The key role of coligands in novel ruthenium(II)-cyclopentadienyl bipyridine derivatives: Ranging from non-cytotoxic to highly cytotoxic compounds

Article abstract of DOI:10.1016/j.jinorgbio.2015.06.015

A new family of eight ruthenium(II)-cyclopentadienyl bipyridine derivatives, bearing nitrogen, sulfur, phosphorous and carbonyl sigma bonded coligands, has been synthesized. Compounds bearing nitrogen bonded coligands were found to be unstable in aqueous

Full text of DOI:10.1016/j.jinorgbio.2015.06.015

The key role of coligands in novel ruthenium(II)-cyclopentadienyl bipyridine
derivatives: ranging from non-cytotoxic to highly cytotoxic compounds
Leonor C oˆ rte-Real, M. Paula Robalo, Fernanda Marques, Guilherme
Nogueira, Fernando Avecilla, Tiago J.L. Silva, Filipa C. Santos, A. Isabel
Tomaz, M. Helena Garcia, Andreia Valente
PII:
DOI:
Reference:
S0162-0134(15)30020-9
doi: 10.1016/j.jinorgbio.2015.06.015
JIB 9745
To appear in:
Journal of Inorganic Biochemistry
Received date:
Revised date:
Accepted date:
5 February 2015
15 June 2015
15 June 2015
Please cite this article as: Leonor C ˆo rte-Real, M. Paula Robalo, Fernanda Marques, Guil-
herme Nogueira, Fernando Avecilla, Tiago J.L. Silva, Filipa C. Santos, A. Isabel Tomaz,
M. Helena Garcia, Andreia Valente, The key role of coligands in novel ruthenium(II)-
cyclopentadienyl bipyridine derivatives: ranging from non-cytotoxic to highly cytotoxic
compounds, Journal of Inorganic Biochemistry (2015), doi: 10.1016/j.jinorgbio.2015.06.015
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

ACCEPTED MANUSCRIPT
The key role of coligands in novel ruthenium(II)-cyclopentadienyl bipyridine
derivatives: ranging from non-cytotoxic to highly cytotoxic compounds
a
b,c
d
a
Leonor Côrte-Real, M. Paula Robalo, Fernanda Marques, Guilherme Nogueira, Fernando
e
a
a
a
,a
Avecilla, Tiago J. L. Silva, Filipa C. Santos, A. Isabel Tomaz, M. Helena Garcia,** and
*
,a
Andreia Valente
a
Centro de Química Estrutural, Faculdade de Ciências da Universidade de Lisboa, Campo
Grande, 1749-016 Lisboa, Portugal.
b
Área Departamental de Engenharia Química, Instituto Superior de Engenharia de Lisboa, Rua
Conselheiro Emídio Navarro 1, 1959-007 Lisboa, Portugal.
c
Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco
Pais, 1049-001 Lisboa, Portugal.
d
Centro de Ciências e Tecnologias Nucleares, Instituto Superior Técnico, Universidade de
Lisboa, Estrada Nacional 10 (km 139.7) 2695-066 Bobadela LRS, Portugal.
e
Departamento de Química Fundamental, Universidade da Coruña, Campus de A, Zapateria
1
5071, A Coruña, Spain.
*
Corresponding author: Tel.: +351 217500898.
e-mail address: amvalente@fc.ul.pt (A. Valente).
**
Corresponding author: Tel.: +351 217500088.
e-mail address: mhgarcia@fc.ul.pt (M.H. Garcia).
1

ACCEPTED MANUSCRIPT
Abstract: A new family of eight ruthenium(II)-cyclopentadienyl bipyridine derivatives, bearing
nitrogen, sulphur, phosphorous and carbonyl sigma bonded coligands, has been synthesized.
Compounds bearing nitrogen bonded coligands were found to be instable in aqueous solution,
while the others presented appropriate stabilities for the biologic assays and pursued for
determination of IC values in ovarian (A2780) and breast (MCF7 and MDAMB231) human
5
0
cancer cell lines. These studies were also carried out for the [5:HSA] and [6:HSA] adducts (HSA
Human Serum Albumin) and a better performance was found for the first case. Spectroscopic,
=
electrochemical studies by cyclic voltammetry and Differential Functional Theory calculations
allowed to get some understanding on the electronic flow directions within the molecules and to
find a possible clue concerning the structural features of coligands that can activate bipyridyl
5
ligands towards an increased cytotoxic effect. X-ray structure analysis of compound [Ru(η -
C H )(bipy)(PPh )][PF ] (7; bipy = bipyridine) showed crystallization on C2/c space group with
5
5
3
6
5
+
two enantiomers of the [Ru(η -C H )(bipy)(PPh )] cation complex in the racemic crystal
5
5
3
packing.
1
. Introduction
Cancer claims the lives of millions of people worldwide every year. The high mortality inherent
to cancer conditions together with the very debilitating side effects caused by the drugs in
clinical use (mainly platinum-based drugs such as cisplatin) are self-explicative about the
urgency to find new chemotherapeutic options. In this frame, ruthenium has been seen as a
promising alternative metal due its unique chemistry in aqueous solution[1,2]. The two most
well-known examples of this class of compounds, that entered Phase II clinical trials, are
imidazolium trans-[tetrachloro(dimethyl sulfoxide)(1H-imidazole)ruthenate(III)] (NAMI-A)[3]
2

ACCEPTED MANUSCRIPT
and indazolium trans-[tetrachlorobis(1H-indazole)ruthenate(III)] (KP1019)[4]. Though these two
inorganic ruthenium(III) compounds present similar chemical structures, they exhibit completely
different chemotherapeutic behaviour. While NAMI-A shows a marked efficiency against the
formation of metastases[5], KP1019 has cytotoxic activity against a wide panel of human tumour
models[4]. During the last decade other structurally different families of ruthenium compounds
have been synthesized, some of them also exhibiting interesting potential. This is the case of the
II
6
organometallic mononuclear piano-stool ‘Ru (η -C H )’ complexes[6,7]. The literature
6
6
II
5
concerning the isoelectronic ‘Ru (η -C H )’ fragment has been less explored, although important
5
5
results have been obtained combining this fragment with selective kinase inhibitors[8,9]. Our
research group has been engaged in the synthesis of new ruthenium compounds based on the
II
5
+
[
Ru (η -C H )(PP)(L)] structure (PP = monodentate or bidentate phosphane; L = monodentate
5 5
or bidentate heteroaromatic ligand)[10 -16]. The results obtained for this family of compounds
show cytotoxic activities in the nano- and sub-micromolar range against several human cancer
cell lines (e.g. MiaPaCa, LoVo, PC3, HL-60, MCF7, HT29, A2780, A2780cisR) [10-16], placing
II
them among the best cytotoxic Ru -arene complexes. In addition, these complexes present, in
most cases, lower IC values than cisplatin. Factors that modulate the cytotoxic activity of Ru-
5
0
based drugs are numerous, and seem to be dependent on the family compounds under scrutiny.
We are interested in pinpointing/understanding whether (and how) structural factors control and
potentiate the anti-tumour activity of ruthenium-cyclopentadienyl complexes. In this frame, the
II
5
structure-activity performance of a new family of complexes of general formula [Ru (η -
C H )(bipy)(L)][PF ] (bipy = 2,2’-bipyridine; L = imidazole (1), 1-benzyl-1-imidazole (2), 4-
5
5
6
(1H-imidazol-1-yl)phenol (3), 1-(4-methoxyphenol)-1H-imidazole (4), dimethyl sulfoxide (5),
II
5
carbon monoxide (6) and triphenylphosphane (7)) was designed using TM34, [Ru (η -
3

