Tuning of redox potentials for the design of ruthenium anticancer drugs - An electrochemical study of [trans-RuCl4L(DMSO)]- and [trans-RuCl4L2]- complexes, where L = imidazole, 1,2,4-triazole, indazole

Article abstract of DOI:10.1021/ic049479c

The electrochemical behavior of [trans-RuCl₄L(DMSO)]^- (A) and [trans-RuCl₄L₂]^- (B) [L = imidazole (Him), 1,2,4-triazole (Htrz), and indazole (Hind)] complexes has been studied in DMF, DMSO, and aqueous media by cyclic voltammetry and controlled potential electrolysis. They exhibit one single-electron Ru^III/Ru^II reduction involving, at a sufficiently long time scale, metal dechlorination on solvolysis, as well as, in organic media, one single-electron reversible Ru ^III/Ru^II oxidation. The redox potential values are interpreted on the basis of the Lever's parametrization method, and particular forms of this linear expression (that relates the redox potential with the ligand E_L parameter) are proposed, for the first time, for negatively (1-) charged complexes with the Ru^III/IV redox couple center in aqueous phosphate buffer (pH 7) medium and for complexes with the Ru ^III/IV couple in organic media. The E_L parameter was estimated for indazole showing that this ligand behaves as a weaker net electron donor than imidazole or triazole. The kinetics of the reductively induced stepwise replacement of chloride by DMF were studied by digital simulation of the cyclic voltammograms, and the obtained rate constants were shown to increase with the net electron donor character (decrease of E_L) of the neutral ligands (DMSO < indazole < triazole < imidazole) and with the basicity of the ligated azole, factors that destabilize the Ru^II relative to the Ru^III form of the complexes. The synthesis and characterization of some novel complexes of the A and B series are also reported, including the X-ray structural analyses of (Ph₃PCH ₂Ph)[trans-RuCl₄(Htrz)(DMSO)], [(Ph₃P) ₂N][trans-RuCl₄(Htrz)(DMSO)], (H₂ind)-[trans- RuCl₄(Hind)(DMSO)], and [(Hind)₂H][trans-RuCl ₄(Hind)₂].

Full text of DOI:10.1021/ic049479c

Inorg. Chem. 2004, 43, 7083−7093
Tuning of Redox Potentials for the Design of Ruthenium Anticancer
-
Drugs
−
an Electrochemical Study of [trans-RuCl4L(DMSO)] and
-
[trans-RuCl4L2] Complexes, where L
) Imidazole, 1,2,4-Triazole,
Indazole
Erwin Reisner,†,‡ Vladimir B. Arion,*,‡ M. Fa´tima C. Guedes da Silva,†,§ Roman Lichtenecker,‡
Anna Eichinger,‡ Bernhard K. Keppler,*,‡ Vadim Yu. Kukushkin,| and Armando J. L. Pombeiro*,†
Centro de Qu´ımica Estrutural, Complexo I, Instituto Superior Te´cnico, AV. RoVisco Pais,
1049-001 Lisbon, Portugal, Institute of Inorganic Chemistry, UniVersity of Vienna,
Wa¨hringerstrasse 42, A-1090 Vienna, Austria, UniVersidade Luso´fona de Humanidades e
Tecnologias, AV. Campo Grande 376, 1749-024 Lisbon, Portugal, and Department of Chemistry,
St. Petersburg State UniVersity, 198504 Stary Petergof, Russian Federation
Received April 23, 2004
The electrochemical behavior of [trans-RuCl4L(DMSO)]- (A) and [trans-RuCl4L2]- (B) [L
) imidazole (Him), 1,2,4-
triazole (Htrz), and indazole (Hind)] complexes has been studied in DMF, DMSO, and aqueous media by cyclic
voltammetry and controlled potential electrolysis. They exhibit one single-electron RuIII/RuII reduction involving, at
a sufficiently long time scale, metal dechlorination on solvolysis, as well as, in organic media, one single-electron
reversible RuIII/RuIV oxidation. The redox potential values are interpreted on the basis of the Lever’s parametrization
method, and particular forms of this linear expression (that relates the redox potential with the ligand EL parameter)
/
II
are proposed, for the first time, for negatively (1−
) charged complexes with the RuIII redox couple center in
/
IV
aqueous phosphate buffer (pH 7) medium and for complexes with the RuIII couple in organic media. The EL
parameter was estimated for indazole showing that this ligand behaves as a weaker net electron donor than
imidazole or triazole. The kinetics of the reductively induced stepwise replacement of chloride by DMF were studied
by digital simulation of the cyclic voltammograms, and the obtained rate constants were shown to increase with the
net electron donor character (decrease of EL) of the neutral ligands (DMSO < indazole < triazole < imidazole) and
with the basicity of the ligated azole, factors that destabilize the RuII relative to the RuIII form of the complexes. The
synthesis and characterization of some novel complexes of the A and B series are also reported, including the
X-ray structural analyses of (Ph3PCH2Ph)[trans-RuCl4(Htrz)(DMSO)], [(Ph3P)2N][trans-RuCl4(Htrz)(DMSO)], (H2ind)-
[trans-RuCl4(Hind)(DMSO)], and [(Hind)2H][trans-RuCl4(Hind)2].
Introduction
trials of two ruthenium complexes, i.e., (H2im)[trans-RuCl4-
(Him)(DMSO)] (NAMI-A, Him ) imidazole),5-10 as an
efficient anti-metastatic drug, and (H2ind)[trans-RuCl4-
The search for novel ruthenium-based antitumor drugs has
been stimulated by clinical success of metal-containing drugs
in general1-4 and, in particular, by recent reaching of clinical
(4) Guo, Z.; Sadler, P. J. Angew. Chem., Int. Ed. 1999, 38, 1512-1531.
(5) Sava, G.; Gagliardi, R.; Bergamo, A.; Alessio, E.; Mestroni, G.
Anticancer Res. 1999, 19, 969-972.
(6) Bergamo, A.; Gagliardi, R.; Scarcia, B.; Furlani, A.; Alessio, E.;
Mestroni, G.; Sava, G. J. Pharmacol. Exp. Ther. 1999, 289, 559-
564.
* Authors to whom correspondence should be addressed. E-mail:
arion@ap.univie.ac.at (V.B.A.); keppler@ap.univie.ac.at (B.K.K.); pombeiro@
ist.utl.pt (A.J.L.P.)
† Centro de Qu´ımica Estrutural, Complexo I, Instituto Superior Te´cnico.
‡ Institute of Inorganic Chemistry, University of Vienna.
§ Universidade Luso´fona de Humanidades e Tecnologias.
| Department of Chemistry, St. Petersburg State University.
(1) Wong, E.; Giandomenico, C. M. Chem. ReV. 1999, 99, 2451-2466.
(2) Rosenberg, B.; VanCamp, L. Cancer Res. 1970, 30, 1799-1802.
(3) Rosenberg, B.; VanCamp, L.; Krigas, T. Nature 1965, 205, 698-
699.
(7) Sava, G.; Capozzi, I.; Clerici, K.; Gagliardi, G.; Alessio, E.; Mestroni,
G. Clin. Exp. Met. 1998, 16, 371-379.
(8) Bergamo, A.; Cocchietto, M.; Capozzi, I.; Mestroni, G.; Alessio, E.;
Sava, G. Anti-Cancer Drugs 1996, 7, 697-702.
(9) Sava, G.; Clerici, K.; Capozzi, I.; Cocchietto, M.; Gagliardi, R.;
Alessio, E.; Mestroni, G.; Perbellini, A. Anti-Cancer Drugs 1999, 10,
129-138.
10.1021/ic049479c CCC: $27.50
Published on Web 09/21/2004
© 2004 American Chemical Society
Inorganic Chemistry, Vol. 43, No. 22, 2004 7083

Reisner et al.
(Hind)2] (KP1019, Hind ) indazole),11,12 as an anticancer
drug that exhibits excellent activity especially against au-
tochtonous colorectal tumors. Moreover, quite recently we
observed that both trans- and cis-isomers of [RuCl4(Htrz)2]-
(Htrz ) 1,2,4-triazole) exhibit a time-dependent response of
three human cell lines, i.e., SW480, HT29 and SK-BR-3,
and, in addition, the trans-isomer displays a higher anti-
proliferative activity than the cis-species.13
Although the mode of antitumor action of the ruthenium-
(III)-containing species has not yet been established, it was
suggested that a mechanism which originates from an
actiVation by reduction might be responsible for the activity
of NAMI-A and KP1019.14-17 According to this hypothesis
the reduction of the RuIII center would produce more labile,
toward substitution, RuII-Cl species, which would rapidly
react with specific sites of proteins altering their activity.16
Thus, to be active in vivo the RuIII complexes, acting as
prodrugs, should possess biologically accessible reduction
potentials (from -0.4 to +0.8 V vs NHE).5,14-16,18
Figure 1. Anticancer ruthenium(III) complexes of the types A [trans-
RuCl4L(DMSO)]-, and B [trans-RuCl4L2]-; underlined complexes have
been characterized in this work by X-ray crystallography.
