Novel dinuclear luminescent compounds based on iridium(III) cyclometalated chromophores and containing bridging ligands with ester-linked chelating sites

Article abstract of DOI:10.1021/ic000212x

The syntheses and study of the spectroscopic, redox, and photophysical properties of a new set of species based on Ir(III) cyclometalated building blocks are reported. This set includes three dinuclear complexes, that is, the symmetric (with respect to the bridging ligand) diiridium species [(ppy)₂Ir(μ-L-OC(O)-C(O)O-L)Ir(ppy)₂] [PF₆]₂ (5; ppy = 2-phenyl-pyridine anion; L-OC(O)-C(O)O-L = bis[4-(6′-phenyl-2,2′-bipyridine-4′-yl)phenyl] benzene-1,4-dicarboxylate), the asymmetric diiridium species [(ppy)₂Ir(μ-L-OC(O)-L)Ir(ppy)₂] [PF₆]₂ (3; L-OC(O)-L = 4-{[(6′-phenyl-2,2′-bipyridine-4′-yl) benzoyloxy]phenyl}-6′-phenyl-2,2′-bipyridine), and the mixed-metal Ir-Re species [(ppy)₂Ir(μ-L-OC(O)-L)Re(CO)₃Br][PF₆] (4). Syntheses, characterization, and spectroscopic, photophysical, and redox properties of the model mononuclear compounds [Ir(ppy)₂(L-OC(O)-L)][PF₆] (2) and [Re(CO)₃(L-COOH)Br] (6; L-COOH = 4′-(4-carboxyphenyl)-6′-phenyl-2,2′-bipyridine) are also reported, together with the syntheses of the new bridging ligands L-OC(O)-L and L-OC(O)-C(O)O-L. The absorption spectra of all the complexes are dominated by intense spin-allowed ligand-centered (LC) bands and by moderately intense spin-allowed metal-to-ligand charge-transfer (MLCT) bands. Spin-forbidden MLCT absorption bands are also visible as low-energy tails at around 470 nm for all the complexes. All the new species exhibit metal-based irreversible oxidation and bipyridine-based reversible reduction processes in the potential window investigated (between +1.80 and -1.70 V vs SCE). The redox behavior indicates that the metal-based orbitals are only weakly interacting in dinuclear systems, whereas the two chelating halves of the bridging ligands exhibit noticeable electronic interactions. All the complexes are luminescent both at 77 K and at room temperature, with emission originating from triplet MLCT states. The luminescence properties are temperature- and solvent-dependent, in accord with general theories: emission lifetimes and quantum yields increase on passing from acetonitrile to dichloromethane fluid solution and from room-temperature fluid solution to 77 K rigid matrix. In the dinuclear mixed-chromophore species 3 and 4, photoinduced energy transfer across the ester-linked bridging ligands seems to occur with low efficiency.

Full text of DOI:10.1021/ic000212x

Inorg. Chem. 2001, 40, 1093-1101
1093
Novel Dinuclear Luminescent Compounds Based on Iridium(III) Cyclometalated
Chromophores and Containing Bridging Ligands with Ester-Linked Chelating Sites§
Francesco Neve,*,† Alessandra Crispini,† Scolastica Serroni,‡ Fre´de´rique Loiseau,‡ and
Sebastiano Campagna*,‡
Dipartimento di Chimica, Universita` della Calabria, I-87030 Arcavacata di Rende (CS), Italy, and
Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, Universita` di Messina,
via Sperone 31, I-98166 Messina, Italy
ReceiVed February 25, 2000
The syntheses and study of the spectroscopic, redox, and photophysical properties of a new set of species based
on Ir(III) cyclometalated building blocks are reported. This set includes three dinuclear complexes, that is, the
symmetric (with respect to the bridging ligand) diiridium species [(ppy)2Ir(µ-L-OC(O)-C(O)O-L)Ir(ppy)2]-
[PF6]2 (5; ppy ) 2-phenylpyridine anion; L-OC(O)-C(O)O-L ) bis[4-(6′-phenyl-2,2′-bipyridine-4′-yl)phenyl]-
benzene-1,4-dicarboxylate), the asymmetric diiridium species [(ppy)2Ir(µ-L-OC(O)-L)Ir(ppy)2][PF6]2 (3; L-
OC(O)-L ) 4-{[(6′-phenyl-2,2′-bipyridine-4′-yl)benzoyloxy]phenyl}-6′-phenyl-2,2′-bipyridine), and the mixed-
metal Ir-Re species [(ppy)2Ir(µ-L-OC(O)-L)Re(CO)3Br][PF6] (4). Syntheses, characterization, and spectroscopic,
photophysical, and redox properties of the model mononuclear compounds [Ir(ppy)2(L-OC(O)-L)][PF6] (2) and
[Re(CO)3(L-COOH)Br] (6; L-COOH ) 4′-(4-carboxyphenyl)-6′-phenyl-2,2′-bipyridine) are also reported,
together with the syntheses of the new bridging ligands L-OC(O)-L and L-OC(O)-C(O)O-L. The absorption
spectra of all the complexes are dominated by intense spin-allowed ligand-centered (LC) bands and by moderately
intense spin-allowed metal-to-ligand charge-transfer (MLCT) bands. Spin-forbidden MLCT absorption bands are
also visible as low-energy tails at around 470 nm for all the complexes. All the new species exhibit metal-based
irreversible oxidation and bipyridine-based reversible reduction processes in the potential window investigated
(between +1.80 and -1.70 V vs SCE). The redox behavior indicates that the metal-based orbitals are only weakly
interacting in dinuclear systems, whereas the two chelating halves of the bridging ligands exhibit noticeable
electronic interactions. All the complexes are luminescent both at 77 K and at room temperature, with emission
originating from triplet MLCT states. The luminescence properties are temperature- and solvent-dependent, in
accord with general theories: emission lifetimes and quantum yields increase on passing from acetonitrile to
dichloromethane fluid solution and from room-temperature fluid solution to 77 K rigid matrix. In the dinuclear
mixed-chromophore species 3 and 4, photoinduced energy transfer across the ester-linked bridging ligands seems
to occur with low efficiency.
Introduction
Most of the systems investigated so far are based on Ru(II)
building blocks, with Os(II) and Re(I) species also playing
The design of luminescent and redox-active multinuclear
polypyridine metal complexes continues to attract great interest
in light of the key role that these species currently play in the
development of multicomponent assemblies featuring photoin-
duced made-to-order properties.1-6 In particular, multinuclear
polypyridine metal complexes contribute extensively to the
development of artificial antennas for the photochemical
conversion of solar energy7-10 and as active elements for the
elaboration and storage of information at the molecular level.11-15
leading roles.2 However, Ir(III) cyclometalated complexes have
recently been the object of increasing interest because the
excited-state properties of these species can be competitive with
respect to those of ruthenium systems16-20 (for example,
(7) Balzani, V.; Juris, A.; Venturi, M.; Campagna, S.; Serroni, S. Acc.
Chem. Res. 1998, 31, 26.
(8) Slate, C. A.; Striplin, D. R.; Moss, J. A.; Chen, P.; Erickson, B. W.;
Meyer, T. J. J. Am. Chem. Soc. 1998, 120, 4885.
(9) Hu, Y.-Z.; Tsukiji, S.; Shinkai, S.; Oishi, S.; Hamachi, I. J. Am. Chem.
Soc. 2000, 122, 241.
(10) Campagna, S.; Serroni, S.; Puntoriero, F.; Di Pietro, C. In Electron
Transfer in Chemistry; Balzani, V., Ed.; VCH-Wiley; Vol. 5, in press.
(11) Beer, P. D.; Szemes, F.; Balzani, V.; Sala`, C. M.; Drew, M. G.; Dent,
S. W.; Maestri, M. J. Am. Chem. Soc. 1997, 119, 11864.
(12) Barigelletti, F.; Flamigni, L.; Collin, J.-P.; Sauvage, J.-P. Chem.
Commun. 1997, 333.
(13) Waldmann, O.; Hassmann, J.; Mu¨ller, P.; Hanan, G. S.; Volkmer, D.;
Schubert, U. S.; Lehn, J.-M. Phys. ReV. Lett. 1997, 78, 3390.