ACCEPTED MANUSCRIPT
C H )(bipy)(PPh )][CF SO ] [12], as a model. Thus, the introduction of coligand diversity was
5
5
3
3
3
achieved by σ-coordination of different atoms, namely N, S, C and P. In addition, the precursor
II
5
of these compounds, [Ru (η -C H )(bipy)(NCCH )][PF ] 8 was isolated and characterized for
5
5
3
6
comparison and better understanding of the spectroscopic and electrochemical data.
2
2
. RESULTS AND DISCUSSION
.1. Synthesis of the Ru(II) complexes
5
Mononuclear complexes of the general formula [Ru( -C H )(bipy)(L)][PF ] with L = imidazole
5
5
6
(1), 1-benzyl-1-imidazole (2), 4-(1H-imidazol-1-yl)phenol (3) and 1-(4-methoxyphenol)-1H-
imidazole (4), were prepared, as shown in Scheme 1, by ligand substitution from the parent
5
cationic complex [Ru( -C H )(bipy)(NCCH )][PF ] 8 in dichloromethane, at room temperature,
5
5
3
6
5
in the presence of a slight excess of the corresponding ligand. For complex 5, [Ru( -
C H )(bipy)(NCCH )][PF ] was dissolved in ca. 200 µL of DMSO and 15 mL of water. After
5
5
3
6
solvent removal, the complex was recovered as an orange compound. In the case of compound 6,
a flow of CO was passed through the solution for ca. 15 min, allowing the formation of the
complex as an orange-brownish product. Compound 7 was synthesized by halide abstraction
5
followed by ligand substitution from [Ru(η -C H )(PPh ) Cl]. The new compounds were
5
5
3 2
recrystallized by slow diffusion of n-hexane or diethyl ether in dichloromethane, acetonitrile or
methanol solutions, giving crystalline orange to red compounds. All the compounds are fairly
stable to air and moisture in the solid state and were obtained in good yields (50-90%).
The formulation and purity of all the new compounds is supported by analytical data, FT-IR
1
13
31
spectroscopy, H, C, P NMR spectroscopic data and elemental analyses. The solid state FT-
4

ACCEPTED MANUSCRIPT
IR spectra (KBr pellets) of the complexes presented a large number of bands which identify the
presence of the various fragments of the molecules. Characteristic bands were used to confirm
-1
the presence of the cyclopentadienyl (ca. 3140-3000 cm ), the bipyridine ligand (ca. 1520-1400
-1
-
-1
cm ) and the PF anion (≈ 840 and 560 cm ) in all the studied complexes. The infrared
6
spectrum of compound 5 confirms the coordination of DMSO ligand through the sulphur atom
-1
due to the presence of a band at 1094 cm attributed to 
and the absence of any significant
S=O
-1
vibration in the 920-930 cm range, characteristic of coordination by the oxygen atom. It has
been reported in the literature that in ruthenium complexes, a sulphur–bounded DMSO exhibits a
-1
distinctive S=O band between 1080 and 1150 cm , depending on the electronic nature of the
other ligands in the coordination sphere, while an oxygen-bounded form shows a signal of
-1
slightly lower energy between 900 and 1000 cm [17-20]. The positive shift in the  of ca. 44
S=O
-1
cm can be explained by the inversed polarization of the sulfoxide -bond effect [21]. Also, in
the case of compound 6, the coordination of carbon monoxide can be confirmed by the infrared
-1
-1
stretching vibration occurring at 1970 cm . Besides, the observed shift of ca. -200 cm in the
-1
νC≡O is indicative of a strong π-backdonation effect (νC=O, gas = 2143 cm [23] vs. ν
= 1970
C≡O
-1
1
cm in compound 6) which was further corroborated by our H NMR studies (see below).
Scheme 1.
1
Analysis of the overall H NMR results presented on Table S1 showed that, comparing with
5
[
Ru( -C H )(NCMe) ][PF ], the substitution of the acetonitrile ligands by bipyridine and L
5
5
3
6
5
ligands leads to a general deshielding on  -C H protons, which extension is related with the π-
5
5
acceptor capability of the L coligands. For compounds 1-4 the cyclopentadienyl ring displayed
5

ACCEPTED MANUSCRIPT
signals in the characteristic range of monocationic ruthenium(II) (≈ 4.5 ppm). In the case of
compounds 5-7 a marked deshielding of 0.6-1.2 ppm on these protons was observed, especially
in the case of compound 6, presenting the CO coligand. This observation clearly corroborates the
π-backdonation effect found in our FT-IR studies. The chemical shifts observed for the protons
of coordinated bipyridine on compounds 1-6 are very close of the uncoordinated ligand ones
revealing that the deshielding expected upon a dative σ coordination of bipyridine ligand is
compensated by the shielding due to π-back electronic flow towards the bipyridyl coordinated
molecule (Table S1). Surprisingly compound 7 revealed a shielding of 0.33 ppm on the bipyridyl
ortho protons. This observation might be explained by the effect of one of the phenyl ring
current on the ortho bipy ligand, which adequate orientation and close proximity can be observed
5
+
on the ORTEP plot for the cation complex [Ru(η -C H )(bipy)PPh ] (see below). Relatively to
5
5
3
the effect observed on the coordination of the coligands, complexes 1-4, showed a shielding of
0.5 ppm on the H proton of the coordinated imidazole ring in all cases, while the remaining
~
4
protons are almost unchanged. This electronic density at the imidazole ligands also suggests the
existence of some π-backdonation effect to this coligand. For complex 5 (bearing the S-bound
1
13
DMSO coligand), a downfield shift of H and C resonances for the methyl groups upon
coordination is consistent with the sulfur coordination mode[19] and corroborates the obtained
5
FT-IR results. Globally it is observed that  -C H ring releases the electronic flow through the
5
5
ruthenium centre toward coligands and bipyridyl ligand. Thus, a general trend on this electronic
flow towards the L coligand was observed for compounds 1-6 being this effect stronger in the
case of CO. The additional electronic effect originated by PPh on compound 7 leads to an
3
asymmetric charge distribution on the N-heteroaromatic bipyridyl ligand, and consequently an
important shielding on the bipyridyl ortho protons is observed.
6

ACCEPTED MANUSCRIPT
2
.2. UV-visible (UV-Vis) studies
5
Optical absorption spectra of these new seven [Ru( -C H )(bipy)(L)][PF ] complexes together
5
5
6
-6
-3
with all ligands were recorded in 10 – 10 M dichloromethane solutions (see Experimental
Section). Figure 1a shows the spectrum of compound 1 in dichloromethane and typifies the
general behaviour of all the compounds. All the studied complexes showed two intense bands in
the UV region attributed to π – π* electronic transitions occurring in the aromatic rings and a
3
-1
-1
quite complex pattern occurring in the visible region (400-600 nm) with max ~ 5x10 M cm
that can be related to several metal to ligand charge transfer bands (MLCT), from Ru 4d to π* N-
heteroaromatic rings for compounds 1-4. This pattern was also observed for compounds 5-7 and
it is also compatible with the occurring MLCT effect originated by DMSO and CO coligands. In
order to infer about the charge transfer character of these bands, further electronic spectra were
obtained in acetonitrile and DMSO (see experimental section). Although no significant
solvatochromic effect had been observed within the used polarity solvent range, an elucidation
on the complexity on the lower energy band was noticeably found. To get some understanding
about the influence of the several coligands presenting quite different donor-acceptor abilities
(
N-heteroaromatic derivatives, CO, DMSO and PPh ) a detailed comparison was carried out to
3
infer about the shift of these MLCT bands (Figure 1b). Compounds 1-4 containing N-
heteroaromatic coligands presented MLCT bands occurring at the lowest energy. In Figure 1b
one can compare the behaviour of compounds 1 , 5 and 7 (with medium π acceptor coligand) and
6
(with the best π acceptor coligand). The shifts observed on the energy of the MLCT band seem
to be related with the π-backdonation from Ru to the coligand L (in accordance with Cp chemical
7