While the qualitative aspects of the hypothesis of activation
by reduction are already well-recognized,14,15 little was done
to systematically investigate the electrochemical behavior of
the anticancer agents. To provide a methodological approach
on how to tune the RuIII/RuII redox potential to design
prodrugs of the type [trans-RuCl4L(DMSO)]- (A) and [trans-
RuCl4L2]- (B) (L ) azole ligand), which could be activated
in tumor cells and remain inactive in normal tissues, we
performed an electrochemical study of a series of these
complexes (Figure 1) where the ligands imidazole, 1,2,4-
triazole, and indazole (Figure 2) were chosen due to the good
antitumor activity of their corresponding RuIII complexes and
their effect on ruthenium redox potential.
Figure 2. Heterocyclic azole ligands.
ligand parameter) and in which SM and IM are dependent
upon the metal and redox couple, the spin state, and the
stereochemistry, we have tentatively estimated the EL
parameter for one of the N-ligands of this study (indazole),
identified some new products, and predicted the reduction
potentials for ruthenium(III) anticancer compounds.
The kinetics of the cathodically induced chloride ligand
substitution on complexes A and B were investigated by
digital simulation of the cyclic voltammograms and shown
to correlate with the basicity of the azo-N atom of the
heterocyclic ligand, which provides a useful tool for further
design of Ru-based antitumor complexes. In addition, on the
basis of the general expression proposed by Lever (eq 1),19
which assumes an additive contribution to the redox potential
of all the ligands (as measured for each of them by the EL
E ) S ‚ E + I
M
(1)
∑
M
L
Moreover, in pursuit of our interest in the establishment
of redox potential-structure relationships in coordination
compounds,20-26 we have extended the application of Lever’s
equation (eq 1) to the reduction of the negatively (1-)
charged RuIII complexes in neutral aqueous phosphate buffer
medium, and to the oxidation of such RuIII complexes in an
organic medium, by obtaining, for the first time, SM and IM
for such types of systems. The thus-proposed particular forms
of eq 1 can also give a valuable contribution toward the
possibility of tuning the redox potentials of RuIII complexes
with pharmacological significance.
(10) Sava, G.; Alessio, E.; Bergamo, A.; Mestroni, G. In Metallopharma-
ceuticals; Clarke, M. J., Sadler, P. J., Eds.; Topics in Biological
Inorganic Chemistry series, vol. 1; Springer: Berlin, 1999; pp 143-
169.
(11) Keppler, B. K.; Lipponer, K. G.; Stenzel, B.; Kratz, F. In Metal
Complexes in Cancer Chemotherapy; Keppler, B. K., Ed.; VCH:
Weinheim, 1993; pp 187-220.
(12) Galanski, M.; Arion, V. B.; Jakupec, M. A.; Keppler, B. K. Curr.
Pharm. Des. 2003, 9, 2078-2089.
(20) Pombeiro, A. J. L. New J. Chem. 1997, 21, 649-660.
(21) Zang, L.; Guedes da Silva, M. F. C.; Kuznetsov, M. L.; Gamasa, M.
P.; Gimeno, J.; Frau´sto da Silva, J. J. R.; Pombeiro, A. J. L.
Organometallics 2001, 20, 2782-2793.
(22) Guedes da Silva, M. F. C.; Trzeciak, A. M.; Zio´lkowski, J. J.;
Pombeiro, A. J. L. J. Organomet. Chem. 2001, 620, 174-181.
(23) Guedes da Silva, M. F. C.; Frau´sto da Silva, J. J. R.; Pombeiro, A. J.
L.; Amatore, C.; Verpeaux, J. N. Inorg. Chem. 1998, 37, 2344-2350.
(24) Almeida, S. S. P. R.; Pombeiro, A. J. L. Organometallics 1997, 16,
4469-4478.
(25) Pombeiro, A. J. L.; Costa, M. T. A. R. S.; Wang, Y.; Nixon, J. F. J.
Chem. Soc. Dalton Trans. 1999, 3755-3758.
(26) Guedes da Silva, M. F. C.; Martins, L. M. D. R. S.; Frau´sto da Silva,
J. J. R.; Pombeiro, A. J. L. Collect. Czech. Chem. Commun. 2001,
66, 139-154.
(13) Arion, V. B.; Reisner, E.; Fremuth, M.; Jakupec, M. A.; Keppler, B.
K.; Kukushkin, V. Yu.; Pombeiro, A. J. L. Inorg. Chem. 2003, 42,
6024-6031.
(14) Clarke, M. J. Coord. Chem. ReV. 2003, 236, 209-233, and references
therein.
(15) Clarke, M. J.; Zhu, F.; Frasca, D. R. Chem. ReV. 1999, 99, 2511-
2533, and references therein.
(16) Clarke, M. J. Prog. Clin. Biochem. Med. 1989, 10, 25-39.
(17) Alessio, E.; Balducci, G.; Lutman, A.; Mestroni, G.; Calligaris, M.;
Attia, W. M. Inorg. Chim. Acta 1993, 203, 205-217.
(18) Kirlin, W. G.; Cai, J.; Thompson, S. A.; Diaz, D.; Kavanagh, T. J.;
Jones, D. P. Free Radical Biol. Med. 1999, 27, 1208-1218.
(19) Lever, A. B. P. Inorg. Chem. 1990, 29, 1271-1285.
7084 Inorganic Chemistry, Vol. 43, No. 22, 2004

Tuning of Redox Potentials for Ruthenium Anticancer Drugs
Synthesis of Complexes. (H2im)[trans-RuCl4(Him)(DMSO)]
(A1),33 [(DMSO)2H][trans-RuCl4(DMSO)2],34 (H2im)[trans-RuCl4-
(Him)2] (B1),35 (H2trz)[trans-RuCl4(Htrz)2] (B2),13 (H2ind)[trans-
RuCl4(Hind)2] (B3),36 and Na[trans-RuCl4(Hind)2]37 (B3b) were
prepared as described elsewhere.
All these results along with the syntheses (A1a, A2, A2a,
A2b, A3, A3a, and B3a) and structural characterization (A2a,
A2b, A3, and B3a) for some of the complexes are reported
herein.
Experimental Section
(H2trz)[trans-RuCl4(Htrz)(DMSO)] (A2). 1,2,4-Triazole (2.00
g, 29.0 mmol) was added to [(DMSO)2H][trans-RuCl4(DMSO)2]
(1.00 g, 1.8 mmol) in ethanol (96%, 40 mL) and the suspension
was vigorously stirred for complete dissolution. The orange mixture
was left standing overnight at room temperature. The orange
precipitate formed was filtered off, washed with two 10-mL portions
of diethyl ether, and dried at room temperature in vacuo. Yield:
0.35 g, 42% based on Ru. Anal. Calcd for C6H13N6Cl4ORuS (Mr
) 462.17 g/mol), %: C, 15.66; H, 2.85; N, 18.26; S, 6.97. Found,
%: C, 15.71; H, 2.82; N, 18.36; S, 7.21. ESI-MS(-ve), m/z: 391
[RuCl4(Htrz)(DMSO)]-; 322 [RuCl4(DMSO)]-; 244 [RuCl4]-. mp
) 142 °C. TLC on SiO2, Rf ) 0.49 (eluent CH2Cl2/MeOH ) 70:
30). IR spectrum in KBr, selected bands, cm-1: 1115 s, 1083 vs
ν(S ) O), 432 s ν(Ru - S).
(H2ind)[trans-RuCl4(Hind)(DMSO)] (A3). Indazole (0.47 g, 4.0
mmol) was added to [(DMSO)2H][trans-RuCl4(DMSO)2] (1.12 g,
2.0 mmol) in acetone (30 mL). The mixture was heated at 45 °C
over 0.75 h, and allowed to cool to room temperature. The brick-
red product formed was filtered off, washed with acetone, and dried
at room temperature in vacuo. Yield: 0.60 g, 53% based on Ru.
Anal. Calcd for C16H19N4Cl4ORuS (Mr ) 558.28 g/mol), %: C,
34.42; H, 3.43; N, 10.04; S, 5.74. Found, %: C, 34.46; H, 3.37; N,
9.99; S, 5.86. ESI-MS(-ve), m/z: 440 [RuCl4(Hind)(DMSO)]-; 322
[RuCl4(DMSO)]-; 244 [RuCl4]-. mp ) 170 °C (dec). TLC on SiO2,
Rf ) 0.51 (eluent CH2Cl2/MeOH ) 75:25). IR spectrum in KBr;
selected bands, cm-1: 1126 m, 1097 s, 1061 vs V(S ) O), 424 vs
ν(Ru - S). Single crystals suitable for X-ray diffraction study were
selected directly from the isolated product.
(Ph3PCH2Ph)[trans-RuCl4L(DMSO)] [L ) Him (A1a), Htrz
(A2a), Hind (A3a)], and (PPN)[trans-RuCl4(Htrz)(DMSO)]
(A2b). A solution of (Ph3PCH2Ph)Cl (0.09 g, 0.22 mmol) or of
(PPN)Cl (0.13 g, 0.22 mmol) in EtOH (96%, 1 mL) was added to
(HL)[trans-RuCl4L(DMSO)] (0.22 mmol) in H2O (2 mL) with
stirring. The immediately formed powder was filtered off, washed
with two 2-mL portions of H2O, two 2-mL portions of EtOH (96%),
and two 2-mL portions of diethyl ether, and dried at room
temperature in vacuo. The synthesis of (Ph3PCH2Ph)[trans-RuCl4-
(Hind)(DMSO)] (A3a) was carried out in ethanol (96%, 8 mL).
