A unique family of stable and water-soluble S-nitrosothiol complexes

Article abstract of DOI:10.1021/ic7024999

In this work, we present a complete and detailed experimental characterization and theoretical study of a variety of coordinated S-nitrosothiols (RSNOs), such as cysteine derivatives, mercaptosuccinic acid, benzyl thiol, and phenyl thiol. Some of them are extremely unstable and sensitive in free form. Strikingly, in contrast with free S-nitrosothiols, we found that, upon coordination to iridium, they become very stable even in aqueous solutions. The study of these coordinated complexes provides further insight on the elucidation of structural aspects dealing with the nature of the S-N bond in RSNOs, a fact which still remains a matter of controversy.

Full text of DOI:10.1021/ic7024999

Inorg. Chem. 2008, 47, 4723-4733
A Unique Family of Stable and Water-Soluble S-Nitrosothiol Complexes
Laura L. Perissinotti,† Gregory Leitus,‡ Linda Shimon,‡ Dario Estrin,† and Fabio Doctorovich*,†
Departamento de Química Inorgánica, Analítica y Química Física/INQUIMAE, Facultad de
Ciencias Exactas y Naturales, UniVersidad de Buenos Aires, Ciudad UniVersitaria, Pabellón II,
(C1428EHA) Buenos Aires, Argentina, and Department of Chemical SerVices, The Weizmann
Institute of Science, RehoVot 76100, Israel
Received December 26, 2007
In this work, we present a complete and detailed experimental characterization and theoretical study of a variety
of coordinated S-nitrosothiols (RSNOs), such as cysteine derivatives, mercaptosuccinic acid, benzyl thiol, and phenyl
thiol. Some of them are extremely unstable and sensitive in free form. Strikingly, in contrast with free S-nitrosothiols,
we found that, upon coordination to iridium, they become very stable even in aqueous solutions. The study of
these coordinated complexes provides further insight on the elucidation of structural aspects dealing with the nature
of the S-N bond in RSNOs, a fact which still remains a matter of controversy.
Introduction
A long- and well-established method that has been used to
prepare nitrosothiols complexes in situ8–11 consists of the
reaction of a free thiolate anion and the ligand NO+ coordinated
to [Fe(CN)5]3- or other coordinating moiety. This reaction of
nitroprusside with thiols has been extended to a great variety
of nitrosyl complexes of formula type {(X)5MNO}n, with X
comprising ligands of different donor–acceptor abilities and M
) Fe, Ru, and Os. In most cases, these complexes are unstable
and decompose spontaneously to metal complexes and disul-
fides, the lifetimes depending strongly on the thiol structure.12
Recently, in a previous work, we presented the reaction
between a free thiol, benzyl mercaptan, and K[IrCl5NO],
leading to the formation of the surprisingly stable coordinated
nitrosothiol trans-K[IrCl4(CH3CN)N(O)SCH2Ph].
S-nitrosothiols (RSNOs) have been steadily attracting
increasing interest in view of their role in ViVo as potential
biocatalysts and reagents for the storage and transport of
nitric oxide.1 For that reason, the chemistry and biochem-
istry of S-nitrosothiols have been surveyed several
times.2–6
In general, RSNO compounds may be prepared from
treatment of the precursor thiol (RSH) with nitrosating
agents,7 including nitric oxide in the presence of electron
acceptors, nitrosonium salts, nitrous acid, inorganic nitrites,
metal nitrosyl complexes, and many others.5 In the cases
when the nitrosating agent is a metal nitrosyl, the thiol interacts
with the nitrosyl ligand, forming an S-nitrosothiol ligand
(N(O)SR) coordinated to the metal center, {M-N(O)SR}.
CH3CN
K[IrCl5NO] + PhCH2SH
* Corresponding author e-mail: doctorovich@qi.fcen.uba.ar.
trans-K[IrCl4(CH3CN)N(O)SCH2Ph] + H+ + Cl-
†
Universidad de Buenos Aires.
The Weizmann Institute of Science.
‡
This was the first time that a nitrosothiol complex was
isolated and fully characterized, including a structural
determination by X-ray crystallography. This finding repre-
(1) Williams, D. L. H. Acc. Chem. Res. 1999, 32 (10), 869–876.
(2) Jourd’heuil, D.; Hallen, K.; Feelisch, M.; Grisham, M. B. Free Radical
Biol. Med. 2000, 28 (3), 409–417.
(3) Stamler, J. S.; Jia, L.; Eu, J. P.; McMahon, T. J.; Demchenko, I. T.;
Bonaventura, J.; Gernert, K.; Piantadosi, C. A. Science 1997, 276
(5321), 2034–2037.
(8) Butler, A. R.; Calsyharrison, A. M.; Glidewell, C.; Sorensen, P. E.
Polyhedron 1988, 7 (13), 1197–1202.
(4) Stamler, J. S.; Jaraki, O.; Osborne, J.; Simon, D. I.; Keaney, J.; Vita,
J.; Singel, D.; Valeri, C. R.; Loscalzo, J. Proc. Natl. Acad. Sci. U. S. A.
1992, 89 (16), 7674–7677.
(9) Johnson, M. D.; Wilkins, R. G. Inorg. Chem. 1984, 23 (2), 231–235.
(10) Mulvey, D.; Waters, W. A. J. Chem. Soc., Dalton Trans. 1975, 10,
951–959.
(5) Szacilowski, K.; Stasicka, Z. Prog. React. Kinet. Mech. 2001, 26 (1),
1–58.
(11) Szacilowski, K.; Oszajca, J.; Stochel, G.; Stasicka, Z. J. Chem. Soc.,
Dalton Trans. 1999, 14, 2353–2357.
(6) Mannick, J. B.; Hausladen, A.; Liu, L. M.; Hess, D. T.; Zeng, M.;
Miao, Q. X.; Kane, L. S.; Gow, A. J.; Stamler, J. S. Science 1999,
284 (5414), 651–654.
(12) Szacilowski, K.; Wanat, A.; Barbieri, A.; Wasielewska, E.; Witko,
M.; Stochel, G.; Stasicka, Z. New J. Chem. 2002, 26 (10), 1495–1502.
(13) Perissinotti, L. L.; Estrin, D. A.; Leitus, G.; Doctorovich, F. J. Am.
Chem. Soc. 2006, 128 (8), 2512–2513.
(7) Oae, S.; Shinhama, K. Org. Prep. Proced. Int. 1983, 15 (3), 165–
198.
10.1021/ic7024999 CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 11, 2008 4723
Published on Web 05/09/2008

Perissinotti et al.
mL of acetonitrile + 20 µL of dried methanol) was added under
an Ar atmosphere to 10 mg (2.28 × 10-2 mmol) of K[IrCl5(NO)]
in 0.25 mL of acetonitrile. For complexes 3a, 4a, and 7a, a solvent
mixture was used: 0.25 mL of acetonitrile + 20 µL of dried
methanol in the case of 3a and 4a; in the case of 7a, the thiol was
dissolved in a minimum amount of water with a stoichiometric
amount of trifluoroacetic acid. The nitrosyl iridate(III) complex is
sensitive to light, so all of the procedures were carried out under
light protection.
sents a good example of showing that the coordination
chemistry of RSNOs uncovers new aspects of nitrosation
processes; for instance, coordination to metal centers changes
considerably the stability of the RSNOs.5
In this work, we present a complete and detailed experi-
mental characterization and theoretical study of these novel
compounds bearing a variety of coordinated nitrosothiols.
This variety consists of cysteine derivaties, mercaptosuccinic
acid, benzyl thiol, and phenyl thiol. The free forms of some
of them are extremely unstable and sensitive. Strikingly, in
contrast with free S-nitrosothiols, upon coordination, they
become very stable even in aqueous solution.
It is important to remark that the study of these coordinated
complexes is especially relevant because it allows the
elucidation of structural aspects dealing with the nature of
the S-N bond in RSNOs, a fact which still remains a matter
of controversy.
Complex 1a. KH[IrCl5N(O)SCH2Ph]·1/3CH3CN. The reaction
was done at -20 °C; the product precipitated immediately as a
green powder. Reaction yield: 92.0%. 1H NMR (D2O, 500 MHz):
δ 4.63 (s, 2H, -CH2-aryl), 7.34-7.40 (m, 5H, aryl). Elem anal.
found (calcd): C, 15.9 (15.9); N, 3.3 (3.2); H, 1.6 (1.5); S, 5.4 (5.5).
