Widely differing photochemical behavior in related octahedral {Ru-NO} 6 compounds: Intramolecular redox isomerism of the excited state controlling the photodelivery of NO

Article abstract of DOI:10.1021/ic100491g

trans-[(NC)Ru(py)₄(μ-CN)Ru(py)₄(NO)]³⁺ (py = pyridine) is a stable species in aqueous solution. It displays an intense absorption in the visible region of the spectrum (λ_max = 518 nm; ε_max = 6100 M^-1 cm^-1), which turns this compound into a promising agent for the photodelivery of NO. The quantum yield for the photodelivery process resulting from irradiation with 455 nm visible light was found experimentally to be (0.06 ± 0.01)×10 ^-3 mol einstein^-1, almost 3 orders of magnitude smaller than that in the closely related cis-[RuL(NH₃)₄(μ-pz) Ru(bpy)₂(NO)]⁵⁺ species (L = NH₃ or pyridine, pz = pyrazine, bpy = 2,2′-bipyridine; φ_NO = 0.02-0.04 mol einstein^-1 depending on L) and also much smaller than the one in the mononuclear compound trans-[ClRu(py)₄(NO)]²⁺ (φ_NO = (1.63 ± 0.04)×10^-3 mol einstein^-1). DFT computations provide an electronic structure picture of the photoactive excited states that helps to understand this apparently abnormal behavior.

Full text of DOI:10.1021/ic100491g

Inorg. Chem. 2010, 49, 6925–6930 6925
DOI: 10.1021/ic100491g
Widely Differing Photochemical Behavior in Related Octahedral {Ru-NO}6 Compounds:
Intramolecular Redox Isomerism of the Excited State Controlling the Photodelivery of NO
Ariel G. De Candia, Juan P. Marcolongo, Roberto Etchenique, and Leonardo D. Slep*
´
ꢀ
´
´
´
Departamento de Quımica Inorganica, Analıtica y Quımica Fısica, INQUIMAE, Facultad de Ciencias Exactas
ꢀ
y Naturales, Universidad de Buenos Aires-CONICET, Ciudad Universitaria Pabellon 2, 3er Piso, C1428EHA
Buenos Aires, Argentina
Received March 15, 2010
trans-[(NC)Ru(py)4( μ-CN)Ru(py)4(NO)]3þ (py = pyridine) is a stable species in aqueous solution. It displays an
intense absorption in the visible region of the spectrum (λmax = 518 nm; εmax = 6100 M-1 cm-1), which turns this compound
into a promising agent for the photodelivery of NO. The quantum yield for the photodelivery process resulting from irradiation
with 455 nm visible light was found experimentally to be (0.06 ( 0.01) ꢀ 10-3 mol einstein-1, almost 3 orders of magnitude
smaller than that in the closely related cis-[RuL(NH3)4(μ-pz)Ru(bpy)2(NO)]5þ species (L = NH3 or pyridine, pz = pyrazine,
bpy = 2,20-bipyridine; φNO = 0.02-0.04 mol einstein-1 depending on L) and also much smaller than the one in the mono-
nuclear compound trans-[ClRu(py)4(NO)]2þ (φNO = (1.63 ( 0.04) ꢀ 10-3 mol einstein-1). DFT computations provide an
electronic structure picture of the photoactive excited states that helps to understand this apparently abnormal behavior.
Introduction
electronic configurations.5-7 On the basis of recent estima-
tions,7 ∼80% of all known nitrosyl species with n=6 are pro-
bably involved in photoinduced linkage-isomerism processes.
In (aqueous) solution, {M-NO}6 species are also known
to get implicated in photochemical processes but of a differ-
ent kind, leading to the photodissociation of a neutral nitric
oxide molecule and the formation of an M(III) solvento
species, as in eq 1.8,9
Transition metal nitrosyl compounds have become the focus
of many studies during the last 20 years once the physiological
role of nitric oxide was unraveled.1 These apparently simple
molecules are fascinating systems that display diverse reactivity
modes, depending on the coordination geometry and overall
electron counting at the {M-NO}n fragment.2 The electronic
structure and properties of several group 8 octahedral species
with n=6 are compatible with an MII-NOþ limiting electronic
description.3 The interaction of these species with light in the
solid state at low temperature has been described since the
1970s. These studies triggered the discovery of the light-induced
linkage isomerism of the metal nitrosyl bond.4 The activity in
this field has been remarkable, and nowadays the family of
compounds displaying this sort of behavior has grown to several
dozens, including examples with other metal centers and
hν
n
n
½X5M- NOꢁ
½X M- OH ꢁ þ NO•
ð1Þ
f
5
2
þ H2O
(5) (a) Fomitchev, D. V.; Furlani, T. R.; Coppens, P. Inorg. Chem. 1998, 37,
1519–1526. (b) Cheng, L.; Novozhilova, I.; Kim, C.; Kovalevsky, A.; Bagley, K. A.;
Coppens, P.; Richter-Addo, G. B. J. Am. Chem. Soc. 2000, 122, 7142–7143. (c)
Kim, C.; Novozhilova, I.; Goodman, M. S.; Bagley, K. A.; Coppens, P. Inorg. Chem.