(Ph3PCH2Ph)[trans-RuCl4(Him)(DMSO)] (A1a). Yellow pow-
der. Yield: 0.14 g, 83% based on Ru. Anal. Calcd for C30H32N2-
Cl4OPRuS (Mr ) 742.51 g/mol), %: C, 48.53; H, 4.34; N, 3.77;
S, 4.32. Found, %: C, 48.57; H, 4.35; N, 3.74; S, 4.34. ESI-MS-
(-ve), m/z: 390 [RuCl4(Him)(DMSO)]-; 322 [RuCl4(DMSO)]-; 244
[RuCl4]-. ESI-MS(+ve), 353 (Ph3PCH2Ph)+. mp ) 191 °C. TLC
on SiO2, Rf ) 0.53 (eluent CH2Cl2/MeOH ) 85:15). IR spectrum
in KBr; selected bands, cm-1: 1112 vs, 1063 vs V(S ) O), 421 m
ν(Ru - S). UV-vis (MeOH), λmax, nm (ꢀ, mΜ-1 cm-1): 466
(0.51), 398 (4.08), 295 (1.20).
Physical Measurements. Elemental analyses were carried out
by the Microanalytical Service of the Instituto Superior Te´cnico in
Lisbon or on a Carlo Erba microanalyzer at the Institute of Physical
Chemistry of the University of Vienna. Infrared spectra were
recorded on a Perkin-Elmer FTIR 2000 spectrometer in KBr pellets
(4000-400 cm-1). UV-vis spectra were recorded on a Perkin-
Elmer Lambda 9 UV-vis spectrophotometer using samples dis-
solved in water or methanol. Electrospray ionization mass spec-
trometry was carried out in methanol with a Bruker Esquire 3000
instrument (Bruker Daltonic, Bremen, Germany). The given m/z
values, originating from the most intense isotopes, were obtained
by the mass linearization procedure. Expected and experimental
isotope distributions were compared. Melting points were measured
on a Kofler-table (Leica Galen III). For TLC, Merck UV 254 SiO2
plates were used.
Cyclic voltammograms were measured in a two-compartment
three-electrode cell using a 1.0-mm-diameter glassy-carbon disk
(or a 0.5-mm-diameter platinum disk) working electrode, probed
by a Luggin capillary connected to a silver-wire pseudo-reference
electrode, and a platinum auxiliary electrode. Measurements were
performed by cyclic voltammetry (CV) at room temperature using
an EG & G PARC 273A potentiostat/galvanostat. Controlled
potential electrolyses (CPE) were carried out in a two-compartment
three-electrode cell with a carbon plate working electrode and a
platinum gauze counter electrode separated by a glass frit; a Luggin
capillary, probing the working electrode, was connected to a silver-
wire pseudo-reference electrode. Deaeration of solutions was
accomplished by passing a stream of high-purity nitrogen through
the solution for 10 min prior to the measurements and then
maintaining a blanket atmosphere of nitrogen over the solution
during the measurements. The potentials were measured in 0.15
M [nBu4N][BF4]/DMF or DMSO and in 0.2 M phosphate buffer
at pH 7 or aqueous 0.2 M KNO3, using the [Fe(η5-C5H5)2]0/+ (E1/2
ox
) +0.72 V or +0.68 V vs NHE in DMF or DMSO, respec-
ox
tively),27,28 or methyl viologen (E1/2 ) -0.44 V28 vs NHE in
water), respectively, as internal standards, and are quoted relative
to NHE.
Crystallographic Structure Determination. X-ray diffraction
measurements were performed on a Nonius Kappa CCD diffrac-
tometer at 120 K. Single crystals were positioned at 30, 30, 35,
and 25 mm from the detector, and 376, 389, 488, and 240 frames
were measured, each for 100, 100, 70, and 220 s over 2, 1.5, 1.5,
and 2° for A2a, A2b, A3, and B3a, respectively. The data were
processed using the Denzo-SMN software. The structures were
solved by direct methods by using the SHELXS-97 package and
refined by full-matrix least-squares techniques with SHELXL-
97.29,30 All hydrogens were inserted in calculated positions and
refined using a riding model. Drawings were made with ORTEP.31,32
Chemicals. Imidazole, 1H-1,2,4-triazole, indazole, methyl violo-
gen, (Ph3PCH2Ph)Cl, (PPN)Cl, i.e. [(PPh3)2N]Cl [bis(triphenylphos-
phine)iminium chloride], ferrocene, and solvents were commercially
available from Aldrich and used as purchased.
(Ph3PCH2Ph)[trans-RuCl4(Htrz)(DMSO)] (A2a). Yellow pow-
der. Yield: 0.14 g, 88% based on Ru. Anal. Calcd. for C29H31N3-
Cl4OPRuS (Mr ) 743.50 g/mol), %: C, 46.85; H, 4.20; N, 5.65;
S, 4.31. Found, %: C, 46.69; H, 4.15; N, 5.62; S, 4.37. ESI-MS-
(29) Sheldrick, G. M. SHELXS-97, Program for Crystal Structure Solution;
University of Go¨ttingen, Germany, 1997.
(30) Sheldrick, G. M. SHELXL-97, Program for Crystal Structure Refine-
ment; University of Go¨ttingen, Germany, 1997.
(31) Johnson, C. K. Report ORNL-5138; Oak Ridge National Laboratory:
Oak Ridge, TN, 1976.
(27) Barette, W. C., Jr.; Johnson, H. W., Jr.; Sawyer, D. T. Anal. Chem.
1984, 56, 1890-1898.
(28) Guedes da Silva, M. F. C.; Pombeiro, A. J. L.; Geremia, S.; Zangrando,
E.; Calligaris, M.; Zinchenko, A. V.; Kukushkin, V. Yu. J. Chem.
Soc., Dalton Trans. 2000, 1363-1371.
Inorganic Chemistry, Vol. 43, No. 22, 2004 7085

Reisner et al.
(-ve), m/z: 391 [RuCl4(Htrz)(DMSO)]-; 322 [RuCl4(DMSO)]-; 244
[RuCl4]-. ESI-MS(+ve), 353 (Ph3PCH2Ph)+. mp ) 177 °C. TLC
on SiO2, Rf ) 0.53 (eluent CH2Cl2/MeOH ) 85:15). IR spectrum
in KBr; selected bands, cm-1: 1119 vs, 1078 vs V(S ) O), 426 m
ν(Ru - S). UV-vis (MeOH), λmax, nm (ꢀ, mΜ-1 cm-1): 478
(0.39), 406 (3.72), 298 (1.06). Orange crystals suitable for X-ray
diffraction study were grown in an H-shaped tube by the solvent
diffusion method upon reacting equimolar amounts of A2 in ethanol
and benzyltriphenylphosphonium chloride in water.
(PPN)[trans-RuCl4(Htrz)(DMSO)] (A2b). Yellow powder.
Yield: 0.17 g, 81% based on Ru. Anal. Calcd for C40H39N4Cl4-
OP2RuS (Mr ) 928.66 g/mol), %: C, 51.73; H, 4.23; N, 6.03; S,
3.45. Found, %: C, 51.93; H, 4.03; N, 5.90; S, 3.49. ESI-MS(-ve),
m/z: 391 [RuCl4(Htrz)(DMSO)]-; 322 [RuCl4(DMSO)]-; 244
[RuCl4]-. ESI-MS(+ve), 538 (PPN)+. mp ) 202 °C. TLC on SiO2,
Rf ) 0.50 (eluent CH2Cl2/MeOH ) 85:15). IR spectrum in KBr,
selected bands, cm-1: 1119 vs, 1078 vs V(S ) O), 428 m ν(Ru -
S). UV-vis (MeOH), λmax, nm (ꢀ, mΜ-1 cm-1): 478 (0.42), 406
(3.97), 300 (1.13). Yellow crystals suitable for X-ray diffraction
study were grown in an H-shaped tube by the solvent diffusion
method upon reacting equimolar amounts of A2 in ethanol and
(PPN)Cl in water.
Table 1. UV-vis Absorption Bands [Wavelength and Molar
Absorptivity (in brackets)] for Complexes A and B in Water at Room
Temperature
λmax
(nm)
ꢀ
a
complex
(mM-1 cm-1
)
(H2im)[trans-RuCl4(Him)(DMSO)] A1
452
(0.49)
462
(0.45)
462
(0.65)
396
(0.50)
418
390
(3.64)
396
(3.76)
396
(4.33)
348
(3.12)
362
288
293
(H2trz)[trans-RuCl4(Htrz)(DMSO)] A2
(H2ind)[trans-RuCl4(Hind)(DMSO)] A3
(H2im)[trans-RuCl4(Him)2] B1
(H2trz)[trans-RuCl4(Htrz)2] B2
Na[trans-RuCl4(Hind)2] B3b
233
235
(0.37)
422
(2.75)
364
(1.24)
(4.37)
a An intense UV absorption band prevented the accurate determination
of the molar absorptivity for the shorter wavelength band, as well as the
resolution of this band for complexes A3 and B3b.