(14) Bassani, D. M.; Lehn, J.-M.; Fromm, K.; Fenske, D. Angew. Chem.,
Int. Ed. Engl. 1998, 37, 2364.
§ Dedicated to Professor Alan L. Balch on the occasion of his 60th
birthday.
* To whom correspondence should be addressed. E-mail: f.neve@unical.it
or photochem@chem.unime.it.
† Universita` della Calabria.
‡ Universita` di Messina.
(1) Sauvage, J.-P.; Collin, J.-P.; Chambron, J.-C.; Guillerez, S.; Coudret,
C.; Balzani, V.; Barigelletti, F.; De Cola, L.; Flamigni, L. Chem. ReV.
1994, 94, 993.
(2) Balzani, V.; Juris, A.; Venturi, M.; Campagna, S.; Serroni, S. Chem.
ReV. 1996, 96, 759.
(15) Zahavy, E.; Fox, M. A. Chem. Eur. J. 1998, 4, 1647.
(16) Crosby, G. A. Acc. Chem. Res. 1975, 8, 231.
(17) Ayala, N. P.; Flynn, C. M., Jr.; Sacksteder, L. A.; Demas, J. N.; De
Graff, B. A. J. Am. Chem. Soc. 1990, 112, 3837.
(18) van Diemen, J. H.; Hage, R.; Lempers, H. E. B.; Reedijk, J.; Vos, J.
G.; De Cola, L.; Barigelletti, F.; Balzani, V. Inorg. Chem. 1992, 31,
3518.
(3) Harriman, A.; Ziessel, R. Chem. Commun. 1996, 1707.
(4) Shaw, J. R.; Sadler, G. S.; Wacholtz, W. F.; Ryu, C. K.; Schmehl, R.
H. New J. Chem. 1996, 20, 749.
(5) De Cola, L.; Belser, P. Coord. Chem. ReV. 1998, 177, 301.
(6) Barigelletti, F.; Flamigni, L. Chem. Soc. ReV. 2000, 29, 1.
10.1021/ic000212x CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/08/2001

1094 Inorganic Chemistry, Vol. 40, No. 6, 2001
Neve et al.
Chart 1
luminescence quantum yields of iridium cyclometalated species
are often larger than those of analogous ruthenium compounds
having the same excited-state energy and the former compounds
are usually better photoreductants than the latter) and new
synthetic strategies have become available.21,22
3, and the mixed-metal Ir-Re species 4. Syntheses, character-
ization, and spectroscopic, photophysical, and redox properties
of the model mononuclear compounds 2 and 6 are also reported.
We point out that, to the best of our knowledge, 4 is the first
luminescent mixed-metal Ir-Re species reported so far.
As bridging ligands for the dinuclear species, we used spacers
containing ester linkages connecting phenylbipyridine sites. Ester
linkages have not been exploited so far for the syntheses of
multimetallic luminescent species, so this study is also useful
for investigating whether these connections, potentially very
versatile from a synthetic viewpoint, guarantee enough electronic
communication between the photo- and redox-active subunits
to allow efficient photoinduced energy-transfer processes.23
We report here the syntheses and a thorough study of the
spectroscopic, redox, and photophysical properties (both in
butyronitrile rigid matrix at 77 K and in acetonitrile and
dichloromethane fluid solutions at room temperature) of a new
set of species based on Ir(III) cyclometalated building blocks,
2-5 (for the structural formulas of the new compounds, see
Chart 1. The peripheral ligands of the Ir(III) subunits are
2-phenylpyridine anions). This set includes three dinuclear
complexes, that is, the symmetric (with respect to the bridging
ligand) diiridium species 5, the asymmetric diiridium species
Experimental Section
General Procedures. The 1H NMR (300.13 MHz) spectra were
recorded at room temperature with a Bruker AC 300 spectrometer;
(19) Schmid, B.; Garces, F. O.; Watts, R. J. Inorg. Chem. 1994, 33, 9.
(20) Calogero, G.; Giuffrida, G.; Serroni, S.; Ricevuto, V.; Campagna, S.
Inorg. Chem. 1995, 34, 541.
(21) Mamo, A.; Stefio, I.; Parisi, M. F.; Credi, A.; Venturi, M.; Di Pietro,
C.; Campagna, S. Inorg. Chem. 1997, 36, 5947.
(22) Collin, J.-P.; Dixon, I. M.; Sauvage, J.-P.; Williams, J. A. G.;
Barigelletti, F.; Flamigni, L. J. Am. Chem. Soc. 1999, 121, 5009.
(23) It has been demonstrated that even electronic interactions as small as
10 cm-1 are enough to allow fast energy and electron-transfer processes
(see: Scandola, F.; Indelli, M. T.; Chiorboli, C.; Bignozzi, C. A. Top.
Curr. Chem. 1990, 158, 73.).

Luminescent Ir(III) Compounds
Inorganic Chemistry, Vol. 40, No. 6, 2001 1095
chemical shifts are referenced to internal SiMe4. FT-IR spectra were
recorded on a Perkin-Elmer 2000 spectrophotometer for KBr pellets.
Elemental analyses were performed using a Perkin-Elmer 2400 mi-
croanalyzer. Electrochemical measurements were carried out in argon-
purged acetonitrile at room temperature with a PAR 273 multipurpose
equipment interfaced to a PC. The working electrode was a glassy
carbon (8 mm2, Amel) electrode. The counter electrode was a Pt wire,
and the reference electrode was an SCE separated with a fine glass
frit. The concentrations of the complexes were about 5 × 10-4 M.
Tetrabutylammonium hexafluorophosphate was used as supporting
electrolyte, and its concentration was 0.05 M. Cyclic voltammograms
were obtained at scan rates of 20, 50, 200, and 500 mV/s. For reversible
processes, half-wave potentials (vs SCE) were calculated as the average
of the cathodic and anodic peaks. The criteria for reversibility were
the separation between cathodic and anodic peaks, the close-to-unity
ratio of the intensities of the cathodic and anodic currents (ratios
between 0.95 and 1.05 were considered acceptable), and the constancy
of the peak potential on changing scan rate. The number of exchanged
electrons was measured with differential pulse voltammetry (DPV)
experiments performed with a scan rate of 20 mV/s, a pulse height of
75 mV, and a duration of 40 ms. Details of the method used for
determining the number of the electrons, as well as the cyclic
voltammograms and DPV curves for the complexes studied, are given
in Supporting Information. For irreversible processes, the values
reported are the peaks estimated by DPV. Absorption spectra were
recorded with a Kontron Uvikon 860 spectrophotometer. Luminescence
spectra were performed with a Spex-Jobin Yvon Fluoromax-2 spec-
trofluorimeter equipped with a Hamamatsu R3896 photomultiplier and
were corrected for photomultiplier response using a program purchased
with the fluorimeter. Emission lifetimes were measured with an
Edinburgh FL-900 single-photon counting device (nitrogen discharge;
pulse width, 3 ns). Emission quantum yields were measured at room
temperature (20 °C) using the optically dilute method.24 [Ru(bpy)3]2+
(bpy ) 2,2′-bipyridine) in aerated aqueous solution was used as a
quantum yield standard, assuming a value of 0.028.25 Experimental
uncertainties were as follows: absorption maxima, (2 nm; molar
absorption coefficient, 10%; emission maxima, (4 nm; luminescence
lifetimes, 10%; luminescence quantum yields, 20%; redox potentials,
(10 mV.
Materials. Re(CO)5Br and 1,4-bis(chlorocarbonyl)benzene were
purchased from Aldrich and used as received. [Ir(ppy)2Cl]2 was prepared
as reported in the literature.26 The syntheses of the ligands 4′-(4-
hydroxyphenyl)-6′-phenyl-2,2′-bipyridine (L-OH)27 and 4′-(4-carbox-
yphenyl)-6′-phenyl-2,2′-bipyridine (L-COOH),28 as well as that of the
complex [Ir(ppy)2(L-OH)][PF6] (1),28 have been reported previously.
All other reagents and solvents (including dry solvents) were used as
received from Aldrich. However, the supporting electrolytes and the
glassware employed for the electrochemical experiments were stored
in an oven for at least 24 h before use.