ACCEPTED MANUSCRIPT
shifts).Thus, compounds presenting the highest π-backdonation present the lowest energy MLCT
band (1-4 < 7 < 5 < 6).
Figure 1.
The newly synthesized complexes are soluble in dichloromethane, acetonitrile, acetone and
DMSO. Their solubility in water/aqueous media is low to moderate, but this drawback is easily
overcome by adding a low amount of DMSO as a co-solvent.
Envisaging the use of these new compounds as cytotoxic agents and their study in human cancer
cell lines, their stability and behaviour in aqueous solution was studied in HEPES buffer at pH
7
.4, using 2-3% DMSO, by UV-Vis spectroscopy. For compounds 1-4, one can readily observe
an immediate colour change once the solution is prepared: firstly the compounds were dissolved
in DMSO and when upon dilution in water or aqueous buffered medium they changed from their
original colour (orange or red) to yellow. This transformation in compound 2 was somehow
slower and it was monitored by UV-Vis spectroscopy. As showed in Figure S1a, the MLCT
-1
band occurring at ~ 455 cm completely vanishes after 45 minutes. Moreover, the electronic
spectrum of this solution was found superposed with the one of compound 5 showing the
substitution of the imidazole ring by DMSO (Figure S1b). Due to instability of complexes 1-4 in
DMSO/water system, our in vitro biological studies were only carried on using compounds 5-7,
which were found to present adequate stability. Compound 7 showed some precipitation after ca.
4
h at concentrations suitable for UV-Vis spectroscopy; however, when working at lower
concentrations (e.g. fluorescence studies and cellular viability studies), no precipitation was
observed.
8

ACCEPTED MANUSCRIPT
5
2
.3. Single crystal structure of [Ru(η -C H )(PPh )(bipy)][PF ] 7
5
5
3
6
5
[
Ru(η -C H )(bipy)(PPh )][PF ]·(CH ) CO 7 crystallized from acetone-d solution (used for the
5 5 3 6 3 2 6
NMR studies) as orange prisms (crystal dimensions 0.56 x 0.34 x 0.23 mm). Figure 2 shows an
5
+
ORTEP representation of [Ru(η -C H )(bipy)(PPh )] cation 7.
5
5
3
Figure 2.
-
The asymmetric unit contains one cationic ruthenium complex, one PF anion and one acetone
6
molecule. In the molecular structure, the ruthenium centre adopts the expected "piano stool"
distribution formed by the ruthenium-Cp unit bound to the nitrogen atoms of the bipy ligand.
One phosphane group occupies the remaining coordination position. The distance for Ru-P bond
is Ru(1)-P(1) = 2.3021(5) Å, and distances for Ru-N bonds are Ru(1)-N(1) 2.0731(14) Å and
Ru(1)-N(2) 2.0712(14) Å. The distance between Ru and the centroid of the π-bonded
cyclopentadienyl moiety is 1.821 Å to Ru centre (ring slippage 0.045 Å). The mean value of the
Ru-C bond distance is 2.1848(18) Å. Table S2 contains selected bond lengths and angles for
compound 7. π-π stacking interactions are absent in the structure since the distance between the
bipyridine ligand ring N(1)-C(24)-C(25)-C(26)-C(27)-C(28) and the phosphane phenyl ring
C(12)-C(13)-C(14)-C(15)-C(16)-C(17) is 4.140 Å. Contrarily, for the reported structure of cation
5
+
-
complex [Ru(η -C H )(bipy)(PPh )] [CF SO ] ,[12] - stacking interactions can be observed.
5
5
3
3
3
5
In addition, X-ray structure of 7 shows two enantiomers for the cation complex [Ru(η -
+
C H )(bipy)PPh ] in the racemic crystal (space group C2/c). The chirality is due to a twist of the
5
5
3
5
+
PPh and Cp units. The cation complex [Ru(η -C H )(bipy)(PPh )] presents a mirror plane
3
5
5
3
9

ACCEPTED MANUSCRIPT
which contain P, Ru and the centroid of Cp ring (see Figure S2)[25]. The two enantiomers
5
+
present in the [Ru(η -C H )(bipy)PPh ] crystal are herein reported for the first time.
5
5
3
2
.4. Electrochemical experiments
5
The electrochemical behaviour of this family of ruthenium complexes [Ru( -
C H )(bipy)(L)][PF ] was studied by cyclic voltammetry in dichloromethane and in acetonitrile
5
5
6
solutions containing ammonium hexafluorophosphate as supporting electrolyte (Table S3 and
Table S4, see supplementary information).
No redox processes for the free ligands were observed in dichloromethane, with exception of 1-
(4-methoxyphenyl)-1H-imidazole (1-MPI, L for complex 4), which showed a reductive process
5
+
at E = -0.84V. The cyclic voltammogram of complex [Ru(η -C H )(bipy)(4-IMP)] 3 (4-IMP =
pc
5
5
4
-(1H-imidazol-1-yl)phenol) (Figure 3) is representative of the general behaviour of this family
in dichloromethane. All the complexes showed one quasi-reversible redox process at positive
potentials, attributed to the Ru(II)/Ru(III) redox pair, with a lower intensity for the cathodic wave
when compared with the anodic current. Nevertheless, when the scan rate direction is reversed
immediately after the oxidation potential this decrease is less pronounced; this might indicate
that the ruthenium oxidation is followed by a chemical reaction leading to non reductible species.
For complexes 2 and 5, an irreversible reductive process at negative potentials (E = -1.13 and -
pc
1
.26 V, respectively) was observed and attributed to a ligand-based reductive process.
The effect of the different coligands in the electronic environment of the ruthenium centre was
assessed by comparison of the oxidation potentials for complexes 1-4 with complexes 5 and 7,
bearing the dimethyl sulfoxide and triphenylphosphane ligands respectively, instead of an
imidazole derivative. In general, the presence of the monodentate imidazole based ligand
lowered the oxidation potential of ruthenium centre, up to 500 mV. However, the substitution in
1
0

ACCEPTED MANUSCRIPT
the imidazole ring did not lead to a substantial change on the electronic environment of the metal
centre, since the Ru(II)/Ru(III) potential remained almost unchanged (E1/2 ~ 0.65 to 0.68 V).
5
Finally, for complex [Ru(η -C H )(bipy)(CO)][PF ] 6, the absence of any redox processes at the
5
5
6
positive potentials range is in agreement with our spectroscopic data. The strong -backdonation
effect pulls the electronic density from the ruthenium center to CO coligand, shifting the
ruthenium oxidation process out of the solvent potential window. In accordance, a quasi-
reversible redox process within the bipyridine ligand was observed at E1/2 = -1.32 V.
Figure 3.
Concerning the electrochemical behaviour in acetonitrile, for the imidazole-based ligands (1-
benzyl-1-imidazole (1-BI), 4-(1H-imidazol-1-yl)phenol) (4-IMP)) and 2,2’-bipyridine no redox
processes were observed. Imidazole (ImH) and 1-(4-methoxyphenyl)-1H-imidazole (1-MPI)
showed two irreversible processes each one.
For the ruthenium(II) complexes, the behaviour is much more complicated leading generally to
several undefined redox processes (some of them less intense than the one attributed to the metal
centre), which indicates that the complexes are less stable in this solvent. In general, all the
complexes exhibited one irreversible or quasi-reversible oxidation process at the positive
potentials, attributed to the Ru(II)/Ru(III) redox pair (Figure 4) with a less intense cathodic
wave, in accordance with the previous behaviour in dichloromethane. At the negative potential
range, complexes 1-4 display one or two irreversible reductive processes which can be attributed
to reductions within the N-heteroaromatic coligands. For complexes 5-8, the irreversible (5 and
1
1