UV–visible [λmax, nm (ꢀ, M-1 cm-1)]: 328(2465), 397(4409),
511(876), 556(855). FTIR (KBr), cm-1: υNO ) 1431, υSN ) 779.
Complex 1b. trans-K[IrCl4(CH3CN)N(O)SCH2Ph] ·CH3CN. The
reaction was done at room temperature. After some minutes, a red
1
product precipitated. Reaction yield: 80.3%. H NMR (D2O, 500
MHz): δ 2.92 (s, 3H, CH3CN-), 4.67 (s, 2H, -CH2-aryl),
7.36–7.37 (m, 5H, aryl). Elem anal. found (calcd): C, 21.6 (21.7);
N, 6.9 (6.9); H, 2.1 (2.1); S, 5.1 (5.3). UV–visible [λmax, nm (ꢀ,
M-1 cm-1)]: 325(1809), 387(3283), 475(1059), 525(919). FTIR
(KBr), cm-1: υNO ) 1432, υSN ) 794.
Experimental Section
General Procedures. Materials. Potassium pentachloronitrosyl
iridate(III) was purchased from Strem Chemicals, benzylmercaptan
and thiophenol from Fluka, mercaptosuccinic acid and N-acetyl-
L-cystein from Merk, and L-cystein ethyl ester from Sigma Aldrich;
all of them were used as received.
Acetonitrile was purchased from Aldrich; it was dried with CaH2
and distilled over it. DMSO-d6 and D2O were purchased from
Aldrich and used as received. The water used in all reactions was
MilliQ water obtained from deionized water.
Complex 1c. trans-(P(Ph)4)[IrCl4(CH3CN)N(O)SCH2Ph]·CH3-
CN. The reaction was done in an analogous manner to that of
complex 1b. Complex 1c was obtained by adding a saturated
aqueous solution of the crude product to an ethanol saturated
solution of tetraphenylphosphonium chloride. FTIR(KBr), cm-1
:
υNO ) 1436, υSN ) 852.
Complex 2a. K2[IrCl5N(O)SPh]. The reaction was done at -20
°C. The product precipitated immediately as a green powder.
Reaction yield: 98.8%. 1H NMR (D2O, 500 MHz): δ 7.49 (d, 2H,
aryl), 7.65–7.76 (m, 3H, aryl). Elem anal. found (calcd): C, 12.2
NMR spectra were acquired on a Bruker AM 500 spectrometer.
All measurements were carried out in D2O and DMSO-d6.
UV/visible absorption spectra were acquired on a Hewlett-
Packard HP8453 diode array spectrometer. The half-lives were
measured in milliQ water (pH ) 4–5, T ) 25 °C, λ ) 475–581
nm) without buffering the solution (pH did not change during the
decomposition).
(12.2); N, 2.5 (2.4); H, 1.0 (1.0); S, 5.4 (5.4). UV–visible [λmax
,
nm (ꢀ, M-1 cm-1)]: 337(2967), 396(4054), 506(849), 556(822).
FTIR(KBr), cm-1: υNO ) 1438, υSN ) 763.
Complex 2b. trans-K[IrCl4(CH3CN)N(O)SPh] ·CH3CN. The
reaction was done at room temperature. After some minutes, a
green-brown product precipitated. Reaction yield: 84.5%. 1H NMR
(D2O, 500 MHz): δ 2.95 (s, 3H, CH3CN-), 7.49 (s, 2H, aryl),
7.65–7.76 (m, 3H, aryl). Elem anal. found (calcd): C, 20.4 (20.2);
N, 7.0 (7.1); H, 1.8 (1.8); S, 5.3 (5.3). UV–visible [λmax, nm (ꢀ,
M-1 cm-1)]: 325(2821), 387(4728), 478(1174), 525(986). FTIR
(KBr), cm-1: υNO ) 1453, υSN ) 786.
FTIR spectra were acquired with a Nicolet Avatar FTIR
spectrophotometer. Solid spectra were obtained as KBr pellets.
Mass spectrometric measurements were performed using a high-
resolution hybrid quadrupole (Q) and orthogonal time-of-flight
(TOF) mass spectrometer (QTof from Micromass, U.K.) operating
in the negative ion electrospray ionization (ESI) mode set from
2100 to 3500 V. Samples dissolved in appropriate solvents were
injected through an uncoated fused-silica capillary, using a syringe
pump (Harvard Apparatus, Pump 11, 15 µL min-1). The temper-
ature of the nebulizer and desolvation gas was set at 100 °C, and
the cone voltage was set between 25 and 35 V. Tandem mass
spectra (ESI-MS/MS) were acquired using the product ion scan
mode via mass selection of the ion of interest, followed by collision-
induced dissociation (CID) with Ar using energies varying from
15 to 35 eV with high-accuracy orthogonal TOF mass analysis of
the CID ionic fragments. For comparison with experimental data,
isotopic patterns were calculated using the MassLynx software.
Coordinated RSNOs are soluble in water, DMSO, and DMF. The
solvent mixture chosen as the most appropriate for the ESI-MS
experiments was a 1:1 mixture of acetonitrile and DMSO. ESI-
MS and ESI-MS/MS spectra were collected using a needle voltage
of 2500 V and a cone voltage of 25 V, and the collision energy
was 15 eV for CID using argon as the collision gas.
Complex 3a. KH[IrCl5N(O)SCH2CH(NHCOCH3)COOH]. The
reaction was done at -10/-20 °C. The product precipitated
1
immediately as a green powder. Reaction yield: 82.0%. H NMR
(D2O, 500 MHz): δ 2.0 (s, 3H, -COCH3), 3.78–3.98 (dd, 2H,
-SCH2-), 4.65–4.70 (m, 1H, -SCH2CH). Elem anal. found
(calcd): C, 9.7 (9.9); N, 4.4 (4.6); H, 1.9 (1.5); S, 5.0 (5.3).
UV–visible [λmax, nm (ꢀ, M-1 cm-1)]: 331(2246), 398(4375),
572(944). FTIR (KBr), cm-1: υNO ) 1431, υSN) 781.
Complex 4a. H[IrCl5N(O)SCH2CH(COOCH2CH3)NH3]. The
reaction was done at -10/-20 °C. The product precipitated
immediately as a green powder. Reaction yield: 78.96%. 1H NMR
(D2O, 500 MHz): δ 1.33 (t, 3H, -COOCH2CH3), 3.88–4.06 (dd,
2H, -SCH2-), 4.36 (m, 2H, -COOCH2-), 4.5 (t, 1H, SCH2CH).
Elem anal. found (calcd): C, 11.0 (10.9); N, 4.9 (5.1); H, 2.3 (2.2);
S, 5.6 (5.8). UV–visible [λmax, nm (ꢀ, M-1 cm-1)]: 327(1836),
400(3235), 581(658). FTIR(KBr), cm-1: υNO ) 1438, υSN ) 784.
Complex 5a. K2[IrCl5N(O)SCH(COOH)CH2COOH] or KH-
[IrCl5N(O)SCH(COOH)CH2COOH]. The reaction was done at
All complexes were synthesized as follows: 2.28 × 10-2 mmol
of corresponding thiol dissolved in 0.25 mL of acetonitrile (for the
case of complexes 3a and 4a, a solvent mixture was used: 0.25
4724 Inorganic Chemistry, Vol. 47, No. 11, 2008

Stable and Water-Soluble S-Nitrosothiol Complexes
-20 °C. Immediately, a green product appeared, but its color
changed to brown in a few minutes. H NMR (DMSO-d6, 500
V ) 3764.1(6) Å3, T ) 120(2) K, Z ) 8, Dc ) 1.178 mg/m3, µ )
7.393 mm-1. Refinement method: full-matrix least squares on F2,
R ) 0.0700, Rw ) 0.1962 (observed data with I > 2σ(I)).
Complex 8. Formula min, Mr ) 537.32, specimen 0.20 × 0.20 ×
0.05 mm, orthorhombic, space group Pbca, a ) 7.5310(2) Å, b )
19.3820(6) Å, c ) 23.7470(7) Å, ꢀ ) 104.463(3)°, F(000) ) 2016,
V ) 3466.25(17) Å3, T ) 120(2) K, Z ) 8, Dc ) 2.059 mg/m3, µ
) 8.549 mm-1. Refinement method: full-matrix least squares on
F2, R ) 0.0576, Rw ) 0.0802 (observed data with I > 2σ(I)).
