond ) 1-2 × 10-5 s-1 for the second one,
involving the formation and decay of the intermediate I1,
respectively. Figure 3b shows the Specfit spectra calculated
for the reactant, the intermediate, and the product. Table 1
includes the calculated absorbance maxima for I1. Similar
(b) pH 6-8. Decomposition of Mixtures of
[Fe(CN)5NO]3- and [Fe(CN)4NO]2-. Figure 3a shows the
(26) Schwane, J. D.; Ashby, M. T. J. Am. Chem. Soc. 2002, 124, 6822-
6823.
(27) Manoharan, P. T.; Gray, H. B. Inorg. Chem. 1966, 5, 823-839.
(28) Gray, H. B.; Beach, N. A. J. Am. Chem. Soc. 1963, 85, 2922-2926.
(29) Sampath, V.; Rousseau, D. L.; Caughey, W. S. In Methods in Nitric
Oxide Research; Feelisch, M., Stamler, J. S., Eds.; Wiley: Chichester,
U.K., 1996; Part VI, Chapter 29.
2784 Inorganic Chemistry, Vol. 44, No. 8, 2005

Release of NO from Reduced NP Ion
Figure 4. Chronoamperograms for the release of NO after addition of
3 mg of Na2S2O4 to 10 mL of a 3.4 mM nitroprusside solution at different
pHs.
spect to initial reduced complex), together with some ferrous
and ferric ions and a white precipitate, presumably
Fe2[Fe(CN)6].21 The relative yields of NP and N2O were 40%
and 20%, respectively, with respect to initial reduced NP
concentration.
The successive IR spectra in Figure 5a (pH 7, D2O
solution) show a decrease in the intensity of the bands for
both the [Fe(CN)5NO]3- and [Fe(CN)4NO]2- ions along with
an absorption increase at 2038 cm-1 ([Fe(CN)6]4- and/or
other Fe-cyano-aqua complexes). The absorption increase
at 2142 and 1936 cm-1 corresponding to NP is again clearly
observed, and the peak at 2230 cm-1 for N2O is apparent at
ca. 3-6 h after the start of decomposition (cf. Figure SI2,
measurements in H2O). Another outstanding feature is given
by the appearance of a weak absorption at 1695 cm-1 (with
a maximum intensity attained around 2.5 h) which is absent
at the beginning and end of the process revealing its
intermediate character (we assign these IR features to I1; see
Discussion). Finally, Figure 5b again shows the peak at
1695 cm-1 and a weaker absorption at ca. 1737 cm-1 after
a complementary experiment with a controlled addition of
NO to [Fe(CN)4NO]2- was performed. These peaks disappear
under a further excess of NO.
In the IR labeling experiments with [Fe(CN)415NO]2-,
Figure SI3 shows that the intermediate I1 appears ca. 2 h
after the beginning of the spontaneous decomposition. As
expected, it shows that all of the relevant peaks associated
with νNO stretchings (Table 1) are shifted downward by ca.
40 cm-1, including the band previously assigned to the I1
complex at 1695 cm-1 which now develops at 1655 cm-1.
In a second complementary experiment with unlabeled NO
bubbling through a fresh solution of [Fe(CN)415NO]2-, Figure
SI4 shows that a shoulder assignable to I1 appears at ca.
1670 cm-1. Interestingly, two absorptions for the emerging
NP complex can be observed. In addition to the one at
1897 cm-1, corresponding to [Fe(CN)515NO]2-,24 another one
shows up at 1936 cm-1, corresponding to [Fe(CN)514NO]2-.
Figure 3. (a) UV-vis spectral changes observed during the decomposition
of a solution ca. 0.25 mM in reduced NP (reductant: Na2S2O4) at pH 6.0
(I ) 0.1 M, T ) 25.5 °C, 0.01 M bis-tris buffer) with a cycle time of
7200 s. Inset: kinetic traces at 336 and 618 nm fitted to a double-exponential
model by SPECFIT, kfirst ) (5.4 ( 0.2) × 10-5 s-1, ksecond ) (9.3 ( 0.5)
× 10-6 s-1. (b) Spectra of the reactants, intermediates, and products obtained
from the SPECFIT analysis.