ACCEPTED MANUSCRIPT
8
) or quasi-reversible (6 and 7) reductive processes found around -1.40 V, can be assigned to the
bipy ligand.
The lability observed for the heteroaromatic ligands during the stability studies (see Figure S1)
make us consider the possible replacement of these ligands by the acetonitrile coordinative
solvent during the electrochemical experiments. Thus, the electrochemical behaviour of the
5
5
parent
compound
[Ru(η -C H )(NCMe) ][PF ]
and
the
precursor
[Ru(η -
5
5
3
6
C H )(bipy)(NCMe)][PF ] (8), (which was isolated and characterized for this purpose, see
5
5
6
Experimental Section) was also studied in the same experimental conditions. Comparison of the
5
cyclic voltammograms of complexes 1 and 3 with [Ru(η -C H )(NCMe) ][PF ] showed that
5
5
3
6
substitution of all ligands by acetonitrile solvent molecules did not occur. Nevertheless,
5
comparison with [Ru(η -C H )(bipy)(NCMe)][PF ] 8 showed that the second oxidative process is
5
5
6
due to the formation of these latter species by replacement of the N-heteroaromatic coligand by
5
one acetonitrile molecule (Figure 4). Complex [Ru(η -C H )(bipy)(CO)][PF ] 6 did not show
5
5
6
any oxidation process in the solvent window compatible with its redox behaviour in
dichloromethane.
Our electrochemical results showed that substitution of the imidazole based ligands by DMSO,
CO, acetonitrile or PPh provides an increase on the oxidation potential of the ruthenium centre.
3
Figure 5 compares the oxidation potentials of Ru(II)/Ru(III) pair and shows the trend ability of
the ruthenium centre for oxidation: ImH ~ 1-BI > 4-IMP ~ 1-MPI > NCCH > DMSO > PPh
3
3
>>CO. This trend revealed the net electronic density at the ruthenium centre which is a result of
the different σ and π ability of the ligands. Therefore compounds possessing the most demanding
coligands (CO, PPh and DMSO) present the highest redox oxidation potentials.
3
1
2

ACCEPTED MANUSCRIPT
Figure 4.
Figure 5.
2
.5. Density Functional Theory (DFT) Studies
In order to further support our spectroscopic data we performed DFT calculations in cationic
complexes 1 and 5-7. The optimized structures are shown in Figure S3 and the relevant
structural parameters are shown in Table 1. Comparison of the crystallographic data of 7 with its
DFT optimized structure showed a good agreement in the Ru-Cp and Ru-N bond distances, with
differences of only +0.08 Å and +0.05 Å, respectively. DFT calculations predicted a slightly
longer Ru-P bond distance, where a difference of +0.17 Å was observed. This value is however
in the same range found for other CpRu related complexes[24]. Hence, it is fair to state that the
rather modest 6-31G(d,p)/M06L-DFT level of theory is adequate to compute the geometries of
this family of complexes.
Table 1.
π-backdonations from the organometallic fragment to the Cp, L and bipy ligands were assessed
by charge decomposition analysis (CDA) and are also showed in Table 1. Noteworthy is the fact
that both the triphenylphosphane and CO ligands in complexes 7 and 6, respectively, originate
relatively high estimated π-backdonations to the corresponding coligands L, proving the
enhanced stability of these two compounds. π-backdonations to the bipy fragment
+
(
[CpRuL] →bipy) were estimated in 0.543, 0.658, 0.598 and 0.584 electrons for 1, 5, 6 and 7,
respectively, showing that DMSO ligand (poor π-acceptor) leads to the highest π-backdonation
1
3

ACCEPTED MANUSCRIPT
to the bipy fragment. Nevertheless this effect is not so noticeable in the proton NMR chemical
shifts of the coordinated bipy.
The electronic spectra of the molecules were computed under the Time-dependent density
functional theory (TD-DFT) formalism, with the inclusion of solvation effects by the PCM
model. Acetonitrile was the solvent chosen for these studies. As an example, the TD-DFT
spectrum of compound 7, alongside with its experimental spectrum in acetonitrile, are depicted
in Figure 6. In Table 2 are selected main vertical optical transitions that contribute to the overall
calculated spectrum.
Figure 6.
Our TD-DFT results confirmed the overall behaviour of the complexes, where an intense UV
band was obtained, together with several low intensity bands in the visible region. The estimated
wavelength of the UV absorption band is very close to the experimental value (λexp = 290 nm;
λcalc = 281 nm). An analysis of the molecular orbitals involved allowed the attribution of this
band to a combination of a πphosphane - π*bipy transition (inter ligand transition) with a d – π*
Ru
bipy
transition (MLCT). The overall shape of the spectra in the visible region is similar to the
calculated spectra. The calculated maxima of absorption of the three visible bands are also in
good agreement with the experimental maxima absorption, for which differences of only 3, 9 and
3
nm were obtained when going from the highest to the lowest energy transitions (Table 2),
respectively. These bands were all attributed to MLCT charge transfers, dRu – π*bipy transitions,
together with dRu – π*phosphane transitions. In fact, it is not uncommon that UV-Vis bands of
organometallic complexes cover several transitions. This is the case of the present complex
1
4

ACCEPTED MANUSCRIPT
where the calculated transition at 356 nm comprises four vertical excitations, as can be seen in
Table 2.
Table 2.
2
.6. Biologic Studies
2
.6.1. In vitro cytotoxicity
The cytotoxic activity of compounds 5-7 was assessed by the colorimetric MTT assay with three
human tumour cell lines, namely A2780 (ovarian carcinoma), MCF7 (breast adenocarcinoma)
and MDAMB231 (breast adenocarcinoma derived from metastatic site). This panel of cell lines
was selected in view of their different responses to cisplatin, the drug in clinical use, and taking
also into account that they represent the most frequent cancer diseases diagnosed in women, i.e.,
breast (MCF7, ERα+ and MDAMB231, ER-, PR-, HER2-) and ovarian (A2780) cancer
conditions. The IC values were obtained from experiments performed after a 72 h treatment
5
0
with the complexes within the 0.1 µM-200 µM concentration range. The IC values obtained for
50
cisplatin were used for comparison. Results are summarized in Table 3. As can be observed, the
presumably “spectator” coligand imparts impressive differences on the cytotoxic potency of the
compounds. Compound 5, bearing the DMSO coligand, was non-cytotoxic; replacing this ligand
by CO, we obtained compound 6 that showed a moderate cytotoxicity against the ovarian cancer
cells; finally, when the coligand is the triphenlyphosphane, the resulting compound 7, was highly
cytotoxic against all three cancer cell lines tested, as expected from our previous results with its
5
analogue TM34, [Ru( -C H )(bipy)(PPh )][CF SO ] [12,14]. Phosphane coligands, and
5
5
3
3
3
particularly PPh , have been related to the increasing cytotoxicity on ruthenium based systems as
3
recently reported for [Ru(Cl)(PPh )(Lig-N)], [Ru(Cl) (Lig-N)] (where Lig-N = pyridine derivate)
3
2
1
5

ACCEPTED MANUSCRIPT
and [Ru(Cl)(PPh ) ] derivatives[26]. In this frame, the cytotoxic effect of the uncoordinated
3
2
triphenylphosphane, was studied on MCF7 cancer cell line in the same experimental conditions
as for complexes 5-7. The results show that IC50 for uncoordinated PPh is 112.6 ± 33 µM
3
(Figure S4), showing that the strong cytotoxicity of compound 7 is, in fact a synergy between all
5
the compound components, where the fragment {Ru( -C H )} plays a crucial role on the
5
5
electronic interactions. To sum up, the differences on the electronic activation of the bipy ligand
see above) imparted by the co-ligand structural diversity seem to have a direct correlation with
the cytotoxic properties of the complexes.
(
Table 3.
2
.6.2. In vitro studies with Human Serum Albumin (HSA)
Binding to HSA
HSA is the most important non-specific transport vehicle in the human blood plasma. Its
important ability to bind both endogenous metabolic compounds and exogenous therapeutic
drugs is well-known and strongly affects drug bioavailability and modifies the retention time in
vivo[29-31]. The assessment of complex binding to HSA is an important first approach to drugs
pharmacokinetics. In addition, due to the known enhanced permeability and retention effect
(EPR effect) selective extravasation and retention of macromolecular drugs by solid tumours
could be achieved[32]. In this context, HSA could potentially increase the cytotoxicity of
complexes 5 and 6 through selective extravasation and retention by the tumour tissue[32].
Binding to HSA was investigated using fluorescence spectroscopy, using the intrinsic emission
of HSA mainly due to its tyrosine and tryptophan residues. HSA contains a single tryptophan
1
6