Theoretical Calculations. All calculations performed in this
work were carried out using the Gaussian03 package.15 Geometry
optimizations were carried out at the PBE1PBE/SDD level of
theory; this functional proved to yield results of higher quality
compared to the previously used mPW1k.16 On the optimized
structures, single-point energy calculations at the PBE1PBE/SDB-
cc-pVDZ level of theory were performed, to increase the accuracy
of the results. Each stationary point in the gas phase was
characterized by performing a normal modes analysis.
1
MHz): δ 2.65–2.72 (m, 2H, -CH2COOH), 4.57–4.59 (m, 1H,
-SCH-). UV–visible [λmax, nm (ꢀ, M-1 cm-1)]: 332(2592),
401(5141), 581(1063). FTIR (KBr), cm-1: υNO ) 1435, υSN ) 781.
Complex 6a. K2[IrCl5N(O)SCH(COOCH3)CH2COOCH3] ·
½CH3CN. The reaction was done at -20 °C. After 15 min, a green
product appeared. Reaction yield: 80.1%. 1H NMR (DMSO-d6, 500
MHz):
δ 2.76–2.95 (m, 2H, -CH2COO-), 3.65 (s, 6H,
-COOCH3), 4.77–4.78 (m, 1H, -SCH-). Elem anal. found (calcd):
C, 12.2 (12.4); N, 2.9 (3.1); H, 1.7 (1.6); S, 5.0 (4.7). UV–visible
[λmax, nm (ꢀ, M-1 cm-1)]: 322(4801), 397(5674), 574(1358).
FTIR(KBr), cm-1: υNO ) 1437, υSN ) 790.
Complex 7a. The reaction was done at -20 °C. To a solution
of nitrosyl iridate(III) complex was added the solution containing
the thiol, resulting in a color change to green. After centrifugation,
the solution became separated into two phases; the aqueous phase
containing the product was poured over cooled acetone, and the
green product precipitated. 1H NMR (D2O (40µL) + CD3CN (0.5
mL), 500 MHz): δ 2.08–2.23 (m, 2H, -CH2CH(NH2)-), 2.38–2.49
(m, 2H, -NH(CO)CH2-), 3.60–3.80 (dd, 2H, -SCH2-), 3.89 (d,
2H, -CH2COOH), 4.10 (m, 1H, -SCH2CH-), 4.73 (m, 1H,
-CH(NH2)COOH). UV–visible [λmax, nm (ꢀ, M-1 cm-1)]: 329, 398,
510, 573. FTIR (KBr), cm-1: υNO ) 1422, υSN ) 770.
Solvent effects were modeled using the polarized continuum
model (PCM)17 scheme. The PCM implementation, in which the
self-consistency between solute wave function and solvent polariza-
tion is achieved during the self-consistent field cycle, has been
employed.
All 15N labeled complexes, that is, K2[IrCl515N(O)SR] and trans-
K[IrCl4(CH3CN)15N(O)SR] ·CH3CN, were obtained in an analogous
manner, using K[IrCl5(15NO)] prepared from Na15NO2 and K3[IrCl6]
according to literature procedures. 14In order to avoid decomposition
of the K[IrCl5(15NO)] complex and increase the obtained yield, a
modification from the literature procedure was made. Instead of
washing the labeled complex with methanol and ether as suggested,
we dissolved it in dry acetonitrile; separated the sodium salt, which
is not soluble in this solvent; concentrated the complex under argon;
and finally precipitated it with dry toluene. The pentachloronitrosyl
iridate(III) complex is light-sensitive, so all of the procedures were
done under light protection.
Results and Discussion
1. Reaction of K[IrCl5NO] with Thiols: Synthesis
and Structure. Free RSNOs are usually thermally unstable
and undergo spontaneous decomposition, while iridium-
coordinated ones turn out to be very stable even in water. It
is known that, for free RSNOs, their stability is affected by
the thiol alkyl chain structure, operating mainly via electronic
effects, through electron-releasing or -withdrawing substit-
uents.5 In this regard, we decided to synthesize different
coordinated RSNOs in order to evaluate the effect of the
thiol structure on their stability and reactivity.
X-ray Analysis for Structures 1b, 1c, and 8. Complex 1b was
crystallized from CH3CN at room temperature to give red crystals.
Crystal 1c was obtained by adding a saturated aqueous solution of
the crude product to an ethanol-saturated solution of tetraphe-
nylphosphonium chloride. After slow evaporation of the supernatant,
green crystals appeared. Complex 8 (orange crystals) was obtained
by leaving complex 2b dissolved in acetonitrile for several weeks.
The crystals were mounted on the nylon loop and flash-frozen in a
nitrogen stream at 120 K. Data were collected on a Nonius
KappaCCD diffractometer mounted on an FR590 generator equipped
with a sealed tube with Mo KR radiation (λ ) 0.710 73 Å) and a
graphite monochromator. The structures were solved using direct
methods with SHELXS-97 and refined by the full-matrix least-
squares technique with SHELXS-97 based on F2.
In all cases, it was found that, when the thiol was added
to an acetonitrile solution of K[IrCl5(NO)] in the dark, at
room temperature, immediate formation of the corresponding
coordinated nitrosothiol was observed as a precipitate, of
which product and color depend on the reactioning thiol and
temperature. All products obtained turned out to be very
soluble in water. Most of them are very stable in the solid
form and in solution.
Acetonitrile selection as reaction solvent lay in the fact
that the reactant complex K[IrCl5NO] did not decompose or
react with the solvent during the reaction time in the absence
of light.
As can be seen in Scheme 1, complexes b have an acetonitrile
molecule coordinated to the trans position. The coordination
of acetonitrile in the trans possition resulted as a consequence
of the labilization of the corresponding chloride, which is
replaced by a molecule of acetonitrile when the reaction takes
place in this solvent.
Complex 1b. Formula min, Mr ) 608.40, specimen 0.3 × 0.3
× 0.02 mm, monoclinic, space group P2(1)/n, a ) 17.548(4) Å, b
) 6.8190(14) Å, c ) 18.261(4) Å, ꢀ ) 115.40(3)°, F(000) ) 1152,
V )1972.8(7) Å3, T ) 120(2) K, Z ) 4, Dc ) 2.047 mg/m3, µ )
7.624 mm-1. Refinement method: full-matrix least squares on F2,
R ) 0.0447, Rw ) 0.0691 (observed data with I > 2σ(I)).
Complex 1c. Formula min, Mr ) 899.67, specimen 0.60 × 0.30
(15) Frisch, M. J. et al. Gaussian03, Revision C.02; Gaussian, Inc.:
Wallingford, CT, 2004. See Supporting Information for remaining
authors.
(16) Quintal, M. M.; Karton, A.; Iron, M. A.; Boese, A. D.; Martin, J. M. L.
J. Phys. Chem. A 2006, 110 (2), 709–716.
(17) Cossi, M.; Barone, V.; Cammi, R.; Tomasi, J. Chem. Phys. Lett. 1996,
255 (4–6), 327–335.
j
× 0.10 mm, triclinic, space group P1, a )20.053(2) Å, b )
7.2500(5) Å, c ) 26.738(3) Å, ꢀ ) 104.463(3)°, F(000) ) 1176,
(14) Bottomley, F.; Clarkson, S. G.; Tong, S. B. J. Chem. Soc., Dalton
Trans. 1974, (21), 2344–2346.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4725

Perissinotti et al.
Scheme 1
Table 1. Energy of Formation for Complexes 1a,b and 2a,b in the Gas
Phase and in Acetonitrilea
complex
energy (kcal/mol)
energy (kcal/mol) PCM
1a
1b
2a
2b
15.36
-18.51
32.10
-39.79
-44.09
-31.59
-34.29
-3.77
a Zero-point and thermal corrections are included.
plex 2b) complex was obtained. When carrying the same
reaction at a lower temperature, clean formation of K2-
[IrCl5N(O)SPh] (complex 2a) also occurred.