Table 2. UV-Vis Kinetic Data Obtained for the Spontaneous
Decomposition of the Mixture of [Fe(CN)5NO]3- and [Fe(CN)4NO]2- as
a Function of pHa
pH
k
first (×105 s-1
)
ksecond (×105 s-1
)
10.0
9.0
8.0
7.0
6.0
5.0
4.0
2.5 ( 0.5
3.5 ( 0.5
5.3 ( 0.5
2.7 ( 0.2
5.5 ( 0.1
20 ( 0.5
200 ( 0.5
-
-
1.5 ( 0.5
1.1 ( 0.5
0.9 ( 0.3
1.4 ( 0.3
not measured
a Reported values correspond to average values of at least two indepen-
dent measurements. Errors were estimated from the difference between those
values. T ) 25.5 °C, 0.01 M buffer, I ) 0.1 M (NaCl). See text for the
determination of rate constants.
spectral and kinetic data were obtained by using either
dithionite or tetrahydroborate as reducing agents. Table 2
reports the kinetic results in all of the studied pH ranges.
The chronoamperograms recorded in Figure 4 show that
NO is generated rapidly during the first minutes with the
rate increasing as the pH decreases. The concentration of
NO remains constant or decreases with time. For N2O, the
mass-spectrometric measurements reveal a continuous ex-
ponential increase in the pressure in the reactor with
kobs ) 1.4 × 10-5 s-1 for the formation rate.
At the end of the reaction, [Fe(CN)6]4- was found in low
yields, in contrast with free cyanides (60-200% with re-
Inorganic Chemistry, Vol. 44, No. 8, 2005 2785

Roncaroli et al.
Figure 5. (a) IR spectral changes observed during the decomposition of
a solution ca. 17 mM in reduced NP (reductant: Na2S2O4) at pH 7
(T ) 25 °C, 0.5 M phosphate buffer) in D2O, kobs ≈ 5 × 10-5 s-1 (decay
of νNO). (b) Spectrum obtained after bubbling NO trough a reduced
nitroprusside solution until first color change (17 mM in D2O). A detailed
description of the spectral changes is given in the text. Peak numbers
correspond to the following (as detailed in Table 1): 1, [Fe(CN)5NO]2-; 2,
Figure 6. (a) EPR spectral changes recorded during the decomposition
of a solution ca. 0.26 mM in reduced nitroprusside (reductant: NaBH4) at
pH 5.0 (I ) 0.1 M, T ) 25 °C, 0.01 M acetate buffer). (b) Spectrum obtained
after bubbling NO through a reduced nitroprusside solution (0.26 mM,
I ) 0.1 M, 0.01 M acetate buffer).
[Fe(CN)4NO]2-; 3, [Fe(CN)5NO]3-; 4, [Fe(CN)6]4-; 5, [Fe(CN)4(OH)NO]2-
and 6, [Fe(CN)4(NO)2]2-
;
.
and Prussian-Blue type precipitates. N2O is also observed
in the IR spectra (Figure SI5).
In the EPR measurements (not shown), the intensity of
the reactant’s signal decreases exponentially, with a rate
constant of ca. 5 × 10-5 s-1, affording an EPR-silent product.
Values on the same order of magnitude (k ) ca. 5-8 ×
10-5 s-1) may be estimated from the IR data in Figure 5a
Figure 6a shows the successive EPR spectra obtained at
pH 5. The signals of the reactant decay30 evolve through an
apparently silent intermediate (second spectrum from bottom)
to reach a final different spectrum (which also decays for
longer times). The last spectrum is similar to those reported
in the literature for the four-coordinate {Fe(L)2(NO)2} com-
pounds (S ) 1/2, L ) thiolates, imidazole, CO, etc.) called
“g ) 2.03” dinitrosyl iron complexes (DNIC).31 It is also
similar to the one obtained after bubbling NO through the
reduced NP solution (Figure 6b), and we identify it as I2.
by measuring the intensity decrease of νNO
.
(c) pH 4-5. Fast Decomposition of Predominant
[Fe(CN)4NO]2-. The UV-vis spectral runs reveal a signifi-
cant increase in the decomposition rate for the [Fe(CN)4NO]2-
ion, as detailed in Table 2. This is confirmed by the elec-
trochemical experiments on release of NO. Figure 4 shows
a consistent increase in the rate and amount of NO produced
at pH 4 within the first minutes. From the exponential
display, we estimate a value for k-NO, in fair agreement with
the kinetic data in Table 2, of ca. 2 × 10-3 s-1.