ACCEPTED MANUSCRIPT
moiety, Trp214 (located in the subdomain II.A), which can be selectively excited at λexc = 295
nm and used as an intrinsic structural probe in the protein[29,33]. Trp214 is very sensitive to its
local environment and its emission easily reflects small changes in the vicinity of the indole ring
[
33], which makes it an attractive probe for detecting interactions between HSA and a
metallodrug. In this study, fatty acid-free HSA was used to evaluate the binding ability of the
protein for the ruthenium complexes 5-7. The low solubility of these complexes required the use
of a small amount of DMSO as a co-solvent. A DMSO content as low as 2% quenches the
fluorescence emission of Trp214 in ca. 10% (no spectral shifts being observed). In this study,
care was taken to add, at the same time, the same volume of DMSO to both “HSA alone” and
“complex-HSA” samples, to ensure the same DMSO content in “HSA alone” solutions (prepared
from stock solutions with no DMSO), in samples and in blanks. In HSA, the maximum emission
intensity for albumin Trp214 is observed at 334 nm[33]. As an example, Figure 7a shows the
effect of increasing concentrations of compound 5 on the protein emission: a quenching of the
fluorescence intensity is observed, with no shift in the maximum λ . The same behaviour was
em
observed for complexes 6 and 7. All complexes exhibit a strong absorption band at 295 nm (λexc)
that tails to the visible range, and absorbance is still considerably high at 340 nm (λem max, see
Figure 1). Hence, data must be corrected for inner filter effects (IFE) at both excitation and
emission wavelengths to account for a decrease in the fluorescence intensity that is not due to a
real binding interaction[54,33]. The relative intensity emitted at 340 nm in the presence of
complexes 5-7 and corrected for IFE is depicted in Figure 7b. This wavelength was chosen
because it is near of the maximum emission wavelength (345 nm) but after the water Raman
scattering peak, and the signal-to-background ratio is much more favourable than it would be if
the actual maximum emission wavelength was used. It can be clearly seen in Figure 7b that the
1
7

ACCEPTED MANUSCRIPT
presence of complexes 5-7 quenches the Trp214 emission in a concentration-dependent manner,
which is consistent with binding to the protein (although real saturation conditions could not be
attained due to solubility constrains). In addition, the extent of quenching observed is greater for
complex 6 bearing the CO coligand.
The very small changes detected in the UV-Vis electronic spectrum of these compounds in the
presence of the protein support the fact that the parent complex binds as a whole, with no loss of
any of its coligands.
Figure 7.
Effect of HSA of the cytotoxicity of compounds 5 and 6
As explained above, serum proteins play a crucial role in the transport and delivery of
antitumoral metallodrugs, and are frequently involved in their mechanism of action. In the
prospective that the majority of metallodrugs are administered intravenously, interactions with
HSA can be determinant for their biodistribution and can affect the drugs’ biological activity. In
this frame, the cytotoxic activity of the complexes 5 and 6 when bound to HSA was evaluated
aiming at obtaining an increase in their anticancer activity. Experiments were carried out with
MCF7 cells at 72 h incubation time. Several controls were used in these assays, namely cells
with no treatment (negative control), cells treated with HSA alone and cells treated with
complexes 5 and 6 in the absence of HSA. On the case of complex 5 (Figure 8), binding to HSA
resulted in an increase of ca. 38% on the cytotoxic activity against MCF7 cells, probably as a
result of a better uptake into the cells. However, one should not neglect that since this compound
1
8

ACCEPTED MANUSCRIPT
presents a high IC , the quantities of HSA needed to achieve a 1:0.5 ratio are also high, causing
5
0
by its turn a decrease on the cellular viability of the MCF7 cells. For complex 6, the incubation
with HSA had no effects on the cytotoxicity of the compound (Figure 8), indicating that the
activity of the complex remains relatively unaffected when bound to HSA. These results show
that the interaction of these complexes with HSA in vivo will likely have a positive outcome in
what concerns activity.
Figure 8.
3
. Conclusions
5
A new family of [Ru(η -C H )(bipyridine)(L)][PF ] complexes has been designed and
5
5
6
synthesized in view to understand the role of the σ bonded coligands (L) in the cytotoxicity of
the new compounds and to find possible structure-activity correlations. A set of compounds
presenting N-bonded (1-4, imidazole derivatives), S-bonded (5, DMSO), C-bonded (6, CO) and
P-bonded (7, PPh ) coligands was studied and the spectroscopic, electrochemical and DFT data
3
were correlated to understand the electronic features of these molecules. Taking all the results in
consideration, it has been clearly shown that the coligand L has a crucial role in the fine-tuning
of the electronic density at ruthenium centre as a result of their different π acceptor character.
Thus, N-imidazole derivatives, DMSO and CO coligands pulled the electronic density from the
5
+
{
Ru(η -C H )(bipy)} fragment via a π-backdonation interaction (the most drastic case was CO).
5 5
The ability of coligand PPh as π acceptor is not far from the other coligands, with the obvious
3
exception of CO, when experimental data are compared, but surprisingly compound 7 showed
higher electronic density on bipy rings. An explanation for this effect can possibly be accounted
by the influence a phenyl ring current of the phosphane in the protons of bipyphenyl (see X ray
1
9

ACCEPTED MANUSCRIPT
section) leading to a higher electronic density of at the bipy ligand. As consequence, the
bipyridyl ligand might be more activated for expected electronic interactions with the charged
cell membranes, which have been on the basis of the good anticancer performance of our
5
{
Ru(η -C H )} based family of compounds[28,34]. In good agreement with this evidence, from
5
5
compounds 5-7, compound 7 showed the best IC values for the ovarian (A2780) and breast
50
(MCF7 and MDAMB231) human cancer cell lines. Instability in HEPES buffer excluded
compounds 1-4 from this cytotoxic evaluation study. Although our studies do not lead to a
straightforward conclusion about structure – activity relations one can foresee that for this
5
+
general family of [Ru(η -C H )(L)(bipy)] the best cytotoxic effect could be associated with a
5
5
better electronic activation of the bipy as already observed in analogous compounds[55]. Thus,
the coligand L can play an important role on this activation and it should certainly present
remarkable σ donor and poor π acceptor abilities to boost the electronic activation at the
bipyridyl ligand.
4
4
. Experimental Section
.1. General procedures
Fatty acid-free HSA (approx. 99%, lyophilized powder, A3782) was purchased from Sigma-
Aldrich. Millipore® water was used for the preparation of all aqueous solutions, and 10 mM 4-
(
2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer (Sigma‐Aldrich) was used in
all experiments involving spectroscopic measurements. This buffer system was adjusted to pH
.4 using KOH and/or HCl solutions. All syntheses were carried out under dinitrogen
7
atmosphere using current Schlenk techniques and the solvents used were dried using standard
5
methods[35]. Starting material [Ru(η -C H )(PPh ) Cl] was prepared following the methods
5
5
3 2
described in the literature[36]. FT-IR spectra were recorded in a Shimadzu IRAffinity-1
2
0

ACCEPTED MANUSCRIPT
spectrophotometer controlled by Shymadzu’s IRsolution software (version 1.60) with KBr
1
13
31
pellets; only significant bands are cited in text. H, C and P NMR spectra were recorded on a
1
13
Bruker Avance 400 spectrometer at probe temperature. The H and C chemical shifts are
3
1
reported in parts per million (ppm) downfield from internal Me Si and the P NMR spectra are
4
reported in ppm downfield from external standard, 85% H PO . Elemental analyses were
3
4
obtained at Laboratório de Análises from Instituto Superior Técnico. Electronic spectra were
recorded at room temperature on a Jasco V-660 spectrophotometer controlled by Jasco’s spectra
manager (version 2.09) in the range of 190-900 nm.
4
.2. Complexes syntheses
5
[
Ru(η -C H )( NCCH ) ][PF ] synthesis was adapted from [37]. Briefly, benzeneruthenium(II)
5
5
3 3
6
chloride dimer (1.0 g, 2.0 mmol) was added to a 500 mL round bottomed flask and dissolved in
300 mL of acetonitrile. To the brick red solution was added freshly synthesized sodium
≈
cyclopentadiene (464 mg, 4.0 mmol). After stirring over night at r.t., the reaction was filtered
through Celite and the solvent evaporated. The resulting compound was then taken up in the
minimum volume of water. To this solution a saturated aqueous solution of KPF was added and
6
the product was vigorously extracted into dichloromethane. The organic phase was evaporated to
give a light brownish solid which was dissolved in ≈ 800 mL of acetonitrile and irradiated with a
UV-light for 4 h. Finally, the solvent was removed under vacuum to give a brown product (yield
≈
50%).
General procedure applied to the synthesis of complexes 1-4
Ru(NCCH ) ][PF ] (0.3 mmol) were dissolved in dichloromethane (≈ 20 mL). The solution was
[
3
3
6
cooled down to 0 ºC and then 0.3 mmol of 2,2-bipyridine were added. After stirring the solution
for ≈ 5 min at this temperature, the ice-bath was removed and the solution was stirred for an
2
1