2000, 39, 5791–5795. (d) Boulet, P.; Chermette, H.; Weber, J. Inorg. Chem. 2001, 40,
7032–7039. (e) Coppens, P.; Novozhilova, I.; Kovalevsky, A. Chem. Rev. 2002, 102,
861–883. (f) Guida, J. A.; Ramos, M. A.; Piro, O. E.; Aymonino, P. J. J. Mol. Struct.
2002, 609, 39–46. (g) Kovalevsky, A. Y.; Bagley, K. A.; Coppens, P. J. Am. Chem.
Soc. 2002, 124, 9241–9248. (h) Kovalevsky, A. Y.; Bagley, K. A.; Cole, J. M.;
Coppens, P. Inorg. Chem. 2003, 42, 140–147.
(6) Boulet, P.; Buchs, M.; Chermette, H.; Daul, C.; Furet, E.; Gilardoni,
F.; Rogemond, F.; Schlapfer, C. W.; Weber, J. J. Phys. Chem. A 2001, 105,
8999–9003.
(7) Bitterwolf, T. E. Coord. Chem. Rev. 2006, 250, 1196–1207.
(8) (a) Wolfe, S. K.; Swinehart, J. H. Inorg. Chem. 1975, 14, 1049–1053. (b)
De Leo, M.; Ford, P. C. J. Am. Chem. Soc. 1999, 121, 1980–1981. (c) Adachi, H.;
Sonoki, H.; Hoshino, M.; Wakasa, M.; Hayashi, H.; Miyazaki, Y. J. Phys. Chem.
A 2001, 105, 392–398. (d) Togniolo, V.; da Silva, R. S.; Tedesco, A. C. Inorg.
Chim. Acta 2001, 316, 7–12. (e) Videla, M.; Braslavsky, S. E.; Olabe, J. A.
Photochem. Photobiol. Sci. 2005, 4, 75–82.
*To whom correspondence should be addressed. E-mail: slep@qi.fcen.uba.ar.
(1) (a) Ignarro, J. L. Nitric Oxide in Biology and Photobiology; Academic
Press: San Diego, CA, 2000. (b) Stamler, J. S.; Singel, D. J.; Loscalzo, J. Science
1992, 258, 1898–1902. (c) Richter-Addo, G. B.; Legzdins, P. Metal Nitrosyls;
Oxford University Press: New York, 1992.
(2) (a) Enemark, J. H.; Feltham, R. D. Coord. Chem. Rev. 1974, 13, 339–406.
(b) Enemark, J. H.; Feltham, R. D. Top. Stereochem. 1981, 12, 155–215. (c) Olabe,
J. A.; Slep, L. D. Reactivity and Structure of Complexes of Small Molecules: Nitric and
Nitrous Oxide. In Comprehensive Coordination Chemistry II, from Biology to
Nanotechnology; Mc Cleverty, J. A., Meyer, T. J., Eds.; Elsevier: Oxford, U.K., 2004;
Vol. 1, pp 603-623.
(3) Roncaroli, F.; Videla, M.; Slep, L. D.; Olabe, J. A. Coord. Chem. Rev.
2007, 251, 1903–1930.
(4) (a) Carducci, M. D.; Pressprich, M. R.; Coppens, P. J. Am. Chem. Soc.
1997, 119, 2669–2678. (b) Coppens, P.; Fomitchev, D. V.; Carducci, M. D.; Culp,
K. J. Chem. Soc., Dalton Trans. 1998, 865–872.
(9) Tfouni, E.; Krieger, M.; McGarvey, B. R.; Franco, D. W. Coord.
Chem. Rev. 2003, 236, 57–69.
r
2010 American Chemical Society
Published on Web 06/25/2010
pubs.acs.org/IC

6926 Inorganic Chemistry, Vol. 49, No. 15, 2010
De Candia et al.