Table 2. Crystallographic Data for A2a, A2b, A3, and B3a
A2a
A2b
A3
B3a
chemical formula C29H31N3Cl4O C40H39N4Cl4O- C16H19N4Cl4O- C28H25N8-
PRuS
743.50
120
P2RuS
928.66
120
RuS
558.28
120
Cl4Ru
716.42
120
formula weight
T, K
(Ph3PCH2Ph)[trans-RuCl4(Hind)(DMSO)] (A3a). Yield: 0.14
g, 79% of orange-yellow powder based on Ru. Anal. Calcd for
C34H34N2Cl4OPRuS (Mr ) 792.57 g/mol), %: C, 51.52; H, 4.32;
N, 3.53; S, 4.05. Found, %: C, 51.49; H, 4.12; N, 3.51; S, 4.02.
ESI-MS(-ve), m/z: 440 [RuCl4(Hind)(DMSO)]-; 322 [RuCl4-
(DMSO)]-; 244 [RuCl4]-. ESI-MS(+ve), 353 [Ph3PCH2Ph]+. mp
) 190 °C. TLC on SiO2, Rf ) 0.55 (eluent CH2Cl2/MeOH ) 85:
15). IR spectrum in KBr; selected bands, cm-1: 1112 vs, 1077 vs
V(S ) O), 423 ν(Ru - S). UV-vis (MeOH), λmax, nm (ꢀ, mΜ-1
cm-1): 476 (0.51), 406 (3.92), < 300 (absorption of benzene ring).
[(Hind)2H][trans-RuCl4(Hind)2] (B3a). Na[trans-RuCl4(HInd)2]
(0.30 g, 0.6 mmol) in water (260 mL) was added to an aqueous
solution (260 mL) containing indazole (0.19 g, 1.6 mmol) and 0.2
M HCl (8 mL). The reaction mixture was filtered off, and left to
stand for 48 h at 4 °C. The brown-red plates were separated by
filtration, washed with water, and dried at room temperature in
vacuo. Yield: 0.25 g, 58%. Anal. Calcd. for C28H25N8Cl4Ru (Mr
) 716.42 g/mol), %: C, 46.94; H, 3.52; N, 15.64. Found: C, 46.54;
H, 3.36; N, 15.45. ESI-MS(-ve), m/z: 480 [RuCl4(Hind)2]-; 244
[RuCl4]-. ESI-MS(+ve): 119 [Hind]+. mp ) 170 °C (dec). TLC
on SiO2, Rf ) 0.58 (eluent CH2Cl2/MeOH ) 50:50). IR spectrum
in KBr; selected bands, cm-1: 1628, 1357, 744. UV-vis (MeOH),
space group
a/Å
b/Å
P1h
P21/c
P21/n
P1h
9.599(2)
12.680(3)
14.476(3)
112.15(3)
109.16(3)
90.90(3)
1521.8(6)
2
17.029(3)
16.428(3)
14.653(3)
10.619(2)
8.691(2)
22.667(5)
6.602(1)
9.962(2)
11.671(2)
92.78(3)
102.57(3)
104.75(3)
720.1(2)
1
c/Å
R/deg
â/deg
γ/deg
V/Å3
Z
λ/Å
103.32(3)
103.00(3)
3989.0(14)
4
0.71073
1.546
8.32
0.0261
0.0705
2038.3(8)
4
0.71073
1.819
14.11
0.0296
0.0709
0.71073
1.622
10.17
0.0304
0.0780
0.71073
1.652
9.51
0.0355
0.0817
F
calcd/g cm-3
µ
calcd/cm-1
R1a
wR2b
2
2
2 2
a R1 ) ∑||Fo| - |Fc||/∑|Fo |. b wR2 ) [∑w|Fo | - |Fc |)2/∑w|Fo | ]1/2
.
[trans-RuCl4(DMSO)2] with an excess of the corresponding
azole heterocycle. The corresponding (Ph3PCH2Ph)- or
(PPN)-complex salts were obtained by metathesis of (HL)-
[trans-RuCl4L(DMSO)] and (Ph3PCH2Ph)Cl or (PPN)Cl.
Complexes B were prepared starting from an HCl-acidic
solution of RuCl3 and an excess of azole ligand.38
Their electronic absorption spectra in water are similar.
One main band with an absorption maximum in the 396-
348 nm range, a shoulder at 462-396, and a third absorption
at 293-233 in A1, A2, B1, and B2 are observed (Table 1).
In the spectra of A3 and B3b an intense absorption band of
the condensed benzene ring prevented the resolution of the
latter band. Replacement of the strong π-acceptor DMSO
by an azole ligand shifts the absorption maximum to shorter
wavelengths. The azole ligands follow the same tendency,
where the absorption maximum for the more electron
donating and basic imidazole, pKa (H2im+) ) 7.00,39 is blue
shifted compared to 1,2,4-triazole, pKa (H2trz+) ) 2.55,40
or indazole, pKa (H2ind+) ) 1.25.41
λ
max, nm (ꢀ, mΜ-1 cm-1): 435 (0.68), 370 (4.30). Single crystals
suitable for X-ray diffraction study were selected directly from the
isolated product.
Results and Discussion
Synthesis and Characterization of the Complexes.
Complexes A were synthesized by reaction of [(DMSO)2H]-
(32) International Tables for X-ray Crystallography; Kluwer Academic
Press: Dordrecht, The Netherlands, 1992; Vol. C, Tables 4.2.6.8 and
6.1.1.4.
(33) Mestroni, G.; Alessio, E.; Sava, G. Intl. Patent PCT C07F 15/00,
WO98/00431, 1998.
(34) (a) Alessio, E.; Balducci, G.; Calligaris, M.; Costa, G.; Attia, W. M.;
Mestroni, G. Inorg. Chem. 1991, 30, 609-618. (b) Jaswal, J. S.; Rettig,
S. J.; James, B. R. Can. J. Chem. 1990, 68, 1808-1817.
(35) Keppler, B. K.; Rupp, W.; Juhl, U. M.; Endres, H.; Niebl, R.; Balzer,
W. Inorg. Chem. 1987, 26, 4366-4370.
(36) Lipponer, K. G.; Vogel, E.; Keppler, B. K. Met.-Based Drugs 1996,
3, 243-260.
(37) Peti, W.; Pieper, T.; Sommer, M.; Keppler, B. K.; Giester, G. Eur. J.
Inorg. Chem. 1999, 1551-1555.
(38) Kralik, F.; Vrestal, J. Collect. Czech. Chem. Commun. 1961, 26, 1298-
1304.
(39) Reedijk, J. Heterocyclic Nitrogen-Donor Ligands. In ComprehensiVe
Coordination Chemistry; Wilkinson, G.; Gillard, R. D.; McCleverty,
J. A., Eds.; Pergamon Press: Elmsford, NY, 1987; Vol. 2, pp 73-98.
(40) Potts, K. T. Chem. ReV. 1961, 61, 87-127.
7086 Inorganic Chemistry, Vol. 43, No. 22, 2004

Tuning of Redox Potentials for Ruthenium Anticancer Drugs
Table 3. Bond Lengths (Å) and Angles (deg) for A2a, A2b, A3, and
B3b
Atom1-Atom2
A2a
A2b
A3
B3a
Ru1-N4
Ru1-N1
Ru1-S1
Ru1-Cl1
Ru1-Cl2
Ru1-Cl3
Ru1-Cl4
S1-O1
2.1207(17) 2.1042(12)
2.0901(13) 2.062(3)
2.2924(7)
2.2730(8)
2.3373(9)
2.2691(6)
2.3705(5)
Figure 3. Structure of [trans-RuCl4(Htrz)(DMSO)]- in A2a.
2.3497(7)
2.3964(9)
2.3460(9)
2.3571(7)
2.3602(11)
2.3616(10)
2.3449(12) 2.3462(5)
2.3730(9) 2.3421(5)
2.3613(12) 2.3603(5)
1.4742(16) 1.4741(10) 1.4951(12)
Atom1-Atom2-Atom3
A2a
A2b
A3
B3a
N4-Ru1-S1
N1-Ru1-S1
N1#-Ru1-N1
N4-Ru1-Cl1
N1-Ru1-Cl1
N4-Ru1-Cl2
S1-Ru1-Cl1
Ru1-S1-O1
178.34(4) 176.65(3)
175.87(3)
180.00
89.29(5)
91.59(4)
88.12(3)
88.27(8)
86.82(5)
90.61(3)
87.57(4)
89.89(2)
Figure 4. Structure of [trans-RuCl4(Htrz)(DMSO)]- in A2b.