ACCEPTED MANUSCRIPT
additional 30 min, followed by the addition of 0.3 mmol of the adequate ligand (L = ImH; L =
1
2
1
-BI; L = 4-IMP; L = 1-MPI). The resulting solution was filtered through Celite and reduced in
3 4
volume. n-Hexane was added to precipitate the compounds as orange/red products.
5
[
Ru(η -C H )(2,2-bipy)(imidazole)][PF ], 1
5
5
6
-1
Orange; Yield = 71 %. Recrystallized from CH Cl /n-hexane. IR (KBr, cm ): ν(C-H aromatics)
2
2
-
1
3
140-3000, ν(C-C aromatics) 1600-1400, ν(C-N) 1400-1200, ν(PF ) 842 and 557. H NMR
6
3
3
(
(CD ) CO, Me Si, /ppm): 11.68 (s, 1, NH); 9.88 (d, 2, H ; J = 5.6 Hz); 8.47 (d, 2, H ; J
3
2
4
d
HH
a
HH
3
3
=
8.4 Hz); 8.07 (t, 2, H ; J = 7.8 Hz); 7.72 (s, 1, H ); 7.63 (t, 2, H ; J = 6.6 Hz); 7.07 (s, 1,
b HH 2 c HH
5
13
H ); 6.60 (s, 1,H ); 4.48 (s, 5,  -C H ); C NMR ((CD ) CO, Me Si, /ppm): 156.7 (C ); 156.6
5
4
5
5
3 2
4
e
5
(
C ); 139.0 (C ); 137.3 (C ); 130.4 (C ); 127.0 (C ); 123.9 (C ); 118.5 (C ); 71.9 ( -C H );
d
2
b
4
c
a
5
5
5
3
1
-
-1
-1
P((CD ) CO, /ppm): -144.27 (setp, PF ). UV-Vis in CH Cl , max/nm (/M cm ): 244
3
2
6
2
2
-1
-
(
15280), 296 (24472), 345 (4788), 465 (3991), 511 (Sh). UV-Vis in NCMe, max/nm (/M cm
1
): 209 (Sh), 242 (14577), 295 (23124), 341 (4428), 462 (3704). UV-Vis in DMSO, max/nm
-1
-1
(
/M cm ): 298 (23300), 345 (4591); 469 (3754). Elemental analysis (%) Found: C 39.4, H 3.2,
N 9.8. Calc. for C H F N PRu•⅓CH Cl : C 39.1, H 3.2, N 9.9.
1
8
17
6
4
2
2
5
[
Ru(η -C H )(2,2-bipy)(1-Benzylimidazole)][PF ], 2
5 5 6
-1
Dark red; Yield: 80 %. Recrystallized from CH Cl /n-hexane. IR (KBr, cm ): ν(C-H aromatics)
2
2
-
1
3
130-3000, ν(C-C aromatics) 1600-1400, ν(C-N) 1400-1200, ν(PF ) 839 and 555. H NMR
6
3
3
(
(CD ) CO, Me Si, /ppm): 9.83 (d, 2, H ; J = 4 Hz); 8.46 (d, 2, H ; J = 7.6 Hz); 8.07 (t, 2,
3
2
4
d
HH
a
HH
3
3
H ; J = 7 Hz); 7.83 (s, 1, H ); 7.61 (t, 2, H ; J = 7 Hz); 7.29 (m, 3, H +H +H ); 7.03 (m,
b
HH
2
c
HH
5
8
12
5
13
3
, H +H +H ); 6.51 (s, 1, H ); 5.12 (s, 2, H ); 4.47 (s, 5,  - C H ); C NMR ((CD ) CO,
9
10
11
4
6
5
5
3 2
Me Si, /ppm): 157.11 (C ); 156.65 (C ); 140.79 (C ); 137.44 (C ); 137.35 (C ); 131.29 (C );
4
e
d
2
7
c
4
2
2

ACCEPTED MANUSCRIPT
1
29.83 (C , C ); 129.16 (C ); 128.35 (C , C ); 127.14 (C ); 123.89 (C ); 121.86 (C ); 72.01
8
12
10
9
11
b
a
5
5
31
-
(
 - C H ); 51.76 (C ). P RMN ((CD ) CO, /ppm): -144.24 (setp, PF ). UV-Vis in CH Cl ,
5
5
6
3 2
6
2
2
-
1
-1



max/nm (/M cm ): 244 (15978); 296 (23176); 345 (4781); 465 (3702). UV-Vis in NCMe,
-1
-1
max/nm (/M cm ): 242 (15575); 294 (23377); 341 (4555); 461 (3711); UV-Vis in DMSO,
-1
-1
max/nm (/M cm ): 298 (23596); 345 (4559); 466 (3802). Elemental analysis (%) Found: C
4
7.1, H 3.6, N 8.7. Calc. for C H F N PRu• CH Cl : C 47.1, H 3.7, N 8.7.
25 23 6 4 2 2
5
Ru(η -C H )(2,2-bipy)(4-(1-imidazolyl)-phenol)][PF ], 3
5
5
6
Dark red; Yield: 79%. Recrystallized from NCMe/diethyl ether and CH Cl /n-hexane. IR (KBr,
2
2
-1
cm ): ν(C-H aromatics) 3140-3000, ν(C-C aromatics) 1600-1400, ν(C-N) 1400-1200, ν(C-O)
-
1
3
1
247, ν(PF ) 842 and 555. H NMR ((CD ) CO, Me Si, /ppm): 9.90 (d, 2, H ; J = 5.6 Hz);
6
3 2
4
d
HH
3
3
8
.76 (s, 1, OH); 8.48 (d, 2, H ; J = 8.0 Hz); 8.19 (s, 1, H ); 8.08 (t, 2, H ; J = 7.8 Hz); 7.63
a HH 2 b HH
3
3
(
t, 2, H ; J = 6.6 Hz); 7.36 (s, 1, H ); 7.25 (d, 2, H +H ; J = 8.4 Hz); 6.90 (d, 2, H +H ;
c
HH
5
8
10
HH
7
11
3
5
13
JHH = 8.8 Hz); 6.55 (d, 1, H ); 4.51 (s, 5,  - C H ). C NMR ((CD ) CO, Me Si, /ppm):
4
5
5
3 2
4
1
1
58.37 (C ); 157.10 (C ); 156.70 (C ); 139.09 (C ); 137.36 (C ); 131.25 (C ); 129.47 (C );
9 e d 2 b 4 6
5
31
27.17 (C ); 123.88 (C +C ); 120.82 (C ); 117.16 (C +C ); 72.00 ( - C H ). P NMR
c
8
10
5
7
11
5
5
-
-1
-1
(
(CD ) CO, /ppm): -144.21 (setp, PF ). UV-Vis in CH Cl , max/nm (/M cm ): 242 (24500),
3
2
6
2
2
-1
-1
2
95 (29700), 345 (5380), 464 (4390), 508 (Sh). UV-Vis in NCMe, max/nm (/M cm ): 241
-1
-1
(
23300), 293 (28900), 339 (4750), 460 (3840). UV-Vis in DMSO, max/nm (/M cm ): 297
(27800), 341 (4920), 466 (3820). Elemental analysis (%) Found: C 45.5, H 3.3, N 8.8. Calc. for
C H F N OPRu: C 45.9, H 3.4, N 8.9.
24
21
6
4
5
[
Ru(η -C H )(2,2-bipy)(1-(4-methoxyphenyl)-1H-imidazole)][PF ], 4
5
5
6
2
3

