Complexes of functionalized dipyrido[3,2-a:2′,3′-c]-phenazine: A synthetic, spectroscopic, structural, and density functional theory study

Article abstract of DOI:10.1021/ic050179k

The ligands 11-bromodipyrido[3,2-a:2′,3′-c]phenazine and ethyl dipyrido[3,2-a:2′,3′-c]phenazine-11-carboxylate have been prepared and coordinated to ruthenium(ll), rhenium(l), and copper(l) metal centers. The electronic effects of substitution of dipyrido[2,3-a:3′,2′-c] phenazine (dppz) have been investigated by spectroscopy and electrochemistry, and some photophysical properties have been studied. The crystal structures of [Re(L)(CO)₃Cl] (L = ethyl dipyrido-[3,2-a:2′,3′-c] phenazine-11-carboxylate or 11-bromodipyrido[3,2-a:2′,3′-c] phenazine) are presented. Density functional theory calculations on the complexes show only small deviations in bond lengths and angles (most bonds within 0.02 A, most angles within 2°) from the crystallographic data. Furthermore, the vibrational spectra of the strongest Raman and IR bands are predicted to within an average 6 cm^-1 for the complexes [Re(L)(CO)₃Cl] and [Cu(L)-(triphenylphosphine)₂]BF ₄ (in the 1000-1700 cm^-1 region). Spectroscopic and electrochemical evidence suggest that reduction of the complex causes structural changes across the entire dppz ligand. This is unusual as dppz-based ligands typically have electrochemical properties that suggest charge localization with reduction on the phenazine portion of the ligand. The excited-state lifetimes of the complexes have been measured, and they range from ca. 200 ns for the [Ru(L)(2,2×-bipyridine)₂](PF₆)₂ complexes to over 2 μs for [Cu(11-bromodipyrido[3,2-a:2′,3′-c]phenazine) (PPh₉)₂](BF₄) at room temperature. The emission spectra suggest that the unusually long-lived excited states of the copper complexes result from metal-to-ligand charge transfer (MLCT) transitions as they are completely quenched in methanol. Electroluminescent films may be fabricated from these compounds; they show MLCT state emission even at low doping levels [<0.1% by weight in poly(vinylcarbazole) polymer matrix].

Full text of DOI:10.1021/ic050179k

Inorg. Chem. 2005, 44, 3551−3560
Complexes of Functionalized Dipyrido[3,2-a:2′,3′-c]-phenazine: A
Synthetic, Spectroscopic, Structural, and Density Functional Theory
Study
Natasha J. Lundin,† Penny J. Walsh,† Sarah L. Howell,† John J. McGarvey,‡ Allan G. Blackman,*,† and
Keith C. Gordon*,†
Department of Chemistry, MacDiarmid Institute for AdVanced Materials and Nanotechnology,
UniVersity of Otago, Union Place, P.O. Box 56, Dunedin, New Zealand, and School of Chemistry,
Queen’s UniVersity Belfast, Belfast BT9 5AG, Northern Ireland
Received February 2, 2005
The ligands 11-bromodipyrido[3,2-a:2
been prepared and coordinated to ruthenium(II), rhenium(I), and copper(I) metal centers. The electronic effects of
substitution of dipyrido[2,3-a:3 ,2 -c]phenazine (dppz) have been investigated by spectroscopy and electrochemistry,
and some photophysical properties have been studied. The crystal structures of [Re(L)(CO)3Cl] (L ethyl dipyrido-
[3,2-a:2 ,3 -c]phenazine-11-carboxylate or 11-bromodipyrido[3,2-a:2 ,3 -c]phenazine) are presented. Density functional
theory calculations on the complexes show only small deviations in bond lengths and angles (most bonds within
′,3′-c]phenazine and ethyl dipyrido[3,2-a:2′,3′-c]phenazine-11-carboxylate have
′
′
)
′
′
′ ′
0.02 Å, most angles within 2
°
) from the crystallographic data. Furthermore, the vibrational spectra of the strongest
1
Raman and IR bands are predicted to within an average 6 cm- for the complexes [Re(L)(CO)3Cl] and [Cu(L)-
1
(triphenylphosphine)2]BF4 (in the 1000
−
1700 cm- region). Spectroscopic and electrochemical evidence suggest
that reduction of the complex causes structural changes across the entire dppz ligand. This is unusual as dppz-
based ligands typically have electrochemical properties that suggest charge localization with reduction on the
phenazine portion of the ligand. The excited-state lifetimes of the complexes have been measured, and they range
from ca. 200 ns for the [Ru(L)(2,2′-bipyridine)2](PF6)2 complexes to over 2 µs for [Cu(11-bromodipyrido[3,2-a:2′,3′-
c]phenazine)(PPh3)2](BF4) at room temperature. The emission spectra suggest that the unusually long-lived excited
states of the copper complexes result from metal-to-ligand charge transfer (MLCT) transitions as they are completely
quenched in methanol. Electroluminescent films may be fabricated from these compounds; they show MLCT state
emission even at low doping levels [<0.1% by weight in poly(vinylcarbazole) polymer matrix].