It has not been until recently that the advantages of using
coordination species behaving as cage compounds for the
photodelivery of biomolecules were acknowledged.10,11 Co-
ordination compounds are thermally stable; they can be
designed to display high extinction coefficients in the visible
region of the spectrum and high quantum yields for the
photodissociation process. The ubiquitous role of NO in
many biochemical processes together with the photochemical
reactivity described by eq 1 for {M-NO}6 species fueled-up
the work with nitrosyl containing transition metal complexes
in pharmacology.9,11,12 Unluckily, most of the {M-NO}6
group 8 octahedral species normally display absorption
profiles shifted to the UV region of the spectrum because
of the stabilization of the metal centered orbitals that arises
from strong metal-nitrosyl π-interactions. This feature pre-
cludes to some extent the use of these molecules with living
tissues that are sensitive and opaque to irradiation with low
wavelength light sources. Two approaches are described in
the literature to cope with this complication: substitution of
the metal centers to explore other electronic configurations13
or modifications of the coordination sphere of the {M-NO}6
complexes to broaden the absorption spectrum to the visible
region.14 This can be performed by means of conjugated
heterocyclic ligands that can act as light antennae or employ-
ing a second metal center that introduces long-range charge-
transfer excited states.15,16 We focus here on this last alter-
native. A recent set of publications15,16 describes the photo-
delivery of an NO molecule from a set of pyrazine bridged
dinuclear species of general formula cis-[RuL(NH3)4(μ-pz)-
Ru(bpy)2(NO)]5þ (pz=pyrazine, bpy=2,20-bipyridine, L=
pyridine or NH3 (compound 15þ)). These compounds display
quantum yields for the photodelivery of NO (φNO) close to
0.03-0.04 mol einstein-1 depending on L and red-shifted ab-
sorption profiles with high molar absorbance in the visible
region. In these reports,15,16 the lowest energy excited state res-
ponsible for the experimentally found photochemical beha-
vior was tentatively described as one of the two possible elec-
tronic isomers cis-[RuIIL(NH3)4(μ-pz)RuIII(bpy)2(NO•)]5þ or
Some time ago, some of us reported the preparation
and characterization of trans-[(NC)Ru(py)4(μ-CN)Ru(py)4-
(NO)]3þ (py=pyridine, 23þ).17 This dinuclear complex dis-
plays an intense absorption in the visible region (λmax
=
518 nm, ε518=6100 M-1 cm-1) arising from a charge transfer
process involving [Ru(py)4(CN)2] as the donor fragment and
[Ru(py)4(NO)]3þ as the acceptor. We now explored the
photochemical release of NO in our CN-bridged species.
To our surprise, and despite the similarities with the
pz-bridged species, we measured a rather small φNO ≈ 10-5
mol einstein-1, even when the mononuclear trans-[ClRu-
(py)4(NO)]2þ species (32þ) itself photoreleases NO much
more efficiently (see later). This initially disappointing result
prompted us to explore in greater detail the electronic
structure of the lowest energy excited state of the dinuclear
compounds. Our findings indicate important differences in
the electronic distribution and constitute a piece of warning
for the future design of NO releasing compounds.
Experimental Section
Materials. The compounds trans-[RuCl(py)4(NO)](PF6)2 and
trans-[(NC)Ru(py)4(μ-CN)Ru(py)4(NO)](PF6)3 were prepared
according to previously published procedures17,18 and fully
1
characterized by IR (KBr pellets), H- and 13C-NMR spectro-
scopies, and elemental analyses. All other reagents employed in
this work were obtained commercially and used as supplied.
Physical Determinations. Microanalytical data for C, H, and
N were obtained with a Carlo Erba EA 1108 analyzer. IR
spectral measurements (KBr pellets) were carried out using
alternatively one of two Fourier transform (FT) spectrophoto-
meters, a Nicolet 150P and a Thermo Nicolet AVATAR 320.
1H- and 13C-NMR spectra were measured with a 500 MHz
Bruker AM 500 spectrometer. UV-vis spectra were recorded
with an HP 8452A diode array spectrophotometer in a gastight
stoppered 1 cm path-length quartz cuvette cell. Visible light
irradiation of the samples was performed with a 455 nm, 5 mm
diameter light-emitting diode (LED) with a spectral half-width
of 20 nm. The light power of this source (1.44 ꢀ 10-8 einstein s-1
cm-3) was calibrated in a test photolysis of cis-[Ru(bpy)2-
(py)2]2þ (quantum yield of photosubstitution =0.26).19
The quantum yields for the photorelease of nitrogen mon-
oxide were determined by measuring the concentration of NO at
low conversion. The quantitative determination was performed
with a commercial NO detector (inNO, Nitric Oxide Measuring
System, Harvard Apparatus GmbH). The system consisted
of an 800 mV potentiostat (TEQ-03) and a combined electrode
(700 μm diameter, amino series of nitrogen monoxide sensors).