Synthesis of the Ligands. 4-{[(6′-Phenyl-2,2′-bipyridine-4′-yl)-
benzoyloxy]phenyl}-6′-phenyl-2,2′-bipyridine (L-OC(O)-L). The
acid L-COOH (0.061 g, 0.173 mmol) was suspended in benzene
(3 mL) and heated to reflux with an excess of thionyl chloride (1 mL)
for 16 h. The solvent (and the excess of thionyl chloride) was removed
under reduced pressure, and the crude acid chloride L-COCl was then
used without further purification. The acid chloride and L-OH (0.055
g, 0.173 mmol) were added to degassed dichloromethane (50 mL) and
stirred under a nitrogen atmosphere to give a white suspension. A
colorless solution was formed just after the subsequent addition of
triethylamine (5-6 drops). Stirring was continued for 5 h. After this
time, the solvent was concentrated and washed with water (3 × 25
mL). The organic phase was dried over Na2SO4 and the solvent removed
in vacuo. The resulting white residue was separated from minor
impurities by column chromatography over silica with CH2Cl2/3%
MeOH as eluant (Rf ) 0.82). The product was recovered as a white
solid (0.095 g, 84%). IR (KBr, cm-1): ν(COO) 1740. 1H NMR
(CDCl3): δ (ppm) 8.73-8.61 (m, 6 H), 8.41 (d, J ) 8.5 Hz, 2 H, Hc),
8.30-8.21 (m, 4 H′), 8.03 (s, 1 H), 8.00 (s, 1 H), 7.98 (d, J ) 8.5 Hz,
2 H, Hd), 7.93 (d, J ) 8.5 Hz, 2 H, Ha), 7.91-7.85 (m, 2 H,), 7.59-
7.47 (m, 6 H), 7.44 (d, J ) 8.5 Hz, 2 H, Hb), 7.39-7.34 (m, 2 H).
Anal. Calcd for C45H30N4O2: C, 82.04; H, 4.59; N, 8.51. Found: C,
82.88; H, 4.71; N, 8.44.
Bis[4-(6′-phenyl-2,2′-bipyridine-4′-yl)phenyl]-benzene-1,4-dicar-
boxylate (L-OC(O)-C(O)O-L). A solution of L-OH (0.065 g, 0.2
mmol) in dry THF (16 mL) was treated under nitrogen with 1,4-bis-
(chlorocarbonyl)benzene (0.02 g, 0.1 mmol) and a few drops of
triethylamine. The clear, pale-yellow solution turned into a suspension
within 1-2 min. After it was stirred for 24 h, the solid was separated
by filtration and the filtrate rotary-evaporated to dryness. The solid
residue was redissolved in chloroform (30 mL), and the solution was
consecutively washed with water (2 × 25 mL), 0.1 M HCl (2 × 25
mL), and water (2 × 25 mL). The organic phase was dried over
Na2SO4 and the solvent removed in vacuo. The crude product was
purified through chromatography over silica (CH2Cl2/10% MeOH). The
pure product was obtained as a white solid (0.067 g, 86%). IR (KBr,
1
cm-1): ν(COO) 1738. H NMR (CDCl3): δ (ppm) 8.73 (br d, 1 H,
H6), 8.71 (d, J ) 8.8 Hz, 1 H, H3), 8.67 (br s, 1 H, H3′), 8.40 (s, 2 H,
Hc), 8.23 (d, J ) 6.9 Hz, 2 H, H2′′,6′′), 8.00 (s, 1 H, H5′), 7.93 (d, J )
8.8 Hz, 2 H, Ha), 7.89 (td, J ) 7.9, 7.5, 2.0 Hz, 1 H, H4), 7.58-7.47
(m, 3 H, H3′′,4′′,5′′), 7.43 (d, J ) 8.8 Hz, 2 H, Hb), 7.34 (m, 1 H, H5).
Anal. Calcd for C52H34N4O4: C, 80.19; H, 4.40; N, 7.19. Found: C,
79.16; H, 4.36; N, 6.82.
Synthesis of the Complexes. [Ir(ppy)2(L-OC(O)-L)][PF6] (2).
The acid chloride L-COCl was freshly prepared as reported above
for the synthesis of L-OC(O)-L. Solid [(ppy)2Ir(L-OH)][PF6] (1)
(0.069 g, 0.071 mmol) was added to a suspension of L-COCl (0.026
g, 0.071 mmol) in degassed dichloromethane (50 mL) under nitrogen.
The resulting orange suspension turned clear upon addition of a few
drops of triethylamine. The solution was stirred at room temperature
for 48 h, then filtered. Evaporation of the solvent under reduced pressure
gave a crude orange product (0.085 g), which was purified by
chromatography over neutral alumina with a mixture of methanol (3%)
in dichloromethane as eluant (Rf ) 0.89). Yield, 0.078 g (84%). IR
(KBr, cm-1): ν(COO) 1737, ν(PF) 845. 1H NMR (CD3CN): δ (ppm)
8.84 (d, J ) 1.8 Hz, 1 H, H3′), 8.77-8.69 (m, 3 H), 8.74 (d, J ) 1.5
Hz, 1 H, H3′ of free L moiety), 8. 39-8.32 (m, 4 H), 8.25 (d, J ) 1.5
Hz, 1 H, H5′ of free L moiety), 8.19-8.09 (m, 6 H), 8.01-7.80 (m, 7
H), 7.79 (d, J ) 1.8 Hz, 1 H, H5′), 7.66 (d, J ) 5.1 Hz, 1 H), 7.62-
7.53 (m, 5 H), 7.49-7.44(m, 2 H), 7. 31 (d, J ) 7.8 Hz, 1 H), 7.18-
7.09 (m, 2 H), 6.98-6.92 (m, 2 H), 6.83 (td, J ) 7.4, 7.4, 1.4 Hz, 1
H), 6.77 (t, J ) 7.7, 2 H), 6.65 (vbr s, 2 H), 6.58 (t, J ) 7.7 Hz, 1 H),
6.38 (td, J ) 7.4, 7.4, 1.4 Hz, 1 H), 5.97 (d, J ) 7.3 Hz, 1 H), 5.59 (d,
J ) 7.4 Hz, 1 H). Anal. Calcd for C67H46 F6IrN6O2: C, 61.69; H, 3.55;
N, 6.44. Found: C, 62.12; H, 3.67; N, 6.32.
[(ppy)2Ir(µ-L-OC(O)-L)Ir(ppy)2][PF6]2 (3). [Ir(ppy)2Cl]2 (0.011
g, 0.0103 mmol) and complex 2 (0.027 g, 0.0206 mmol) were
suspended in a CH2Cl2/MeOH (4 mL/6 mL) mixture. Reflux of the
stirred reaction mixture for 3 h afforded an orange solution. After it
was cooled to room temperature, an excess of NH4PF6 dissolved in
methanol (1 mL) was added and stirring continued for 30 min. The
solution was then filtered and concentrated in vacuo until a dark-orange
solid formed. The solid was separated by filtration, washed with diethyl
ether, and vacuum-dried. The analytically pure product was obtained
upon chromatography over neutral alumina with acetonitrile as eluant
(Rf ) 0.95). Yield, 0.030 g (74%). IR (KBr, cm-1): ν(COO)1736, ν-
(PF) 842. 1H NMR (CD3CN): δ (ppm) 8.87 (d, J ) 2.4 Hz, 1 H), 8.83
(br s, 1 H), 8.75 (m, 2 H), 8.41 (d, J ) 8.4 Hz, 2 H), 8.27-8.10 (m,
6 H), 7.99-7.87 (m, 10 H), 7.85 (d, J ) 2.4 Hz, 1 H), 7.80 (s, 1 H),
7.69 (t J ) 6.1 Hz, 2 H), 7.62 (d, J ) 7.3 Hz, 3 H), 7.55 (d J ) 8.5
Hz, 3 H), 7.47 (m, 2 H), 7.31 (d, J ) 7.3 Hz, 2 H), 7.17-7.08 (m, 4
H), 6.96 (m, 4 H), 6.86-6.76 (m, 6 H), 6.66 (vbr s, 4 H), 6.58 (t, J )
7.3 Hz, 2 H), 6.38 (t, J ) 7.3 Hz, 2 H), 5.97 (d, J ) 8.5 Hz, 2 H), 5.59
(d, J ) 7.3 Hz, 2 H). Anal. Calcd for C89H58F12Ir2N8O2P2: C, 54.94;
H, 3.00; N, 5.76. Found: C, 55.11; H, 3.10; N, 5.84.