ACCEPTED MANUSCRIPT
-1
Dark red; Yield: 82%. Recrystallized from NCMe/diethyl ether. IR (KBr, cm ): ν(OH) 3530,
-
ν(C-H aromatics) 3140-3000, ν(C-C aromatics) 1600-1400, ν(C-N) 1400-1200, ν(PF ) 840 and
6
1
3
3
5
8
60. H RMN ((CD ) CO, Me Si, /ppm): 9.90 (d, 2, H ; J = 5.6 Hz); 8.49 (d, 2, H ; J
=
3
2
4
d
HH
a
HH
3
3
.0 Hz); 8.27 (s, 1, H ); 8.08 (t, 2, H ; J = 7.8 Hz); 7.64 (t, 2, H ; J = 6.4 Hz); 7.42 (s, 1,
2
b
HH
c
HH
3
3
H ); 7.40 (d, 2, H +H ; J = 8.8 Hz); 7.02 (d, 2, H +H ; J = 8.8 Hz); 6.55 (s, 1, H ); 4.52
5
8
10
HH
7
11
HH
4
5
13
(
s, 5,  - C H ); 3,81 (s, 3, H ). C NMR ((CD ) CO, Me Si, /ppm): 160.46 (C ); 157.07 (C );
5
5
12
3 2
4
9
e
1
1

56.69 (C ); 139.15 (C ); 137.35 (C ); 131.28 (C ); 130.05 (C ); 127.16 (C ); 123.87 (C +C );
d 2 b 4 6 c 8 10
5
31
23.66 (C ); 120.75 (C ); 115.79 (C +C ); 72.00 ( - C H ); 56.07 (C ). P RMN ((CD ) CO,
a
5
7
11
5
5
12
3 2
-
-1
-1
/ppm): -144.26 (setp, PF ). UV-Vis in CH Cl , max/nm (/M cm ): 241 (24900), 295
6
2
2
-
1
-1
(
(
(
27400), 345 (4630), 464 (3680), 513 (Sh). UV-Vis in NCMe, max/nm (/M cm ): 241
-
1
-1
26700), 293 (29200), 339 (4990), 460 (3940). UV-Vis in DMSO, max/nm (/M cm ): 297
30000), 340 (4800), 465 (3540). Elemental analysis (%) Found: C 44.9, H 3.6, N 8.1. Calc. for
C H F N OPRu•½CH Cl : C 44.8, H 3.5, N 8.2.
25
23
6
4
2
2
5
Synthesis of [Ru(η -C H )(2,2-bipy)(dmso)][PF ], 5
5
5
6
5
0
.3 mmol of [Ru(η -C H )(NCCH ) ][PF ] were dissolved in dichloromethane (≈ 20 mL). The
5 5 3 3 6
solution was cooled down to 0 ºC and then 0.3 mmol of 2,2-bipyridine were added. After stirring
the solution for ≈ 5 min at this temperature, the ice-bath was removed and the solution was
stirred for an additional 30 min. Then, the dichloromethane was evaporated and the remaining
product was dissolved in 200 µL of DMSO. After, ca. of 15 mL of deionized water were added
and the solution was stirred for 30 min. The solution was filtered and the water was co-
evaporated with diethyl ether. A recrystallization from CH Cl /diethyl ether was performed using
2
2
the remaining solution, allowing the precipitation of the product.
2
4

ACCEPTED MANUSCRIPT
-1
5
-
Yellow; Yield: 70%. IR (KBr, cm ): ν(C-H η -C H ) 3130, ν(CH ) 2920, ν(S=O) 1094, ν(PF )
5
5
3
6
1
3
8
40 and 557. H NMR ((CD ) CO, Me Si, /ppm): 9.57 (d, 2, H ; J = 5.6 Hz); 8.60 (d, 2, Ha;
3
2
4
d
HH
3
3
3
5
JHH = 8.4 Hz); 8.20 (t, 2, H ; J = 8.0 Hz); 7.64 (t, 2, H ; J = 6.6 Hz); 4.99 (s, 5, η -C H );
b
HH
c
HH
5
5
1
3
3
1
.12 (s, 6, -CH ). C NMR ((CD ) CO, Me Si, /ppm): 157.59 (C ); 157.57 (C ); 138.72 (C );
3
3 2
4
e
d
b
5
31
27.02 (C ); 124.30 (C ); 78.11 ( -C H ); 49.78 (-CH ). P NMR ((CD ) CO, /ppm): -144.27
a
c
5
5
3
3 2
-
-1
-1
(
setp, PF ). UV-Vis in CH Cl , λmax/nm (/M cm ): decomposition along time. UV-Vis in
6
2
2
-1
-1
NCMe, max/nm (/M cm ): 238 (6820), 288 (11099); 385 (1743), 445 (Sh). UV-Vis in
-
1
-1
DMSO, max/nm (/M cm ): 292 (16314); 389 (2776), 445 (Sh). Elemental analysis (%)
Found: C 37.2, H 3.5, N 5.1. Calc. for C H F N OPRuS: C 37.4, H 3.5, N 5.1.
1
7
19
6
2
5
Synthesis of [Ru(η -C H )(2,2-bipy)(CO)][PF ], 6
5
5
6
5
0
.3 mmol of [Ru(η -C H )(NCCH ) ][PF ] were dissolved in dichloromethane (≈ 20 mL). The
5 5 3 3 6
solution was cooled down to 0 ºC and then 0.3 mmol of 2,2-bipyridine were added. After stirring
the solution for ≈ 5 min at this temperature, the ice-bath was removed and the solution was
stirred for an additional 30 min, followed by the addition of CO as a slow stream of this gas
passing through the solution for ≈ 20 min at r.t.. The resulting solution was filtered through
Celite and reduced in volume. n-Hexane was added to precipitate the compound as an orange
product.
-1
5
Orange; Yield: 60%. Recrystallized from MeOH/diethyl ether. IR (KBr, cm ): ν(C-H η -C H )
5
5
-
1
3
3
125, ν(C=O) 1970, ν(PF ) 840 and 557. H NMR (MeOD, Me Si, /ppm): 9.10 (d, 2, H ; J
6
4
d
HH
3
3
3
=
5.6 Hz); 8.54 (d, 2, H ; J = 8.0 Hz); 8.22 (t, 2, H ; J = 8.0 Hz); 7.56 (t, 2, H ; J = 6.6
a
HH
b
HH
c
HH
5
13
Hz); 5.34 (s, 5, η -C H ). C NMR (MeOD, Me Si, /ppm): 158.82 (C ); 157.91 (C ); 140.46
5
5
4
d
e
5
1
(
C ); 127.59 (C ); 125.18 (C ); 84.52 ( -C H ). H NMR (CD CN, Me Si, /ppm): 9.00 (d, 2,
b
c
a
5
5
3
4
5
31
H ); 8.38 (d, 2, H ); 8.15 (t, 2, H ); 7.51 (t, 2, H ); 5.26 (s, 5, η -C H ). P NMR (MeOD,
d
a
b
c
5
5
2
5