Calibration of the electrode in the range of 10-400 nM was
performed by generating NO according to the reaction in eq 2,
under an Ar atmosphere
cis-[RuIIIL(NH3)4(μ-pz)RuII(bpy)2(NO•)]5þ
.
(10) (a) Salierno, M.; Marceca, E.; Peterka, S.; Yuste, R.; Etchenique, R.
J. Inorg. Biochem. 2010, 104, 418–422. (b) Fino, E.; Araya, R.; Peterka, D. S.;
Salierno, M.; Etchenique, R.; Yuste, R. Front. Neural Circuits 2009, 3, 2. (c) Rial
Verde, E. M.; Zayat, L.; Etchenique, R.; Yuste, R. Front. Neural Circuits 2008
2, 2. (d) Zayat, L.; Noval, M. G.; Campi, J.; Calero, C. I.; Calvo, D. J.; Etchenique,
R. ChemBioChem 2007, 8, 2035–2038. (e) Zayat, L.; Salierno, M.; Etchenique,
R. Inorg. Chem. 2006, 45, 1728–1731. (f) Nikolenko, V.; Yuste, R.; Zayat, L.;
Baraldo, L. M.; Etchenique, R. Chem. Commun. 2005, 1752–1754. (g) Zayat, L.;
Calero, C. I.; Albores, P.; Baraldo, L. M.; Etchenique, R. J. Am. Chem. Soc. 2003,
125, 882–883.
(11) Rose, M. J.; Mascharak, P. K. Curr. Opin. Chem. Biol. 2008, 12, 238–
244.
1
H2SO4 þ NaNO2 þ NaI f NO þ I2 þ H2O þ Na2SO4 ð2Þ
2
(12) Ford, P. C.; Bourassa, J.; Miranda, K.; Lee, B.; Lorkovic, I.; Boggs,
S.; Kudo, S.; Laverman, L. Coord. Chem. Rev. 1998, 171, 185–202.
(13) (a) Eroy-Reveles, A. A.; Leung, Y.; Beavers, C. M.; Olmstead,
M. M.; Mascharak, P. K. J. Am. Chem. Soc. 2008, 130, 4447–4458. (b) Ford,
P. C.; Wecksler, S. Coord. Chem. Rev. 2005, 249, 1382–1395. (c) Hoffman-Luca,
C. G.; Eroy-Reveles, A. A.; Alvarenga, J.; Mascharak, P. K. Inorg. Chem. 2009,
48, 9104–9111.
(14) (a) Halpenny, G. M.; Mascharak, P. K. Inorg. Chem. 2009, 48, 1490–
1497. (b) Rose, M. J.; Mascharak, P. K. Chem. Commun. 2008, 3933–3935. (c)
Rose, M. J.; Mascharak, P. K. Inorg. Chem. 2009, 48, 6904–6917. (d) Rose, M. J.;
Olmstead, M. M.; Mascharak, P. K. Polyhedron 2007, 26, 4713–4718. (e) Rose,
M. J.; Patra, A. K.; Alcid, E. A.; Olmstead, M. M.; Mascharak, P. K. Inorg. Chem.
2007, 46, 2328–2338.
For each calibration, aliquots (20 μL) of aqueous NaNO2
(∼10 μM) were added to 3.0 mL of a 0.03 M solution of NaI in
0.1 M H2SO4. Cronoamperograms were registered at a fixed
temperature (25.0 ( 0.1 ꢀC, Lauda RC 20 thermostat) while
stirring the solution in order to maintain a constant rate of
oxidation of the produced NO at the electrode surface. The
typical sensitivity of the electrode was about 170 pA/nM. For
the quantum yield determinations, chronoamperograms were
(17) Roncaroli, F.; Baraldo, L. M.; Slep, L. D.; Olabe, J. A. Inorg. Chem.
2002, 41, 1930–1939.
(18) Bottomley, F.; Mukaida, M. J. Chem. Soc., Dalton Trans. 1982,
1933–1937.
(19) Pinnick, D. V.; Durham, B. Inorg. Chem. 1984, 23, 1440.
(15) Sauaia, M. G.; de Lima, R. G.; Tedesco, A. C.; da Silva, R. S. J. Am.
Chem. Soc. 2003, 125, 14718–14719.
(16) Sauaia, M. G.; de Lima, R. G.; Tedesco, A. C.; da Silva, R. S. Inorg.
Chem. 2005, 44, 9946–9951.

Article
Inorganic Chemistry, Vol. 49, No. 15, 2010 6927
registered upon irradiation of 3.0 mL of an acidified (HCl 0.01
M) aqueous solution of each complex in steps of 10-120 s. Each
experiment was carried out under Ar. During the photolysis
measurements, the NO sensor was positioned outside the light
path to avoid any photoelectric interference.