(30) The initial EPR spectrum in Figure 6a shows the presence of both the
tetracyanonitrosyl and pentacyanonitrosyl species. The percentage of
each species is strongly affected by the temperature. Only the
tetracyanonitrosyl ion gives an EPR signal at 22 °C.12d The final
spectrum in Figure 6a and the one in Figure 6b agree with those
reported for the four-coordinated dinitrosyls.31
(31) (a) Lee, M.; Arosio, P.; Cozzi, A.; Chasteen, N. D. Biochemistry 1994,
33, 3679-3687. (b) Vanin, A. F.; Serezhenkov, V. A.; Mikoyan, V.
D.; Genkin, M. V. Nitric Oxide 1998, 2, 224-234. (c) Reginato, N.;
McCrory, C. T. C.; Pervitsky, D.; Li, L. J. Am. Chem. Soc. 1999,
121, 10217-10218. (d) Butler, A. R.; Glidewell, C.; Hyde, A. R.;
Walton, J. C. Polyhedron 1985, 4, 797-809. (e) Costanzo, S.; Menage,
S.; Purrello, R.; Bonomo, R.; Fontecave, M. Inorg. Chim. Acta 2001,
318, 1-7. (f) Burlamaschi, L.; Mart´ın, G.; Tiezzi, E. Inorg. Chem.
1969, 8, 2021-2025. (g) McDonald, C. C.; Phillips, W. D.; Mower,
H. F. J. Am. Chem. Soc. 1965, 87, 3319-3326. (h) Ueno, T.;
Yoshimura, T. Jpn. J. Pharmacol. 2000, 82, 95-101. (i) Foster, M.
W.; Cowan, J. A. J. Am. Chem. Soc. 1999, 121, 4093-4100. (j) Bryar,
T. R.; Eaton, D. R. Can. J. Chem. 1992, 70, 1917-1926.
Although our experimental setup had a headspace, we
believe that this does not alter our conclusions. Since all of
the experiments were done under the same conditions, we
believe that a comparison between the results at different
pH levels can still be made. At pH 4, the value of k-NO
(ca. 2 × 10-3 s-1) could be affected to some degree by the
loss of NO on the minute time scale. However, the order of
magnitude of k-NO is in good agreement with that measured
spectrophotometrically.
NP is also formed in a 50% yield at the end of reaction
along with high yields of cyanide (300%), aqueous Fe(III),
2786 Inorganic Chemistry, Vol. 44, No. 8, 2005

Release of NO from Reduced NP Ion
Figure 8. (a) IR spectral changes observed during the decomposition of
a solution ca. 17 mM in reduced nitroprusside (reductant: Na2S2O4) at
pH 4 (T ) 25 °C, 0.5 M acetate buffer) in D2O, kobs ≈ 1 × 10-3 s-1 (decay
of νNO). (b) Spectrum obtained after bubbling NO through a reduced
nitroprusside solution (17 mM, 0.5 M acetate buffer) in D2O. A detailed
description of the spectral changes is given in the text. Peak numbers
correspond to the following (as detailed in Table 1): 1, [Fe(CN)5NO]2-; 2,
Figure 7. (a) UV-vis spectral changes recorded during the decomposition
of a solution ca. 0.32 mM in reduced nitroprusside (reductant: NaBH4) at
pH 5.0 (I ) 0.1 M, T ) 25.5 °C, 0.01 M acetate buffer) with cycle times
of 900 and 3600 s after 18000 s. Inset: kinetic traces at 330 and 618 nm
fitted to a three-exponential model by SPECFIT, kfirst ) (2.4 ( 0.2) ×
10-4 s-1, ksecond ) (5.1 ( 0.5) × 10-5 s-1, and kthird ) (2.4 ( 0.7) ×
10-5 s-1. (b) Spectra of the reactants, intermediates, and products obtained
from the SPECFIT analysis.
[Fe(CN)4NO]2-
;
and 7, [Fe(CN)2(NO)2]-. The broad envelope at
2000-2100 cm-1 corresponds to a mixture of soluble and/or colloidal Fe-
(II)-cyano or Fe(II)-cyano-aqua complexes.
i.e., lower decomposition rates were observed, and the I1
intermediate could be detected again in the experiment with
dithionite.
In Figure 7, both dinitrosyl intermediates, I1 and I2, can
be detected in the UV-vis spectra when tetrahydroborate is
used following a SPECFIT analysis with a three exponential
model. Only I2 is observed when dithionite is used (Figure
SI 6). Table 1 shows the calculated UV-vis absorption
parameters for I2.