95.964(13)
118.88(7) 117.69(4) 115.09(5)
in A2a and A2b, correspondingly. This latter bond is
markedly shorter than the one in free DMSO [1.492(1) Å]43
indicating greater S-O double bond character for the
S-bonded molecule. Both Ru1-S1 and S1-O1 are slightly
shorter than those found in Na[trans-RuCl4(NH3)(DMSO)]
[2.2797(7) and 1.479(3) Å, respectively] and significantly
shorter than those in Na[trans-RuCl4(Him)(DMSO)] [2.2956-
(6) and 1.487(2) Å].17 The N1-H group of the triazole ring
appears to form a bifurcated hydrogen bond involving O1′
(x, 0.5 - y, z - 0.5) as well as Cl3′ (x, 0.5 - y, z - 0.5)
(O1′‚‚‚H 2.37(1) Å, H-N1 0.90(1) Å, and O1′‚‚‚H-N1
112.0(1)°; Cl3′‚‚‚H 2.57(1) Å, H-N1 0.90(1) Å, Cl3′‚‚‚H-
N1 149.1(1)°) (Figure 4). Bond lengths and angles in the
(Ph3PCH2Ph)+ and (PPN)+ cations are similiar to those found
elsewhere.44,45
Crystal Structures. Crystallographic data for A2a, A2b,
A3, and B3a are summarized in Table 2, and selected bond
lengths and angles are presented in Table 3.
The solid-state structures of A2a and A2b feature an
S-bonded DMSO ligand trans to the N-donor atom of Htrz
(Figures 3 and 4). To our knowledge these are the first
ruthenium complexes containing one triazole ligand to be
crystallographically characterized. From isotropic displace-
ment parameters it was impossible to distinguish between
carbon and nitrogen atoms in the heterocyclic ring of A2a.
Moreover, the distribution of electron density over the
triazole ring does not allow the localization of double bonds.
All ring bond lengths are essentially the same. Three different
models (coordination via N2, coordination via N2 with
rotational (180°) disorder, or coordination via N4) were
refined but no model could be favored over the two others.
However, analysis of the crystal structure of A2a showed
the presence of interanion contacts (Figure 3) which can be
described as hydrogen bridges of the H-N1-N2 parts of
the triazole rings with the following parameters: N1-H )
0.86 Å, H‚‚‚N2′ ) 2.656 Å, N1‚‚‚N2′ ) 3.262 Å, N1HN2′
) 128.6°. In addition, the interplanar separation between the
interacting triazole rings is 0.89 Å. Similar but stronger
interactions have been observed for Cd(NCS)2(Htrz)2.42
Hence, coordination of triazole via N4 is proposed, and the
same coordination mode has been revealed in A2b. In
contrast, however, in this complex a remarkable difference
in ring bond lengths of triazole was found. In particular the
N2-C3 bond at 1.3176(18) Å is about 5σ shorter than the
same bond in A2a at 1.335(3) Å. On the other hand, the
bond angles in the triazole ring, both in A2a and A2b, are
very similar, providing an additional piece of evidence for
the same coordination mode of Htrz in A2a and A2b. The
Ru1-S1 bond distances are 2.2730(8) and 2.2691(6) Å, and
the S1-O1 bond lengths are 1.4742(16) and 1.4741(10) Å
The proposed binding mode agrees well with the higher
basicity of N4 in comparison to that of N2.46 In [MnII(Htrz)-
(H2O)4(SO4)],47 [Cd(Htrz)2(NCS)2],42 and [Fe(bpy)(Htrz)-
Cl3]48 coordination via N4 to the metal ion has been
established by X-ray crystallography. The same coordination
mode of triazole was proposed for Zn(II) in human carbonic
anhydrase,48,49 as a result of hydrogen bonding.
The ruthenium(III) atom in A3 has the expected distorted
octahedral geometry (Figure 5), with four chloride ligands
in the equatorial positions and a DMSO molecule bound
through its sulfur atom trans to the indazole ligand in axial
positions. The average value of the Ru-Cl bond lengths
[2.362(12) Å] is comparable with those found in Na[trans-
[RuCl4(Him)(DMSO)]‚H2O‚Me2CO17 and [(DMSO)2H][RuCl4-
(DMSO)2]34 [2.342(7) and 2.348(4) Å, respectively]. The
(43) Calligaris, M.; Carugo, O. Coord. Chem. ReV. 1996, 153, 83-154.
(44) Cifuentes, M. P.; Waterman, S. M.; Humphrey, M. G.; Heath, G. A.;
Skelton, B. H.; White, A. H.; Seneka Perera, M. P.; Williams, M. L.
J. Organomet. Chem. 1998, 565, 193-200.
(45) Bartczak, T. J.; Wolowiec, S.; Latos-Grazynski, L. Inorg. Chim. Acta
1998, 277, 242-246.
(46) Meot-Ner, M.; Liebman, J. F.; Del Bene, J. E. J. Org. Chem. 1986,
51, 1105-1110.
(41) Catala´n, J.; Claramunt, R. M.; Elguero, J.; Laynez, J.; Mene´ndez, M.;
Anvia, F.; Quian, J. H.; Taagepera, M.; Taft, R. W. J. Am. Chem.
Soc. 1988, 110, 4105-4111.
(42) Haasnoot, J. G.; De Keyzer, G. C. M.; Verschoor, G. C. Acta
Crystallogr. 1983, C39, 1207-1209.
(47) Gorter, S.; Engelfriet, D. W. Acta Crystallogr. 1981, B37, 1214-
1218.
(48) Haasnoot, J. G. Coord. Chem. ReV. 2000, 200-202, 131-185.
(49) Mangani, S.; Liljas, A. J. Mol. Biol. 1993, 232, 9-14.
Inorganic Chemistry, Vol. 43, No. 22, 2004 7087

Reisner et al.
Figure 5. Structure of [trans-RuCl4(Hind)(DMSO)]- in A3.
Figure 8. Multiple scan cyclic voltammograms of 5 mM (H2im)[trans-
RuCl4(Him)(DMSO)] A1 in DMF (a and b) or DMSO (c) and (H2ind)-
[trans-RuCl4(Hind)2] B3 (d) in DMF, with 0.15 M [nBu4N][BF4], at a carbon
disk working electrode and at a scan rate of 2 (a), 0.2 (b), or 1 (c and d) V
s-1
.
symmetry. The Ru-N1 bond distance [2.062(3) Å] is
intermediate between those found in (Ph4P)[trans-RuCl4-
(Hind)2] [2.0517(21) and 2.0711(22) Å]. The Ru-Cl bond
distances [2.3602(11) and 2.3616(10) Å] are comparable to
those in the ruthenium tetraphenylphosphonium salt [average
2.3639(8) Å].37 Both indazole ligands lie in one plane in
contrast to what was found in (Ph4P)[trans-RuCl4(Hind)2]
where they were significantly twisted. The plane through both
indazole ligands crosses the equatorial plane between Cl1
and Cl2, the torsion angle Cl1-Ru-N1-N2 and the angle
Cl1-Ru-Cl2 being 63.5(2)° and 90.33(4)°, respectively. The
distribution of electron density over the trans indazole ligands
in both complexes is also similar.
Very strong H-bond(s) between the two indazole units of
the cation were found. This interaction can be classified as
a homonuclear centered H-bond of the type A in accordance
with Speakman classification,50 with the proton located in a
center of symmetry. The interatomic separation N3‚‚‚N3′
[2.664 Å] is significantly shorter than the van der Waals
radius of the nitrogen atom [3.10 Å].50 Crystallographic data
on the hydrogen-bonded cations are relatively scarce.51
Electrochemical Studies. Cathodic Behavior in Aprotic
Media. The cyclic voltammograms of 0.15 M [nBu4N][BF4]/
DMF or DMSO solutions of the complexes (HL)[trans-
RuCl4L(DMSO)] (A) and (HL)[trans-RuCl4L2] (B) [L )
imidazole (Him), 1,2,4-triazole (Htrz) or indazole (Hind)],
at a glassy carbon working electrode and at a scan rate of
0.2 V s-1 display (Figures 8-10) one quasireversible (for
A) or irreversible (for B) single-electron (as confirmed by
cathodic controlled potential electrolyses (CPE) of A1 and
B3) reduction wave, Ired, which is assigned to the RuIIIfRuII
process. The reduction potentials (Table 4) for compounds
A (E1/2red ) -0.04 to -0.23 V) are considerably higher than
those for B (Ep/2red ) -0.41 to -0.74 V) which agrees with
the stronger net electron-acceptor character of the S-
Figure 6. Structure of [trans-RuCl4(Hind)2]- in B3a.
Figure 7. Structure of (Hind)2H+ in B3a.
Ru-N1 bond [2.0901(13) Å] is significantly longer than
those in (Ph4P)[trans-RuCl4(Hind)2]37 [2.0517(21) and 2.0711-
(22) Å] and 8.5 σ longer than that in [(Hind)2H][trans-RuCl4-
(Hind)2] (see below) [2.062(3) Å], which is expected taking
into account the stronger trans-influence of the S atom of
DMSO when compared to the N atom of the azole ligand.17
However, the Ru-S1 bond (sulfur being trans to the indazole
nitrogen) [2.2924(7) Å] is markedly shorter than that found
in [(DMSO)2H][RuCl4(DMSO)2] [2.348(1) Å] with mutually
trans dimethyl sulfoxide ligands.34 Thus, replacement of one
of the DMSOs by an indazole ligand results in a shortening
of the other Ru-DMSO bond due to a diminished π-back-
bonding competition between the two trans-ligands or
decreased σ-directed trans influence of nitrogen ligand. The
SdO bond length of 1.4951(12) Å is slightly longer than
that of free DMSO [1.492(1) Å]43 indicating a greater S-O
single bond character for the S-bonded molecule.