Theoretical Calculations. We employed density functional
theory (DFT) computations to fully optimize the geometries
of the ground singlet and first-excited triplet states of 15þ, 23þ
,
and 32þ in vacuum, without symmetry constraints. In order to
test some of the conclusions, the computations were also per-
formed on a small set of mononuclear ruthenium-nitrosyl species
that also undergo efficient photodissociation of NO, namely,
trans-[Ru(pz)(NH3)4(NO)]3þ and trans-[Ru(py)(NH3)4(NO)]3þ
.
We also included cis-[(pz)Ru(bpy)2(NO)]3þ, a fragment related
to 15þ. The calculations were carried out with the Gaussian 03
package,20 at the B3LYP level of theory using restricted and
unrestricted approximations of the Kohn-Sham equations
depending on the total number of electrons. In all cases, we
employed the effective core potential basis set LanL2DZ, tight
SCF convergence criteria, and default settings in the geometry
optimizations. This combination proved to be suitable for
geometry predictions in other ruthenium centered nitrosyl
species.21,22 The true nature of the stationary points, which
correspond to (local) minima in the potential energy surface,
was confirmed bynumerical vibrational frequency computations.
In all cases the absence of negative frequencies confirmed that the
optimized geometries correspond to stable configurations in the
potential energy surfaces. The spin density of the triplet states was
calculated on the basis of the density matrix given by Gaussian 03.
Time dependent (TD)DFT computations at the ground state
equilibrium geometry involving at least 70 singlet and 70 triplet
excited states were employed as an assistant tool in the inter-
pretation and assignment of the electronic spectra and photo-
chemical properties.
Figure 1. Electronic spectra of 32þ (dashed line) and 23þ (solid line) at
the actual concentrations used in the photolysis experiments (3.4 ꢀ 10-4
M and 1.5 ꢀ 10-5 M in aqueous solution, respectively). Inset: Chrono-
amperograms for the photorelease of NO of 32þ (dashed line) and 23þ
(solid line) upon irradiation steps of 20 s. The typical sensitivity of the NO
detector was about 170 pA/nM.
reliable measurements at low conversion. Figure 1 displays
the chronoamperogram of NO released by photolysis of a
1.5ꢀ10-5 M solution of 23þ. The experimental φNO result
is (0.06 ( 0.01) ꢀ 10-3 mol einstein-1, a value that is sur-
prisingly smaller than the typical ones measured in other
{RuNO}6.
We attempted to compare this value with the one in 32þ
,
but to our knowledge this system was never studied quanti-
tatively in solution. A recent publication dealing with the
light-induced linkage isomerism of 32þ suggested that even in
the solid state this compound may undergo photodecompo-
sition, probably due to the photorelease of NO.23 Therefore
we also explored this molecule under the same conditions
employed for the dinuclear compound. In this case, the
455 nm irradiation wavelength (see Figure 1) overlaps with
a weak absorption band centered at 453 nm (ε453=140 M-1
cm-1). This same transition was observed in acetonitrile
Further explorations of the potential energy surface (PES) of
the triplet states were done using the same functional and level of
theory as in the optimization jobs. In each run, the Ru-N bond
˚
length was increased in steps of 0.02 to 0.2 A starting from the
equilibrium distance. The rest of the internal degrees of freedom
were then fully optimized. The dissociation curves obtained for
the different compounds allow us to correlate qualitatively the
dissociation energy (De) to the photochemical quantum yield
determined experimentally.
solution (λmax=450 nm, ε450=150 M-1 cm-1 24
) and ascribed
to a metal to ligand dRu f π*NO charge transfer transition
(see later). The observed quantum yield for product forma-
tion obtained spectroscopically was in this case 30 times
higher reaching (1.63 ( 0.04) ꢀ 10-3 mol einstein-1. This
value, though still small for any practical applications, lies in
the lower end of the usual range observed in other mono-
nuclear {RuNO}6 systems,9,25 suggesting that the [Ru(py)4-
(NO)]3þ moiety is not responsible by itself for the overall
Results and Discussion
The deeply violet dinuclear compound23þ owes its color to
an intense Gaussian-shaped absorption band (λmax=518 nm;
ε
max=6100 M-1 cm-1, Figure 1). We already reported that
thisspecies seems toreleaseNO upon irradiation inthe visible
region of the spectrum, but only now, and inspired by the
photochemical properties of 15þ we decided to examine this
process quantitatively. We explored the photochemistry in
aqueous solution irradiating at 455 nm on the upper energy
side of the absorption band (ε455 = 3335 M-1 cm-1) and
employing low light power to avoid side reactions. The
quantum yield for NO-formation (φNO) was obtained by
electrochemically sensing the NO produced upon irradiation,
a technique already employed with many nitrosyl com-
pounds. This is a simple and fast method for the detection
and quantification of the NO in the 10-9 M scale, allowing
decrease in φNO
.