Figure 8a shows the IR spectra obtained in the D2O
solution at pH 4. The decay of the bands from [Fe(CN)4NO]2-
can be observed, and the product formation is associated with
increasing absorptions apparently corresponding to the fol-
lowing three processes: (i) formation of NP consistent with
bands at 1936 and 2142 cm-1 and with the chemical analysis
at the end of reaction, (ii) Fe(II) and Fe(III) Prussian-Blue
type species corresponding to a broad envelope in the
2000-2080 cm-1 region,21 and (iii) an increase and further
decrease of a new absorption at 1737 cm-1, assigned to I2.
This process correlates with the shift of the band from 1800
to 1810 cm-1. The absorptions of the I2 intermediate at 1810
and 1737 cm-1 can also be observed in Figure 8b after
bubbling NO through a solution of the reactant.
(d) Influence of Added Metal Ions. Figure SI7 shows
the influence of different metal ions on the rate of release
of NO. A rapid increase (Cu > Co > Fe) is observed for
the short times after addition followed by a slower increase
for the last two ions and a distinctive decrease for copper.
The influence of metal ions probably occurs because of
association effects on bound cyanides, promoting the de-
composition of [Fe(CN)4NO]2-. It is also related to the ability
to form stable complexes with cyanide. This is equivalent
to the pH effect noted above. The presence of metal ions is
highly feasible in biological media, thus allowing for a fast
decomposition process even at pH conditions around 7 which
are insufficiently low to promote the complete release of
cyanide. We also checked the influence of added ligands on
the release of NO. We found that thiocyanate, o-phen-
anthroline, and EDTA enhanced the decomposition rate of
the complex (viz., cyanide release), but surprisingly, no effect
on the release of NO was observed.
Discussion
An interesting feature in the experiments at pH 4-5 was
related to the influence of the addition of cyanide. This led
to a behavior similar to that reported above for higher pHs;
The observed rate constant for the decay of reduced NP
at pH 10 is 2-3 times larger than the value previously
Inorganic Chemistry, Vol. 44, No. 8, 2005 2787

Roncaroli et al.
13,32
measured for [Fe(CN)5NO]3-, reaction 2 (k2 ) k-NO
)
complexes and have been characterized by UV-vis, NMR,
and IR spectroscopies, showing in the latter case an intense
band at 1695 cm-1 and a much weaker band at 1777 cm-1.
The presently reported labeling results for I1 lead to similar
conclusions about its structural identity. The shift of 40 cm-1
is close to that observed for trans-Fe(P)(NO)2,35 consistent
with a doubly labeled dinitrosyl species. As a result of the
theoretical calculations, a trans-syn (C2V) conformation has
been proposed for Fe(TPP)(NO)235,36 and similar geometries
have been calculated for hypothetical model analogues, such
as [M(NH3)4(NO)2]2+ (M ) Fe, Ru).35 We anticipate, here,
some preliminary results from DFT calculations for the
proposed [Fe(CN)4(NO)2]2- complex which also reproduce
the trans-syn geometry and provide fairly consistent νNO
values at 1725 and (weaker) 1826 cm-1 for the singlet
dinitrosyl species.
[Fe(CN)5NO]3- + H2O { k2 } [Fe(CN)5H2O]3- + NO (2)
k-2
The difference arises from the fact that, under our reaction
conditions, a small amount of [Fe(CN)4NO]2- is present in
the equilibrium mixtures described by reaction 1 (cf. UV-
vis and IR spectra). The formation of hexacyanoferrate(II)
observed occurs through a follow-up to reaction 2, viz. the
thermal decomposition of [Fe(CN)5H2O]3-, which evolves
according to the stoichiometry in eq 3
[Fe(CN)5H2O]3- f 5/6[Fe(CN)6]4- + 1/6Fe2+ + H2O (3)
Reaction 3 has previously been analyzed mechanistically
on the basis of a set of competitive-consecutive reactions.33
The rate-determining first-order decay of [Fe(CN)5H2O]3-
leads to [Fe(CN)4(H2O)2]2- with k-CN ) 1.25 × 10-4 s-1
(25.0 °C, I ) 1 M). The successive release of the remaining
cyanides is faster as is the reverse coordination of CN- to
the reacting [Fe(CN)5H2O]3- ion.33 The products in reaction
3 lead to the formation of Prussian-Blue type precipitates,
as observed in some of our experiments, unless the solutions
are sufficiently dilute.21,33
The decomposition of I1 occurs through a disproportion-
ation process, as described by eq 6.
