Photolabile Ru Model Complexes with Chelating Diimine Ligands for Light-Triggered Drug Release

Article abstract of DOI:10.1002/ejic.201701392

A series of water-soluble photolabile model complexes of the general formula [Ru([9]aneS₃)(chel)(py)]Cl₂ ([9]aneS₃ = 1,4,7-trithiacyclononane, chel = chelating diimine) was prepared and fully characterized. The phototriggered release of pyridine with visible light as a function of the nature of the diimine {chel = 2,2′-bipyridine (6) 1,10-phenanthroline (7), 4,7-diphenil-1,10-phenanthroline (8), dipyrido-[3,2-a:2′,3′-c]phenazine (dppz, 9), 2,2′-biquinoline (bq, 10)} was investigated. Our aim is to establish whether this type of complexes might be realistically used in the photo-uncaging strategy of photoactivated chemotherapy (PACT) in the future. Compounds 6–9 present a MLCT absorption in the blue region of the visible spectrum. When irradiated with light at 470 nm, they rapidly and quantitatively release the coordinated pyridine. Complex 10 turned out to be quite different from the others in the series. Structure-wise, in 10 the average plane of coordinated bq – owing to its steric demand – is remarkably tilted relative to the equatorial coordination plane [37.43(4)°, with the “front“ of the ligand pointing towards the axial py] and the orientation of py is nearly orthogonal compared to that found in 6 and 7 for minimizing steric clashes with bq. The low-lying acceptor orbitals of bq induce a red-shift of the MLCT absorption maximum to ca. 500 nm. Contrary to the expectations, complex 10 is more photostable compared to 6–9 and photo-dissociation of both py and bq, in nearly equal amounts, occurs. A detailed theoretical investigation was performed on 10 (and on 6 for comparison), for explaining its peculiar spectral features and photochemical behavior.

Full text of DOI:10.1002/ejic.201701392

A Journal of
Accepted Article
Title: Photolabile [Ru([9]aneS3)(chel)(py)][Cl]2 (chel = chelating
diimine) model complexes for light-triggered drug release
Authors: Federica Battistin, Gabriele Balducci, Jianhua Wei, Anna
Renfrew, and Enzo Alessio
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.201701392
Link to VoR: http://dx.doi.org/10.1002/ejic.201701392

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
Photolabile [Ru([9]aneS3)(chel)(py)]Cl2 (chel = chelating diimine)
model complexes for light-triggered drug release.
Federica Battistin,[a] Gabriele Balducci,[a] Jianhua Wei,[b] Anna K. Renfrew,*[b] and Enzo Alessio*[a]
Abstract:
A
series of water-soluble photolabile model
are inactive and non-toxic in the dark, whereas they are locally
activated in vivo upon irradiation with visible light. In this context,
photodynamic therapy (PDT), a clinically approved treatment for
some skin diseases, age-related macular degeneration and
some cancers, is the most well-known application. PDT uses a
photosensitizer (PS) at non-toxic concentrations that, in the most
common type II mechanism, upon light-excitation catalytically
complexes of the general formula [Ru([9]aneS3)(chel)(py)]Cl2
([9]aneS3 = 1,4,7-trithiacyclononane, chel = chelating diimine)
was prepared and fully characterized. The photo-triggered
release of pyridine with visible light as a function of the nature of
the diimine (chel = 2,2′-bipyridine (6) 1,10-phenanthroline (7),
4,7-diphenil-1,10-phenanthroline
(8),
dipyrido-[3,2-a:2′,3′-
c]phenazine (dppz, 9), 2,2′-biquinoline (bq, 10)) was investigated. generates singlet oxygen (1O2) or other highly cytotoxic ROS
Our aim is that of establishing if this type of complexes in the
future might be realistically used in the photo-uncaging strategy
of photoactivated chemotherapy (PACT). Compounds 6 – 9
present a MLCT absorption in the blue region of the visible
spectrum. When irradiated with light at 470 nm, they rapidly and
quantitatively release the coordinated pyridine. Complex 10
turned out to be quite different from to the others in the series.
Structure-wise, in 10 the average plane of coordinated bq –
owing to its steric demand – is remarkably tilted relative to the
equatorial coordination plane (37.43 (4)°, with the “front” of the
ligand pointing towards the axial py) and the orientation of py is
ca. orthogonal compared to that found in 6 and 7 for minimizing
steric clashes with bq. The low-lying acceptor orbitals of bq
induce a red-shift of the MLCT absorption maximum to ca. 500
nm. Contrary to the expectations, complex 10 is more photo-
stable compared to 6 – 9 and photo-dissociation of both py and
bq, in ca. equal amounts, occurs. A detailed theoretical
investigation was performed on 10 (and on 6 for comparison),
for explaining its peculiar spectral features and photochemical
behavior.
such as superoxide radical anions.[1] Another phototherapy
approach is the so-called photoactivated chemotherapy (PACT)
in which a kinetically inert and biologically non-active prodrug is
irreversibly activated by irradiation with visible light that induces
the cleavage of
a
photolabile protecting group.[2] The
photoactivation process is also called photo-uncaging.
Compared to PDT, PACT is a stoichiometric rather than catalytic
process, but has the advantage of not depending on the
presence of molecular oxygen. Thus, in principle, PACT agents
are active also in hypoxic tumor tissues. In general, ideal PDT or
PACT agents are water-soluble and resistant to photobleaching.
In addition, they should be activated within the phototherapeutic
window ( > 600 nm), where light is more penetrating into the
tissues and less harmful.
By virtue of their peculiar light absorption properties and rich
photoreactivity, d-block metal compounds are attracting rapidly
increasing interest as potential PDT and PACT agents.[3-7]
Among them, Ru(II) compounds are extensively investigated
due to their superior photophysical and photochemical
properties.[8-10] For example, even though most PDT
photosensitizers used in clinic are based on porphyrin
derivatives,[11]
a Ru(II)-polypyridyl complex (TLD-1433) is
undergoing a phase I clinical trial in Canada as PDT agent in
patients with bladder cancer.[12,13]
1. Introduction
A typical reaction that can occur in inorganic PACT agents is the
photo-induced release of ligands from coordinatively-saturated
and inert prodrugs. The most extensively investigated class of
ruthenium PACT agents is that of polypyridyl complexes of the
[Ru(bpy)3]2+ family that contain (at least) one sterically hindering
diimine such as 6,6′-dimethyl-2,2′-bipyridine (dmbpy).[14] The
strain caused by such a ligand in the coordination sphere
promotes the light-induced population of low-lying dissociative
metal-centered triplet excited states (3MC) and consequently its
release. Coordination and organometallic Ru(II) compounds in
which visible light triggers the release of a single monodentate
ligand have also been investigated.[15,16] A careful design of the
metal prodrug can lead to metal complexes with dual activity, i.e.
photo-triggered ligand release (PACT) and singlet oxygen
production (PDT).[17] The activation of coordinatively-saturated
cytotoxic Ru complexes through the photo-deprotection of a
ligand has been also been reported.[18]
In therapy, light-triggered treatments are appealing since – in
principle – they can generate a drug with high spatial and
temporal selectivity, resulting in a greater specificity of action.
Such treatments require light-activated prodrugs that – ideally –
[a]
[b]
F. Battistin, Dr. G. Balducci, Prof. Dr. E. Alessio
Department of Chemical and Pharmaceutical Sciences
University of Trieste
Via L. Giorgieri 1, 34127 Trieste, Italy
E-mail: alessi@units.it
http://dscf.units.it/en/department/people/alessio-enzo/1055
Dr. J. Wei, Dr. A. K. Renfrew
School of Chemistry
University of Sydney, Sydney, NSW, 2006, Australia.
Supporting information for this article is given via a link at the end of
the document. CCDC 1588006, 1588009, 1588010, and 1588014
contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via
In photo-labile metal complexes the focus can be on the
activated metal fragment, that may bind to biomolecules such as
www.ccdc.cam.ac.uk/data_request/cif.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
DNA through its newly generated coordination sites,[14,19] or on
the released ligand if itself
a
pharmacologically-active
molecule,[20] or on the combined action of both.[21] In polypyridyl
Ru(II) complexes the increased cytotoxicity is generally
attributed to the intracellular formation of the bis-aqua complex
cis-[Ru(bpy)2(OH2)2]2+ species. However, a very recent paper by
Bonnet and coworkers demonstrated that in the case of
[Ru(bpy)2(dmbpy)]2+ the photo-released dmbpy ligand, rather
than the ruthenium bis-aqua fragment, is responsible for the
observed phototoxicity.[22]
Figure 1. The diimine ligands used in this work with proton labelling scheme
for NMR purposes: 2,2′-bipyridine (bpy), 1,10-phenanthroline (phen), 4,7-
diphenil-1,10-phenanthroline (4,7-Ph2phen), dipyrido-[3,2-a:2′,3′-c]phenazine
(dppz), 2,2′-biquinoline (bq).