At this point there is a crucial issue that has to be considered
in order to explain the differences in the quantum yields of NO
photorelease, namely, the electronic description of the lowest
energy excited states of the different compounds involved in the
primary photochemical process. In the past we have success-
fully employed density functional theory (DFT) electronic
structure calculations to understand the photochemical beha-
vior of [Ru(bpy)(AN)4]2þ (bpy =2,20-bipyridyl, AN=aceto-
nitrile).22 We apply here the same methodology to explore the
electronic structure of the three species in an attempt to evaluate
(20) Frisch, M. J. Gaussian 03, versions C.02 and D.01, Gaussian Inc.:
Wallingford, CT, 2004.
(21) (a) Videla, M.; Jacinto, J. S.; Baggio, R.; Garland, M. T.; Singh, P.;
Kaim, W.; Slep, L. D.; Olabe, J. A. Inorg. Chem. 2006, 45, 8608–8617. (b) De
Candia, A. G.; Marcolongo, J. P.; Slep, L. D. Polyhedron 2007, 26, 4719–4730.
(22) Petroni, A.; Slep, L. D.; Etchenique, R. Inorg. Chem. 2008, 47, 951–
956.
(23) Schaniel, D.; Cormary, B.; Malfant, I.; Valade, L.; Woike, T.; Delley,
B.; Kramer, K. W.; Gudel, H. U. Phys. Chem. Chem. Phys. 2007, 9, 3717–
3724.
(24) Coe, B. J.; Meyer, T. J.; White, P. S. Inorg. Chem. 1995, 34, 593–602.
(25) Rose, M. J.; Mascharak, P. K. Coord. Chem. Rev. 2008, 252, 2093–
2114.

6928 Inorganic Chemistry, Vol. 49, No. 15, 2010
De Candia et al.
Figure 2. Dissociation curves calculated by DFT for the lowest energy
excited triplets of the three compounds (15þ, 23þ, and 32þ). See the
Experimental Section for details in the calculations.
Figure 3. MO diagrams showing the most relevant frontier orbitals for
15þ, 23þ, and 32þ. The arrows identify the main orbital contributions to
the lowest energy intensity-carrying electronic transitions. For the sake of
clarity, the orbitals that are not relevant to the discussion of the photo-
reactivity were omitted.
the intrinsic differences between 32þ, 23þ, and 15þ that justify
φNO values that span a 2 orders of magnitude range.
The ground state geometry optimization of the three species
renders essentially linear Ru-NO fragments with typical
geometrical parameters for octahedral species (Supporting
Information). The evaluation of the photochemistry also
requires a geometry optimization of the lowest energy excited
state and a complementary potential energy surface (PES)
scan of the relevant internal coordinates (in this case the
Ru-NO bond-length) in order to evaluate the dissociation
of the NO moiety.
Figure 2 shows the dissociation curves of the lowest energy
triplet states of the three species, calculated in vacuum. The
dissociation energy follows the trend 15þ < 32þ < 23þ. The
strength of the Ru-N bond is also reflected in the excited
state computed equilibrium Ru-NO bond-lengths (Support-
ing Information). Though the energy values are probably
overestimated because of the absence of solvent in the
computations, the results are still valuable from the qualita-
tive point of view. In this set of compounds it appears that the
rupture of the Ru-N bond determines the quantum yield of
photorelease. The computed dissociation energy (De) in other
mononulcear {RuNO}6 species known to also undergo
efficient photodissociation9 (see the Supporting Information)
are comparable to the values obtained for 15þ and 32þ. This
suggests that the abnormally low φNO (or high De) of 23þ does
not originate on a subtle effect but most probably on a large
difference in the electronic structure of its photoreactive state.