Introduction
interest because of their intense fluorescence upon association
with DNA. Complexes of this type have been tagged
“molecular light switches” because, although the fluorescence
of these complexes in aqueous solutions is negligible,
increases of greater than 10 000-fold in the presence of DNA
Dipyrido[3,2-a:2′,3′-c]phenazine (dppz) (Figure 1) is a
heterocyclic aromatic ligand whose complexes have received
much attention in recent years, primarily because the
planarity and extended aromatic nature of dppz gives it the
ability to intercalate with DNA.1-9 Ruthenium-based poly-
pyridyl complexes containing dppz3 have been of specific
(4) Yam, V. W.-W. V.; Lo, K. K.-W.; Cheung, K.-K.; Kong, R. Y.-C. J.
Chem. Soc., Dalton Trans. 1997, 2067.
(5) Friedman, A. E.; Chambron, J. C.; Sauvage, J. P.; Turro, N. J.; Barton,
J. K. J. Am. Chem. Soc. 1990, 112, 4960.
(6) Coates, C. G.; McGarvey, J. J.; Callaghan, P. L.; Coletti, M.; Hamilton,
J. G. J. Phys. Chem. B 2001, 105, 730.
(7) Coates, C. G.; Callaghan, P.; McGarvey, J. J.; Kelly, J. M.; Jacquet,
L.; Kirsch-De Mesmaeker, A. J. Mol. Struct. 2001, 598, 15.
(8) Coates, C. G.; Olofsson, J.; Coletti, M.; McGarvey, J. J.; Oenfelt, B.;
Lincoln, P.; Norden, B.; Tuite, E.; Matousek, P.; Parker, A. W. J.
Phys. Chem. B 2001, 105, 12653.
* Authors to whom correspondence should be addressed. Fax: +64-3-
479-7906. E-mail: kgordon@alkali.otago.ac.nz (K.C.G.); blackman@
alkali.otago.ac.nz (A.G.B.).
† University of Otago.
‡ Queen’s University Belfast.
(1) Stoeffler, H. D.; Thornton, N. B.; Temkin, S. L.; Schanze, K. S. J.
Am. Chem. Soc. 1995, 117, 7119.
(2) Hiort, C.; Lincoln, P.; Norden, B. J. Am. Chem. Soc. 1993, 115, 3448.
(3) Turro, C.; Bossmann, S. H.; Jenkins, Y.; Barton, J. K.; Turro, N. J. J.
Am. Chem. Soc. 1995, 117, 9026.
(9) Coates, C. G.; Callaghan, P. L.; McGarvey, J. J.; Kelly, J. M.; Kruger,
P. E.; Higgins, M. E. J. Raman Spectrosc. 2000, 31, 283.
10.1021/ic050179k CCC: $30.25
Published on Web 04/21/2005
© 2005 American Chemical Society
Inorganic Chemistry, Vol. 44, No. 10, 2005 3551

Lundin et al.
donating group at the same position is expected to have the
opposite effect.18 One way to probe these substituent effects
is to model the complexes of interest using computational
methods such as density functional theory (DFT).
Computational chemistry allows the possibility of model-
ing metal polypyridyl systems, providing structural and
energetic information. One of the early examples in this field
was a study by Broo and Lincoln.19 They used semiempirical
intermediate neglect of differential overlap methods to model
a number of ruthenium(II) polypyridyl complexes, including
dppz complexes. These calculations were compared with the
ab initio Hartree-Fock (HF) method, with the ruthenium-
(II) center approximated by the effective core potential of
Stevens et al.,20 which includes 28 core electrons and the
valence double-ú basis set. Basis sets for the other atoms
were the minimal STO-3G or the 6-31G basis set. The STO-
3G basis set calculations are the least accurate because this
is the smallest basis set one can use in HF methods. The
6-31G basis set is also limited because it does not contain
any polarization functions, which have been found to be very
important in systems containing larger atoms.21 The HF
method has the fundamental shortcoming of not including
electron correlation appropriately, which is important in
extended π systems. DFT offers a method that includes
correlation and is generally more accurate than HF methods.22
DFT has been used to study metal polypyridyl com-
plexes.23-25 Dattelbaum et al.23 used DFT/B3LYP calcula-
tions to model the geometries and associated energies of [Re-
(CO)3(4,4′-X2bpy)(4-ethylpyridine)]+ complexes (X ) CH3,
H, COOEt) and to identify mixing of π*(4,4′-X2bpy) and
π*(CO) MOs. They described the rhenium(I) center using
the LANL2 effective core potential and associated basis set
and other atoms using the 6-31G(d) basis set. In another
study, Dattelbaum et al.24 used the DFT/B3LYP theory to
determine geometries and vibrational frequencies of both the
ground singlet state and the lowest triplet state of fac-[Re-
(CO)3(pp)(L)]+ complexes [pp ) polypyridyl ligand L )
4-ethylpyridine, 1-methyl-6-oxyquinone, or 10-(4-picolyl)-
phenothiazine]. They used the “small core” LANL2 relativ-
istic effective core potential and associated basis set to
describe the rhenium(I) center26 and the 6-31G(d) basis set
for the ligands. Time-dependent DFT calculations were used
to calculate the energies of excited states from the ground-
state geometry. Calculated vibrational frequencies of excited
states were compared with time-resolved infrared measure-
ments. Changes in the carbonyl absorptions upon excitation
Figure 1. Structure of dppz, showing ring labels and IUPAC numbering
(left) and the bond and angle labeling of the cis rotamer of 1 (right).