In order to gain further insight into the thermodynamic
stability of the different products obtained from each thiol,
DFT calculations were done in the gas phase and in CH3CN
using the PCM model (see Table 1). We have computed the
energy associated with the following reactions:
It should be noted that most of these complexes are easily
obtained in excellent yields (80–99%), except for the case
RS- + [IrCl5NO]- f [IrCl5N(O)SR]2- (complexes a)
(1)
of 5; where unidentified products were also obtained.
Thiols with Aromatic Rings. In the case of thiols with
aromatic rings (benzylmercaptan and phenylthiol), both
products were observed, complex b as a red precipitate and
complex a as a green precipitate (not soluble in acetonitrile).
The ratio of products obtained depends mainly on the reaction
temperature. When the reaction was carried out at room
temperature, the product obtained was only complex b, but
when the reaction was done at low temperatures (-20 °C),
a green precipitate immediately appeared (complex a).
When the reactioning thiol was benzylmercaptan, the
complex trans-K[IrCl4(CH3CN)N(O)SCH2Ph] (complex 1b)
seemed to be easier to obtain and characterize than
K2[IrCl5N(O)SPh] (complex 1a). In fact, after recrystallyza-
tion of 1b from acetonitrile solution, dark red crystals suitable
for X-ray diffraction were obtained13 (Figure 1a).
The tetraphenylphosphonium salt was also obtained; a
saturated aqueous solution of the crude product was added
to an ethanol-saturated solution of tetraphenylphosphonium
chloride. After slow evaporation of the supernatant, green
crystals appeared (1c; Figure 1b). Two molecules of water
per molecule of 1c could be seen in the X-ray structure.
In the case of phenylthiol, when the reaction was done at
room temperature, the trans-K[IrCl4(CH3CN)N(O)SPh] (com-
RS- + [IrCl5NO]- + CH3CN f
trans-[IrCl4(CH3CN)N(O)SR]- + Cl- (complexes b) (2)
The differences in energy in the gas phase and in
acetonitrile indicate that, for both thiols, the formation of
complexes b is thermodynamically favored over that of
complexes a. A more realistic picture is obtained taking into
account solvation effects. While in the gas phase the energy
difference between complexes b and a is 33.8 kcal/mol for
1 and 35.8 for 2, in acetonitrile this difference is much
smaller, being only 4.3 and 2.7 kcal/mol, respectively.
Although complexes a are less stable than complexes b,
at low temperatures it is possible to obtain them in the pure
form due to their rapid precipitation. At room temperature
(∼25 °C), and after a long time (1–2 h), only complexes b
are obtained, and at temperatures within -20 and +25 °C,
mixtures of both a and b products are obtained.
Biologically Relevant Thiols. We have chosen N-acetyl-
L-cysteine and L-cysteine ethyl ester, because of their better
solubility in organic solvents than the parent compound
cysteine. In addition, these thiols are biologically relevant
model thiols. We have recently obtained the coordinated
S-nitrosogluthatione (complex 7a); we have characterized it
1
through UV–visible, FTIR, and H NMR spectroscopy, but
we have not performed a complete characterization yet.
The reactions were done in a mixture of solvents to
improve solubility of the thiols. In the case of N-acetyl-L-
cysteine and L-cysteine ethyl ester, the reaction solvent
consisted of a mixture of dry MeOH and CH3CN (∼20 µL/
0.25 mL). When the reactions were done at low temperatures,
the complexes KH[IrCl5N(O)SCH2CH(NHCOCH3)COOH]
(complex 3a) and H[IrCl5N(O)SCH2CH(COOCH2CH3)NH3]
(complex 4a) were the only products obtained as green
precipitates. When the reactions were done at room temper-
ature, the same products were obtained. These products were
insoluble in acetonitrile and very soluble in water. When
the reaction was done with glutathione, the mixture of solvent
used consisted of water and acetonitrile. A stoichiometric
amount of trifluoracetic acid was used in order to diminish
Figure 1. X-ray crystal structures for complexes 1b with an acetonitrile
molecule as solvate and 1c with two molecules of water. Thermal ellipsoids
at the 50% level.
4726 Inorganic Chemistry, Vol. 47, No. 11, 2008

Stable and Water-Soluble S-Nitrosothiol Complexes
Table 2. Energy of Formation for Complexes 5a and 6a in the Gas
same trend is observed for the selected angles (Table SI1,
Supporting Information).
According to the X-ray data, several noticeable features
arise:
Phase and in Acetonitrilea
complex
energy (kcal/mol)
energy (kcal/mol) PCM
5a
6a
18.96
20.13
-37.29
-37.17
When the counterion is K+, the crystal lattice is less sterically
crowded than when the counterion is the larger PPh4+. This is
consistent with the contracted geometry for complex 1c as
compared to 1b. In addition, in the case of 1c, the complex has
almost a Cs symmetry, while 1b has a lower symmetry (see
Figure 2). The difference in color could be ascribed to the
interaction of two of the chloride ligands, with the acceptor K+
in the case of 1b (Cl--K+: 3.1 Å) and one chloride ligand
with the donor oxygen atoms of two water molecules in the
case of 1c (Cl--OH2: 3.2 Å)(Figure 1).
a Zero-point and thermal corrections are included.
the nucleofilic character of the amines and carboxylic
moieties. The complex isolated as a green powder (complex
7a) turned out to be very stable in the solid form, insoluble
in acetonitrile, and very soluble in water.
Mercaptosuccinic Acid and Mercaptosuccinic Acid
Methyl Ester. The [Fe(CN)5N(SR)O]3- complex (complex
5n, SR ) mercaptosuccinate) is reported to be the most
stable of the ones derived from nitroprusside ([Fe-
(CN)5NO]3-), with a lifetime of 36 h.18 In our case, the
reaction of K[IrCl5NO] with mercaptosuccinic acid dis-
solved in acetonitrile resulted in a very low yield of
product [IrCl5N(O)SCH(COOH)CH2COOH]2- (complex
5a), among other unidentified products. It is possible that
the thiol also reacts with K[IrCl5NO] via the -COOH
moieties, leading to the formation of a byproduct. The
formation of dimers could also occur.
Esterification of the -COOH moieties and reaction with
K[IrCl5NO] in acetonitrile led to the clean formation of
K2[IrCl5N(O)SCH(COOCH3)CH2COOCH3] (complex 6a) as
the only product, which precipitated as a green powder. As
in the case of the cysteine derivatives, this complex was very
soluble in water and insoluble in acetonitrile.
In Table 2 are shown the energies corresponding to the
reaction shown in eq 1 for complexes 5a and 6a from
[IrCl5NO]2- and RS- in the gas phase and in CH3CN. The
values obtained show that the thermodynamic stability in
the gas phase and in solution is almost the same for both
complexes. These results confirm our suggestion that de-
rivatization decreases the reactivity of the product by
blocking side reactions in which the carboxylic moieties may
be involved, but the intrinsic stability is almost the same.
The values also show that formation of the complexes is
more favorable in solution than in the gas phase, as the
product is more strongly solvated than the reactants.
In the case of complex 1c, it is important to notice the
+
way in which the PPh4 cation accommodates; instead of
placing one above the other as in the case PPh4[IrCl5NO],19
it prefers to form a helix via π-π ring interactions with itself
and the complex.
UV–Visible Characterization. In all cases, the UV/visible
absorption consists of four bands: two bands in the UV and
two in the visible region; in Figure 3a, the complete spectra
for complexes 1a (green) and 1b (red) and the reactant
K[IrCl5NO] (ochre) can be seen on a molar absortivities
scale. The complex bands are clearly different from the
reactant ones (Figure 3a, in ochre). In Figure 3b is shown
the UV–visible absorption spectrum for the reactant
K[IrCl5NO], together with the assignation for the bands. The
band assignation for the [IrCl6]-3 complex20 is shown in
Figure 3c as a reference. The extinction coefficients are
shown for each band inside brackets.
As can be seen in Table 3, all complexes exhibit similar and
characteristic spectra. The intense band at 205–260 nm (ꢀ ∼104
M-1 cm-1) is attributed to the π f π* transition in the ligand
and the Ir-Cl ligand-to-metal charge transfer transition. The
next band, centered at about 325–330 nm (ꢀ within 1800–4000
M-1 cm-1) is assigned to an allowed nO f π* transition in the
ligand. There are two other bands in the visible region, one at
about 380–400 nm (ꢀ within 3000–5000 M-1 cm-1) and the
other at about 470–580 nm (ꢀ within 600–1000 M-1 cm-1),
both assigned to metal-to-ligand charge transfer (MLCT)
transitions. DFT calculations predict the highest occupied
molecular orbital (HOMO) to be mainly located on the metal,
but with an important contribution of equatorial chlorides (dxy,
p(Cl)) and the lowest unoccupied molecular orbital (LUMO)
mainly located on the ligand with some choride and acetonitrile
contribution; see Figure 4.