trans-[Fe(CN)4(NO)2]2- + CN- + H+ f
[Fe(CN)5NO]2- + HNO (6)
2HNO a N2O + H2O
(7)
The stoichiometry found for NP/N2O (2:1 mole ratio)
agrees with reactions 6 and 7. The ability to detect I1 is
consistent with the values of the rate constants kfirst and ksecond
in Table 1 and also agrees with the UV-vis and IR evidence
on the time scale of the detection of the intermediate.
Particularly revealing is the systematic appearance of rate-
constant values around 10-5 s-1, obtained through different
types of measurements, either for the decay of [Fe(CN)4NO]2-
or for the formation of the products, including the rate of
formation of N2O. By considering the concentration of NO
under steady-state conditions for reactions 4-6, we find
kfirst ) 2k4. The mechanistic picture is consistent with NO
dissociation controlling the nitrosyl transfer to other accep-
tors, as previously observed with intermolecular NO transfer
from Mn- to Fe-tropocoronand complexes.37
In reaction 6, we propose an intramolecular electron
transfer in I1 leading to bound NO+ and NO-/HNO, followed
by the release of the latter species. Fast coupling of HNO
generates N2O, eq 7, as described elsewhere.38 More evidence
on this decomposition mode for I1 is given by the comple-
mentary results obtained when unlabeled NO reacts with
[Fe(CN)415NO]2-. The intermediate shift (ca. 20 cm-1)
corresponds to the presence of coupled oscillators of 14NO
and 15NO in I1. Most importantly, not only labeled NP but
also unlabeled NP is formed as a product. This occurs
because there is a probability for each nitrosyl to be oxidized
or reduced in the intramolecular process.
The results at pH 10 show the formation of NP and N2O
as additional products, although in minor amounts when
compared to those obtained at pH 6-7. In this pH range,
the UV-vis, IR, and EPR spectra reveal the distinctive
formation of intermediate I1. We propose it to be a trans-
dinitrosyl species which forms after the dissociation of NO
from the [Fe(CN)4NO]2- ion, eq 4, and the subsequent
coordination to the same complex, eq 5
[Fe(CN)4NO]2- + 2H2O a [Fe(CN)4(H2O)2]2- + NO (4)
[Fe(CN)4NO]2- + NO a trans-[Fe(CN)4(NO)2]2- (5)
The EPR-silent properties of I1 can be ascribed to a low-
spin Fe(II) center containing two antiferromagnetically
coupled NO ligands. Support for the product identification
in eq 5 is provided by kinetic evidence: the SF experiment
in which excess NO reacts with [Fe(CN)4NO]2- shows a
first-order rate law in each reactant with k5 ) 4.3 ×
104 M-1 s-1. The IR evidence obtained in the absence or
presence of additional NO shows a transient new absorption
at 1695 cm-1 which can be assigned to an N-O stretching
vibration in I1. The decay of the IR band upon further NO
bubbling indicates that I1 is also unstable under these
conditions. Complementary evidence is provided by recent
reports on the coordination of NO to ferrous nitrosyl
porphyrins (P) giving trans-Fe(P)(NO)2 in low-temperature
solutions (P ) TPP, meso-tetraphenylporphinato, and TmTP,
meso-tetra-m-tolylporphinato).34 These are also EPR-silent
The stoichiometric results strongly support the intra-
molecular reaction path in the absence of excess NO. This
(32) An excess of cyanide was used in the work reported in ref 13 providing
for the near exclusion of [Fe(CN)4(NO)]2- as well as for the fast
scavenging of the [Fe(CN)5H2O]3- ion in the ensuing reaction 2. This
allowed us to obtain k-NO ) 1.6 × 10-5 s-1 at 25 °C from the
[Fe(CN)5NO]3- ion in a clean manner.
(33) Olabe, J. A.; Zerga, H. O. Inorg. Chem. 1983, 22, 4156-4158.
(34) Lorkovic, I.; Ford, P. C. J. Am. Chem. Soc. 2000, 122, 6516-6517.
(35) Patterson, J. C.; Lorkovic, I. M.; Ford, P. C. Inorg. Chem. 2003, 42,
4902-4908.