The crystal structure of B3a consists of the complex anions
[trans-RuCl4(Hind)2]- (Figure 6) and the unusual [(Hind)‚
‚‚H‚‚‚(Hind)]+ cations (Figure 7). It displays a distorted
octahedral coordination with four chloride ions in the
equatorial plane and two indazole ligands in axial positions.
The structure of the anion is stabilized by two symmetry
related intramolecular hydrogen bonds of the type N-H‚‚‚
Cl (N2-H20 0.76 Å, H20‚‚‚Cl2 2.68 Å, N2‚‚‚Cl2 3.15 Å,
N2H20Cl2 121.9°). In contrast to (Ph4P)[trans-RuCl4-
(Hind)2]37 the ruthenium atom in B3a lies at a center of
(50) Emsley, J. Chem. Soc. ReV. 1980, 9, 91-124.
(51) James, B. R.; Morris, R. H.; Einstein, F. W. B.; Willis, A. J. Chem.
Soc. Chem. Commun. 1980, 31-32.
7088 Inorganic Chemistry, Vol. 43, No. 22, 2004

Tuning of Redox Potentials for Ruthenium Anticancer Drugs
Table 4. Cyclic Voltammetric Data for (HL)[trans-RuCl4L(DMSO)]
(A) and (HL)[trans-RuCl4L2] (B) [L ) Him (A1 and B1), Htrz (A2 and
B2), or Hind (A3 and B3)] and Estimated Redox Potential Values
Experimentala
Calculatedb
complex Ered/Ired Eox/IIox Eox/ac Eox/bc Eox/a′c E/Ired
E/a
E/b
A1d
A2
A3
B1
-0.22
(-0.23)
-0.06
(-0.09)
-0.04
(-0.10)
-0.72*
1.45
1.51
1.55*
0.02
(0.66)
0.13
(0.75)
0.17
-0.22 0.04
(0.57)
-0.16 0.10
(0.63)
Figure 9. Cyclic voltammogram of 5 mM (H2trz)[trans-RuCl4(Htrz)-
(DMSO)] A2, in DMF with 0.15 M [nBu4N][BF4], at a carbon disk working
electrode and at a scan rate of 0.2 V s-1, showing two consecutive scan
cycles [first (black line) and second (gray line) scan].
(0.78)
0.96 -0.49 -0.15 1.29 -0.66 -0.39 -0.12
(-0.74)* (0.99)
B2
-0.48*
1.18 -0.21
0.03 1.43 -0.54 -0.28
0.00
(-0.44)* (1.19)
B3e
-0.43*
1.27 -0.15
0.16 1.55
(-0.41)* (1.27)
a Potentials in V ( 0.02 vs NHE measured at a scan rate of 0.2 V s-1
,
in 0.15 M [nBu4N][BF4]/DMF (or DMSO, in brackets). E1/2 values are given
for the reversible waves, whereas for the irreversible ones, the Ep/2 values
are indicated by an asterisk. b Potentials in V vs NHE estimated from eq 1
with SM and IM values of 0.97 and 0.04, respectively, and known EL for
the ligands (see text). Values in brackets refer to the DMSO derivative.
c Oxidation waves for the complexes formed upon the first [(a) (RuII/RuIII)
and (a′) (RuIII/RuIV)] and the second [(b) (RuII/RuIII)] chloride replacement
by solvent molecules detected upon scan reversal after the formation of
wave Ired (see text, Figures 8-10, and Schemes 1 and 2). d Wave b and the
additional waves c and d of the complexes (see text) formed by further
chloride replacement by solvent are observed after CPE, at Eox values of
0.31, 0.67, and 0.97 V, respectively; the corresponding estimated values
are 0.30, 0.56, and 0.83 V vs NHE. e Waves c and d of the complexes (see
text) formed by further chloride replacement by solvent are observed after
CPE, at Eox values of 0.43 and 0.62 V, respectively; the corresponding
estimated values are 0.35 and 0.64 V vs NHE.
Figure 10. Cyclic voltammograms of a 5 mM (H2ind)[trans-RuCl4(Hind)2]
B3 solution in DMF with 0.15 M [nBu4N][BF4] at a scan rate of 0.2 V s-1
and at a platinum disk (black line) or glassy carbon (gray line) working
electrode with scan reversal following wave * or the reduction wave Ired
(a), or showing the effect of replacing the latter by the former electrode
(b).
the oxidation potentials of [trans-RuCl4(Hind)(DMSO)]-
(A3) and [trans-RuCl4(Hind)2]- (B3) in DMF and DMSO
solutions, the known values of EL ligand parameters and of
SM and IM (see above), enabled the estimate of the yet
unknown EL ligand parameter for indazole as an average of
four values (two measured in DMF and further two in
DMSO) obtained in this work: EL(Hind) ) 0.26 V vs NHE.
It is higher than those of triazole and imidazole, indicating
that indazole is a weaker net electron-donor than the other
ligands.
Upon potential scan-reversal following the formation of
wave Ired [but not of wave * (Figure 10a)], one (wave a, for
A, Figures 8a-c and 9) or two (waves a and b, for B, Figures
8d and 10) quasireversible oxidation waves are detected,
being due to the oxidation of products formed in the cathodic
processes. CPE at the cathodic wave Ired generates several
species as indicated by monitoring the electrolysis by cyclic
voltammetry (CV) (apart from waves a and b, two other
waves, c and d, are detected at higher potentials, Figure S1
in the Supporting Information). Attempts to isolate these
products failed, but on the basis of the agreement between
the measured values of the oxidation potentials of these
waves from each complex (0.02, 0.31, 0.67, and 0.96 V from
A1 and -0.15, 0.16, 0.43, and 0.62 V from B3) and the
predicted ones [0.04, 0.30, 0.56, and 0.83 V and -0.14, 0.12,
0.35, and 0.64 V, respectively, by applying Lever’s equation
(eq 1) and the known EL values (see above) for chloride,
DMSO and Him, as well as for DMF (0.03 V)53] we propose
the following formulations, derived from chloride ligand dis-
coordinated DMSO ligand as compared with the azole ones
(see below) and with the more effective σ-donor ability of
the latter ligands. In addition, in both series of compounds,
the reduction potentials of the Htrz complexes are less
cathodic than those of the Him ones, in accord with the
relative electron-donor character of the N-ligands.
By replacing the vitreous carbon disk by a platinum disk
working electrode, the cathodic pattern for both types (A
and B) of complexes is as described above, apart from the
detection of an additional irreversible reduction wave [at
red
potential values of Ep ) -0.52 V (A1 and B1), -0.20 V
(A2 and B2), -0.14 V (A3 and B3), or -0.15 V (B3a, wave
* in Figure 10)] which is due to the protic reduction of the
azolium counterions H2im+, H2trz+, H2ind+, or (Hind)2H+,
respectively. The values for triazolium (H2trz)+ and imida-
zolium (H2im)+ are in agreement with those found in the
literature.13,52
red
The E1/2 values for complexes A and B are in good
agreement with those expected on the basis of the general
equation (eq 1) proposed by Lever, by using the known
values of SM and IM (0.97 and 0.04, respectively)19 for the
RuIII/RuII redox couple and the EL values for the various
ligands [EL(Cl-) ) -0.24,19 EL(DMSO) ) 0.57,28 EL(Him)
) 0.12,19 and EL(Htrz) ) 0.1819] (Table 4). Moreover, the
application of eq 1, considering the experimental values of
(52) Serli, B.; Zangrando, E.; Iengo, E.; Mestroni, G.; Yellowlees, L.;
Alessio, E. Inorg. Chem. 2002, 41, 4033-4043.
(53) http://www.chem.yorku.ca/profs/lever (homepage of A. B. P. Lever).
Inorganic Chemistry, Vol. 43, No. 22, 2004 7089

Reisner et al.
Scheme 1. Anodic and Proposed Cathodic Processes of Complexes
A. L′ ) DMF or DMSO
at Ru(II), itself a good π-donor where all t2g orbitals are
filled.59
This is also consistent with the known promotion of
hydrolysis of chloride ligands for compound A1 upon
chemical reduction by biological reductants (e.g., ascorbic
acid), in aqueous physiological conditions, as shown by UV-
vis and 1H NMR studies (but not by electrochemical
methods).10,60,61
The high-scan-rate-limiting behaviors of Schemes 1 and
2 should correspond to the quasireversible single-electron
reduction of A or B since there would be no time for
appreciable dechlorination. For complexes A cyclic voltam-
I
ox
Scheme 2. Anodic and Proposed Cathodic Processes of Complexes
B. L′ ) DMF
mograms at scan rates higher than 10 V s-1 reveal a ip
/
Iipred ratio equal to 1, but such a limiting behavior is not yet
achieved for complexes B up to 40 V s-1. Upon decreasing
the scan rate, there occurs an increasing conversion of A or
B into the solvolysis products (Figures 8a and b), and
multiple scans at a given scan rate result in an increase of
wave a (Figures 8a and c) or b (Figure 8d) with time. No
significant variation of the above behaviors was detected
upon changing the concentration of the complexes, thus
indicating the involvement of first-order chemical steps.