2. Spectroscopic Properties and Structure of S-Nitro-
sothiol Complexes. Crystallographic Structure. We were
able to crystallize the complex bearing 1 with two different
counterions: potassium (K+), when the crystals were obtained
from slow evaporation of an acetonitrile solution (complex
+
1b), and tetraphenylphosphonium (PPh4 ), when the cris-
tallyzation was done in water (complex 1c). It is interesting
to mention that the same anion crystallized with a different
counterion has a different color, being red when the
Consistently with our tentative assignation, the lowest
energy band in the UV–visible spectrum could be qualita-
tively assigned as (MLCT) transition. This last band is
responsible for the color of the S-nitrosothiol; the energy of
this transition is slightly affected by the thiol structure and
by the nature of the solvent. The shoulder observed on this
band can be attributed to the existence of two rotational
counterion is K+ and green for the case of PPh4 . Figure 1
+
shows clearly that both complexes are almost identical from
a structural point of view. In addition, all bond lengths are
shorter for the PPh4+ complex. This fact is in agreement with
the corresponding FTIR frequency values (Table 4). The
(19) Doctorovich, F.; Di Salvo, F.; Escola, N.; Trapani, C.; Shimon, L.
Organometallics 2005, 24 (20), 4707–4709.
(18) Szacilowski, K.; Chmura, A.; Stasicka, Z. Coord. Chem. ReV. 2005,
249 (21–22), 2408–2436.
(20) Lever A. B. Inorganic Electronic Spectroscopy; 2nd ed.; Elsevier: New
York, 1984.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4727

Perissinotti et al.
Table 3. Selected Experimental Bond Lengths (Å) and IR Frequencies for Complexes Bearing 1-6a
IR frequencies
S-N
species
S1-N2
N2-O2
Ir-N1
Ir-N2
O1-N1-S-C
N-O
CN
K[IrCl5NO]
Na[IrCl5NO]
1-NO
1.12019 (1.170)
1.153
(1.222)
1.76019 (1.755)
1.750
200619 (1927)
1986
(1.905)
1500–153023 (1548) - (647)
complex 1a (1.990)
(1.238)
(1.894)
(16.28)
1431 (1491)
779 (727)
complex 1b 1.734(8) (1.855) 1.219(10) (1.242) 1.963(8) (1.925) 2.032(10) (1.990) -0.3(8) (0.335) 1432 (1479)
794 (784) 2327 (2446)
852
787
complex 1c 1.659(13)
complex 1d
1.208(14)
1.944(12)
1.983(13)
1.7(13)
1436
1405
2-NO
(1.918)
(1.216)
(1.224)
(1.229)
(1.211)
(1.240)
1580–167023 (1560) - (649)
complex 2a (2.074)
complex 2b (1.917)
(1.887)
(1.923)
(13.28)
(0.644)
1438 (1453)
1453 (1433)
(1597)
1431 (1483)
1556 (1616)
1438 (1459)
(1567)
763 (688)
786 (756) 2324 (2448)
(622)
781 (730)
637 (649)
784 (720)
(659)
781 (732)
(664)
790 (736)
770
(1.989)
3-NO
complex 3a (1.934)
4-NO
complex 4a (1.999)
5-NO (1.921)
complex 5a (2.007)
6-NO (1.916)
(1.945)
(1.910)
(1.892)
(1.886)
(1.889)
(40.31)
(11.95)
(33.82)
(27.13)
1.769(3)27 (1.980) 1.191(4)27 (1.206)
(1.244)
(1.218)
(1.239)
(1.219)
(1.240)
1435 (1500)
(1562)
1437 (1493)
1422
complex 6a (1.995)
complex 7a
CH3CN
2254 (2326)
a The DFT calculated values are given in parentheses. X-NO, where X ) 1-7 denotes the free S-nitrosothiol derived form the corresponding thiols 1-7.
Note: the FTIR assignation made for 1b in the previous work13 corresponds to complex 1a instead of 1b.
the lowest-energy band changes. In addition, a solvatochromic
effect could also be observed. In the inset of Figure 3a are
shown the spectra for complex 1b in acetonitrile and in acetone;
the lowest-energy band is shifted to the red on going from water
to acetonitrile (573 nm) and acetone (583 nm). Another
interesting point is that there is only one band in that region of
the spectrum, while in water, two bands are observed. The same
features are observed for the other complexes.
FTIR Characterization. The experimental and calculated
structural parameters and IR vibrations for all complexes
studied are listed in Table 4. All complexes studied exhibit
two characteristic IR signals. The FTIR spectra were obtained
in the solid state, and the assignment of the characteristic
Figure 2. X-ray crystal structures for complexes 1b (a) and 1c (b) showing
the symmetry adopted.
isomers of the molecule as have been shown to exist in free
S-nitrosothiols.21 These visible bands present some solvent
dependence, shifting to the blue in the presence of polar
solvents (Figure 3a, inset).
The observed solvatochromic effect can be ascribed to an
interaction of the chloride ligands with the solvent, similarly
to the previously suggested interaction of the cyanide ligands
with the solvent in the case of [Fe(CN)5NO]2-,22 where the
cyanides donate charge to the water molecules.
bands due to the -SNO group was confirmed by preparing
the 15N-labeled derivatives [IrCl515N(O)SR]2-. The difference
spectra revealed isotope-sensitive vibrations at about
1405–1438 cm-1 and 761–794 cm-1. These bands were
attributed to υ(NO) and υ(NS) stretching vibrations, respec-
tively. An important observation is the complete loss of the
υ(SH) stretch at approximately 2600 cm-1 upon the forma-
tion of the S-nitroso complex.
Because of their high reactivity, some of the free RSNOs
studied here have never been isolated, and in some cases,
only a characteristic broad, strong IR band has been assigned
to the stretching vibration of N-O. This υ(NO) stretching
vibration becomes 70–200 cm-1 lower upon coordination.
This trend is well described by the calculations that predict
a 60-157 cm-1 shift toward lower frequencies for the υ(NO)
stretching vibration and 40-137 cm-1 toward higher fre-
quencies for the υ(SN) stretching vibration.
As can be seen in Figure 3a, the spectra for complex 1a
differs slightly from that of complex 1b; the bands in the visible
region are shifted to the blue for complex 1b, and the shape of
(21) Barrett, J.; Debenham, D. F.; Glauser, J. Chem. Commun. 1965, 12,
248.
(22) Estrin, D. A.; Hamra, O. Y.; Paglieri, L.; Slep, L. D.; Olabe, J. A.
Inorg. Chem. 1996, 35 (23), 6832–6837.
(23) Oae, S.; Kim, Y. H.; Fukushima, D.; Shinhama, K. J. Chem. Soc.,
Perkin Trans. 1 1978, 9, 913–917.
(24) Patel, R. P.; McAndrew, J.; Sellak, H.; White, C. R.; Jo, H. J.; Freeman,
B. A.; rley-Usmar, V. M. Biochim. Biophys. Acta 1999, 1411 (2–3),
385–400.
(25) Moynihan, H. A.; Roberts, S. M. J. Chem. Soc., Perkin Trans. 1 1994,
7, 797–805.
(26) Clancy, R.; Cederbaum, A. I.; Stoyanovsky, D. A. J. Med. Chem. 2001,
44 (12), 2035–2038.
Complexes bearing 1 and 2 display special features; when
acetonitrile is coordinated in the position trans to the RSNO
ligand (complex 1b), the peaks that correspond to the
coordinated CN stretching mode vibration are observed at
about 2327 cm-1, while for free acetonitrile, they are at 2254
cm-1. These peaks are not observed for complexes a. The
(27) Yi, J.; Khan, M. A.; Lee, J.; Richter-Addo, G. B. Nitric Oxide 2005,
12 (4), 261–266.