The photoreactivity of this kind of species has been ratio-
nalized in the past as arising from the presumably long-lived
lowest energy excited state assumed to be mostly RuIII-NO•
in character. Figure 3 collects the relevant frontier orbitals in
the three species. In all cases, the main contribution to the
lowest unoccupied molecular orbital (LUMO) comes from the
nitrosyl π*NO, with partial delocalization onto the metal
centered orbitals. A lowest energy excited state with some
degree of charge transfer to the nitrosyl fragment is therefore a
reasonable choice to start the analysis. The nature and energy
spacing of the highest occupied molecular orbital (HOMO)
and other occupied MOs close in energy are, on the contrary,
markedly different. This fact will prove determinant when
exploring the excited state behavior. In none of the three cases
Figure 4. Spin density for the lowest energy triplet states at ground state
(top) and at the excited state equilibrium geometry (bottom) of the three
compounds (15þ, 23þ, and 32þ).
the nature of the thermally equilibrated lowest energy excited
state is straightforward. We compare here the three cases.
trans-[ClRu(py)4(NO)]2þ. (TD)DFT computations at
the ground state equilibrium geometry suggest that the
lowest energy absorption band at 450 nm results from the
convolution of two weak charge tranfer transitions. The first
one is dRu f π*NO in origin and involves orbitals 99 and 103/
104 (calculated λmax = 508 nm, f = 0.0015). The second one
corresponds to the HOMO f LUMO transition and can
be viewed as a πCl f π*NO charge transfer (calculated λmax
=
438 nm, f=0.0020). In any case, intersystem crossing occur-
ring after the absorption of light populates a triplet state,
which is mostly dRu f π*NO in origin, as can be seen on a
plot of the computed spin density of the lowest excited triplet
state at the ground state equilibrium geometry (Figure 4).
The orbitally degenerate excited state molecule undergoes a
Jahn-Teller distortion6 that removes the orbital degeneracy
and leads to bending of the Ru-NO unit and lengthening
˚
of the N-O bond (—RuNO 179.5ꢀ and dRu-N=1.77 A at the
˚
ground state (GS) equilibrium geometry; dRu-N=1.95 A and
—
RuNO=141.6ꢀ at the minimum of the PES of the 3MLCT
state). In parallel with these geometrical changes, the species
undergoes electronic rearrangements that influence the total
spin density (Figure 4). This geometrical and electronic changes

Article
Inorganic Chemistry, Vol. 49, No. 15, 2010 6929
leave an NO• molecule weakly coordinated to a Ru(III) center.
Dissociation of this fragment occurs via further lengthen-
ing of the Ru-N bond. We observed the same kind of
behavior for the computed lowest energy triplet state of other
mononuclear {RuNO}6 species (Supporting Information).
trans-[(NC)Ru(py)4(μ-CN)Ru(py)4(NO)]3þ. The pre-
sence of an extra ruthenium center introduces new transitions
in the visible region of the spectrum. The lowest energy
absorption band responsible of its deep violet color has
been tentatively associated to a long-range donor-acceptor
charge transfer transition (DACT).17 A molecular orbital
picture built upon DFT electronic structure computations
confirms this assignment (Figure 4). The t2g orbital set of the
metal centers splits under C4 symmetry into b2 and e. The
HOMO (orbital 203) is nonbonding, transforms as b2 and is
mostly located on the dicyano fragment. Immediately below
in energy is a first set of degenerate e orbitals essentially
centered on the same fragment (orbitals 201 and 202). The
LUMO is located on the other end of the molecule and is
largely π*(NO) in character (orbitals 204 and 205). Both
b2 feandefe transitions are symmetry allowed. However,
the orientation of the b2 orbital of the dicyano fragment
makes the former transition overlap forbidden, leaving a
unique assignment for the strong visible band. The lowest
energy triplet state at the GS equilibrium geometry still arises
from a long distance CT but involving a different metal
orbital on the donor fragment. Both (TD)DFT and single
point computations of the triplet state show that this state is
actually b2 f e in character. Jahn-Teller distortion during
the lifetime of the excited state leads to lengthening of the
Ru-NO bond and bending of the Ru-NO angle. The com-
puted relaxed triplet state geometry shows an NO mole-
becomes responsible for the richer electronic spectro-
scopy. (TD)DFT computations performed at the ground
state geometry in vacuo indicate that the strong absorp-
tion in the visible region of the spectrum, observed experi-
mentally at 530 nm, is better described as a dπ((H3N)5-
RuII) f π*pz MLCT (λcalcd = 449 nm; f = 0.420). The
electronic distribution of this state can therefore be
described as cis-[(H3N)5RuIII(μ-pz•-)RuII(bpy)2(NO)]5þ
.
However, the lowest energy excited triplet state dis-
plays a completely different electronic distribution, better
described as cis-[(H3N)5RuIII(μ-pz)RuII(bpy)2(NO•)]5þ
(Figure 4). This is the outcome of electronic isomerism
processes typically associated with ground-state mixed-
valence systems, which can also occur between electro-
nically coupled excited states of comparable energy. Also
in this case, the Ru-NO fragment distorts during the
lifetime of the excited state to yield a minimum in the PES.