(36) Conradie, J.; Wondimagegn, T.; Ghosh, A. J. Am. Chem. Soc. 2003,
125, 4968-4969.
(37) Franz, K. J.; Lippard, S. J. Inorg. Chem. 2000, 39, 3722-3723.
(38) Shafirovich, V.; Lymar, S. V. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
7340-7345.
2788 Inorganic Chemistry, Vol. 44, No. 8, 2005

Release of NO from Reduced NP Ion
means that NO+ and NO- must be formed as bound species
in the first event. Direct evidence is provided for NO+ but
not for NO- (or HNO). This is understandable because of
its high lability. However, HNO has been well established
in the literature as a precursor to N2O. The labeling
experiments reinforce the identity of I1 as well as its
decomposition mode.
The process described by eqs 4-7 appears to be the first
reported example of dinitrosyl formation and decomposition
starting from a mononitrosyl complex without external NO
assistance. It takes advantage of the trans effect of bound
NO (eq 1) and thus promotes ligand interchange between
cyanide and NO itself, coming from autodissociation. The
intramolecular character of I1 disproportionation (eq 6) is
also in contrast with external NO-assisted processes reported
in the literature. These have been observed following
coordination of NO to several metal centers (Mn, Re, Fe,
Ru, Os, Ni, Cu) forming, successively, mono- and dinitrosyls,
which decompose in the presence of more NO to give N2O
and nitrite. A different picture appears for each of the metals
considered with some ambiguity remaining on the detailed
mechanistic issues.15
In contrast with previous studies regarding disproportion-
ation,39 the above-mentioned Fe(P)(NO)2 complex remains
stable for hours even at ambient temperature when compared
with the Ru analogue which readily leads to bound nitrite
and N2O.40 Remarkably, the disproportionation process is
observed with Fe(TPP)(NO) only if trace amounts of NO2
are present.41 The contrast with our reactive [Fe(CN)4(NO)2]2-
complex may be related to the low value for the Fe(NO+)/
Fe(NO) redox couple (-0.37 V vs SCE)7 which makes NP
a very stable product when one of the NO ligands in I1
induces an electron transfer to the opposite NO. On the other
hand, a potential of 0.74 V has been reported for the formal
Fe(III)/Fe(II) couple in Fe(TPP)NO42 suggesting a high
comparative stability of the Fe(II)NO fragment in the
dinitrosyl species.
Finally, the experiments at pH 4-5 show some common
as well as distinctive features with respect to those at the
higher pHs. The formation of NP (50% yield) and N2O
suggest a disproportionation route similar to that previously
described. There is a remarkable increase in the decomposi-
tion of [Fe(CN)4NO]2- at pH 4 (2 orders of magnitude faster,
cf. Table 2). As the decomposition is also enhanced by some
metal ions, we propose that this is because of the labilization
of cyanide. Our results show that the addition of cyanide
suppresses the pH effect. Then, equilibrium 8 can be
proposed
[Fe(CN)4NO]2- + H+ a [Fe(CN)3NO]- + HCN (8)
Some evidence for the existence of [Fe(CN)3(NO)]- has
been reported.7 This species decomposes according to eq 9,
releasing cyanide and NO to the reaction medium under pH
conditions where HCN is stabilized. A conversion of Fe(II)
from low-spin to high-spin is probably operative at a given
stage of the decomposition process thus favoring the faster
release of cyanide and NO.
[Fe(CN)3NO]- + 3H+ f Fe2+ + 3HCN + NO (9)
A novel issue introduced by the results at pH 4-5 is the
appearance of intermediate I2. Remembering that the EPR-
silent intermediate in Figure 6 (pH 5) is the previously
described I1, we propose that the latter either may evolve
through disproportionation, as already discussed, or may also
react leading to the new four-coordinate dinitrosyl species,
I2, as is evident from the final spectrum in Figure 6a
coincident with the one obtained after reacting [Fe(CN)4NO]2-
with NO (Figure 6b, see Results). The IR results in Figure
8 provide additional evidence for the identity of I2,
[Fe(CN)2(NO)2]-. The NO-stretching frequencies are re-
markably close to those reported for [Fe(CO)2(NO)2]+,31c viz.