4728 Inorganic Chemistry, Vol. 47, No. 11, 2008

Stable and Water-Soluble S-Nitrosothiol Complexes
Figure 3. (a) UV–vis spectra for product 1a ([1a] ) 3.5 × 10-4 M, green line) and 1b ([1b] ) 5.1 × 10-4 M, red line) in water and reactant K[IrCl5NO]
([K[IrCl5NO]] ) 1 × 10-3 M, ochre line) in acetonitrile. Inset: product 1b in water (red line), in acetonitrile (red dot line), and in acetone (red dashed line).
(b) UV–vis absorption spectra for the reactant complex K[IrCl5NO] in acetonitrile (λmax are depicted in black, extinction coefficients inside brackets). (c)
Spectroscopic assignation for K3[IrCl6] (charge transfer in pink, d-d transitions in green).
Table 4. UV–Visible Spectroscopic Properties of Complexes Bearing 1-6 and the Corresponding Free S-Nitrosothiols in Watera
complex
λ
max, nm (ꢀmax, M-1 cm-1
)
S-nitrosothiol
λ )
max, nm (ꢀmax, M-1 cm-1
1a
1b
2a
2b
3a
4a
5ab
6a
328(2465), 397(4409), 511(876), 556(855)
325(1809), 387(3883), 475(1059), 525sh(919)
337(2967), 396(4054), 506(849), 556(822)
325(2821), 387(4728), 478(1174), 525sh(986.6)
331(2246), 398(4375), 572(944)
327(1836), 400(3235), 581(658)
332(2592), 401(5141), 581(1063)
322(4801), 397(5674), 574(1358)
1
340(1020), 530sh(17), 560(26)23
2
374(1776), 570(53)24
3
4
5
6
334(507), 545(16)25
343(1019), 544(36)26
336(840), 546(13)5
a See Supporting Information for UV–visible spectrum of complex 7a. b Measured in acetonitrile.
longer. Thus, at the PBE1PBE/SDD level of theory, the
degree of S-NO interaction is underestimated. It is interest-
ing to remark that, when adding polarization functions,
geometrical parameters and calculated frequencies matched
worse than when using a base such as SDD (without
polarization functions; see Supporting Information). It is
important to notice some features from this table: upon
coordination, the NO distance increases from 1.17 for the
K[IrCl5SN] to almost 1.24 for the RSNO complexes. The
S-N distance increases for a complexes and decreases for
b ones compared to the free RSNO. Noticeably, the corre-
sponding stretching vibration does not seem to reflect the
changes. This fact could be explained taking into account
that, upon coordination, the new vibration mode involves
not only the S-N bond but also the Ir-NO bond. As
depicted in the table, the Ir-NO bond distance is lower for
the a complexes than for the b ones; as mentioned before,
the opposite trend is observed for the S-N bond distances
in these complexes.
Figure 4. Calculated HOMO and LUMO orbitals for complex 1b.
υ(NO) stretching vibration has almost the same value for
complexes a and b, while the υ(SN) stretching vibration for
complexes a is shifted toward lower values, indicating a
weakening of the SN bond; this finding is consistent with
the calculated thermodynamic stability (Table 2).
In addition, the vibration modes related to S-N and N-O
bonds are very sensitive to a change of counterion. As can
+
be observed in Table 4 for complex 1c (P(Ph)4 is the
counterion), the N-O and S-N frequencies are shifted to
higher values as compared to complexes 1a and 1b (K+ is
the counterion). If the counterion is Cs+ (complex 1d), the
NO frequency is shifted to a lower value (1405 cm-1), while
the SN turned out to be 787 cm-1.
The experimental S-N bonds are about 0.1 Å shorter than
calculated (Table 4), while the calculated N-O bonds are
In complexes b, the calculations predict that the CN
stretching frequency for the coordinated acetonitrile molecule
is shifted 120 cm-1 toward higher frequencies upon coor-
dination. This is consistent with the experimental value of
70 cm-1 toward higher frequencies. As was observed in the
Inorganic Chemistry, Vol. 47, No. 11, 2008 4729

Perissinotti et al.
Table 5. Half-Lives Measured for Complexes of the Type
case of the S–N bond, this shift to higher frequencies does
not correlate with the measured bond distances for free versus
coordinated acetonitrile (1.129 Å versus 1.138 Å (1b) and
1.131 Å (1c)). When acetonitrile is coordinated to iridium,
the CN bond distance is slightly longer than the same
distance for the free one. Thus, in this case, the FTIR
frequencies do not correlate with the X-ray bond distances;
this is probably related to the fact that, upon coordination,
the normal mode includes not only the C-N stretching but
also the Ir-N one.
trans-[IrCl4(CH3CN)N(O)SR]– and [Fe(CN)5N(O)SR]3-, in Water
Solution and in the Absence of Oxygen
τ½ [IrCl5N(O)SR]2-
τ½ [Fe(CN)5N(O)SR]3-
thiol (R)
(min)a
(min)b
1
1a, 1.60 × 104
1b, 1.99 × 104
2a, 1.63 × 103
2b, 5.55 × 103
3a, 7.75 × 103
4a, 2.15 × 102
5a, 6.14
2
3
4
5
6
3n, 7.529
4n, 1.1629
5n, >216629
6a, 9.60
ESI-MS Spectra Determination (Complexes 1, 2, 4).
Short-lived adducts from the reaction of [Fe(CN)5NO]2 with
thiolates, in particular, mercaptosuccinate, have been pre-
viously detected through the ESI-MS technique.28 The
a Isolated complex in MilliQ water; pH ) 5 for 1, 2, and 3; pH ) 4 for
4, 5, and 6 (T ) 25 °C, trans-[IrCl4(CH3CN)N(O)SR]2- ∼ 3 × 10-4 M).
Decomposition was followed at the lowest UV–visible energy band.
b Complex generated in situ from [Fe(CN)5NO]2- ) 2.5 × 10-3 M; thiolate,
1.25 × 10-2 M; Na2CO3/H3BO3; buffer pH ) 10.
{M+ [Fe(CN)5NO(C4H4O4S)]4-}- ion itself has been identi-
3
[IrCl5(NO)]- (m/z 399.800), and other anions related to it
such as [IrCl4]- (m/z 332.850) have been observed as minor
ions (see Supporting Information). The ESI-MS spectra of
these complexes also show that the loss of NO and [IrCl4]-
is hardly noticeably.
fied by MS and also through its decomposition products.
In contrast to the above-mentioned case, where the
complex is an unstable adduct which could only account
for its existence as a minor species, several diagnostic
ions could be observed for the iridium complexes, and
these ions clearly confirm the identity of the organic
ligands and the proposed structures.
1
1H NMR Characterization. S-nitrosation results in H
NMR chemical shift changes for the R and ꢀ hydrogens with
reference to pure thiols. The signals due to the hydrogen
atoms located in the R position are shifted downfield,
approximately by 1 ppm, while the influence on ꢀ hydrogen
atoms depends on the structure of the carbon backbone
(0.1–0.2 ppm).
Clean 1H NMR spectra were obtained for complexes 1a,b,
2a,b, and 3a in D2O. For complexes 1b and 2b, coordinated
CH3CN was clearly observed. In the case of complexes 4a
and 7a, the spectra showed a small amount of disulfide, a
product from decomposition. For complexes bearing 5 and
6, it was not possible to obtain clean spectra in D2O, since
NMR signals attributable to decomposition products im-
mediately appeared and several signals attributable to
coordinated CH3CN were also observed. Cleaner spectra
were obtained in DMSO-d6, where decomposition products
could still be observed but in a lower ratio than in D2O, see
Supporting Information.
3. Stability in Water. Half-Lives Measurements. All
synthesized complexes turned out to be very stable
in the solid form, in the absence of light and oxygen. The
thermal stability in water solution depends strongly on the
thiol structure. The half-lives of these complexes, in the
absence of light, vary from minutes to days even in the
presence of oxygen. In all cases, except for complexes 1b
and 2b, the decomposition obeys first-order kinetics. Oxygen
seems to have no effect on the reaction rate in all cases.
Table 5 depicts the half-lives measured for the iridium
complexes together with previously measured values5,18 for
[Fe(CN)5N(O)SR]3-complexes.
All complexes (1, 2, 4) with coordinated RSNOs provide
clean ESI-MS nearly free of fragment ions.