ACCEPTED MANUSCRIPT
-
-1
-1

/ppm): -144.57 (setp, PF ). UV-Vis in CH Cl , max/nm (/M cm ): 245 (10200); 287 (9000);
6
2
2
-1
-1
3
12 (6110); 341 (Sh); 460 (Sh). UV-Vis in NCMe, max/nm (/M cm ): 245 (9610); 288
-
1
-1
(
8500); 311 (5820); 349 (Sh); 460 (Sh). UV-Vis in DMSO, max/nm (/M cm ): 293 (8010);
15 (6040); 354 460 (Sh). Elemental analysis (%) Found: C 39.7, H 2.9, N 5.0. Calc. for
C H F N OPRu: C H F N OPRu•1/5C H O: C 39.6, H 3.0, N 5.5.
3
16
13
6
2
16 13
6
2
4
10
5
Synthesis of complex [Ru(η -C H )(2,2-bipy)(PPh )][PF ], 7
5
5
3
6
5
To a solution of [Ru(η -C H )(PPh ) Cl] (150 mg, 0.21 mmol) in ≈ 30 mL of CH Cl bipyridine
5
5
3 2
2
2
(
45 mg; 0.21 mmol) was added, followed by AgPF (53 mg, 0.21 mmol). The reaction proceeded
6
under reflux for 3 h leading to a dark orange solution and a white precipitate of AgCl. The
solution was separated from the precipitate of AgCl by cannula-filtration and the solvent was
removed by vacuum. The residue was recrystallized from dichloromethane/n-hexane.
-1
Orange; Yield: 90%. IR (KBr, cm ): ν(C-H aromatics) 3140-3000, ν(C-C aromatics) 1600-1400,
-
1
ν(C-N) 1400-1200, ν(PF ) 839 and 556. H NMR ((CD ) CO, Me Si, /ppm): 9.50 (d, 2, H ;
6
3 2
4
d
3
3
3
JHH = 5.6 Hz); 8.17 (d, 2, H ; J = 8.0 Hz); 7.90 (t, 2, H ; J = 7.8 Hz); 7.42 (m, 3,
a
HH
b
HH
5
13
Hpara(PPh )); 7.32 (m, 8, H +H (PPh )), 7.12 (m, 6, H (PPh )); 4.92 (s, 5, η -C H ). C
3
c
meta
3
orto
3
5
5
NMR ((CD ) CO, Me Si, /ppm): 157.11 (C ); 156.65 (C ); 137.07 (C ); 133.85, 133.74 (CH,
3
2
4
d
e
b
PPh ); 132.63, 132.22 (C , PPh ); 129.41, 129.32 (CH, PPh ); 125.93 (C ); 124.22 (C ); 79.39
3
q
3
3
c
a
5
31
-
(
 -C H ). P NMR ((CD ) CO, /ppm): 51.32 (s, PPh ); -144.25 (setp, PF ). UV-Vis in
5
5
3 2
3
6
-1
-1
CH Cl , max/nm (/M cm ): 238 (Sh) 291 (20785); 341 (Sh); 422 (3782); 470 (Sh); UV-Vis in
2
2
-1
-1
NCMe, max/nm (/M cm ): 228 (Sh); 290 (17440); 342 (Sh); 414 (3209); 468 (Sh); UV-Vis in
-1
-1
DMSO, max/nm (/M cm ): 293 (20330); 342 (Sh); 418 (3619); 470 (Sh). Elemental analysis
(
%) Found: C 51.0, H 3.8, N 3.4. Calc. for C H F N P Ru•0.8CH Cl : C 50.9, H 3.7, N 3.5.
3
3
28
6
2
2
2
2
2
6

ACCEPTED MANUSCRIPT
5
Synthesis of complex [Ru(η -C H )(2,2-bipy)(NCCH )][PF ], 8
5
5
3
6
This compound, used as intermediate in the synthesis of compounds 1-7 was isolated and
charaterized in view to confirm some observations found in the electrochemical experiments. 0.3
5
mmol of [Ru(η -C H )(NCCH ) ][PF ] were dissolved in dichloromethane (≈ 20 mL). The
5
5
3 3
6
solution was cooled down to 0 ºC and then 0.3 mmol of 2,2-bipyridine were added. After stirring
the solution for ≈ 5 min at this temperature, the ice-bath was removed and the solution was
stirred for an additional 30 min. After solvent evaporation under vacuum the compound was
1
recrystallized from acetonitrile/diethyl ether to give an orange product. Yield: 40%. H NMR
(
CD CN, Me Si, /ppm): 9.47 (d, 2, H ); 8.28 (d, 2, H ); 8.01 (t, 2, H ); 7.49 (t, 2, H ), 4.42 (s, 5,
3
4
d
a
b
c
5
31
-
η -C H ), 1.96 (s, 3, CH ). P NMR (CD CN, /ppm): -144.64 (setp, PF ).
5
5
3
3
6
4
.3. X‐ray crystal structure determination
5
Three-dimensional X-ray data for [Ru(η -C H )(bipy)PPh ][PF ]·(CH ) CO 7 were collected on
5
5
3
6
3 2
a Bruker SMART Apex CCD diffractometer at 100(2) K, using a graphite monochromator and
Mo-K radiation ( = 0.71073 Å) by the -ω scan method. Reflections were measured from a

hemisphere of data collected of frames each covering 0.5 degrees in ω. Of the 265907 reflections
measured in 7, all of which were corrected for Lorentz and polarization effects, and for
absorption by semi-empirical methods based on symmetry-equivalent and repeated reflections,
8
335 independent reflections exceeded the significance level F/(F) > 4.0, respectively.
Complex scattering factors were taken from the program package SHELXTL[38]. The structure
2
was solved by direct methods and refined by full-matrix least-squares methods on F . The non-
hydrogen atoms were refined with anisotropic thermal parameters in all cases. Hydrogen atoms
were located in difference Fourier map and left to refine freely, except for C(2S), C(3SA),
2
7

ACCEPTED MANUSCRIPT
C(3SB), C(3) and C(4), which were included in calculation positions and refined in the riding
-
mode. A final difference Fourier map showed no residual density outside: 0.986 and -0.783 e.Å
3
2
2
2
2
.
A weighting scheme w = 1/[σ (F ) + (0.049500 P) + 7.468500 P] for 7, where P = (|F | +
o o
2
2
|F | )/3, were used in the latter stages of refinement. The crystal presents a slight disorder on the
c
acetone molecule. This disorder has been refined and two atomic sites for one carbon atom and
oxygen atom of the acetone molecule have been observed and refined with the anisotropic
atomic displacement parameters. The site occupancy factor was 0.45910 for C(3SA) and
O(1SA). CCDC No. &&& contain the supplementary crystallographic data for 7. These data can
be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44)
1
223-336-033; or e-mail: deposit@ccdc.cam.ac.uk. Crystal data and details of the data collection
and refinement for the new compounds are collected in Table 4.
Table 4.
4
.4. Electrochemical experiments
The electrochemical experiments were performed on an EG&G Princeton Applied Research
Model 273A potentiostat/galvanostat and monitored with the Electrochemistry PowerSuite v2.51
software from Princeton Applied Research. Cyclic voltammograms were obtained in 0.1 M or
0
.2 M solutions of [NBu ][PF ] in NCMe or CH Cl respectively, using a three-electrode
4 6 2 2
configuration cell with a platinum-disk working electrode (1.0 mm diameter) probed by a Luggin
capillary connected to a silver-wire pseudo-reference electrode and a Pt wire counter electrode.
The electrochemical experiments were performed under a dinitrogen atmosphere at room
temperature. The redox potentials were measured in the presence of ferrocene as the internal
2
8

ACCEPTED MANUSCRIPT
standard and the redox potential values are normally quoted relative to the SCE by using the
ferrocenium/ferrocene redox couple (E1/2 = 0.46 or 0.40 V vs. SCE for CH Cl or NCMe,
2
2
respectively). The supporting electrolyte was purchased from Fluka (electrochemical grade),
dried under vacuum for several hours and used without further purification. Reagent grade
acetonitrile and dichloromethane were dried over P O and CaH , respectively and distilled
2
5
2
under dinitrogen atmosphere before use.
4
.5. DFT Calculations
All calculations were performed at the DFT level of theory using Thrular’s M06-L functional as
implemented in Gaussian09[39-41]. The choice of this functional was due to its outstanding
performance and low computational cost[42]. The LANL2DZ effective core potential basis set
was used for heavy atoms (S, P and Ru) [43,44]. Geometry optimizations were made in the gas-
phase without any symmetry constrains. The nature of the stationary points was evaluated by
computing the Hessian matrix. No imaginary frequencies were obtained.
TD-DFT was used in combination with the PCM solvation model in order to compute the
electronic spectra of the studied molecules[45-47]. TD-DFT calculations were made applying the
same theory level and basis sets used for the calculation of the geometries. The first 120 lower
excitation energies were computed, and the simulated absorption bands were obtained by
convolution of Gaussian functions centred at the calculated excitation energies using the
QMforge software[48]. The Chemcraft 80 program was used for the visualization of the
remaining computed results[49].
4
.6. Stability studies in HEPES buffer
2
9