have been observed.4 Complexes of dppz derivatives are also
useful as components in organic light-emitting devices
(OLEDs) because their emission color is largely red-shifted
from their absorption wavelength.10,11
Dppz is also interesting from the perspective of its
electronic structure. The properties of dppz suggest that
reduction of the ligand, or complexes containing these
ligands, leads to the formation of a radical anion with a
charge located on the phenazine portion of the ligand.12-16
This may be interpreted in terms of the presence of
unoccupied molecular orbitals (MOs), which are localized
over either the phenanthroline (phen; rings A, B, C) or the
phenazine (phz; rings B, D, E) regions of the ligand.17 This
MO model, which has been used to rationalize the properties
of dppz complexes, was developed in a series of papers by
Kaim et al.12-16 The excited-state electronic properties of
the dppz complexes are also interesting because the metal-
to-ligand charge transfer (MLCT) excited states can be
similar in energy to ligand-centered (LC) states.17 Through
judicious selection of the metal ion coordinated to dppz, one
can tune the nature of the lowest-energy excited state from
MLCT to LC.18 The low orbital overlap between the dπ and
ligand π* MOs results in such complexes being virtually
colorless (in the absence of other chromophoric ligands) as
the MLCT transition is raised in energy from the visible to
the ultraviolet region and lowered in intensity.11
It is possible to tune the energy level of the phz-based
lowest unoccupied MO (LUMO) by substitution of the
aromatic skeleton of dppz. Electron-withdrawing groups
result in less-negative reduction potentials, and an electron-
(10) David, G.; Walsh, P. J.; Gordon, K. C. Chem. Phys. Lett. 2004, 383,
292.
(11) Gordon, K. C.; Walsh, P. J.; McGale, E. M. Curr. Appl. Phys. 2004,
4, 331.
(12) Baumann, F.; Kaim, W.; Posse, M. G.; Katz, N. E. Inorg. Chem. 1998,
37, 658.
(13) Fees, J.; Kaim, W.; Moscherosch, M.; Matheis, W.; Klima, J.; Krejcik,
M.; Zalis, S. Inorg. Chem. 1993, 32, 166.
(14) Fees, J.; Ketterle, M.; Klein, A.; Fiedler, J.; Kaim, W. J. Chem. Soc.,
Dalton Trans. 1999, 2595.
(15) Klein, A.; Kaim, W.; Waldhoer, E.; Hausen, H.-D. J. Chem. Soc.,
Perkin Trans. 2 1995, 2121.
(16) Klein, A.; Scheiring, T.; Kaim, W. Z. Anorg. Allg. Chem. 1999, 625,
1177.
(19) Broo, A.; Lincoln, P. Inorg. Chem. 1997, 36, 2544.
(20) Stevens, W. J.; Basch, H.; Krauss, M. J. Chem. Phys. 1984, 81, 6026.
(21) Foresman, J. B.; Frisch, A. Exploring Chemistry with Electronic
Strucuture Methods, 2nd ed.; Gaussian, Inc.: Pittsburgh, 1995.
(22) Young, D. C. Computational Chemistry: A Practical Guide for
Applying Techniques to Real-World Problems; Wiley: New York,
2001.
(23) Dattelbaum, D. M.; Martin, R. L.; Schoonover, J. R.; Meyer, T. J. J.
Phys. Chem. A 2004, 108, 3518.
(24) Dattelbaum, D. M.; Omberg, K. M.; Hay, P. J.; Gebhart, N. L.; Martin,
R. L.; Schoonover, J. R.; Meyer, T. J. J. Phys. Chem. A 2004, 108,
3527.