[IrCl5N(O)R]-2 Complexes ([IrCl5N(O)SR]2- ) M; K+, H+
) Counterions). For all a complexes, [IrCl5(NO)]- or [M -
(RS-)]- (m/z 399.800) and other anions related to it such as
[IrCl4]- (m/z 332.850) have been observed as major ions (see
Supporting Information). Considering the stability of the +3
oxidation state for iridium, we assume that the metal
remained unaltered as Ir(III).
The major diagnostic anions (those that confirm the
identity of these complexes) are [M - Cl]-, [M - Cl -
NO]-•, and [M + K]-. The ESI-MS spectra of the complexes
bearing 1 and 2 (Table SI2, Supporting Information) indicate
that only complexes 1a and 2a are present due to the single
detection of the [Cl5Ir N(O)SR]2- anion; [Cl4(CH3CN)N(O)-
SR]- was not detected.
According to the spectra, complexes 1a and 2a seem to
be more stable in the gas phase than complex 4a. For this
complex, only a very small signal for [M + K]- could be
observed, and the signal due to [M - Cl - K]- was also
very small.
The ESI-MS/MS spectra of these complexes show a loss
of NO and Cl- in all cases. An interesting point to remark
upon is that, for complexes 1a and 2a, the signal related to
the [IrCl4]- ion is hardly noticeable, but in the case of 4a, it
could be seen with considerable intensity. These findings also
suggest the lower stability of 4a in the gas phase when
compared to 1a and 2a.
trans-[IrCl4(CH3CN)N(O)R]-1 Complexes ([IrCl4(CH3-
CN)N(O)SR]1- ) M; K+ or H+) Counterions). For the b
complexes, in contrast to what has been observed for
complexes a, [M]- and [M - CH3CN]- have been observed
as major ions; [IrCl4 (CH3CN) (NO)]•- (m/z 403.797),
Complexes 1a,b are the most stable of all studied
complexes. When dissolved in water, they can be stored in
the dark and at low temperatures for several months.
Complexes 2a,b and 3a are also very stable. In the case of
(28) Lawrance, G. A.; Maeder, M.; Neuhold, Y. M.; Szacilowski, K.;
Barbieri, A.; Stasicka, Z. J. Chem. Soc., Dalton Trans. 2002, 19, 3649–
3655.
(29) Szacilowski, K.; Stochel, G.; Stasicka, Z.; Kisch, H. New J. Chem.
1997, 21 (8), 893–902.
4730 Inorganic Chemistry, Vol. 47, No. 11, 2008

Stable and Water-Soluble S-Nitrosothiol Complexes
Figure 5. Intermolecular stabilization of the coordinated N-acyl-S-
nitrosothiol 3a.
complexes 3a and 4a, it is important to notice the huge
increase in stability of the coordinated S-nitrosothiol as
compared to [Fe(CN)5N(O)SR]3- complexes. Although both
complexes 3a and 4b involve S-nitrosothiols derived from
cysteine, complex 4a is less stable than 3a; the same trend
is observed for nitroprusside complexes; for the latter, it has
been suggested that the formation of a six-membered ring
by the N-acyl thiols in solution probably12 results in a
strengthening of the S-N bond at the expense of a weaken-
ing of the N-O bond. As alluded to above, this effect may
contribute to the increased stability, see Figure 5.
Figure 6. Optimized geometry of [Fe(CN)5N(SR)O]3- (RS- ) mercap-
tosuccinate) in the gas phase, with three molecules of water-solvating
cyanides.
Table 6. Calculated Structural Parameters and NBO Charges for
Optimized Model Complexes
distances (Å)
NBO charges (e)
As previously mentioned, the decomposition of complexes
1b and 2b did not obey first-order kinetics; these complexes
turned out to be more stable in water than the respective a
complexes.
complex
S-N
N-O
Fe/Ir-N
S
N
O
8n
8a
8b
2.244
1.939
1.854
1.220
1.243
1.240
1.759
1.901
1.920
-0.109
0.263
0.403
0.284
0.035
0.014
-0.318
-0.308
-0.269
Complexes bearing 5 and 6 deserve special attention;
contrary to what has been observed for the [Fe(CN)5-
N(O)SR]3- complex 5n, the iridium ones are by far the most
unstable of all, with half-lives of 6 and 9 min, respectively.
In the case of 5n, the presence of the electron-attracting
carbonyl group ꢀ to the S-nitrosothiol group probably
stabilizes the RSNO ligand by making it less prone to
oxidation. However, in the case of 5a, the carbonyl group
competes with the electrophilic [IrCl5]-2 moiety, diminishing
its stabilizing effect.
In order to gain further insight, we have also performed
calculations on [Fe(CN)5N(SR)O]3- (RS- ) mercaptosuc-
cinate), complex 5n. Importantly, we found that this complex
as such is not a minimum in the gas phase. When optimiza-
tion was done, cleavage of the N-S bond occurred, leading
to thiolate and [Fe(CN)5NO]2-. The situation changes by
adding one or more water molecules close to the cyanide
moieties (see Figure 6). The S-N bond distance in this
complex turns out to be 2.17 Å, while the same parameter
for the iridium complexes is 2.00 and 1.99 (Table 4).
We also performed calculations with methylthiol as a
model thiol, for the nitroprusside complex (8n) and iridium
complexes 8a and 8b. As can be observed in Table 6, the
S-N bond distance is 0.3–0.4 Å larger for the nitroprusside
complex as compared to that of 8a. Additionally, the natural
bond order (NBO) charges, especially on the sulfur atom,
show that in the case of complexes 8a and 8b the S-
nitrosothiol is coordinated, but in the case of 8n, the slightly
negative charge on sulfur indicates that, instead of having
the coordinated S-nitrosothiol, the nitroprusside is interacting
with the thiolate.
distribution in the complex.29,30 This in turn increases the
complex stability in the presence of cations forming stable
ion pairs.
1
Observed Decomposition Products through H NMR,
UV–Visible, and X-ray Crystallography. In all cases, as
the decomposition proceeds, disulfide and complexes of the
type [IrCl5-nXnSolv]-(2-n), where the iridium oxidation state
seems to remain unchanged, are formed. The [IrCl5-
nXnSolv]-(2-n) complexes show UV–visible spectra which
are characteristic of iridium complexes, where the oxidation
state is Ir(III), see Figure SI2 in the Supporting Information.31
Only disulfide formation as the organic product was detected
by 1H NMR after complete decomposition (see Figures SI3
and SI4, Supporting Information). In the case of complex 6,
other unidentified products were also observed. In some
cases, iridium-coordinated acetonitrile could also be observed
(Figure SI4, Supporting Information).
In the case of 2b, a crystal suitable for X-ray analysis was
obtained, and cis-[IrCl4(CH3CN)2]K · 2CH3CN (complex 8)
was observed as the decomposition product in acetonitrile
in the absence of light (Figure 7).
It is surprising that the obtained product is the cis and not
the trans bis-acetonitrile as expected, since CH3CN in 2b is
presumably located trans to the RSNO ligand, as in the case
of 1b. This could be rationalized in terms of a mechanism
which involves the [IrCl4(CH3CN)]- intermediate with a
trigonal bipyramid geometry.
4. General Remarks. Up to this point, a complete
characterization and discussion of several properties of these
complexes has been done. As a result of trying to integrate
and rationalize the observed properties, general remarks arise.
These findings again point toward the fact that the solvent
plays an important role in the stability of the n complex; in
fact, it has been suggested that interaction of cations with
complex 5n stabilizes the S-N bond, changing the charge
(30) Szacilowski, K. Chem.sEur. J. 2004, 10 (10), 2520–2528.
(31) Chang, J. C.; Garner, C. S. Inorg. Chem. 1965, 4 (2), 209.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4731

Perissinotti et al.
features. This is consistent with the strong influence of the thiol
structure mentioned previously. All of the complexes turned
out to be more stable than the ironpentacyano ones except for
the case of mercaptosuccinic acid and mercaptosuccinic methyl
ester. This is probably due to the combination of the electrophilic
character of the iridium center and the withdrawing effect of
the carboxylate moieties, as stated before. It is interesting to
remark that, in the case of the ironpentacyano complexes, the
electronegative carboxylates deplete electron density from the
sulfur atom and inhibit electron transfer to the iron center
stabilizing the complex. As a final comment, considering
electronic effects, it can be said that the R substituents that
stabilize coordinated RSNO iridium complexes unstabilize the
ironpentacyano ones.