In parallel with the geometrical changes, one electron
is transferred from the RuII(bpy)2(NO•) to the distant
(H3N)5RuIII leading to a cis-[(H3N)5RuII(μ-pz)RuIII
-
(bpy)2(NO•)]5þspecies whose spin density is displayed in
Figure 4. After this somewhat cumbersome series of
interconvertions between excited state electronic isomers,
this species ends up with a {RuIII-NO•} fragment similar
geometrically and electronically to the one responsible for
NO photoliberation in 32þ. It is not strange therefore that
this compound displays a φNO value of (0.025 ( 0.002)
mol einstein-1 at 532 nm.15
Conclusions
The use of DFT computations helped to understand the
low value of φNO in 23þ. On the basis of the previous
experience with other dinuclear systems, this compound
appeared as a promising cage compound for the photo-
release of NO: it holds an octahedral {RuNO}6 fragment,
it is extremely inert toward thermal decomposition, and it
has a strong absorption band in the visible region of the
spectrum. We have shown that the presence of a second Ru
center, though responsible for the increase in the extinc-
tion coefficient, introduces uncertainty in the prediction of
the electronic distribution upon light absorption. This
situation does not always lead to a lowest energy excited
state with the appropriate electronic distribution to favor
the release of a neutral NO molecule. The example
described in this work and our interpretation show that
the approach based on coupling of metallic fragments to
a knowingly photoreactive nitrosyl carrying center to
enhance its spectral properties may not always lead to
success. At the same time, the high φNO values reported in
the literature for a series of cis-[RuL(NH3)4(μ-pz)Ru-
(bpy)2(NO)]5þ species are the clear demonstration that
the overall strategy can also be successful. Nevertheless,
the election of the appropriate fragment combination
might not be trivial. Electronic structure computations
may eventually assist in this task.
cule coordinated to a RuII center and not to a RuIII as in 32þ
.
A comparison between the computed dissociation energies
for lowest energy triplet states of 23þ and 32þ (Figure 2)
shows a more weakly bounded neutral NO molecule in the
latter. This result is not completely unexpected, provided that
the extra electron in the {RuII-NO•} moiety of 23þ (absent
in the {RuIII-NO•} fragment of 32þ) is associated with a
π-bonding molecular orbital, or in other words the π-back-
donation arising from a RuII metal center is more efficient
than in a RuIII species. In view of the molecular orbital pic-
ture, the electronic distribution for the lowest energy triplet
excited state is in this case better described as trans-[(NC)-
RuIII(py)4(μ-CN)RuII(py)4(NO•)]3þ. In this description, the
NO ligand sits in an essentially equivalent environment than
the one that results from a 1e- reduction of the {RuNO}
fragment to yield a {RuIINO}7 moiety. The actual 1e-
reduced species 22þ has been thoroughly characterized.17
The spectroscopic evidence supports a trans-[(NC)RuII(py)4-
(μ-CN)RuII(py)4(NO•)]2þ formulation, with a non-labile
{RuII-NO•}7 fragment. As a matter of fact, recent studies
have shown that NO• compounds of þ2 metal centers with a
low spin d6 electronic configuration can be much more inert
than expected. For instance, the rate constant for the dis-
sociation of NO• in [Fe(CN)5NO]3- has been reported to be
(1.58 ( 0.06) ꢀ 10-3 s-1 26
. The activation enthalpy reported
.
in this case was ΔHq = 106.4 ( 0.8 kJ mol-1
Acknowledgment. The authors thank the University of
Buenos Aires (UBA), the Consejo Nacional de Investi-
ꢀ
gaciones Cientıficas y Tecnicas (CONICET), the Agen-
cis-[(H3N)5Ru(μ-pz)Ru(bpy)2(NO)]5þ. Analogously to
the case discussed above, the second ruthenium center
ꢀ
ꢀ
cia Nacional de Promocion Cientıfica y Tecnologica
(ANPCyT), and Fundacion Antorchas (Argentina) for
funding. This work was partially supported by the
ꢀ
(26) Roncaroli, F.; Olabe, J. A.; van Eldik, R. Inorg. Chem. 2003, 42,
4179–4189.

6930 Inorganic Chemistry, Vol. 49, No. 15, 2010
De Candia et al.
National Center for Supercomputing Applications
under Grant TG-MCA05S010. L.D.S. and R.E. are
members of the scientific staff of CONICET, and A.
G.D.C. and J.P.M. are postdoctoral and graduate
fellows from CONICET, respectively.
Supporting Information Available: Tables with relevant
structural information and figures showing the DFT opti-
mized structures for the compounds discussed in this work.
This material is available free of charge via the Internet at
http://pubs.acs.org.