In the recent past we reported that dicationic Ru(II) complexes,
such as [Ru([9]aneS3)(bpy)(py)][PF6]2 ([9]aneS3
= 1,4,7-
trithiacyclononane, bpy = 2,2′-bipyridine), are inert in the dark
but rapidly and quantitatively release the pyridine ligand in
aqueous solution when illuminated with blue light ( = 420
2.2. Synthesis of the complexes
nm).[23,24]
Since
the
photo-generated
aqua
species
The Ru(II) compounds were prepared as chloride salts, rather
than as PF6 salts, for improving aqueous solubility. A two-step
procedure was followed (Scheme 1).
[Ru([9]aneS3)(bpy)(OH2)]2+ showed
a
substantial lack of
cytotoxicity (against the MDA-MB-231 human mammary
carcinoma cell line) we suggested that Ru(II) compounds of this
type might be suitable PACT agents for the light-triggered
release of coordinated drugs (photo-uncaging).[25]
In this paper we report our recent work on the model complexes
of the type [Ru([9]aneS3)(chel)(py)]Cl2, where chel is a chelating
diimine. Our aim was that of establishing if complexes of this
series, bearing a pharmacologically active molecule in the place
of pyridine, can be realistically used within the photo-uncaging
strategy. First we investigated if the absorption maxima in the
visible spectrum and the photoinduced release of pyridine can
be tuned by changing the nature of the diimine ligand. For this
purpose, the model complexes with chel = 1,10-phenanthroline
(phen),
4,7-diphenil-1,10-phenanthroline
(4,7-Ph2phen),
dipyrido-[3,2-a:2′,3′-c]phenazine (dppz), 2,2′-biquinoline (bq)
were prepared, fully characterized and investigated. The photo-
induced release of py was qualitatively investigated by H NMR
Scheme 1. Synthetic procedure for the [Ru([9]aneS3)(chel)(py)]Cl2 compounds
6 – 10, chel = bpy (6), phen (7), 4,7-Ph2phen (8), dppz (9), bq (10).
1
and UV-vis spectroscopy. A particularly detailed experimental
and theoretical investigation was performed on the bq derivative,
for explaining its peculiar spectral features and photochemical
behavior.
In the first step, modified from the literature,[28] treatment of the
[Ru([9]aneS3)(dmso-S)Cl2] precursor with a two-fold excess of
chel in refluxing ethanol (3h) afforded the known mono-cationic
intermediates of formula [Ru([9]aneS3)(chel)Cl]Cl (chel = bpy (1),
phen (2), 4,7-Ph2phen (3), dppz (4)) in good isolated yields (65 –
85%). The insertion of bq was more difficult, possibly due also to
the low solubility of the ligand in ethanol. A microwave assisted
reaction in ethanol (140°C, 90 min) was preferred to prolonged
reflux for obtaining [Ru([9]aneS3)(bq)Cl]Cl (5) in good yield. All
complexes, already reported in the literature either as Cl or PF6
salts,[28,29] were characterized by NMR and UV-vis spectroscopy
(Supporting Information, Figure S1), and mass spectrometry.
They are well soluble in ethanol, chloroform, DMSO, and – with
the exception of 3 and 4 – also in water.
2. Results and discussion
2.1. Diimine ligands
The chelating diimines used in this work, with different size and
aromaticity, are shown in Figure 1. The 2,2′-biquinoline ligand,
owing to its steric demand, is known to induce deformation in the
pseudo-octahedral coordination sphere of Ru(II) complexes that
can improve the photo-induced dissociation of ligands.[14b,15] In
addition, its low-lying acceptor orbitals are expected to red-shift
the 1MLCT absorption maximum typical of diimine-Ru(II)
complexes closer to the PDT window.[26,27]
1
As clearly shown by H NMR spectroscopy (Figure S2), in D2O
compounds 1 – 5 are in equilibrium – to different extents – with
the corresponding aqua species [Ru([9]aneS3)(chel)(OH2)]2+
(1aq – 5aq). In 1aq – 4aq the aromatic resonances are slightly
downfield shifted (ca. 0.1 pm or less) compared to those of the
parent complex, and their relative intensity increases upon
diluting the solution and decreases (or disappears altogether)
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
upon adding an excess of NaCl. In the case of 5 a single set of
resonances is observed in D2O, suggesting that no significant
equilibration with the aqua species 5aq occurs at typical NMR
concentrations. The resonances of 5aq appear upon dilution,
and in this case some of them are shifted upfield compared to
those of 5 (e.g. the doublet of H8,8′ falls at 9.26 ppm in 5 and at
diimine ligand. The 1H NMR spectrum of the bq compound 10 is
treated in more detail below.
The electronic absorption spectra of 6 – 9 in the visible region
are characterized by two bands of roughly comparable
intensities ( in the range 3000 - 6000 M-1 cm-1), partially or
completely overlapped, at 350 - 430 nm. The spectrum of 10,
9.09 ppm in 5aq). The resonances of [Ru([9]aneS3)(dppz)Cl]+ (4), instead, shows two rather sharp and more intense bands at 359
that are sharp in CDCl3, are rather broad and have
concentration-dependent shifts in D2O, most likely due to
stacking interactions occurring in solution. Consistent with this
hypothesis, they (and those of 4aq as well) become sharper
upon diluting the solution.
and 378 nm, whereas the lowest energy MLCT band – in good
agreement with the expectations – is red-shifted, with an
absorption maximum at nearly 500 nm (Figure 3).
Treatment of intermediates 1 – 5 with a slight excess of pyridine
in refluxing water afforded the corresponding dicationic
complexes [Ru([9]aneS3)(chel)(py)]Cl2 (chel = bpy (6), phen (7),
4,7-Ph2phen (8), dppz (9), bq (10)) that, with the exception of 6
previously reported by us as PF6 salt,[23] are described here for
the first time. They were fully characterized as 1 – 5 above, and
the single-crystal X-ray structures of 7 (Figure S3) and 10
(Figure 2) were also determined. In 10, as in its precursor 5 and
other bq octahedral complexes,[14b,15,28-30] the distortion in the
geometry induced by the sterically demanding diimine is evident.
In particular, the average plane of bq is remarkably tilted relative
to the equatorial coordination plane (37.43 (4)°, with the “front”
of the ligand pointing towards the axial py), whereas the twist
about the C–C bond between the two quinolines (3.9(2)°) is
negligible. The geometrical features of coordinated bq are
similar also in the trans-RuCl2(bq)(CO)2 (11) complex (Figure 2),
in which the other ligands are sterically undemanding and that
we expressly prepared for the sake of comparison. Another
major structural difference in 10 concerns the rotation of the py
ligand about the Ru–N bond: in 10 py is ca. orthogonal
compared to the other similar complexes (e.g. ca. 76° with
respect to complex 6), most likely for avoiding steric clashes
between the oH atoms of py and H8,8′ of bq.
Figure 3. UV-vis spectra in the visible region of compounds 6 – 10 (ca. in 0.1
mM H2O).
2.3. The 2,2′-biquinoline complexes
1
In the H NMR spectrum of 2,2′-biquinoline the most downfield
resonance is that of H3,3′, followed by that of H4,4′. The anti
conformation assumed by the two quinolines in the free ligand
brings N′ close to H3 (and N to H3′), and the deshielding of H3,3′
was attributed mainly to the electrostatic effect of the lone pairs
(Figure 4).[31] When symmetrically bound to diamagnetic
octahedral metal centers, such as in Re(CO)3(bq)Br,[32] the
proton NMR spectrum of bq undergoes remarkable changes: the
doublet of H8,8′ becomes the most downfield signal ( = 0.71
ppm), whereas that of H3,3′ is shifted to lower frequencies ( =
–0.50 ppm). Such variations are attributable to the
conformational change of the ligand (from anti to syn) and to its
coordination. Regretfully, in the other symmetrical Ru-bq
compounds such as [Ru(phen)2(bq)][PF6]2,14b [Ru(6-p-
cymene)(bq)Cl][PF6],[33] and [Ru(bq)3][PF6]3,[34] the proton NMR
spectra were not assigned.
Figure 2. X-ray molecular structures (50% probability ellipsoids) of
[Ru([9]aneS3)(bq)(py)]Cl2 (10) (left) and of trans-RuCl2(bq)(CO)2 (11) (right).
The two chlorides in 10, and two chloroform crystallization molecules in 11
omitted for clarity. Only one of the two independent molecules of 11 present in
the unit cell is shown.
1
We found that, whereas the H NMR spectrum of 5 is consistent
with such features, the spectrum of 10 is quite different and
more similar to that of the free ligand: the resonance of H8,8′
falls to lower frequencies than those of H3,3′ and H4,4′ (Figure
4).
1
All dicationic complexes are fairly soluble in water. The H NMR
spectra of compounds 6 – 9 (Figures S4 – S7) are unexceptional
and, as for the corresponding precursors, consistent with the CS
symmetry of each complex cation. The most downfield
resonance is that of the protons adjacent to the N atoms of the
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
In the dark, compounds 6 – 10 are stable in D2O (3 mM
solutions) for at least 24h at ambient temperature, no changes in
the NMR spectra were observed.