Figure 7. X-ray crystal structure of cis-[IrCl4(CH3CN)2]K·2CH3CN (7).
Thermal ellipsoids are shown at the 50% level.
On the other hand, it has been suggested that conforma-
tional flexibility could influence compound stability. In
addition, thermal stability may be affected by the steric
interactions involved in the formation of disulfide. It is known
that for free S-nitrosothiols the thermodynamic preferred
conformation is rigidly planar, with the anti conformation
favored. The difference in energy between the syn and anti
conformations appears to be a function of the specific steric
constraints for the moiety. There is, however, a substantial
barrier for the interconversion of these conformers, and
twisting or any disruption of the planarity of this group
destabilizes the fragment toward S-N breaking.Remarkably,
it could be observed that the calculated C-S-N-O torsion
angle is almost zero for 1b and 2b; between 11 and 16° for
complexes 1a, 2a, and 3a; and between 27 and 40° for the
other complexes (see Table 4), indicating more flexibility.
On the basis of this evidence, it is quite possible that the
conformational dynamics of coordinated RSNOs may also
play a role in the stability of these complexes.
Figure 8. Resonance structures for (a) free RSNO and (b) its complexes
with a Lewis acid (A) such as a metal ion.
If we describe the electronic structure of free RSNOs as
a combination of two resonance structures (Figure 8a),32 one
(1) is the conventional RSNO structure with a double N-O
and a single S-N bond and the other (2) is like a zwitterionic
structure that possesses a double S-N and a single N-O
bond. In the case of free RSNOs, an electron distribution
like 2 would be favored by those substituents which donate
electron density, strengthening the S-N bond and weakening
the N-O bond. The opposite effect is achieved with
withdrawing substituents; in this case, the electron density
is shifted within the thiolate moiety due to the inductive effect
of the substituent.33
If we consider the interaction of RSNOs with a Lewis acid,
A (Figure 8b), through its nitrogen atom, the result would
be the same as having an electron-releasing substituent (R).
An electron shift from the lone pairs localized at the sulfur
atom toward A would be favored, thus generating a local
positive charge on the sulfur atom, strengthening the S-N
bond. The observed increase in stability for the coordinated
S-nitrosothiols could be explained in terms of the above-
mentioned facts where A is now the iridate moiety [IrCl5]2-.
The intrinsic stability for each coordinated RSNO is also
related to the structure of the R substituent. If R is an
electron-releasing substituent, the stability would be en-
hanced, whereas the opposite would occur if R is an electron-
withdrawing group.
Although we have not performed a detailed mechanistic
study for the decomposition of the RSNO complexes, the
behavior of the RSNO ligand seems to resemble that of its
parent nitrosothiol in undergoing probably both heterolytic
and homolytic cleavage of the S-N bond: the former is
accompanied by the formation of [IrCl5-nXnNO]-(2-n) and
RS- ions (the inverse reaction of its synthesis), whereas
homolytic rupture leads to different possibilities, as shown
in Scheme 2. Another decomposition pathway for these
complexes could be breakage of the Ir-N bond leading to
free RSNO and [IrCl5-nXnSolv]-(2-n)
.
In Table 7, we show the calculated energies for the possible
decomposition pathways, taking as illustrative examples com-
plexes 1a and 1b.
The results depicted in Table 7 show that, in the case of
a complexes in the gas phase, the S-N heterolytic rupture
is likely to occur to a greater extent (mechanism II), while in
solution, the N-Ir heterolytic rupture, leading to the free
RSNO, seems to be the more feasible pathway. In the case
of b complexes, both in the gas phase and in solution, the
Ir-N heterolytic rupture (mechanism I) seems to be the most
favorable pathway.
As alluded to above, regarding the stability of these com-
plexes in water, we cannot generalize and explain in terms of
a family of compounds because each of them display special
(32) Timerghazin, Q. K.; Peslherbe, G. H.; English, A. M. Org. Lett. 2007,
9 (16), 3049–3052.
ESI-MS results showed that, in the gas phase, both
heterolytic and homolytic decomposition occur. All a
complexes lose NO, leading to the thiol-iridium com-
(33) Arulsamy, N.; Bohle, D. S.; Butt, J. A.; Irvine, G. J.; Jordan, P. A.;
Sagan, E. J. Am. Chem. Soc. 1999, 121 (30), 7115–7123.
4732 Inorganic Chemistry, Vol. 47, No. 11, 2008

Stable and Water-Soluble S-Nitrosothiol Complexes
Scheme 2. Mechanistic Pathways for the Decomposition of Coordinated S-Nitrosothiol Complexesa
a
X: CH3CN molecule; n could be 0 or 1.
Table 7. Calculated Energies for the First Step of the Different Mechanisms in kcal/mola
mechanism I
mechanism II
mechanism III
mechanism IV
complex
gas phase
(PCM) H2O
gas phase
(PCM) H2O
gas phase
(PCM) H2O
gas phase
(PCM) H2O
1a
1b
17.27
2.46
9.65
5.28
-17.28
72.56
37.81
62.68
17.56
19.00
17.66
22.79
32.43
30.80
34.48
39.75
a All values include zero-point corrections.
plexes which could be observed as [M-NO-Cl]•- ions.
The loss of RS• radical leads to ions of the type
[IrCl4NO]•- (362.82 m/z), which were clearly observed
in all cases. While in the gas phase, all of the possible
decomposition pathways seem to be operative, some of
them to a greater extent than others, depending on the
particular complex and kind of product for each complex
(a versus b); in solution, the situation appears to be very
different. When the complexes were dissolved in water,
disulfide (see Supporting Information, Figures SI3–4) and
complexes of the type [IrCl5-nXnSolv]-(2-n) were observed
as decomposition products. We did not find clear evi-
dence for the existence of complexes of the type
[IrCl5-nXnNO]-(2-n) or [IrCl5-nXnSR]-(2-n). It is important
The nature of R appears to be of crucial relevance to the
stability, properties, and type of complex obtained (a vs b).
As a consequence, structural changes in the R group provide
substantial changes in the complex properties. All complexes
turned out to be quite sensitive to environmental effects. For
example, complex 1 crystallized as green or red crystals
depending on the solvent and counterion. The N-O and S-N
bond distances are also very sensitive to changes in the
counterion. In addition, all of the complexes display interest-
ing solvatochromic effects.
A very interesting fact is the obtainment of trans-
[(CH3CN)IrCl4N(O)SR]- complexes; this appears to occur only
for thiols with aromatic rings. In these cases, we were able to
isolate both complexes trans-[(CH3CN)IrCl4N(O)SR]- and
[IrCl5N(O)SR]-2 (R ) benzyl, phenyl). Labilization of the trans
chloride upon ligand coordination may be useful for synthetic
purposes and could give also the possibility to tune the co-
ordinated S-nitrosothiol stability by selecting a proper trans
ligand.
1
to notice that free RSNOs were not observed in the H
NMR spectra as the decomposition proceeded, probably
because their decomposition rate is faster than their
formation rate. In summary, further experiments are
necessary in order to rationalize the decomposition
pathways in water solution in terms of one or various
mechanisms shown in Scheme 2.
Acknowledgment. The authors thank CONICET, UBA,
and ANPCyT for financial support and Professors Marcos
Eberlin for MS analysis, David Milstein for X-ray analysis,
and especially Jan Martin for providing the SDB-cc-pVDZ
basis set.
Concluding Remarks
A complete characterization has been done for a variety
of coordinated S-nitrosothiols. This work shows that the very
rich chemistry of coordinated S-nitrosothiols has its source
in the unique properties that result from their coordination
to the iridium center Ir(III)Cln (n ) 4 or 5). All of the
complexes turned out to be very stable in solid form, and
they can be stored for several months. When dissolved in
water in the absence of light, they display a broad range of
lifetimes indicating very different reactivity and a strong
dependence on thiol structure.
Supporting Information Available: Complete UV–visible,
FTIR, and NMR spectra; ESI-MS; and ESI-MS-MS of the most
important ions; computed geometries; geometrical parameters; and
crystallographic information files (CIF) are included. This material
is available free of charge via the Internet at http://pubs.acs.org.
IC7024999
Inorganic Chemistry, Vol. 47, No. 11, 2008 4733