(24) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991.
(25) Nakamaru, K. Bull. Chem. Soc. Jpn. 1982, 55, 2697.
(26) Sprouse, S.; King, K. A.; Spellane, P. J.; Watts, R. J. J. Am. Chem.
Soc. 1984, 106, 6647.
(27) Neve, F.; Ghedini, M.; Francescangeli, O.; Campagna, S. Liq. Cryst.
1998, 24, 673.
(28) Neve, F.; Crispini, A.; Campagna, S.; Serroni, S. Inorg. Chem. 1999,
38, 2250.

1096 Inorganic Chemistry, Vol. 40, No. 6, 2001
Neve et al.
Scheme 1. Synthesis of the Ligandsa
[(ppy)2Ir(µ-L-OC(O)-L)Re(CO)3Br][PF6] (4). A solid mixture
of 2 (0.048 g, 0.037 mmol) and Re(CO)5Br (0.015 g, 0.037 mmol)
was suspended in dry toluene (20 mL) and heated to reflux under argon
for 4 h. The solvent was decanted, and the red sticky solid was
redissolved in dichloromethane to afford a red-orange solution. Rotary
evaporation of the solvent afforded the crude product. The analytically
pure product was obtained as a red-orange solid (0.052 g, 85%) upon
chromatography on neutral alumina with a gradient of 3% methanol in
dichloromethane as eluant (Rf ) 0.7). IR (KBr, cm-1): ν(CO) 2020,
1
1919, 1885; ν(COO) 1735; ν(PF) 842. H NMR (CD3CN): δ (ppm)
9.13 (m, 1 H), 8.84 (d, J ) 2.0 Hz, 1 H), 8.77 (br s, 1 H), 8.73 (m, 3
H), 8.40 (d, J ) 8.8 Hz, 2 H), 8.27 (br t, 1 H), 8.21-8.10 (m, 6 H),
8.05 (d, J ) 2.0 Hz, 1 H), 7.99-7.83 (m, 6 H), 7.81 (s, 1 H), 7.70-
7.60 (m, 7 H), 7.56 (d, J ) 8.8 Hz, 2 H), 7.46 (m, 2 H), 7.31 (d, J )
7.8 Hz, 1 H), 7.17-7.08 (m, 4 H), 6.96 (m, 4 H), 6.83 (t, J ) 7.4 Hz,
2 H), 6.77 (t, J ) 7.8 Hz, 4 H), 6.66 (vbr s, 4 H), 6.58 (t, J ) 7.8 Hz,
2 H), 6.38 (t, J ) 7.0 Hz, 2 H), 5.97 (d, J ) 7.0 Hz, 2 H), 5.59 (d, J
) 7.0 Hz, 2 H). Anal. Calcd for C70H46BrF6IrN6O5PRe: C, 50.82; H,
2.80; N, 5.08. Found: C, 50.17; H, 2.82; N, 5.39.
[(ppy)2Ir(µ-L-OC(O)-C(O)O-L)Ir(ppy)2][PF6]2 (5). Method a.
1,4-Bis(chlorocarbonyl) benzene (0.078 g, 0.038 mmol) was added
under nitrogen to a stirred suspension of 1 (0.074 g, 0.076 mmol) in
degassed dichloromethane (50 mL). Further addition of a few drops of
triethylamine led to the formation of a clear-orange solution. Although
a monitoring of the reaction through TLC (SiO2, CH2Cl2/5% MeOH)
indicated the disappearance of the chloride within 1 h, stirring was
continued at room temperature for 24 h. The solvent was concentrated
to a small volume, affording an orange precipitate, which was separated
by filtration, washed with methanol and diethyl ether, and vacuum-
1
dried (0.077 g, 96%). IR (KBr, cm-1): ν(COO) 1736, ν(PF) 842. H
NMR (CD3CN): δ (ppm) 8.84 (s, 1 H, H3′), 8.76 (d, J ) 8.2 Hz, 1 H,
H3), 8.41 (s, 2 H, Hc), 8.17 (br t, 1 H, H4), 8.13 (d, J ) 8.7 Hz, 2 H
Ha), 7.99-7.84 (m,), 7.81 (s, 1 H, H5′), 7.68 (d, J ) 5.5 Hz, 1 H), 7.62
(d, J ) 7.6 Hz, 1 H), 7.56 (d, J ) 8.7 Hz, 2 H, Hb), 7.47 (m, 1 H),
7.17-7.08 (m, 2 H), 6.95 (m, 2 H), 6.83 (vt, J ) 7.3 Hz, 1 H), 6.77
(vt, J ) 7.3, 2 H), 6.66 (vbr s, 2 H), 6.58 (vt, J ) 7.3 Hz, 1 H), 6.38
(vt, J ) 7.3 Hz, 1 H), 5.97 (d, J ) 7.3 Hz, 1 H), 5.59 (d, J ) 7.3 Hz,
1 H). Anal. Calcd for C96H66F12Ir2N8O4P2: C, 55.70; H, 3.21; N, 5.41.
Found: C, 55.39; H, 3.13; N, 4.75.
Method b. A stirred suspension of [Ir(ppy)2Cl]2 (0.036 g, 0.033
mmol) and L-OC(O)-C(O)O-L (0.025 g, 0.033 mmol) in a CH2-
Cl2/MeOH (10 mL/12 mL) mixture was heated under reflux conditions
for 4 h. After the resulting orange solution was cooled to room
temperature, a methanolic solution (1 mL) of NH4PF6 (10 equiv based
on the ligand) was added to it and stirring continued for 30 min. Partial
evaporation of the solvent under reduced pressure afforded the product
as an orange solid, which was filtered off, washed with methanol and
diethyl ether, and dried in vacuo (0.057 g, 82%). Upon analysis, the
product turned out to be identical to that obtained by method a.
[ReBr(CO)3(L-COOH)] (6). Re(CO)5Br (0.069 g, 0.170 mmol)
and L-COOH (0.060 g, 0.170 mmol) were stirred for 4 h under argon
in refluxing dry toluene (15 mL). The resulting bright-yellow solid was
collected and washed with diethyl ether and hexane before drying in
vacuo (0.114 g, 95%). IR (KBr, cm-1): ν(CO) 2023, 1921, 1902; ν-
(COOH) 1720. 1H NMR (CD3CN-CDCl3): δ (ppm) 9.11 (br d, 1 H,
H6), 8.71 (br s, 1 H, H3′), 8.67 (d, J ) 8.5 Hz, 1 H, H3), 8.24 (br t, 1
H, H4), 8.20 (d, J ) 8.5 Hz, 2 H, Ha), 8.07 (d J ) 8.5 Hz, Hb), 7.97 (d,
J ) 2.4 Hz, 1 H, H5′), 7.71-7.58 (m, 6 H, H2′′,6′′ + H3′′,4′′,5′′ + H5).
Anal. Calcd for C26H16BrIrN2O5Re: C, 44.45; H, 2.30; N, 3.99.
Found: C, 44.78; H, 2.41; N, 4.04.
a Reagents: (i) 1,4-(COCl)2C6H4 (0.5 equiv), NEt3; (ii) SOCl2; (iii)
NEt3.
To prepare Ir-based bimetallic complexes, we followed two
different strategies that both rely on a common building block.
From our recent studies,28 it is known that the complex [(ppy)2Ir-
(L-OH)][PF6] (1, Chart 1) (ppy ) 2-phenylpyridine anion;
L-OH ) 4′-(4-hydroxyphenyl)-6′-phenyl-2,2′-bipyridine) has
two main characteristics: (i) it is a good luminophore and (ii)
it bears a reactive OH group at the periphery of a coordinated
ligand. By means of 1, construction of the bimetallic species
can go either through a single-step reaction to generate sym-
metric species or through a multistep process that leads to
asymmetric species. In the latter case, the key role would be
played by the complex ligand 2 (Chart 1), a potential candidate
for luminescent dyads32 or triads.33
Syntheses of the Ligands. The syntheses of the asymmetric
L-OC(O)-L and symmetric L-OC(O)-C(O)O-L multi-
nucleating ligands based on the 6′-phenyl-2,2′-bipyridine func-
tionality are reported in Scheme 1. In both cases the synthesis
is rather straightforward, implying a condensation under mild
conditions of an acid chloride with a phenolic functional group
in the presence of triethylamine as a base. The ligands were
obtained in very good yield after purification by column
chromatography and characterized by FT-IR and 1H NMR
spectroscopies.