3,3’
4,4’
6,6’
8,8’
5,5’
7,7’
Form our previous work it is already known that, when irradiated
with blue light at 420 nm, [Ru([9]aneS3)(bpy)(py)][PF6]2 rapidly
and quantitatively releases the coordinated pyridine.[23]
Compounds 6 – 9 have a similar behavior: when irradiated with
blue light (LED,  = 470 nm, 40 mW) they release the
coordinated pyridine at comparable rates and extents,
generating selectively the corresponding aqua species (1aq –
4aq) in equilibrium with the chlorido species (1 – 4) (Scheme 2).
No other reaction occurs.
5,5’+
6,6’
3,3’
8,8’
4,4’
7,7’
3,3’+4,4’
7,7’+
6,6’
8,8’+o
m
5,5’
5,5’
p
3,3’+4,4’
8,8’
6,6’
7,7’
*
Figure 4. 1H NMR spectrum (aromatic region) of, from top to bottom: 2,2′-
biquinoline, [Ru([9]aneS3)(bq)Cl]Cl (5), [Ru([9]aneS3)(bq)(py)]Cl2 (10), and
[Ru([9]aneS3)(bq)(NH3)]Cl2 (12). The spectrum of bq is in DMSO-d6, the others
in D2O. In the spectrum of 12 the asterisk indicates residual chloroform.
Scheme 2. Photo-dissociation of pyridine from compounds 6 – 9, exemplified
in the case of chel = bpy.
Since these changes in the chemical shifts of the bq protons
between
5
and 10 could not be attributed to different
The photo-reactions were performed in D2O and quantitatively
monitored by H NMR spectroscopy. An example is reported in
1
conformational strains in the bq frame (see above), we came to
the conclusion that the H8,8′ doublet in 10 is shifted upfield by
the shielding cone of the adjacent axial pyridine. In order to
confirm this hypothesis, that is consistent also with the
orientation of py evidenced by the X-ray structure shown in
Figure 2, we prepared the complex with NH3 in the place of
pyridine, i.e. [Ru([9]aneS3)(bq)(NH3)]Cl2 (12). Indeed, even
though the structural features in 12 (Figure 5) are again similar
to those of 5 and 10 (e.g. the tilt angle of bq is 38.89(3)°), the
NMR spectral pattern of coordinated bq follows the “normal”
order, and the H8,8′ doublet is again the most downfield
resonance.[35]
Figure 6. In the case of the dppz complex 9, the resonances of
the photo-generated aqua and chlorido species – as mentioned
above – are rather broad; the sharp pyridine signals allowed
reliable integration to be performed. Table 1 reports the
percentage amount of photo-released pyridine as a function of
the irradiation time.
7
t = 0
7
7
7
7
2aq
t = 5 min
2
2aq
2aq
py
2aq
py
py
2
2
2
t = 10 min
t = 15 min
Figure 5. X-ray molecular structure (50% probability ellipsoids) of
[Ru([9]aneS3)(bq)(NH3)]Cl2 (12). The two chlorides and
a
methanol
crystallization molecule omitted for clarity.
Figure
6.
Photo-induced
dissociation
of
pyridine
from
[Ru([9]aneS3)(phen)(py)]Cl2 (7) monitored as a function of the irradiation time
( = 470 nm, 40 mW) by 1H NMR spectroscopy in D2O. The positive charges
of the complexes are omitted.
2.4. Photo-induced release of ligands
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
Table 1. Extent of photo-released pyridine, assessed by 1H NMR
spectroscopy, as a function of the irradiation time (LED,  = 470 nm, 40 mW).
Complex (ligand)
6 (bpy)
5 min
10 min
92,8%
15 min]
93,3%
30 min
98,8%
80,0%
60,0%
75,0%
50,0%
78,2%
78,8%
75,0%
87,8%
81,4%
85,0%
96,6%
95,4%
97,7%
7 (phen)
8 (4,7- Ph2phen)
9 (dppz)
Figure 7. 1H NMR spectrum of the reaction mixture obtained upon irradiation
( = 470 nm, 40 mW, 180 min) of a D2O solution of [Ru([9]aneS3)(bq)(py)]Cl2
(10). The positive charges of the complexes are omitted.
Although parallel photo-release of two different ligands has not
been often described, it has been observed recently in the
complex [Ru(bpy)(dmbpy)(L-proline)][PF6], in which substitution
of both dmbpy and L-proline occurred upon illumination.[36]
In summary, the photo-dissociation of py is almost complete
after 30 min of illumination and the bpy complex 6 is the fastest
one, even though it has the smallest absorption coefficient at
470 nm. We found that the dppz complex 9, that has a very
weak absorption at ca. 540 nm (Figure S8), is still photoactive
when irradiated with green light at 530 nm, even though the
photo-release of pyridine is slower: 33% after 15 min ( = 530
nm, 30 mW) compared to 78% when irradiation was performed
at 470 nm with the same power.
2.5. Theoretical calculations
Intrigued by the remarkably different photochemical behavior of
the biquinoline complex 10, we performed a series of theoretical
calculations on it and on the corresponding bpy complex 6,
taken as model for the other diimine compounds 7–9. Structure-
wise, complex 10 has two main geometrical differences
compared to the canonical features of 6, that are likely to be
related to its different photochemistry: the tilted geometry of bq
and the orientation of py (see above).
The behavior of the bq complex 10 upon illumination is quite
different and photo-dissociation of both py and bq in ca. equal
amounts occurs. In general, contrary to the expectations, the
complex is more photo-stable compared to 6 – 9: after 2h of
illumination at 470 nm (40 mW) in D2O ca. 25% of 10 is still
present in solution. The interpretation of the NMR spectra
(Figure 7) was made more difficult by the following facts: i) 2,2′-
biquinoline is insoluble in water, thus the resonances of photo-
released bq cannot be seen; ii) the chemical shifts of the bq
resonances in both 5 and 5aq, i.e. the Ru complexes obtained
upon photo-release of py, as well as the ratio between the two
species, are concentration-dependent; iii) (most of) the released
py binds to the {Ru([9]aneS3)(py)}2+ fragment (thus the
resonances of free py are not clearly seen), affording the
[Ru([9]aneS3)(py)2(OH2)]2+ complex cation that, in addition, is in
equilibrium with [Ru([9]aneS3)(py)2Cl]+. The resonances of these
latter species were unambiguously identified: for this purpose we
made [Ru(9aneS3)(py)2Cl]Cl (13), that in aqueous solution
equilibrates with [Ru([9]aneS3)(py)2(OH2)]2+ (13aq). When a D2O
solution of 10 was irradiated with green light (30 mW) at 530 nm
a similar behavior was observed, but the photo-release of both
bq and py was slower. A similar photochemistry was observed
when the irradiation of 10 was performed in DMSO-d6 where the
resonances of free bq (that is soluble) could be observed: in this
medium the photo-induced dissociation of bq prevails over that
of py (Figure S9).
First of all, our computational protocol (DFT with plane wave
basis set, pseudopotentials and periodic boundary conditions)
was tested on compound
6 that had been previously
investigated using a different computational approach (DFT with
localized basis functions).[23] The results obtained in terms of
optimized geometry, calculated MOs, and electronic transitions
were in excellent agreement with those reported in the literature,
thus confirming the reliability of our protocol. Next, the
calculations were extended to 10.
Figure 8 shows the density of states and its projection onto
selected atomic orbitals of the Ru, py and bq or bpy ligands for
complexes 10 and 6 in the ground state configuration.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
Figure 9. Experimental (top) and calculated (with the “turbo_lanczos” program,
bottom) absorption spectra for complex 10. The vertical bars in the simulated
spectrum are the calculated transitions (with the “turbo_davidson” code), with
height equal to the oscillator strength.
Figure 8. Density of states (DOS) and its projection onto selected atomic
orbitals for complexes 10 (top) and 6 (bottom). Vertical lines indicate the
energy of the molecular orbitals in the range from HOMO–2 to LUMO+6. Plots
have been aligned so that the energy of the HOMO’s for the two complexes is
0.0 eV.
Consistent with what previously observed, in 6 the LUMO+4 and
LUMO+5 MOs have a strong metal d-antibonding component; in
addition, they have a significant -antibonding character towards
the bpy and – above all – the pyridine ligands (Figure 10). Thus,
the light-induced population of such orbitals is presumably
responsible for the photo-dissociation of py. Conversely, in 10
the orbitals with the most relevant d* component are the
LUMO+3 and LUMO+4, and they have a relevant bq – rather
than py – contribution (Figure 10). This finding is consistent with
what established for Ru(II)-polypyridyl complexes that contain bq,
in which distortion is known to lower the energy of a dissociative
metal-centred state.[14b,15] In addition, whereas LUMO+4 has a
significant -antibonding character towards bq, in both the
LUMO+3 and LUMO+4 the antibonding character towards py is
mainly of π symmetry (i.e. involving p atomic orbitals normal to
the py plane). This finding, i.e. the increase of π back-bonding
from the filled metal orbital to the π* orbitals of py, is most likely
attributable to the different orientation of py in the bq complex
and might account for the less-pronounced light-induced
dissociation of pyridine in 10.