1810 and 1767 cm-1. The IR absorptions of these dinitrosyl
complexes have been shown to be significantly influenced
by the binding properties of the spectator ligands, L.31
The conversion of I1 to I2 can be formally described by
eq 10
trans-[Fe(CN)4(NO)2]2- + e- f
[Fe(CN)2(NO)2]- + 2CN- (10)
The initial five-coordinate reactant, [Fe(CN)4NO]2- (eq
4), is an {FeNO}7 species, and its configuration has been
described as a d7 Fe(I) with a NO+ ligand.11c Addition of
NO forms intermediate I1, a six-coordinate {Fe(NO)2}8
species (eq 5), for which a distribution of d6 Fe(II) with two
NO radicals as ligands may be envisioned. In reaction 10, I1
reacts to form a product, I2, which is a four-coordinate
species, {Fe(NO)2}9 (S ) 1/2). The description of the
electronic structure for the latter species is still a matter of
discussion, either d7 Fe(I) with two NO ligands31f or d9 Fe-
(-I) with two NO+ ligands.31j Reaction 10 requires a one-
electron reduction which may be accomplished by free NO
or by aqueous ferrous ions.
Conclusions
We present in Scheme 1 an overall description of the
decomposition processes that occur after the one-electron
reduction of NP. The kinetic results show that the dissociation
rates of NO from the penta- or tetracyanonitrosyl ions are
very slow, k-NO ≈ 10-5 s-1, even at pH 7. Evidently, these
rates can hardly account for the rapid vasodilatory action
displayed by NP upon injection in biological fluids. In
contrast to literature reports,12 we show that even the well-
established fast release of the trans-cyano ligand from
[Fe(CN)5NO]3- is not associated with a significant labiliza-
(39) (a) Lin, R.; Farmer, P. J. J. Am. Chem. Soc. 2001, 123, 1143-1150.
(b) Yoshimura, T. Inorg. Chim. Acta 1984, 83, 17-21. (c) Settin, M.
F.; Fanning, J. C. Inorg. Chem. 1988, 27, 1431-1435. (d) Ellison,
M. K.; Schulz, C. E.; Scheidt, W. R. Inorg. Chem. 1999, 38, 100-
108.
(40) Lorkovic, I. M.; Ford, P. C. Inorg. Chem. 1999, 38, 1467-1473.
(41) (a) Lorkovic, I. M.; Ford, P. C. Inorg. Chem. 2000, 39, 632-633.
(b) Lim, M. D.; Lorkovic, I. M.; Wedeking, K.; Zanella, A. W.; Works,
C. F.; Massick, S. M.; Ford, P. C. J. Am. Chem. Soc. 2002, 124, 9737-
9743. (c) Kurtikyan, T. S.; Martirosyan, G. G.; Lorkovic, I. M.; Ford,
P. C. J. Am. Chem. Soc. 2002, 124, 10124-10129.
(42) Olson, L. W.; Schaeper, D.; Lancon, D.; Kadish, K. M. J. Am. Chem.
Soc. 1982, 104, 2042-2044.
Inorganic Chemistry, Vol. 44, No. 8, 2005 2789

Roncaroli et al.
Scheme 1
tively. IR supports the formation of N2O, conclusively, which
correlates with the decomposition of I1. The stoichiometry
for the formation of these products was determined carefully
and found to be [NP]/[N2O] ) 2:1. This is in agreement
with our proposed mechanism for the decomposition of I1.
At the lower pHs, the EPR-silent [Fe(CN)4(NO)2]2-
intermediate becomes a precursor of another dinitrosyl
species, [Fe(CN)2(NO)2]-. We propose it to be a new
member of the well-characterized series of paramagnetic
distorted tetrahedral complexes, {Fe(L)2(NO)2}, with differ-
ent L ligands, described as {Fe(NO)2}9. They are known to
be reversible labile NO carriers effecting the trans-nitro-
sylation processes.16 EPR signals assignable to these dini-
trosyl complexes have been found in tissue of ascite tumors
of mice upon injection with NP.31f Some of them, viz.,
L ) thiolates and imidazole, activate sGC to promote
vasodilation.31h,45
The detection of dinitrosyl compounds as precursors for
disproportionation is of fundamental concern in metal-NO
coordination chemistry, most relevant to the enzymatic
behavior of nitrite- and NO-reductases.15,46 A very recent
report on the NO inhibition of FUR (a non-heme ferric uptake
regulator protein) describes the characterization of two
dinitrosyl species with proposed {Fe(NO)2}8 and {Fe(NO)2}9
structures affording spectroscopic properties surprisingly
similar to those reported here for intermediates I1 and I2.47
Dinitrosyl formation, beyond the mononitrosyl coordination
stage, has been described as being responsible for the
conformational changes associated with NO inhibition of
FUR. Some controversy still exists on related processes that
seem to be present during the activation of sGC leading to
smooth muscle relaxation.8
tion of NO from the metal. A substantial decrease in pH,
down to at least 4, or any other factor favoring the release
of more cyano ligands from [Fe(CN)4NO]2- is required for
the liberation of NO to the medium. Coordinated cyanide is
a strong electron donor because of its free lone pair on the
nitrogen.43 Thus, even in the absence of a decrease in pH,
local situations may provide for protein attachment to
cyanides, as recently proposed,1,44 favoring a complete
decomposition of the complex. This can also be aided by
the ubiquitous presence of appropriate metal ions, such as
copper, as demonstrated in this work.