Results
Assembling photoactive polymetallic species can be a lengthy
process, especially when heteronuclear or asymmetric assemblies
must be generated.1-6 Even in the simplest case, i.e., for
bimetallic species, most synthetic methods are based on a
stepwise approach that ensures control of the target species.29-31
(30) Whittle, B.; Everest, N. S.; Howard, C.; Ward, M. D. Inorg. Chem.
1995, 34, 2025.
(31) Paw, W.; Eisenberg, R. Inorg. Chem. 1997, 36, 2287.
(32) Aspley, C. J.; Lindsay Smith, J. R.; Perutz, R. N. Chem. Commun.
1999, 2269.
(29) Constable, E. C.; Cargill Thompson, A. M. W.; Herveson, P.; Macko,
(33) Barigelletti, F.; Flamigni, L.; Calogero, G.; Hammarstro¨m, L.; Sauvage,
J.-P.; Collin, J.-P. Chem. Commun. 1998, 2333.
L.; Zehnder, M. Chem. Eur. J. 1995, 1, 360.

Luminescent Ir(III) Compounds
Inorganic Chemistry, Vol. 40, No. 6, 2001 1097
Scheme 2
Besides, 6 was also instrumental in the assignment of the
stereochemistry of the Re site in complex 4. The IR spectra of
both 4 and 6 show three intense absorption bands in the spectral
region characteristic of terminal carbonyl ligands (2030-1880
cm-1), a feature that is in agreement with a facial arrangement
of the carbonyl groups of the Re(CO)3 fragment.40 This finding
rules out the possibility for a cyclometalating, tridentate bonding
mode of the L moiety, since involvement of a metalated phenyl
group would likely lead to the formation of a Re(I) octahedral
complex with meridional configuration. The absence of a Re
cyclometalated L moiety in 6 (and by extension in 4) was also
proved by the 1H NMR characterization of the complex, which
revealed the presence of an uncoordinated phenyl group. Finally,
the coordination mode of the ditopic L-OC(O)-L ligand in
2-4 is always N,N to both Ir and Re. For the former, the metal
environment is barely that of complex 1, which has been
previously defined.28
1
The H NMR spectra of 2-4 were recorded in acetonitrile-
d3, affording aromatic regions very rich in signals. Although
no effort has been made to fully assign the spectra, two pairs
of resonances can be easily followed for an indication of the
second metal binding. These refer to the H3′ and H5′ protons of
the central pyridine ring of both termini of the L-OC(O)-L
ligand. When uncoordinated (as in 2), the signals of the free L
terminus are observed at δ 8.74 and 8.25, respectively (see
Experimental Section). Metal coordination triggers downfield
and upfield shifts to δ 8.80 and 7.82 for 3 and δ 8.77 and 8.05
for 4. On the other hand, the remaining pair of signals, which
we assign to the H3′ and H5′ protons of the pyridine always
bound to iridium, are displayed around δ 8.80 and 7.80 for all
three complexes, revealing a common chemical environment.
Complex 1 was also effective in the formation of symmetric
bimetallic species. The room-temperature coupling of 1 with
0.5 equiv of terephthaloyl chloride followed by exchange with
NH4PF6 (Scheme 2, route d) led to the dicationic species 5, in
a way that involved the formation of both a new bridging ligand
and the bimetallic assembly. Interestingly, complex 5 is the only
member of the series of bimetallic species reported here that
can be synthesized by an alternative, more conventional method.
Thus, the direct reaction of the ligand L-OC(O)-C(O)O-L
with the dimer [Ir(ppy)2Cl]2 under reflux conditions (followed
by exchange with excess NH4PF6 at room temperature) also
led to the formation of 5 in good yield. Attempts to obtain Ir-
based trimetallic species through the coupling of 1 with the
electrophile 1,3,5-tris(chlorocarbonyl)benzene41 were unsuc-
cessful. The reaction gave rise to a complex mixture of products
that could not be separated and identified.
Redox Behavior and Spectroscopic and Photophysical
Properties. The absorption spectra of all the novel compounds
exhibit moderately intense absorption in the visible region (ꢀ
in the range 103-104 M-1 cm-1) and a strong absorption in the
UV region (λmax at about 270 nm, ꢀ in the range 104-105 M-1
cm-1). They also show a broad tail that extends toward the red
region. All the compounds exhibit one irreversible oxidation
process at about +1.20 V vs SCE, and 4 also shows an addi-
tional irreversible oxidation process at +1.34 V. The processes
concerning 2, 4, and 6 are monoelectronic, whereas the oxidation
processes of the diiridium species 3 and 5 are bielectronic in
nature. The complexes also show reversible reduction processes
in the potential window investigated (<1.80 V).
Although the free ligands were synthesized and fully char-
acterized for the first time in the present study, they were not
used for the preparation of the metal complexes reported here
(except in one case). Nevertheless, their characterization was
useful to allow the subsequent characterization of the metal
species.
Syntheses of the Metal Complexes. As mentioned above,
our synthetic strategy for the preparation of multimetallic species
took advantage of the available complex [Ir(ppy)2(L-OH)][PF6]
(1) (Chart 1), which possesses a reactive functional group at
the periphery of the chelating L-OH ligand. Condensation of
the phenolic function of 1 with the chlorocarbonyl group of
the L-COCl derivative (ad hoc prepared from L-COOH)
exclusively led to the formation of [Ir(ppy)2(L-OC(O)-L)]-
[PF6] (2) (Scheme 2, route a), thus affording a mononuclear
species that contains a free bipyridyl binding site belonging to
the ligand L-OC(O)-L formed in situ.
Complex 2, which is air-stable and fairly soluble in common
organic solvents, is therefore able to behave as a complex ligand,
and virtually any type of bimetallic system can be assembled
provided the second metal center has neither severe steric
constraints nor high inertness. Although in principle complex
2 may behave as a mono-, bi-,28,34 or tridentate complex
ligand,27,35,36 we chose to pursue its more predictable reactivity
as donor of a bipyridyl fragment. The reaction of 2 with 0.5
equiv of [Ir(ppy)2Cl]2 followed by metathesis with NH4PF6
(Scheme 2, route b) readily afforded the homometallic species
[(ppy)2Ir(µ-L-OC(O)-L)Ir(ppy)2][PF6]2 (3). On the other hand,
the heterometallic species [(ppy)2Ir(µ-L-OC(O)-L)Re(CO)3Br]-
[PF6] (4) could be obtained by reaction of 1 and Re(CO)5Br
following well-known procedures37-39 (Scheme 2, route c).
Prepared in a similar manner, the complex [ReBr(CO)3
(L-COOH)] (6) was intended to serve as a model in the
interpretation of the photophysical and redox behavior of 4.
(34) Neve, F.; Crispini, A. Eur. J. Inorg. Chem. 2000, 1039.
(35) Neve, F.; Ghedini, M.; Crispini, A. Chem. Commun. 1996, 2463.
(36) Neve, F.; Crispini, A.; Campagna, S. Inorg. Chem. 1997, 36, 6150.
(37) Abel, E. W.; Wilkinson, G. J. Chem. Soc. 1959, 1501.
(38) Wrighton, M. S.; Morse, D. L. J. Am. Chem. Soc. 1974, 96, 998.
(39) Juris, A.; Campagna, S.; Bidd, I.; Lehn, J.-M.; Ziessel, R. Inorg. Chem.
1988, 27, 4007.
(40) Stor, G. J.; Morrison, S. L.; Stufkens, D. J.; Oskam, A. Organometallics
1994, 13, 2641.
(41) Davies, P. J.; Grove, D. M.; van Koten, G. Organometallics 1997,
16, 800.

1098 Inorganic Chemistry, Vol. 40, No. 6, 2001
Neve et al.