Consistent with the red-shifted absorption band in the UV-vis
spectrum, and with the general features expected for bq
complexes (i.e. stabilized bq π* orbitals relative to those of
bpy),[14b,15] the HOMO−LUMO energy gap in 10 is smaller than in
6 (2.01 vs 2.32 eV). TDDFT calculations well reproduced the
experimental absorption spectrum of both complexes (Figure 9
and Figure S10) and allowed us to assign also the character of
each band (Table S1),[37] thus confirming that the lowest energy
transition has a ca. 85 – 90% HOMO → LUMO component.[23] In
both complexes the LUMO is almost coincident with a π* MO of
the diimine ligand (Figure S11). However, we notice that
whereas in the bpy complex the three frontier occupied orbitals –
HOMO, HOMO−1 and HOMO−2 (Figure S11) – have an almost
exclusive metal-centered character, i.e. are coincident with the
filled d orbitals of Ru(II), in 10 the HOMO and HOMO−1 get an
appreciable contribution from the atomic orbitals of C and N
atoms of the biquinoline ligand (see also Figure 8). We argue
that the tilted orientation of bq in 10 is responsible for the mixing
of diimine π orbitals with the filled d orbitals of Ru(II). Therefore,
the lowest energy electronic transition can be safely labeled as a
pure MLCT in 6, whereas it has a π − π*component in 10.
The other significant photochemical difference between the two
complexes concerns the photo-induced dissociation of the
chelating diimine, that occurs in 10 (bq) but not in 6 (bpy). We
observe that the binding of bq is arguably weaker compared to
that of bpy because of its tilted coordination geometry that leads
to a smaller overlap in the bonding orbitals.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
the diimine have a facial arrangement. Our aim is that of
establishing if this type of complexes in the future might be
realistically used in photoactivated chemotherapy (PACT).
We found that compounds 6 – 9 behave quite homogeneously
and their photochemical behavior is not particularly affected by
the nature of the diimine: When irradiated with light at 470 nm,
they rapidly and quantitatively release the coordinated pyridine,
generating selectively the corresponding aqua species
[Ru([9]aneS3)(chel)(OH2)]2+ (1aq – 4aq). Even though 6 was
found to be non-phototoxic against the MDA-MB-231 human
mammary carcinoma cells,25 in the future we plan to investigate
1
the phototoxicity also of 7 – 9, as well as the O2 production on
selected compounds. The more lipophilic compounds in the
series might be expected to have better accumulation in cancer
cells and potentially show higher phototoxicity. We also argue
that upon illumination the extended aromatic dppz ligand could
generate singlet oxygen, thus giving to complex
mechanisms of phototoxicity.
9 two
In addition, compounds 6 – 9 would be suitable for the photo-
uncaging of many different pyridine-containing drugs, potentially
for the treatment of both cancer and bacterial infections. This
strategy does not require the concomitant formation of an active
(e.g. cytotoxic) metal fragment. Some examples of py-containing
drugs that have been coordinated to ruthenium previously
(Figure 11) are the antibacterial isoniazid (L1),[39] the P-450
inhibitors metyrapone (L2) and abiraterone (L3),[21,40] and
PARPs inhibitors (PARPs = poly(ADP-ribose) polymerase) such
as nicotinamide (L4), quinazolin-4(3H)-one (L5), and 3-aza-5[H]-
phenanthridin-6-one (L6).[41]
Figure 10. Selected virtual molecular orbitals for 6 (right) and 10 (left) in the
singlet ground state.
In conclusion, contrary to what has been found by Turro and
coworkers for mer-[Ru(tpy)(chel)(py)]2+ species (chel = bpy or
bq), where the distortions induced by the bulky bq led to an
increased photo-induced release of py compared to bpy,[15] in
our case – mainly because of the facial, rather than meridional,
geometry of the complex – the distortions led to the preferential
photo-dissociation of biquinoline itself.[38]
Although triplet states are generally thought to be responsible for
the photochemistry of ruthenium complexes via facile
intersystem crossing, we are confident that our analysis based
on singlet ground and excited states captures the essential
features of complex 10. In fact, the previous work on complex 6
has shown that triplet excited states trace the character and
ground state orbital composition of the singlet counterparts.[23]
Figure 11. Examples of pyridine-containing drugs..
3. Conclusions
Complex 10 turned out to behave quite differently compared to
the others in the series. As expected, the low-lying acceptor
orbitals of bq induce a red-shift in the MLCT absorption
maximum of the complex from ca. 430 to ca. 500 nm. However,
contrary to the expectations, complex 10 turned out to be more
photo-stable compared to 6 – 9 and – upon prolonged
illumination with blue light – photo-dissociation of both py and
bq, in ca. equal amounts, occurs. The single crystal X-ray
structure of 10 showed that in this complex, besides the
expected distortion of coordinated bq due to its steric demand (a
The photo-triggered release of pyridine from the series of water-
soluble model complexes [Ru([9]aneS3)(chel)(py)]Cl2 was
thoroughly investigated as a function of the nature of the
chelating diimine (chel
=
2,2′-bipyridine (bpy, 6) 1,10-
phenanthroline (phen, 7), 4,7-diphenil-1,10-phenanthroline
(Ph2phen, 8), dipyrido-[3,2-a:2′,3′-c]phenazine (dppz, 9), 2,2′-
biquinoline (bq, 10)). In 6 – 10, owing to the face-capping 1,4,7-
trithiacyclononane ligand ([9]aneS3), the leaving ligand (py) and
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
>35° tilt relative to the equatorial coordination plane), the
orientation of py is ca. orthogonal compared to that found in 6
and 7. A detailed theoretical investigation performed on 10 (and
on 6 for comparison), showed that the biquinoline-induced
geometrical distortions lead to differences in the nature of the
excited states that might account for the different photochemical
behavior of this complex.
Solid state infrared spectra were recorded on a Perkin-Elmer
983G spectrometer. Elemental analyses were performed in the
Department of Chemistry of the University of Bologna (Italy).
X-ray diffraction
Data collections were performed at the X-ray diffraction
beamline (XRD1) of the Elettra Synchrotron of Trieste (Italy)
equipped with a Pilatus 2M image plate detector.
In view of the potential investigation of these complexes as
PACT agents, we observe that – unlike the rest of the series –
upon irradiation with visible light complex 10 could generate in
vivo a Ru(II)-aqua species with two or even three coordination
sites that is expected to be more reactive, and thus more
cytotoxic, compared to the mono-aqua species generated from 6
– 9 (e.g. it might be capable of cross linking DNA). Consistent
Collection temperature was 100K (nitrogen stream supplied
through an Oxford Cryostream 700); the wavelength of the
monochromatic X-ray beam was 0.700 Å and the diffractograms
were obtained with the rotating crystal method. The crystals
were dipped in N-paratone and mounted on the goniometer
head with a nylon loop. The diffraction data were indexed,
integrated and scaled using the XDS code.[44] The structures
were solved by the dual space algorithm implemented in the
SHELXT code.[45] Fourier analysis and refinement were
performed by the full-matrix least-squares methods based on F2
implemented in SHELXL.[46] The Coot program was used for
modeling.[47] Anisotropic thermal motion was allowed for all non-
hydrogen atoms. Hydrogen atoms were placed at calculated
positions with isotropic factors U = 1.2×Ueq, Ueq being the
equivalent isotropic thermal factor of the bonded non hydrogen
atom. Crystallographic data and coordination distances and
angles are in the Supporting Material.
with
this
hypothesis,
we
found
that
the
[Ru([9]aneS3)(py)(OH2)2]2+ species binds rapidly the small
amount of photo-released pyridine.
Finally, in the future it would be interesting to make the 6,6′-
dimethyl-2,2′-bipyridine (dmbpy) analogue of these complexes. It
might be expected to behave similarly to the bq complex and the
photo-released dmbpy could induce cell death according to what
shown by Bonnet and coworkers.[22]
Computational methods
Experimental Section
We performed periodic first principle calculations in the frame of
density functional theory (DFT) with the Kohn–Sham orbitals
expanded in a basis of plane waves and the effects of atomic
core regions accounted for by pseudopotentials. The QUANTUM
ESPRESSO suite of codes was used for all the computations.[48]
To model an isolated molecule using a periodic code like
QUANTUM ESPRESSO, a “molecule in the box” approach can
be used: a single molecule is simulated in a unit cell large
enough to minimize any interaction between the molecule itself
and any of its periodic images. A cubic unit cell with edge length
of 19.0 Å for complex 6 and 20.0 Å for complex 10 was found to
give a minimum separation of 10 Å between nearest atoms of
any two contiguous images. Both total energy and scf potential
were corrected for the effect of the fictitious periodicity with the
Martyna-Tuckerman method.[49] Ultrasoft pseudopotentials were
used throughout the calculations.[50] The exchange–correlation
part of the energy functional was modeled with the (spin-
unpolarized) generalized gradient approximation (GGA), in the
PBE parameterization.[51] The plane wave expansion of the
crystalline orbitals was truncated at a cutoff energy of 340 eV
and a corresponding tenfold cutoff was used for the expansion
of the augmentation charge needed by the ultrasoft
pseudopotential method. Integrals over the first Brillouin zone in
reciprocal space were approximated by evaluations of the
integrand functions at the gamma point. Convergence
thresholds for geometry optimization were 1.4 × 10−4 eV for total
energy and 2.6×10−2 eV/Å for the maximum force component
acting on atoms; a threshold of 1.4 × 10−8 eV was imposed for
self-consistency. Excited state calculations and UV–vis spectra
simulation were performed with time dependent DFT (TDDFT).