Kinetic Investigation of the Cathodically Induced
Replacement of Chloride. The mechanisms of Schemes 1
and 2 were investigated in detail by digital simulation
(program ESP)62 of the cyclic voltammograms in the
available range of scan rates (0.01-30 V s-1). A good fit
was obtained (Figures 11 and 12) for the degree of revers-
ibility of the cathodic wave Ired, i.e. Iipox/Iipred, and the extent
of formation of the chloride displacement product, i.e. the
placement at RuII by solvent (Schemes 1 and 2, depending
on the starting complex A or B): (a), [RuIICl3(DMF)L(DM-
SO)]- or [RuIICl3(DMF)L2]-; (b), [RuIICl2(DMF)2L(DMSO)]
or [RuIICl2(DMF)2L2]; (c), [RuIICl(DMF)3L(DMSO)]+ or
[RuIICl(DMF)3L2]+; and (d), [RuII(DMF)4L(DMSO)]2+ or
[RuII(DMF)4L2]2+. In the time-scale of electrolysis and for
both complexes, the formation of species (c) appears to be
dominant as suggested by the predominance of the current
intensity of its wave as compared with the others. Further-
more, for complexes B, the quasireversible second oxidation
(RuIII/RuIV) of [RuCl3(DMF)L2]- (wave a′), formed in the
cathodic process upon RuIIIfRuII reduction (wave Ired), was
also observed by running a simple CV (Table 4). For
complexes A the onset of the solvent-electrolyte discharge
potential prevented the detection of the corresponding
RuIIIfRuIV oxidation waves.
normalized peak-current of wave (a), ipox/Iip (F) (Figure
a
red
12), and wave (b), ipox/Iip (F′) (Figure 12d) as a function
of scan rate. The lowering of the first parameter relative to
the unit reflects the extent of the chemical reaction following
the electron transfer (chloride replacement by solvent),
whereas F and F′ are determined by the relative amount of
the formed corresponding solvent complex. The optimized
values of the homogeneous rate constants for the first (k1),
second (k2), and third (k3, only for B) replacement of chloride
by DMF are quoted in Table 5. The effect of the rate constant
on the reversibility of the reduction wave and on the F
parameter is clearly observed in Figure 12a-c: a decrease
in the former results in (i) an increase of the reversibility
b
red
(54) Salih, T. A.; Duarte, M. T.; Frau´sto da Silva, J. J. R.; Galva˜o, A. M.;
Guedes da Silva, M. F. C.; Hitchcock, P. B.; Hughes, D. L.; Pickett,
C. J.; Pombeiro, A. J. L.; Richards, R. L. J. Chem. Soc., Dalton Trans.
1993, 3015-3023.
A further evidence for chloride loss upon reduction of the
complexes under study is the oxidation wave (e) (Figures 9
and 10a) detected in the subsequent anodic scan upon
formation of wave Ired, at potential values (Ep ) 1.30 - 1.40
V) identical to those measured for the oxidation of Cl- in
solutions of (Ph3PCH2Ph)Cl, (PPN)Cl, or (Me4N)Cl under
the same experimental conditions. The cathodically induced
chloride loss, with the concomitant formation of the above
products, is a well documented type of reaction.28,54-56 In
addition, it is also known that halide is relatively inert to
substitution in the coordination sphere of ruthenium(III),52,57,58
but as a strong σ-donor and a potential π-donor it is labile
(55) Costa, G.; Balducci, G.; Alessio, E.; Tavagnacco, C.; Mestroni, G. J.
Electroanal. Chem. 1990, 296, 57-76.
(56) Duff, C. M.; Heath, G. A. J. Chem. Soc., Dalton Trans. 1991, 2401-
2411.
(57) Zanella, A. W.; Ford, P. C. Inorg. Chem. 1975, 14, 42-47.
(58) Coleman, G. N.; Gesler, J. W.; Shirley, F. A.; Kuempel, J. R. Inorg.
Chem. 1973, 12, 1036-1038.
(59) Ford, P. C. Coord. Chem. ReV. 1970, 5, 75-99.
(60) Sava, G.; Bergamo, A.; Zorzet, S.; Gava, B.; Casarsa, C.; Cocchietto,
M.; Furlani, A.; Scarcia, V.; Serli, B.; Iengo, E.; Alessio, E.; Mestroni,
G. Eur. J. Cancer 2002, 38, 427-435.
(61) Iengo, E.; Mestroni, G.; Geremia, S.; Calligaris, S.; Alessio, E. J. Chem.
Soc. Dalton Trans. 1999, 3361-3371.
(62) Nervi, C. Electrochemical Simulation Package, ESP, Version 2.4;
Dipartimento di Chimica IFM: Torino, Italy, 1994/98; (nervi@lem.
ch.unito.it).
7090 Inorganic Chemistry, Vol. 43, No. 22, 2004

Tuning of Redox Potentials for Ruthenium Anticancer Drugs
Figure 11. Experimental (10 mM, black line) and simulated cyclic
voltammograms (gray line) for (H2ind)[trans-RuCl4(Hind)(DMSO)] A3 in
DMF with 0.15 M [nBu4N][BF4] at a carbon working electrode and at a
scan rate of 0.02 (a), 0.1 (b), 0.5 (c), and 2 (d) V s-1. The simulated
voltammograms were obtained by using the optimized values of the rate
constants given in Table 5.
with a shift of the inflection point of the ipox/Iip curve to
lower scan rates and (ii) a concomitant shift of the maximum
of the F curve. Both rate constants increase with the net
electron donor/acceptor character of the azole ligand (as
expressed by EL) and with its protic basicity (increase of
pKa of the corresponding free azolium acid H2L+).
I
red
The promotion of the rate of the first chloride hydrolysis
by an increase of the basicity of the nitrogen heterocyclic
ligand was recognized,61 although in a qualitative way and
in aqueous medium, for related chloro-RuIII complexes.
However, generalizations for dinuclear complexes with a
bridging N,N-heterocyclic ligand61 should be taken cautiously
in view of the dependence of the rate on the oxidation states
of both metals.
Although the accurate determination of the solvolysis rate
constants of compounds B was hampered by the absence of
wave Iox over a wide range of potential scan rates (Figure
12d), we observed that their values are higher than the
corresponding ones found for A and follow their general
tendency. Dechlorination is thus promoted by an increase
of the electron transfer from the neutral ligands to the metal
with a conceivable concomitant weakening of the metal-
chloride ligand bonds. However, the rate of chloride dis-
placement decreases with the degree of substitution (k1 >
k2 > k3) (Table 5).
Anodic Behavior in Aprotic Media. The cyclic voltam-
mograms of 0.15 M [nBu4N][BF4]/DMF or DMSO solutions
of complexes A and B present one quasireversible (irrevers-
ible in the case of A3) single-electron oxidation wave (wave
IIox, Figures 9 and 10a) at oxidation potential values of 1.45-
1.55 (A) or 0.96-1.27 (B) V vs NHE (Table 4), which is
assigned to the RuIIIfRuIV oxidation. The oxidation potential
of wave IIox reflects, in each set of compounds, the relative
donor/acceptor abilities of the DMSO and the azole ligands
as measured by EL and shown by the plot of Figure 13, which
Figure 12. Experimental (symbols) and theoretical (lines) variations of
the reversibility of the RuIII/RuII reduction wave I, Iipox/Iipred ([, black solid
red
line), of the parameters F ) aipox/Iip (b, gray solid line) and F′ )
bipox/Iipred (9, black dashed line) as a function of scan rate (logarithmic scale)
for compounds A1 (a), A2 (b), A3 (c), and B3 (d). Experimental error bars
are shown at the top right corner. The theoretical lines were obtained by
considering the optimized rate constant values given in Table 5.
includes also the available56 redox potentials for the RuIII/RuIV
couple of the complexes [RuX6-n(RCN)n]z/z- (X ) Cl-, Br-
and R ) Me, Ph). This relationship is expressed by eq 2
(r ) 0.98), which allowed the estimate, for the first time, of
SM and IM (compare with eq 1) for the RuIII/IV redox couple
in organic solvents: SM ) 1.03 and IM ) 1.68 V vs NHE.
E ) 1.03‚ E + 1.68
(2)
∑
L
The slope (SM) is similar to that of the RuIII/II redox couple
and therefore the redox potentials of the RuIII/IV and RuIII/II
centers display a comparable sensitivity to a change of
ligands. However, in view of the limited number of points,
Inorganic Chemistry, Vol. 43, No. 22, 2004 7091

Reisner et al.