Table 1. Half-Wave Potentials in Argon-Purged Acetonitrile Solution, 298 Ka
compound
E1/2(ox),b V vs SCE
E1/2(red), V vs SCE
2 [Ir(ppy)2(L-OC(O)-L)]+
+1.19 [1]
-1.35 [1]; -1.68 [1]
-1.24 [1]
-1.20 [1]; -1.35 [1]
-1.35 [1]; -1.53 [1]
-1.25 [1]
3 [(ppy)2Ir(µ-L-OC(O)-L)Ir(ppy)2]2+
4 [(ppy)2Ir(µ-L-OC(O)-L)Re(CO)3Br]+
5 [(ppy)2Ir(µ-L-OC(O)-(O)CO-L)Ir(ppy)2]2+
6 [Re(CO)3(L-COOH)Br]
+1.18 [2]
+1.19 [1]; +1.34 [1]
+1.19 [2]
+1.34 [1]
a The number of exchanged electrons is reported in brackets. b Irreversible process. In this case, the value reported is the dpv peak.
Table 2. Spectroscopic and Photophysical Dataa
absorptionb
luminescence, 298 K
lumin,d 77 K
λmax, nm τ, µs
λmax, nm
compound
(ꢀ, M-1 cm-1
)
λ
max, nmb τ,b ns
Φb
λ
max,c nm τ,c ns
Φc
2 [Ir(ppy)2(L-OC(O)-L)]+
270 (92500)
377 (7300)
269 (117700)
380 (18500)
271 (83400)
378 (12800)
633
646
636
631
643
75
65
62
69
35
0.016
619
638
623
620
655
192
167
169
188
46
0.059
545
552
550
548
562
3.6
3.5
3.3
3.6
4.8
3 [(ppy)2Ir(µ-L-OC(O)-L)Ir(ppy)2]2+
4 [(ppy)2Ir(µ-L-OC(O)-L)Re(CO)3Br]+
0.006
0.005
0.007
0.002
0.040
0.025e
0.045
0.004
5 [(ppy)2Ir(µ-L-OC(O)-(O)CO-L)Ir(ppy)2]2+ 269 (120500)
380 (15900)
270 (34200)
382 (4400)
6 [Re(CO)3(L-COOH)Br]
a For the absorption, the maxima (or shoulders) of the spin-allowed LC and MLCT bands are given. b In air-equilibrated CH3CN. c In air-
equilibrated CH2Cl2. d In butyronitrile. e This quantum yield is excitation wavelength dependent (see text). The one reported is the average value.
All the compounds are luminescent both at room temperature
in acetonitrile and in dichloromethane fluid solution and at 77
K in butyronitrile rigid matrix. The luminescence maxima are
solvent-dependent; they are shifted to the blue on passing from
acetonitrile to dichloromethane solution, except for 6 which
exhibits the reverse effect. The luminescence spectra always
show relatively broad bands. A vibrational structure is visible
as a shoulder in the red tail of the main band, especially at 77
K. Luminescence decays are monoexponential and are in the
microsecond time scale at 77 K and 1 or 2 orders of magnitude
shorter at 298 K. For all the complexes, luminescence lifetimes
and quantum yields increase on passing from acetonitrile to
dichloromethane. The luminescence properties are in all cases
independent of excitation wavelength (between 320 and 420
nm) when experimental uncertainty is taken into account, with
the exception of the luminescence quantum yield of 4 in
dichloromethane (see Discussion). Table 1 is a listing of the
Figure 1. Absorption spectra of complexes 2 (dashed line), 4 (solid
line), and 6 (dashed-dotted line) in acetonitrile at room temperature.
redox data, while the spectroscopic and photophysical data are
gathered in Table 2. Figure 1 shows the absorption spectra of
2, 4, and 6 in acetonitrile, and Figure 2 displays the emission
spectra of 4 in acetonitrile and dichloromethane fluid solutions
at room temperature and in butyronitrile at 77 K.
Recently, other particular types of orbitals (and states) have also
been proposed to rationalize the spectroscopic and electrochemi-
cal properties of metal compounds containing strong electron-
donor ligands. For example, a covalent metal-C- σ-bonding
(or metal-Si- σ-bonding) orbital has been identified as the orbital
involved in the oxidation process and the lowest-lying charge-
transfer excited state in Ir(III) cyclometalated compounds.45-47
The excited states derived from these σ-bonding orbitals are
usually termed σ-bond to ligand charge-transfer (SBLCT)46,47
or more generally ligand-to-ligand charge transfer (LLCT).48
The species investigated here can be viewed as multicom-
ponent compounds. Even the bridging ligands may be viewed
as multicomponent, segmented species in that their chelating
Discussion
The spectroscopic and photophysical properties and the redox
behavior of transition metal complexes and organometallic
compounds are usually discussed with the assumption that the
ground state, as well as the excited and redox states, can be
described by a localized molecular orbital configuration.42,43
Within this framework, the various spectroscopic transitions and
excited states are classified as metal-centered (MC), ligand-
centered (LC), or charge-transfer (either metal-to-ligand, MLCT,
or ligand-to-metal, LMCT) and the oxidation and reduction
processes are classified as metal- or ligand-centered.42-44
(45) Djurovich, P. I.; Watts, R. J. Inorg. Chem. 1993, 32, 4681.
(46) Didier, P.; Ortmans, L.; Kirsch-DeMesmaeker, A.; Watts, R. J. Inorg.
Chem. 1993, 32, 5239.
(47) Serroni, S.; Juris, A.; Campagna, S.; Venturi, M.; Denti, G.; Balzani,
V. J. Am. Chem. Soc. 1994, 116, 9086.
(48) Vogler, A. In Photoinduced Electron Transfer; Fox, M. A., Chanon,
M., Eds.; Elsevier: Amsterdam, 1988; p 179.
(42) De Armond, M. K.; Carlin, C. M. Coord. Chem. ReV. 1981, 36, 325.
(43) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; von
Zelewsky, A. Coord. Chem. ReV. 1988, 84, 85.
(44) Maestri, M.; Balzani, V.; Deuschel-Cornioley, C.; von Zelewsky, A.
AdV. Photochem. 1992, 17, 1.

Luminescent Ir(III) Compounds
Inorganic Chemistry, Vol. 40, No. 6, 2001 1099
Figure 3. Two chelating subunits of the L-OC(O)-L bridging ligand
(see text).
For the sake of simplicity, we will start the discussion on
the reduction processes from complex 2 (Table 1). The first
reduction process of this compound may be assigned to the
reduction of the coordinated L-OC(O) moiety of the L-
OC(O)-L ligand, and the subsequent reduction may be assigned
to the uncoordinated polypyridine moiety of the same ligand.
This is based on the reported potential values for the reduction
processes of the coordinated ppy ligands, which are known
to be more negative than -1.80 V, 44 and on the fact that
coordinated polypyridines are reduced at less negative potentials
than uncoordinated ones. Compound 5 contains the same type
of coordinated site that is present in 2. Therefore, the first
reduction of 5, occurring at the same potential as the first
reduction of 2, is also assigned to either one of the L-OC(O)
moieties of the L-OC(O)-C(O)O-L bridge. The second
reduction is attributed to the remaining L-OC(O) site of the
same ligand. Because reduction is essentially bpy-centered, the
monoreduced species can be regarded as a mixed-valence, (bis-
phenyl-bpy -)-S-(bis-phenyl-bpy) species, where S repre-
sents the dicarboxylate phenyl spacer (see Chart 1). To evaluate
the stability of the mixed-valence species and as a consequence
the electronic interaction between the sites, the comproportional
constant Kc is useful.49 This can be obtained by the equation
(at 298 K) Kc ) exp(∆E/25.69), where ∆E corresponds to the
difference between the first and second redox potentials in
millivolts. From the reported values, a Kc value of 1100 is
obtained, which indicates a significant interaction between the
two moieties of the bridging ligand across the dicarboxylate
phenyl spacer. This looks somewhat surprising, given the
oxidation properties of 5 (see above), which suggest a negligible
interaction between the metal ions across the bridging ligand.
However, the redox-active sites involved in the reduction process
are closer to one another than those involved in the oxidation
process. Actually, the orbitals involved in the reduction
processes, mainly centered on the bipyridine moieties, are also
most likely partially extended to the phenyl rings directly
connected to the carboxylate spacer, whereas the oxidation
processes involve Ir-C- σ-bonding orbitals where peripheral
ligands have a role. Such differences justify the observed
behavior as far as the interaction between redox sites is
concerned, when Coulombic effects and the mechanism for
superexchange-mediated coupling between the redox-active
sites49 are taken into account.