The QUANTUM ESPRESSO suite offers two codes for this
purpose. The “turbo_davidson” code implements an improved
Materials
All chemicals were purchased from Sigma-Aldrich and used as
received. Solvents were of reagent grade. The ligand dppz
(dipyrido-[3,2-a:2′,3′-c]phenazine) was prepared according to
published procedures.[42] The precursors [Ru([9]aneS3)Cl2(dmso-
S)] was synthesized as described in the literature.[28]
Instrumental methods
Mono- and bi-dimensional (1H-1H COSY, 1H-13C HSQC) NMR
spectra were recorded at room temperature on a Varian 400 or
500 spectrometer (1H: 400 or 500 MHz, 13C:100.5 or 125.7 MHz).
1H and 13C chemical shifts were referenced to the peak of
residual non-deuterated solvent (δ = 7.26 and 77.16 for CDCl3,
2.50 and 39.52 for DMSO-d6) or were measured relative to the
internal standard DSS (δ = 0.00) for D2O. Carbon resonances
were assigned through the HSQC spectra; the resonances of
quaternary carbons were not assigned. ESI mass spectra were
collected in the positive mode on
a Perkin-Elmer APII
spectrometer at 5600 eV. The UV-vis spectra were obtained on
an Agilent Cary 60 spectrophotometer, using 1.0 cm path-length
quartz cuvettes (3.0 mL). A home-made LED apparatus,[43]
a
plastic-coated cylindric (Ømax = 20 mm, h = 110 mm), was used
for performing the photochemical reactions in NMR or test-tubes.
The inside of the well features four pairs of juxtaposed LED
stripes, containing five LEDs of the same color each (emission
maxima:  = 626, 590, 530, 470 nm; band width 10 nm, spectral
range ca. 10 nm). LED stripes of the same color are located
opposite to each other. One or more colors can be activated at
the, with an emission power for each LED that can be regulated
from 1 to 40 mW (30 mW for the green-emitting LEDs,  = 530
nm).
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
Davidson-like algorithm for the computation of individual
excitations clustered around a target energy value, which,
differently from the “conventional” Davidson algorithm, can be
located anywhere in the spectrum. The “turbo_lanczos” program
uses a so-called “pseudo-Hermitian” variant of the recursive
Lanczos scheme to evaluate the whole absorption spectrum in a
given energy range using only the occupied states obtained in a
previous self-consistent field calculation. Both codes rely upon
the formulation of the TDDFT problem in terms of the linearized
quantum Liouville equation and the details about the algorithms
can be found in the original papers.[52,53] 60 trial vectors, a
maximum of 200 basis vectors in the Davidson subspace and a
convergence threshold of 1.0×10−4 for the squared modulus of
the residue were used for the “turbo_davidson” runs; 5000
Lanczos iterations for each of the three directions of the full
dynamical polarizability tensor were used in the UV–vis
spectrum simulation with the “turbo_lanczos” code.
= 685.4): C 52.62; H 4.12; N 4.09. Found: C 52.73; H 4.20; N
4.17. 1H NMR (D2O), δ: 9.52 (d, 2H, H2,9), 8.15 (s, 2H, H5,6),
7.98 (d, 2H, H3,8), 7.69 (br s, 10H, Ph), 2.92 (m, 12H, [9]aneS3,
partially overlapped with the corresponding resonances of 3aq).
1H NMR (CDCl3), δ: 9.40 (d, 2H, H2,9), 8,07 (s, 2H, H5,6), 7.78
(d, 2H, H3,8), 7.57 (m, 10H, Ph), 3.03 (m, 12H, [9]aneS3).
13C{1H} NMR from HSQC (CDCl3), δ: 151.9 (C2,9), 129.2 (Ph),
124.0 (C5,6), 122.4 (C3,8), 35.0 ([9]aneS3). ESI mass spectrum:
649.1 m/z (calcd 649.3) [M]+. UV-vis (H2O): λmax (ε, L mol–1 cm–
1) = 370 (6598), 408 (5786) nm.
[Ru([9]aneS3)(dppz)Cl]Cl (4): Same procedure as for complex
1, using the same amount of precursor and 129.1 mg (2 eq) of
dppz. In this case, after 10 minutes of refluxing the ethanol
solution changed from yellow to red. Yield: 71%. Elemental
analysis calcd for [C24H22N4Cl2RuS3] (MW = 635.5): C 45.42; H
1
3.49; N 8.83. Found: C 45.50; H 3.58; N 8.91. H NMR (CDCl3),
δ: 9.77 (d, 2H, H2,2′), 9.42 (d, 2H, H4,4′), 8.46 (d, 2H, H5,5′),
8.08 (d, 2H, H6,6′), 8.00 (t, 2H, H3,3′), 3.01 (m, 12H, [9]aneS3).
13C{1H} NMR from HSQC (CDCl3), δ: 153.6 (C4,4′), 134.3
(C2,2′), 131.3 (C6,6′), 130.6 (C5,5′), 126.3 (C3,3′), 33.2
([9]aneS3). ESI mass spectrum: 599.0 m/z (calcd 599.1) [M]+.
UV-vis (H2O): λmax (ε, L mol–1 cm–1) = 357 (6875), 423 (4750)
nm.
Synthesis of the complexes
The preparations were performed in light-protected glassware.
[Ru([9]aneS3)(bpy)Cl]Cl (1): The complex was prepared
according to a modified literature procedure.[23] A 100 mg
amount of [Ru([9]aneS3)Cl2(dmso-S)] (0.23 mmol) was partially
dissolved in 15 mL of EtOH. Two equivalents (0.46 mmol) of bpy
(72 mg) were added and the mixture was refluxed for 3 h. After
10 minutes of refluxing the solution changed from yellow to
orange. Precipitation of the product in pure form (according to 1H
NMR spectrum) from the concentrated solution (ca. 8 mL)
occurred upon standing at r.t.. It was removed by filtration,
washed with few mL of EtOH and diethyl ether and dried in
[Ru([9]aneS3)(bq)Cl]Cl
(5).
A
50
mg
amount
of
[Ru([9]aneS3)Cl2(dmso-S)] (0.12 mmol) was partially dissolved in
2 mL of EtOH and two equivalents (0.24 mmol) of 2,2′-
biquinoline (60 mg) were added. The mixture was microwave-
heated at 140°C for 90 min. The white powder (unreacted bq)
was removed by filtration. Evaporation of the solvent afforded a
purple solid that was dissolved in water and filtered to remove
the remaining traces of unreacted bq. The solution was rotary
evaporated to dryness and the solid (pure 5, according to the 1H
NMR spectrum) was dried in vacuo. (Yield 51.1 mg, 66%).
Elemental analysis calcd for [C24H24N2Cl2RuS3] (MW = 607.9): C
vacuo.
Yield:
72%.
Elemental
analysis
calcd
for
[C16H20N2Cl2RuS3] (MW = 509,2): C 37.79; H 3.96; N 5.51.
1
Found: C 37.68; H 4.05; N 5.60. H NMR (D2O), δ: 9.05 (d, 2H,
H6,6′), 8.47 (d, 2H, H3,3′), 8.13 (t, 2H, H4,4′), 7.62 (t, 2H, H5,5′),
2.77 (m, 12H, [9]aneS3, partially overlapped with the
corresponding resonances of 1aq). ESI mass spectrum: 473.0
m/z (calcd 473.1) [M]+. UV-vis (H2O): λmax (ε, L mol–1 cm–1) = 361
(2359), 417 (4296) nm.
1
47.36; H 3.97; N 4.60. Found: C 47.28; H 3.88; N 4.51. H NMR
(D2O), δ: 9.10 (d, 2H, H8,8′), 7.95 (t, 2H, H7,7′), 7,65 (d, 2H,
H4,4′), 7,73 (m, 4H, H5,5′ + H6,6′), 7.49 (d, 2H, H3,3′), 2.43 (m,
12H, [9]aneS3). 13C{1H} NMR from HSQC (D2O), δ: 139.0
(C4,4′), 132.8 (C7,7′), 129.6 (C5,5′), 129.1 (C3,3′), 128.6 (C8,8′),
119.4 (C6,6′), 33.1 ([9]aneS3). ESI mass spectrum: 573.1 m/z
(calcd 573.2) [M]+. UV-vis (H2O): λmax (ε, L mol–1 cm–1) = 356
(20106), 373 (19574), 515 (4787) nm.