The fate of the released NO originates from two types of
dinitrosyl intermediates. They have been selectively char-
acterized under different pH conditions. The EPR-silent
trans-dinitrosyl complex [Fe(CN)4(NO)2]2-, described as
{Fe(NO)2},8 is formed through addition of NO to
[Fe(CN)4NO]2- and shows remarkable similarities with a
recently described Fe(TPP)(NO)2 complex, also formed by
coordination of NO to Fe(TPP)NO. Although the latter
complex is stable, the trans-[Fe(CN)4(NO)2]2- complex is
the precursor for disproportionation, regenerating NP and
evolving N2O in a 2:1 mole ratio. This reaction is probably
not significant in biological media relevant to sGC activation
because NO binds to sGC faster (ca. 108 M-1 s-1)8 than it
binds to [Fe(CN)4NO]2- (ca. 104 M-1 s-1, this work).
Under conditions of spontaneous decomposition (no
external NO), the intermediate I1 was characterized through
UV-vis, IR, and EPR spectroscopy. We performed some
preliminary DFT calculations on I1 in which the trans-syn
geometry was obtained, as for related dinitrosyl complexes.
Kinetic evidence (first order in the complex and in NO) also
points to the formation of I1.
Acknowledgment. The authors gratefully acknowledge
financial support from the Deutsche Forschungsgemeinschaft
and the Volkswagenstiftung as well as from the University
of Buenos Aires and the Argentine Government Agencies
ANPCYT and CONICET. We specially thank Prof. Valent´ın
T. Amorebieta for the mass-spectrometric experiments.
J.A.O. and F.R. are a member of the scientific staff and a
fellow of CONICET, respectively.
Supporting Information Available: Additional figures: SI1,
spectral residues of the SPECFIT analysis; SI2-4, IR spectral
changes recorded at pH 7; SI5, IR spectral changes recorded at
pH 4; SI6, UV-vis spectral changes recorded at pH 5.2 with a
SPECFIT output of the kinetic traces; and SI7, chronoamperograms
for the release of NO from reduced NP in the presence of different
metal ions. This material is available free of charge via the Internet
at http://pubs.acs.org.
IC050070C
(45) Vanin, A. F.; Stukan, R. A.; Mabuchkina, E. B. Biochim. Biophys.
Acta 1996, 1295, 5.
Formation of N2O and NP were quantitatiVely detected
through mass spectrometry and chemical analysis, respec-
(46) (a) Wasser, I. M.; de Vries, S.; Moe¨nne-Loccoz, P.; Schro¨der, I.; Karlin,
K. D. Chem. ReV. 2002, 102, 1201-1234. (b) Averill, B. A. Chem.
ReV. 1996, 96, 2951-2964.
(47) D’Autre´aux, B.; Horner, O.; Oddou, J. L.; Jeandey, C.; Gambarelli,
S.; Berthomieu, C.; Latour, J. M.; Michaud-Soret, I. J. Am. Chem.
Soc. 2004, 126, 6005-6016.
(43) Baraldo, L. M.; Forlano, P.; Parise, A. R.; Slep, L. D.; Olabe, J. A.
Coord. Chem. ReV. 2001, 219-221, 881-921.
(44) Kowaluk, E. A.; Seth, P.; Fung, H. L. J. Pharmacol. Exp. Ther. 1992,
262, 916-922.
2790 Inorganic Chemistry, Vol. 44, No. 8, 2005