Figure 2. Corrected luminescence spectra of 4 in acetonitrile (a) and
dichloromethane (b, dashed line) fluid solutions at room temperature
and in butyronitrile (c) at 77 K.
sites are connected to each other by ester linkages. To simplify
the discussion, we will make extensive use of this picture.
Redox Behavior. As mentioned above, oxidations of d6 metal
complexes containing polypyridine and cyclometalating ligands
are usually metal-centered or may involve orbitals that receive
partial contributions from σ bonds. Generally, pure metal-
centered oxidations are reversible, and the contribution of ligand-
centered orbitals via the σ bond leads to irreversible processes.
In light of the irreversibility of the oxidation processes exhibited
by the complexes studied here (Table 1), we assign such
processes to orbitals that receive contributions from M-C-
bonds.
σ
It is interesting to compare the oxidation behavior of 2 and
5, which contain identical redox centers, leaving aside the second
coordination site of the bridging ligand. Because of its bielec-
tronic nature (Table 1), the oxidation of 5 is assigned to two
simultaneous, one-electron processes involving the two metal-
based subunits. This indicates that the electronic interaction
between the two identical redox sites of 5 across the bridge,
usually inferred by the splitting of the redox processes in
symmetric dinuclear complexes, is negligible from an electro-
chemical viewpoint. The identical potentials of the oxidation
processes of 2 and 5 also suggest that the coordination through
the second chelating site of the bridging ligand does not
significantly affect the electronic properties of the other redox
site. Even the oxidation of 3 can be assigned to simultaneous
one-electron processes involving the two metal-based subunits,
although in principle the two redox centers are different, as a
consequence of the bridge asymmetry. Actually, we may
consider the L-OC(O)-L bridging ligand as being formed by
two subunits, one in which the phenylbipyridine chelating site
bears an oxycarbonyl group as substituent (hereafter L-OC-
(O)) and the other in which the substituent is a carboxyl group
(hereafter L-C(O)O) (see Figure 3). Since L-C(O)O is a
slightly better electron-acceptor ligand than L-OC(O), one
might expect that the redox center {(ppy)2Ir(L-C(O)O)} is
oxidized at more positive potentials than the {(ppy)2Ir(L-
OC(O))} fragment. In fact, the simultaneous oxidation of the
two redox centers of 3 indicates that this difference does not
have a sizable effect. As far as 4 is concerned, its first oxidation
process is safely assigned to the iridium center and the second
one to the rhenium center, on the basis of the potential values
of the oxidations reported in Table 1.
The reduction process of 3 is assigned to the L-C(O)O
coordinated moiety of the bridge. This moiety is a better electron
acceptor than L-OC(O), and this agrees with the less negative
potential of the process with respect to the first reduction
processes of 2 and 5. Because of adsorption on the electrode,
further investigation of the reduction behavior of 3 was
prevented. Reduction of 6 is assigned straightforwardly to the
polypyridine ligand (the only reduction site of this complex),
(49) Giuffrida, G.; Campagna, S. Coord. Chem. ReV. 1994, 135-136, 517
and references therein.

1100 Inorganic Chemistry, Vol. 40, No. 6, 2001
Neve et al.
and this assignment allows the attribution of the first reduction
process of 4 to the L-C(O)O moiety coordinated to Re(I) and
of the second reduction to the Ir-coordinated moiety of the
bridge (Table 1).
Absorption Spectra. On the basis of literature data,26,28,44-47
the moderate absorption bands exhibiting a maximum (or most
often a shoulder) at about 380 nm, which is present in all the
new complexes, can be attributed to spin-allowed MLCT
transitions involving the polypyridine ligands, whereas the strong
absorption peaks at about 270 nm can be attributed to spin-
allowed LC transitions (Table 2, Figure 1). This attribution also
indicates that LC transitions involving ppy and the bridging
ligands occur in the same energy region. The tail toward the
red region (Figure 1), which is present in all the complexes
(maxima or shoulders at about 470 nm), is attributed to spin-
forbidden MLCT transitions, which steal intensity from the spin-
allowed bands because of the strong spin-orbit coupling
induced by the heavy metals.
Figure 4. Schematic energy level diagram of 4. Solid and dashed lines
represent radiative and radiationless transitions, respectively. The
peripheral ligands and the effective charges of the complexes are not
reported for convenience.
From the extinction coefficients reported in Table 2, it is clear
that the absorption spectra are essentially additive. For example,
the molar absorption coefficient of the MLCT absorption of 5
is almost twice that of the corresponding MLCT band in 2, in
which a metal ion is missing, whereas the LC band is only
slightly enhanced because of the presence of two additional ppy
ligands in 5. While the LC extinction coefficients of 5 and 3
are roughly the same (only a slight difference in the bridge is
present), the MLCT band is significantly more intense in 3. This
suggests that the Ir f L-C(O)O charge-transfer (CT) transition
has a stronger oscillator strength than the Ir f L-OC(O) CT
transition. Finally, both the LC and MLCT bands of 6 are less
intense than those of the other species, including 2, suggesting
that Re f L-OC(O) CT transition has a lower oscillator
strength than the Ir-based CT ones. This also explains the
extinction coefficient values of the MLCT and LC bands of 4,
which are lower than the corresponding bands of the other
dinuclear species. As we will see later, the differences in molar
absorptivities of the different subunits will be very important
in determining the luminescence properties of the various
compounds.
One of our aims in designing and studying these species was
to clarify whether photoinduced energy transfer across the ester-
linked multichelating bridging ligands could occur in compounds
containing different chromophores. Compounds 3 and 4 are
indeed suitable in this regard in that each of them contains two
intrinsically emitting chromophores with MLCT excited states
that are different in energy; 3 contains an Ir f L-OC(O) CT
excited state (hereafter MLCT1) together with an Ir f L-C(O)O
CT level (hereafter MLCT2). The MLCT1 level is also present
in 4 in which a Re f L-C(O)O CT level (hereafter MLCT3)
is also present. On the basis of the redox properties of the various
subunits and on the photophysical properties of 2, 5, and 6, the
excited-state energy order is MLCT3 < MLCT2 < MLCT1, and
as a consequence, exoergonic photoinduced energy transfer from
MLCT1 to MLCT2 in 3 and from MLCT1 to MLCT3 in 4 could
occur, although in both cases the driving force is small.51 The
situation for complex 4 is schematized in Figure 4.
The discussion on the energy-transfer processes will proceed
starting from compound 4. A cursory look at Table 2 indicates
that the luminescence properties of 4 are dominated by the
MLCT1 level of the upper-lying luminophore both in rigid
matrix and in fluid solution, so energy transfer from MLCT1 to
the lower-lying MLCT3 level seems to occur with low ef-
ficiency. By comparison of the absorption and luminescence
properties of 2 and 6, it appears that the Ir-based chromophore
is both a better absorber and luminophore than the Re-based
one, and this also explains why the luminescence properties of
4 look Ir-based. Following this consideration, it could be
proposed that the two chromophores of 4 are essentially
noninteracting. A useful hint in this regard is the dependence
of the luminescence quantum yield (Φ) of 4 in dichloromethane
on the excitation wavelength. In fact, the Φ value for 4 in
dichloromethane is 0.030 for λexc ) 360 nm (where the
Luminescence Properties. The luminescence of all the
complexes studied here (Table 2, Figure 2) can be safely
3
attributed to MLCT excited states involving the polypyridine
ligands, on the basis of spectral shape, emission energies,
lifetimes, and quantum yields.16-21,26,28,44-47 The red shift of
the emission spectra of the Ir-based complexes on passing from
dichloromethane to acetonitrile also agrees with this assignment
in that the more polar acetonitrile stabilizes the MLCT levels
with respect to the ground state because of the more polar
character of the former levels compared to the latter. The reverse
behavior of the rhenium species (Table 2) is also in line with
expectation in that it is well-known that for Re(I) tricarbonyl
luminophores the excited state is less polar than the ground state,
so these species show a different solvatochromic effect com-
pared with Ir(III) (and also Ru(II) and Os(II)) polypyridine
compounds.37-39 In all cases, luminescence lifetimes and
quantum yields are lower in acetonitrile because this solvent
accelerates radiationless transitions with respect to dichlo-
romethane. The blue shift of the emission spectra on passing
from fluid solution at room temperature to rigid matrix at 77 K
is also typical of charge-transfer emitters, as a consequence of
the absence of the excited-state stabilization due to solvent
reorganization processes, which are hampered in a rigid matrix.50
(51) It is noted that there is an apparent lack of correlation between
spectroscopic and redox properties in 4, for which the HOMO involves
(from redox data, Table 1) an Ir-based orbital while the lowest-lying
MLCT excited state involves (on the basis of the photophysical
properties of the parent mononuclear species 2 and 6; Table 2) a Re-
based orbital. It should be recalled that the spectroscopic/electrochemi-
cal relationship is usually followed in mononuclear species, but it loses
its meaning in multicomponent (supramolecular) systems in which
the HOMO and LUMO of the compound can belong to different
subunits. In 4, the LUMO is centered on the polypyridine moiety of
the bridging ligand connected to the Re(I) center, and as a consequence,
it is not surprising that the MLCT level involving this metal center is
lower-lying than the MLCT level involving the Ir center, where the
HOMO is localized.