Table 5. Kinetic Dataa for (HL)[trans-RuCl4L(DMSO)] (A) and
(HL)[trans-RuCl4L2] (B) [L ) Him (A1 and B1), Htrz (A2 and B2), or
Hind (A3 and B3)], EL Values for the Azole Ligands (L) and PKa
Values for the Free Azolium Acids (H2L+)
compound
k1/s-1
k2/s-1
k3/s-1
EL(L) pKa (H2L+)
A1
A2
A3
B1
B2
B3
3.1 ( 0.2
1.4 ( 0.3
0.7 ( 0.2
> 200
0.20 ( 0.02
0.12 ( 0.03
0.03 ( 0.02
(5 ( 1)‚10
0.12b
0.18b
0.26c
7.00d
2.55d
1.25d
7.00d
2.55d
1.25d
0.4 ( 0.2 0.12b
0.3 ( 0.1 0.18b
0.2 ( 0.1 0.26c
(10 ( 5)‚10 1 ( 0.5
(8 ( 5)‚10 0.5 ( 0.2
Figure 14. Plot of E1/2red (Ired) for complexes (HL)[trans-RuCl4L(DMSO)]
(A) and (HL)[trans-RuCl4L2] (B) in aqueous phosphate buffer, pH 7, against
ΣEL (in V vs. NHE). E ) 0.88‚ΣEL + 0.46 (r ) 0.99). Y1 ) Na[trans-
RuCl4(Hpy)(DMSO)] (E1/2 ) 0.30 V),17 and Y2 ) Na[trans-RuCl4(Meim)-
(DMSO)] (E1/2 ) 0.24 V),17 where EL (Hpy) ) 0.20,19 and EL (Meim) )
0.08 V.19
a The homogeneous rate constants k1, k2, and k3 were determined by
using the ESP simulation program (ref 62) with khet of (5 ( 1)‚10-3 cm
s-1 b From ref 19. c Estimated from eq 1, by using the known (ref 19) values
.
of SM and IM, and EL for the other ligands (see text). d From refs 39-41.
and Na[RuCl4(Meim)(DMSO)] (Meim ) N-methylimida-
zole) whose redox potentials were reported17 by others. This
expression allows estimation of, for the first time, the SM
and IM values for the RuIII/II redox couple for 1- charged
RuIII complexes, in aqueous phosphate buffer at pH 7.
E ) 0.88‚ E + 0.46
(3)
∑
L
However, since in aqueous medium the redox potential
can strongly depend on the pH of the solution,19,63 as well
as on the net charge of the complex,19 this expression is not
expected to be valid when the solution pH or the complex
net charge is different from the above.
Figure 13. Plot of E1/2 (RuIII/RuIV) against ΣEL (in V vs NHE) for the
ox
following complexes: (HL)[trans-RuCl4L(DMSO)] (A) and (HL)[trans-
RuCl4L2] (B) (IIox) in DMF (this work); (Bu4N)2[RuX6] (X ) Br, X1; Cl,
X2), (Bu4N)2[RuX5(RCN)] (X ) Br, R ) Ph, X3; Cl, Me, X4; Cl, Ph,
X5), (Bu4N)[RuX4(RCN)2] (Br, Ph, X6; Cl, Me, X7; Cl, Ph, X8), [RuX3-
(RCN)3] (Cl, Me, X9; Cl, Ph, X10) in CH2Cl2 or CH2Cl2/RCN (R ) Me,
Ph) (ref 56). E ) 1.03‚ΣEL + 1.68 (r ) 0.98).
Despite the quasireversible character of the reduction wave
during the CV time scale of the experiment, exhaustive
cathodic CPE of complex A1 in phosphate buffer, pH 7,
results in the disappearance of the RuIII/RuII wave of the
starting compound, and does not reveal any new electroactive
species. However, CV of a 0.2 M KNO3 solution of
complexes A reveals the Ired wave of the RuIII/RuII couple
and, in the subsequent anodic scan, an additional quasire-
versible oxidation wave (a) of a species formed upon
reduction, at a potential that is 0.12 V higher than that of
the starting RuIII/RuII couple. The similarity of the latter
behavior with that (see above) in DMF or DMSO and earlier
NMR studies60,61 on the hydrolysis of A1 upon reduction in
aqueous medium with biological reductants, suggest ligand
chloride loss with replacement by the solvent (water) as the
initial chemical reaction upon reduction.
Table 6. Cyclic Voltammetric Dataa for (HL)[trans-RuCl4L(DMSO)]
(A) and (HL)[trans-RuCl4L2] (B) [L ) Him (A1 and B1), Htrz (A2 and
B2), or Hind (A3 and B3bc)] in Aqueous Phosphate Buffer, pH 7
compound
E1/2red/Ired a
A1
A2
A3
B1
0.25b
0.30
0.35
-0.16b
0.00
B2
B3bc
0.03
a Values in V ( 0.02 vs NHE. b Values in agreement with those given
in the literature (ref 17 and 63). c Na[trans-RuCl4(Hind)2], B3b, was used
instead of the indazolium salt, B3, due to the low solubility of the latter in
the aqueous electrolyte solution.
further studies on a wider variety of RuIII/RuIV metal
complexes are required before the generality of eq 2 can be
fully recognized.
Conclusions
Lever’s general equation19 was shown to predict well the
reduction potentials of our ruthenium(III) anticancer drugs
with the general formulas [trans-RuCl4L(DMSO)]- and
[trans-RuCl4L2]- in organic media, and a particular form of
that expression was proposed to be applied to (1-) negatively
charged RuIII complexes in aqueous phosphate buffer medium
at pH 7.
Reductively induced stepwise replacement of chloride
ligands by solvent molecules was proposed on the basis of
the identification of the products of substitutions by using
Lever’s equation. Kinetic investigations by digital simulation
Electrochemical Behavior in Aqueous Phosphate Buffer
at pH 7. The cyclic voltammograms of complexes A and B
in 0.2 M phosphate buffer, pH 7, at a carbon disk working
electrode display one quasireversible single-electron reduc-
red
tion wave assigned to the RuIIIfRuII process, at E1/2
potential values ranging from 0.25 to 0.35 (for A) or -0.16
to 0.03 (for B) V vs NHE (Table 6), i.e., ca. 0.4-0.6 V less
cathodic than those measured in the organic solvents.
The linear relationship between the reduction potential and
ΣEL expressed by eq 3 (r ) 0.99) was obtained from the
plot of Figure 14, which includes not only our complexes A
and B, but also Na[RuCl4(Hpy)(DMSO)] (Hpy ) pyrazole)
(63) Ni Dhubhghaill, O. M.; Hagen, W. R.; Keppler, B. K.; Lipponer, K.
G.; Sadler, P. J. J. Chem. Soc., Dalton Trans. 1994, 3305-3310.
7092 Inorganic Chemistry, Vol. 43, No. 22, 2004

Tuning of Redox Potentials for Ruthenium Anticancer Drugs
of the cyclic voltammograms of these reactions revealed that
an increase of the net electron donor character (decrease of
EL) of the neutral ligands (DMSO < indazole < triazole <
imidazole) and of the basicity of the azole ligand promotes
the complex lability toward the solvolytic dechlorination, but
hampers the RuIIIfRuII reduction (shifts cathodically the
reduction potential). Hence, a thermodynamic stabilization
effect of the ligands on the RuIII [relative to RuII] complexes
is concomitant with the kinetic labilization of the corre-
sponding reduced RuII forms.
Replacement of a Cl- ligand by the solvent results in a
positive shift of the reduction potential, and this fact should
also be taken into account to design potential antitumor drugs
able to follow the activation by the reduction pathway.
The slope, SM, and intercept, IM, of Lever’s equation were
also determined, for the first time, for the RuIII/RuIV couple
in organic medium, which will allow the prediction of the
redox potential of other ruthenium systems with such a redox
couple.
drugs with desired redox properties and chloride substitution
lability, in biological systems, which currently constitutes
quite a challenging task in the field of medicinal inorganic
chemistry.
Acknowledgment. We are indebted to FWF (Austrian
Science Fund), University of Vienna (Faculty of Sciences
and Mathematics, International Relations Office), COST, the
Foundation for Science and Technology and its POCTI-
program (FEDER funded) (Portugal), and the European
Community (contract MRTN-CT-2003-503864), for financial
support. We also thank Prof. G. Giester for X-ray data
collection and Dr. T. Weyhermu¨ller for discussions of the
crystallographic part of the work.
Note Added after ASAP: The version of this paper
posted September 21, 2004, contained a number of errors,
primarily in notation. Also, one value in Table 5 was
incorrect. The version posted on September 23, 2004, has
been corrected.
Although the generality of some of the above conclusions
has still to be tested with a wider variety of ruthenium
complexes, they are of significance for the understanding
of the RuIII reduction process which appears to play a key
role in the activation of the ruthenium prodrugs in vivo. They
can also be further applied to design ruthenium antitumor
Supporting Information Available: X-ray crystallographic files
in CIF format for A2a, A2b, A3, and B3a. Cyclic voltammograms
of A1 and B3b (pdf).This material is available free of charge via
the Internet at http://pubs.acs.org.
IC049479C
Inorganic Chemistry, Vol. 43, No. 22, 2004 7093