[Ru([9]aneS3)(phen)Cl]Cl (2): Same procedure as for complex
1, using the same amount of precursor and 83.5 mg (2 eq) of
phen. Also in this case, after 10 minutes of refluxing the ethanol
solution changed from yellow to orange. Yield: 70%. Elemental
analysis calcd for [C18H20N2Cl2RuS3] (MW = 533.3): C 40.60; H
3.79; N 5.26. Found: C 40.52; H 3.68; N 5.18. 1H NMR (D2O), δ:
9.48 (d, 2H, H2,9), 8.78 (t, 2H, H4,7), 8. 22 (s, 2H, H5,6), 8.02
(d, 2H, H3,8), 2.73 (m, 12H, [9]aneS3, partially overlapped with
[Ru([9]aneS3)(bpy)(py)]Cl2 (6): A 60 mg amount of complex 1
(0.13 mmol) was dissolved in 5 mL of H2O and a 50 µL amount
of pyridine (0.6 mmol) was added. The solution was refluxed for
3h, and then it was rotary evaporated to dryness. The obtained
solid was sonicated with 2 mL of acetone, then removed by
filtration, washed with acetone and diethyl ether, and dried in
vacuo. The product was obtained in pure form according to the
1H NMR spectrum. Yield: 66%. Elemental analysis calcd for
[C21H25N3Cl2RuS3] (MW = 588.2): C 42.93; H 4.29; N 7.15.
1
the corresponding resonances of 2aq). H NMR (CDCl3), δ: 9.36
(d, 2H, H2,9), 8.49 (d, 2H, H4,7), 8.05 (s, 2H, H5,6), 7.87 (d, 2H,
H3,8), 2.97 (m, 12H, [9]aneS3).13C{1H} NMR from HSQC
(CDCl3), δ: 152.2 (C2,9), 136.3 (C4,7), 127.8 (C5,6), 126.0
(C3,8), 34.1 ([9]aneS3). ESI mass spectrum: 497.0 m/z (calcd
497.1) [M]+. UV-vis (H2O): λmax (ε, L mol–1 cm–1) = 369 (4711),
415 (4423) nm.
[Ru([9]aneS3)(4,7-Ph2phen)Cl]Cl (3): Same procedure as for
complex 1, using the same amount of precursor and 153.2 mg
(2 eq) of 4,7-Ph2phen. In this case, after 10 minutes of refluxing
the ethanol solution changed from yellow to orange-brown.
Yield: 75%. Elemental analysis calcd for [C30H28N2Cl2RuS3] (MW
1
Found: C 42.65; H 4.25; N 7.08. H NMR (D2O), δ: 9.23 (d, 2H,
H6,6′), 8.66 (d, 2H, o-py), 8.38 (d, 2H, H3,3′), 8.16 (t, 2H, H4,4′),
7.77 (d, 3H, p-py + H5,5′), 7.26 (t, 2H, m-py), 2,90 (m, 12H,
[9]aneS3). ESI mass spectrum: 473.0 m/z (calcd 473.1) [M – py
+ Cl]+. UV-vis (H2O): λmax (ε, L mol–1 cm–1) = 401 (4666) nm.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
[Ru([9]aneS3)(phen)(py)]Cl2 (7): Same procedure as for 6,
using 69 mg of complex 2 (0.13 mmol). The product was
obtained in pure form according to the H NMR spectrum. Yield:
the mixture was sonicated for a few minutes until complete
dissolution of the ligand (yellow solution). After 4h small red
crystals began to form. Crystals were filtered after one day and
washed with CHCl3 and diethyl ether and dried in vacuo (Yield
57.1 mg, 75%). Elemental analysis calcd for [C20H12N2Cl2O2Ru]
(MW = 484.83): C 49.60; H 2.50; N 5.78 Found: C 49.72; H 2.61;
N 5.89. 1H NMR (CDCl3), δ 9.25 (d, 1H, H8,8′), 8.56 (d, 1H,
H3,3′), 8.35 (d, 1H, H4,4′), 8.04 (t, 1H, H7,7′), 7.98 (d, 1H,
H5,5′), 7.79 (t, 1H, H6,6′). 13C{1H} NMR (CDCl3) δ 195.61 (CO),
141.24 (C3,3′), 133.29 (C7,7′), 129.65 (C8,8′), 129.51 (C6,6′),
129.35 (C5,5′), 119.37 (C4,4′). ESI mass spectrum: 448.6 m/z
(calcd 448.8) [M – Cl]+. Selected IR absorption (cm-1): Nujol,
2051 (νCO, s), 1981 (νCO, s).
1
87%. Elemental analysis calcd for [C23H25N3Cl2RuS3] (MW
=
612.4): C 45.17; H 4.12; N 6.87. Found: C 45.25; H 4.20; N 6.98.
1H NMR (D2O), δ: 9.63 (d, 2H, H2,9), 8.76 (d, 2H, o-py), 8.72 (m,
2H, H4,7), 8.15 (s, 2H, H5,6), 8.11 (t, 2H, H3,8), 7.71 (d, 1H, p-
py), 7.20 (t, 2H, m-py), 3.01 (m, 12H, [9]aneS3). 13C{1H} NMR
from HSQC (D2O), δ: 153.3 (C2,9), 152.9 (o-py), 138.2 (C4,7),
138.5 (C3,8 + C5,6), 129.0 (p-py), 126.8 (m-py), 32.2 ([9]aneS3).
ESI mass spectrum: 497.0 m/z (calcd 497.1) [M – py + Cl]+. UV-
vis (H2O): λmax (ε, L mol–1 cm–1) =353 (5153), 405 (4384) nm. X-
ray quality crystals of 7 were obtained by slow diffusion of diethyl
ether into an EtOH solution of the complex.
[Ru([9]aneS3)(bq)(NH3)]Cl2 (12):
A
20 mg amount of
[Ru([9]aneS3)(4,7-Ph2phen)(py)]Cl2 (8): Same procedure as for
6, using 89 mg of complex 3 (0.13 mmol). The product was
obtained in pure form according to the H NMR spectrum. Yield:
[Ru([9]aneS3)(bq)Cl]Cl (5) (0.033 mmol) was dissolved in 1 mL
of H2O; a 45 µL volume of a 25% ammonia solution in water
(0.42 mmol) was added and the mixture was heated in the
microwave at 110°C for 150 min. Then the solvent was rotary
evaporated completely and the resulting oil was crushed with
CHCl3 obtaining a dark red solid that was filtered, washed with
1
84%. Elemental analysis calcd for [C35H33N3Cl2RuS3] (MW
=
764.2): C 55.04; H 4.35; N 5.50. Found: C 55.11; H 4.44; N 5.61.
1H NMR (D2O), δ: 9.73 (d, 2H, H2,9), 8.87 (d, 2H, o-py), 8.09 (d,
2H, H3,8), 7.81 (m, 3H, H5,6 + p-py), 7.60 (m, 10H, Ph), 7.34 (t,
2H, m-py), 3.11 (m, 12H, [9]aneS3). 13C{1H} NMR from HSQC
(D2O), δ: 152.9 (C2,9), 151.9 (o-py), 125.1 (C5,6) 130.5 (C3,8 +
p-py), 126.9 (Ph), 126.7 (m-py), 33.4 ([9]aneS3). ESI mass
spectrum: 649.1 m/z (calcd 649.3) [M – py + Cl]+. UV-vis (H2O):
λmax (ε, L mol–1 cm–1) = 467 (7800), 413 (6600) nm.
1
CHCl3 and diethyl ether and dried in vacuo. According to the H
NMR spectrum the product was pure 12. X-ray quality crystals of
12 were obtained by slow diffusion of diethyl ether into a
methanol solution of the complex. (Yield 14.4 mg, 70%).
Elemental analysis calcd for [C24H27N3Cl2RuS3] (MW = 624.48): C
46.07; H 4.35; N 6.72 Found: C 45.99; H 4.28; N 6.63. 1H NMR
(D2O), δ 8.91 (d, 2H, H8,8′), 8.65 (m, 4H, H3,3′+H4,4′), 8.14 (s,
2H, H5,5′), 8.06 (t, 2H, H7,7′), 7.86 (t, 2H, H6,6′), 3.53 (br s, 3H,
NH3), 2.63 (m, 12H, [9]aneS3). 13C{1H} NMR from HSQC (D2O),
δ: 134.6 (C3,3′), 131.2 (C7,7′), 129.3 (C5,5′), 129.2 (C6,6′),
126.8 (C8,8′), 120.11 (C4,4′), 33.2 ([9]aneS3). ESI mass
spectrum: 573.0 m/z (calcd 573.2) [M – NH3 + Cl]+.