(50) Chen, P.; Meyer, T. J. Chem. ReV. 1998, 98, 1439 and references
therein.

Luminescent Ir(III) Compounds
Inorganic Chemistry, Vol. 40, No. 6, 2001 1101
compound is predominantly excited in the Ir center over the
luminescence properties are typical of the coupled systems. The
luminescence lifetimes and quantum yields are not useful in
clarifying the problem; intrinsic luminescence lifetimes of the
two “isolated” chromophores are expected to be very similar
so that single-photon-counting experiments cannot show the
presence of multiexponential decays, and the same occurs for
the quantum yields. In fact, the luminescence quantum yield
for 3 is independent of the excitation wavelength both in
acetonitrile and in dichloromethane solution within the experi-
mental error because of the similarity in the quantum yields
expected for the “isolated” chromophores, and therefore, such
a result agrees with any interpretation. So we have no definitive
experimental result to choose among the different cases, and
the only way is through a comparison with the properties of 4.
Because of the similarities between the systems, we propose
that even in 3 the energy-transfer process is hardly efficient;
the predominance of MLCT2 luminophore would also in this
case largely be due to a difference in the absorption properties
of the two chromophores, in particular to the better absorption
properties of MLCT2 over those of MLCT1 (see Absorption
Spectra), rather than to an effective energy-transfer process.
Re center; see Figure 1), and it decreases to 0.024 for λexc
)
395 nm (where the two centers are approximately equally
excited) and to 0.020 for λexc ) 410 nm (where the Re center
should be predominantly excited). This suggests that energy
transfer between the chromophores is not an efficient process.
However, on the basis of the quantum yields of the parent
mononuclear species 2 and 6 (Table 2), the difference in
quantum yields on excitation wavelength should be much larger
if energy transfer were totally inefficient, so partial energy
transfer between the chromophores in 4 seems to take place.
The analogous experiment in acetonitrile is not equally useful
because the differences in quantum yields (Φ ) 0.0055 for λexc
) 360 nm and Φ ) 0.0042 for λexc ) 410 nm) are close to the
experimental errors (20%, as given in the Experimental Section),
as also expected because of the closer quantum yields of the
two subunits (see Table 2).
In the absence of energy transfer, in principle the emission
spectra should be wavelength-dependent and the luminescence
decays should be multiexponential. This disagrees with our
findings (see Results and data collected in Table 2). However,
the absorption spectra of the two components are very similar
to each other in shape, with the Ir-based chromophore absorbing
more efficiently, and this makes excitation spectroscopy hardly
capable of identifying multiple emissions. Moreover, the dif-
ferent emission outputs52 made it experimentally difficult to look
for a minor component in the luminescence decay. However,
as also pointed out by one of the referees, two decays
characterized by a 1:10 preexponential factor difference and with
a difference in lifetime larger than 4 (as should be the case for
the two emitting levels of 4 in dichloromethane, on the basis
of the data of the parent complexes 2 and 6) should be
detectable. Because this is not the case, most likely partial energy
transfer takes place, smearing out the lifetime differences and
making detection of the second component much more difficult.
As far as compound 3 is concerned, the luminescence spectra
of this species both in rigid matrix and in fluid solution are
significantly red-shifted with respect to 2 and 5, in which only
the MLCT1 luminophore is present. This suggests that the
emission of 3 is dominated by the MLCT2 luminophore. This
would also imply energy transfer from MLCT1 to MLCT2 across
the bridge. However, comparison of the luminescence properties
of 3 with those of the mononuclear complex [Ir(ppy)2(cpbpy)]+
(7, cpbpy ) 4′-(4-carboxyphenyl)-6′-2,2′-bipyridine; data in
Our results suggest that photoinduced energy transfer across
the ester-linked bridging ligands is barely efficient in the systems
studied here, although it seems to occur to some extent.
However, we would like to point out that this does not mean
that ester-linked connectors are not appropriate for this type of
study. In fact, relatively fast and efficient photoinduced electron
transfer takes place in porphyrin-quinone systems assembled
via ester-linked connectors.53,54 The low driving forces for
energy transfer in our systems (around -0.07 eV in both 3 and
4; the energy levels of MLCT1, MLCT2, and MLCT3 are
approximated to the luminescence maxima of 2, 7, and 6,
respectively; such values do not correspond to the real energy
levels, but their energy gap, the relevant values for our
calculation, should be valid) could be responsible for the
observed behavior, in connection with the reorganization energy
needed for energy transfer in organometallic systems, which is
usually larger than in metal complexes containing only poly-
pyridine ligands.55 Unfavorable energetics (low driving force
and high reorganization energy) coupled with the weak elec-
tronic interaction between the partners, as shown by electro-
chemical results, may justify the low efficiency of the inter-
component energy transfer in the studied cases.
nitrogen-saturated acetonitrile at room temperature are λmax
)
660 nm, τ ) 125 ns, Φ ) 0.017; data for alcoholic matrix at
77 K are λmax ) 565 nm, τ ) 3.3 µs),28 in which an excited
state very similar to MLCT2 is present, indicates that the
emission of 3 does not occur from a pure MLCT2 level. A clear-
cut statement indeed is difficult to make. In principle, three cases
may occur: (a) the two chromophores are only weakly interact-
ing, with energy transfer (if any) occurring only to a minor
extent so that each chromophore essentially contributes almost
independently to the emission properties, as it seems to be the
case for 4; (b) complete energy transfer occurs and the
luminescence properties are dominated by the MLCT2 lumino-
phore; (c) fast equilibration between the two emitting states
occurs (this would imply noticeable interaction) and the
Acknowledgment. We thank the Italian Ministero
dell’Universita` e della Ricerca Scientifica e Tecnologica
(MURST), the Consiglio Nazionale delle Ricerche (CNR), and
the European Community TMR Program (Research Network
on Nanometer Size Metal Complexes) for financial support.
Supporting Information Available: Procedure for determining the
number of the electrons involved in the redox processes and cyclic
and differential pulse voltammograms. This material is available free
of charge via the Internet at http://pubs.acs.org.
IC000212X
(53) Cowan, J. A.; Sanders, J. K. M.; Beddard, G. S.; Harrison, R. J. J.
Chem. Soc., Chem. Commun. 1987, 55.
(54) Irvine, M. P.; Harrison, R. J.; Beddard, G. S.; Leighton, P.; Sanders,
J. K. M. Chem. Phys. 1986, 104, 315.
(55) In the organometallic systems like the ones studied here, the MLCT
states have partial LLCT character so that the dipole changes in the
excited state compared to those in the ground state are more significant
and a relatively large solvent reorganization energy is expected.
(52) Because of the different absorption and emission properties, the ratio
of Ir-based and Re-based contributions to the emission output is more
than 10 in dichloromethane, the solvent in which the possibility to
resolve two lifetimes was higher because of the larger difference in
the intrinsic lifetimes of the two subunits (Table 2). The situation was
even worse in the other solvents, where the lifetimes are closer to
one another.