[Ru([9]aneS3)(dppz)(py)]Cl2 (9): Same procedure as for 6,
using 82 mg of complex 4 (0.13 mmol). The product was
1
obtained in pure form according to the H NMR spectrum. Yield:
65%. Elemental analysis calcd for [C29H27N5Cl2RuS3] (MW
=
714.1): C 48.80; H 3.81; N 9.81. Found: C 48.72; H 3.71; N 9.73.
1H NMR (D2O), δ: 9.76 (d, 2H, H2,2′), 9.03 (d, 2H, o-py), 8.85 (s,
2H, H4,4′), 8.11 (t, 2H, H3,3′), 8.00 (t, 2H, p-py), 7.59 (m, 6H,
H5,5′ + H6,6′ + m-py), 3.05 (m, 12H, [9]aneS3). 13C{1H} NMR
from HSQC (D2O), δ: 155.3 (C2,2′), 152.2 (o-py), 139.6 (C4),
133.3 (C3,3′), 133.2 (p-py), 128.1 (C5,5′ + (C6,6′), 126.8 (m-py),
31.8 ([9]aneS3). ESI mass spectrum: 599.0 m/z (calcd 599.1) [M
– py + Cl]+. UV-vis (H2O): λmax (ε, L mol–1 cm–1) = 355 (11882),
423 (4117) nm.
[Ru([9]aneS3)(py)2Cl]Cl (13):
A
50 mg amount of
[Ru([9]aneS3)Cl2(dmso-S)] (0.13 mmol) was partially dissolved in
10 mL of EtOH; a 36 µL volume of pyridine (0.52 mmol) was
added and the mixture was refluxed for 5 h. During the heating a
clear yellow-orange solution was obtained, from which a yellow
precipitate started to form. After cooling the mixture to r.t., the
product was removed by filtration, washed with EtOH and diethyl
ether and dried in vacuo. (Yield 49.7 mg, 75%). The product was
[Ru([9]aneS3)(bq)(py)]Cl2 (10): Same procedure as for 6, using
83 mg of complex 5 (0.13 mmol). The product was obtained in
pure form according to the 1H NMR spectrum. Yield: 85%.
Elemental analysis calcd for [C29H27N3Cl2RuS3] (MW = 688.87): C
1
pure 13 according to the H NMR spectrum. Elemental analysis
calcd for [C16H22N2Cl2RuS3] (MW = 509.94): C 37.64; H 4.34; N
1
5.49 Found: C 37.78; H 4.44; N 5.58. H NMR (D2O), δ: 8.70 (d,
1
50.65; H 4.25; N 6.11 Found: C 50.73; H 4.32; N 6.21. H NMR
4H, o-py), 7.87 (t, 2H, p-py), 7.43 (t, 4H, m-py), 2.41 (m, 12H,
[9]aneS3 partially overlapped with the corresponding resonances
of 13aq). ESI mass spectrum: 475.0 m/z (calcd 475.1) [M]+.
(D2O), δ: 8.79 (s, 4H, H3,3′ + H4,4′), 8.32 (m, 4H, H8,8′ + o-py),
8.17 (m, 2H, H5,5′), 7.88 (m, 1H, p-py), 7.80 (m, 4H, H7,7′ +
H6,6′), 7.28 (t, 2H, m-py), 2.52 (m, 12H, [9]aneS3). 13C{1H} NMR
from HSQC (D2O), δ: 154.9 (o-py), 141.0 (C3,3′), 139.5 (p-py),
133.3 (C4,4′), 129.7 (C7,7′), 128.8 (C5,5′), 127.0 (m-py), 126.9
(C8,8′), 120.3 (C6,6′), 35.8 ([9]aneS3). ESI mass spectrum:
573.1 m/z (calcd 573.2) [M – py + Cl]+. UV-vis (H2O): λmax (ε, L
mol–1 cm–1) = 359 (14161), 378 (16207), 466 (2854), 493 (3771)
nm. X-ray quality crystals of 10 were obtained by slow diffusion
of diethyl ether into an EtOH solution of the complex.
CCDC 1588006 (for 7), 1588009 (for 10), 1588010 (for 11), and
1588014 (for 12) contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre.
Acknowledgements
trans,cis-[Ru(bq)Cl2(CO)2] (11): To
a 60 mg amount of
trans,cis,cis-RuCl2(CO)2(dmso-O)2 (0.16 mmol) dissolved in 3
mL of CHCl3, 1 eq of 2,2′-biquinoline (43.2 mg) was added and
Financial support from the University of Trieste (FRA2015
G. B. and E. A.), Fondazione Beneficentia Stiftung, and the
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
Thowfeik, D. J. Charboneau, I. Nieto, W. G. Dougherty, W. S. Kassel, T.
J. Dudley, E. J. Merino, E. T. Papish, J. J. Paul, J. Inorg. Biochem.,
2014, 130, 103-111.
Australian Research Council (DE13101650, A.R.) is
gratefully acknowledged. We wish to thank BASF Italia Srl
for a donation of hydrated ruthenium chloride.
[20] a) T. Respondek, R. N. Garner, M. K. Herroon, I. Podgorski, C. Turro, J.
J. Kodanko, J. Am. Chem. Soc., 2011, 133, 17164-17167; b) R.
Sharma, J. D. Knoll, P. D. Martin, I. Podgorski, C. Turro, J. J. Kodanko,
Inorg. Chem. 2014, 53, 3272-3274; c) N. Karaoun, A. K. Renfrew,
Chem. Commun. 2015, 51, 14038-14041; H. Chan, J. B. Ghrayche, J.
Wei,.A. Renfrew, Eur. J. Inorg. Chem. 2017, 1679-1686.
[21] A. Zamora, C. A. Denning, D. K. Heidary, E. Wachter, L. A. Nease, J.
Ruiz, E. C. Glazer, Dalton Trans. 2017, 46,2165-2173.
Keywords: Ruthenium • chelating diimine • photoactivated
chemotherapy • 2,2′-biquinoline • photo-induced ligand release
[1]
a) D. E. J. G. J. Dolmans, D. Fukumura, R. K. Jain, Nat. Rev. Cancer
2003, 3, 380-387; b) A. Juzeniene, Q. Peng, J. Moan, Photochem.
Photobiol. Sci. 2007, 6, 1234-1245; c) T. Debele, S. Peng, H.-C. Tsai,
Int. J. Mol. Sci. 2015, 16, 22094-22136.
[22] J. A. Cuello-Garibo, M. S. Meijer, S. Bonnet, Chem. Commun. 2017,
53 , 6768-6771.
[2]
[3]
G. Mayer, A. Heckel, Angew. Chem. Int. Ed. 2006, 45, 4900-4921.
K. Szaciłowski, W. Macyk, A. Drzewiecka-Matuszek, M. Brindell, G.
Stochel, Chem. Rev. 2005, 105, 2647-2694.
[23] G. Ragazzon, I. Bratsos, E. Alessio, L. Salassa, A. Habtemariam, R. J.
McQuitty, G. J. Clarkson, P. J. Sadler, Inorg. Chim. Acta 2012, 393,
230-238.
[4]
a) N. J. Farrer, L. Salassa, P. J. Sadler, Dalton Trans. 2009, 10690-
10701; b) N. A. Smith, P. J. Sadler, Phil. Trans. R. Soc. A 2013, 371,
20120519.
[24] The photo-triggered ligand release is more efficient compared to the
corresponding photoactivable organometallic compounds of the type
6-arene)(N-N)(py)][PF6]2 previously developed by Sadler and co-
workers (ref 16).
[5]
[6]
U. Schatzschneider, Eur. J. Inorg. Chem. 2010, 1451-1467.
E. Ruggiero, S. Alonso-de Castro, A. Habtemariam, L. Salassa, Struct.
Bond. 2015, 165, 69-108.
[25] I. Finazzi, I. Bratsos, T. Gianferrara, A. Bergamo, N. Demitri, G.
Balducci, E. Alessio, Eur. J. Inorg. Chem. 2013, 4743–4753.
[26] F. Barigelletti, A. Juris, V. Balzani, P. Belser, A. von Zelewsky Inorg.
Chem. 1983, 22, 3335-3339.
[7]
[8]
J. D. Knoll, C. Turro, Coord. Chem. Rev. 2015, 282-283, 110-126.
H. Yin, M. Stephenson, J. Gibson, E. Sampson, G. Shi, T. Sainuddin, S.
Monro, S. A. McFarland, Inorg. Chem. 2014, 53, 4548-4559.
C. Mari, V. Pierroz, S. Ferrari, G. Gasser, Chem. Sci. 2015, 6, 2660-
2686.
[27] E. Baranoff, J.-P. Collin, J. Furusho, Y. Furusho, A.-C. Laemmel, J.-P.
Sauvage Inorg. Chem. 2002, 41, 1215-1222.
[9]
[28] B. J. Goodfellow, V. Félix, S. M. D. Pacheco, J. P. de Jesus, M. G. B.
Drew Polyhedron 1997, 16, 393-401.
[10] V. H. S. van Rixel, B. Siewert, S. L. Hopkins, S. H. C. Askes, A.
Busemann, M. A. Sieglerb, S. Bonnet, Chem. Sci. 2016, 7, 4922-4929.
[11] M. Ethirajan, Y. Chen, P. Joshi, R. K. Pandey, Chem. Soc. Rev. 2011,
40, 340-362. A. B. Ormond, H. S. Freeman, Materials 2013, 6, 817-840.
[12] http://theralase.com/pressrelease/health-canada-approves-clinical-trial-
application-anti-cancer-drug/.
[29] J. Madureira, T. M. Santos, B. J. Goodfellow, M. Lucena, J. P. de Jesus,
M. G. Santana-Marques, M. G. B. Drew, V. Félix J. Chem. Soc., Dalton
Trans. 2000, 4422-4431.
[30] B. Machura, R. Kruszynski, Polyhedron 2007, 26, 3336–3342.
[31] a) J. A. G. Drake, D. W. Jones, Org. Magn. Res. 1982, 18, 42-44; b) A.
G. Osborne, R. Green, I. H. Sadler, D. Reed, Magn. Res. Chem. 1989,
27, 4-12.
http://theralase.com/pressrelease/theralase-research-published-
international-peer-reviewed-journal/.
[13] a) G. Shi, S. Monro, R. Hennigar, J. Colpitts, J. Fong, K. Kasimova, H.
Yin, R. DeCoste, C. Spencer, L. Chamberlain, A. Mandel, L. Lilge and
S. A. McFarland, Coord. Chem. Rev. 2015, 282-283, 127-138; b) J.
Fong, K. Kasimova, Y. Arenas, P. Kaspler, S. Lazic, A. Mandel, L. Lilge,
Photochem. Photobiol. Sci. 2015,14, 2014-2023; c) P. Kaspler, S. Lazic,
S. Forward, Y. Arenas, A. Mandela, L. Lilge, Photochem. Photobiol. Sci.
2016,15, 481-495.
[32] S. A. Moya, J. Guerrero, R. Pastene, R. Schmidt, R. Sariego, R. Sartori,
J. Sanz-Aparicio, I. Fonseca, M. Martinez-Ripoll, Inorg. Chem. 1994, 33,
2341-2346.
[33] R. Lalrempuia, M. R. Kollipara, Polyhedron 2003, 22, 3155–3160.
[34] P. Belser, A. von Zelewsky, Helv. Chim. Acta 1980, 63, 1675-1702.
[35] We found that in the 1H NMR spectra of complexes 5, 10, and 12, the
chemical shifts of the bq protons are remarkably affected by the nature
of the solvent (e.g. DMSO-d6 vs D2O), even though the sequence order
is maintained (ESI, Figure S9).
[14] a) B. S. Howerton, D. K. Heidary, E. C. Glazer, J. Am. Chem. Soc.
2012, 134, 8324-8327; b) E. Wachter, D. K. Heidary, B. S. Howerton, S.
Parkin, E. C. Glazer, Chem. Commun. 2012, 48, 9649-9651; c) D.
Havrylyuk, D. K. Heidary, L. Nease, S. Parkin, E. C. Glazer, Eur. J.
Inorg. Chem. 2017, 1687-1694.
[36] J. A. Cuello-Garibo, E. Perez-Gallent, L. van der Boon, M. A. Siegler, S.
Bonnet, Inorg. Chem. 2017, 56 , 4818-4828.
[37] The calculated UV-vis spectra of complexes 6 and 10 reproduce quite
well the general spectral features, even though the absorption bands
are red-shifted of ca. 60 nm compared to experimental spectra.
Calculations were performed in the vacuo and this might be one of the
reasons for this general shift. To be noted that in ref. 23 the singlet
excited state transitions for complex 6, calculated with a different
approach, had a numerically similar shift – but in the opposite direction
– with respect to the experimental transitions.
[15] J. D. Knoll, B. A. Albani, C. B. Durr, C. Turro, J. Phys. Chem. A 2014,
118, 10603–10610.
[16] a) S. Betanzos-Lara, L. Salassa, A. Habtemariam, P. J. Sadler, Chem.
Commun. 2009, 6622-6624; b) S. Betanzos-Lara, L. Salassa, A.
Habtemariam, O. Novakova, A. M. Pizarro, G. J. Clarkson, B. Liskova,
V. Brabec, P .J. Sadler, Organometallics 2012, 31, 3466-3479.
[17]
a) M. A. Sgambellone, A. David, R. N. Garner, K. R. Dunbar, C. Turro,
J. Am. Chem. Soc. 2013, 135, 11274-11282; b) Y. Chen, W. Lei, G.
Jiang, Y. Hou, C. Li, B. Zhang, Q. Zhou, X. Wang, Dalton Trans. 2014,
43, 15375-15384; c) J. D. Knoll, B. A. Albani, C. Turro, Acc. Chem. Res.
2015, 48, 2280-2287; d) T. Sainuddin, M. Pinto, H. Yin, M. Hetu, J.
Colpitts, S. A. McFarland, J. Inorg. Biochem. 2016, 158, 45-54.
[38] The photo-release of bq from 10 is not attributable to the longer
irradiation times: it occurs (in low amounts) also for short irradiation
times and increases with the irradiation time, i.e. it is not a photo-
bleaching of the complex.
[39] N. A. Smith, P. Zhang, S. E. Greenough, M. D. Horbury, G. J. Clarkson,
D. McFeely, A. Habtemariam, L. Salassa, V. G. Stavros, C. G. Dowson,
P. J. Sadler, Chem. Sci. 2017, 8, 395-404.
[18] a) T. Joshi, V. Pierroz, C. Mari, L. Gemperle, S. Ferrari, G. Gasser,
Angew. Chem. Int. Ed. 2014, 53, 2960-2963; b) C. Mari, V. Pierroz, A.
Leonidova, S. Ferrari, G. Gasser, Eur. J. Inorg. Chem. 2015, 3879-
3891; b)
[40] A. Zara, C. A. Denning, D. K. Heidary, E. Wachter, L. A. Nease, J.
Ruiza, E. C. Glazer, Dalton Trans. 2017, 46, 2165-2173.
[41] Z. Wang, H. Qian, S.-M. Yiu, J. Sun, G. Zhu, J. Inorg. Biochem. 2014,
131, 47-55.
[19] a) A. N. Hidayatullah, E. Wachter, D. K. Heidary, S. Parkin, E. C.
Glazer, Inorg. Chem., 2014, 53, 10030-10032; b) K. T. Hufziger, F. S.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
[42] J. E. Dickeson, L. A. Summers, Aust. J. Chem. 1970, 23, 1023-1027.
[43] F. Battistin, F. Scaletti, G. Balducci, S. Pillozzi, A. Arcangeli, L. Messori,
E. Alessio, J. Inorg. Biochem. 2016, 160, 180-188.
Umari, R. M. Wentzcovitch, J.Phys.:Condens.Matter 2009, 21, 395502;
http://arxiv.org/abs/0906.2569.
[49] G. J. Martyna, M. E. Tuckerman, J. Chem. Phys. 1999, 110, 2810-2821.
[50] D. Vanderbilt, Phys. Rev. B 1990, 41, 7892-7895.
[51] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865-
3868.
[44] W. Kabsch, Acta Cryst. D 2010, 66, 125-132.
[45] G. M. Sheldrick, Acta Cryst. A 2015, 72, 3-8.
[46] G. M. Sheldrick, Acta Cryst. 2008, 64, 112-122.
[47] P. Emsley, K. Cowtan, Acta Cryst. D 2004, 60, 2126-2132.
[48] P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni,
D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I. Dabo, A. Dal Corso, S.
Fabris, G. Fratesi, S. de Gironcoli, R. Gebauer, U. Gerstmann, C.
Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F.
Mauri, R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto, C.
Sbraccia, S. Scandolo, G. Sclauzero, A. P. Seitsonen, A. Smogunov, P.
[52] O. B. Malcioglu, R. Gebauer, D. Rocca, S. Baroni, Comput. Phys.
Commun. 2011, 182, 1744-1754.
[53] X. Ge, S. J. Binnie, D. Rocca, R. Gebauer, S. Baroni, Comput. Phys.
Commun. 2014, 185, 2080-2089.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201701392
European Journal of Inorganic Chemistry
MICROREVIEW
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
The light-induced release of
pyridine from a series of
water-soluble model
complexes of the general
formula
Photolabile Ru(II) complexes
Federica Battistin, Gabriele
Balducci, Jianhua Wei, Anna K.
Renfrew, and Enzo Alessio*
[Ru([9]aneS3)(chel)(py)]Cl2
([9]aneS3 = 1,4,7-
Page No. – Page No.
Photolabile
trithiacyclononane, chel =
chelating diimine) was
investigated with the aim of
establishing if this type of
complexes in the future
might be realistically used
in the photo-uncaging
strategy of photoactivated
chemotherapy (PACT).
[Ru([9]aneS3)(chel)(py)]Cl2
(chel = chelating diimine)
model complexes for light-
triggered drug release .
This article is protected by copyright. All rights reserved.