(17) Waterland, M. R.; Gordon, K. C. J. Raman Spectrosc. 2000, 31, 243.
(18) Waterland, M. R.; Gordon, K. C.; McGarvey, J. J.; Jayaweera, P. M.
J. Chem. Soc., Dalton Trans. 1998, 609.
(25) Wilhelmsson, L. M.; Esbjoerner, E. K.; Westerlund, F.; Norden, B.;
Lincoln, P. J. Phys. Chem. B 2003, 107, 11784.
(26) Russo, T. V.; Martin, R. L.; Hay, P. J. J. Phys. Chem. 1995, 99, 17085.
3552 Inorganic Chemistry, Vol. 44, No. 10, 2005

Functionalized Dipyridophenazine Complexes
Nd3+:YAG laser). Emission signals at various detection wavelengths
were measured with a photomultiplier tube (Model RCA 1P28) in
which several of the dynodes were wired to the anode as in the
circuit described by Hunt and Thomas33 to provide a rapid response.
Kinetic traces were recorded using a digital oscilloscope (Tektronix
model TDS 3032). All measurements were of approximately 4 ×
10-5 M solutions made up in spectrophotometric grade solvents at
room temperature. Solutions were purged with argon for 10 min
prior to measurement. UV-vis spectra were collected before and
after emission lifetime measurements in order to check that no
sample degradation had occurred. Incident pulse energies were
typically ∼10 mJ. The emission lifetime was determined by fitting
a single-exponential function to each signal. Emission signals were
collected at 2-3 wavelengths for each sample, and the reported
lifetimes are the mean of these measurements. FT-Raman spectra
were collected on powder samples using a Bruker IFS-55 interfer-
ometer with an FRA/106S attachment. The excitation source was
a Nd:YAG laser with an excitation wavelength of 1064 nm. An
InGaAs diode (D424) operating at room temperature was used to
detect Raman photons. All spectra were taken with a laser power
of 105 mW at the sample and a resolution of 4 cm-1, using 16-64
scans.
were monitored and compared with characteristic patterns
in the carbonyl absorptions of transient states. This method
was used to assign excited states.
This study investigates the structural and electronic effects
of substituting dppz at the 11 position with a bromide or an
ethyl ester group and, subsequently, the effect of coordination
of a metal center to the substituted ligands. The synthesis of
these ligands and their [Re(L)(CO)3Cl], [Ru(L)(bpy)2](PF6)2,
and [Cu(L)(PPh3)2]BF4 (bpy ) 2,2′-bipyridine, PPh3
)
triphenylphosphine) complexes is reported. The crystal
structures [Re(L)(CO)3Cl] (where L ) ethyl dipyrido[3,2-
a:2′,3′-c]phenazine-11-carboxylate or 11-bromodipyrido[3,2-
a:2′,3′-c]phenazine) are presented and compared with struc-
tures obtained from DFT calculations. The electronic properties
of the ligands and their complexes are investigated using
UV-vis absorption spectroscopy and cyclic voltammetry,
and excited-state properties are explored using fluorimetry
and lifetime measurements.
Experimental Section
Ethyl 3,4-Diaminobenzoate. 3,4-diaminobenzoic acid (500 mg,
3.3 mmol) was suspended in ethanol (30 mL). Concentrated H2-
SO4 (1 mL) was added dropwise, and the mixture was refluxed for
6 h. The solution was concentrated, and then H2O (30 mL) was
added. Aqueous NaOH (1 M) was added until the pH was 10, and
the product was extracted into CHCl3. The organic layer was washed
with H2O, dried with MgSO4, and evaporated to give a light brown
powder (480 mg, 81%). 1H NMR (CDCl3): δH 1.36 (t, 3H, J ) 6.9
Hz), 3.40 (bs, 4H), 4.31 (q, 2H, J ) 6.9 Hz), 6.67 (d, 1H, J ) 8.1
Hz), 7.41 (s, 1H), 7.49 (d, 1H, J ) 8.1 Hz) ppm. Anal. Calcd for
C9H12N2O2: C, 59.98; H, 6.71; N, 15.55. Found: C, 59.84; H, 6.61;
N, 15.24.
Ethyl Dipyrido[3,2-a:2′,3′-c]phenazine-11-carboxylate‚0.5H2O
(1). 1,10-phenanthroline-5,6-dione (585 mg, 2.8 mmol) and ethyl
3,4-diaminobenzoate (500 mg, 2.8 mmol) were suspended in ethanol
(80 mL), and the suspension was refluxed for 1 h. The solvent was
removed, the resulting solid was dissolved in CHCl3, and the
solution was filtered through Celite. The filtrate was purified by
column chromatography on neutral alumina (CHCl3), yielding the
product as a light brown powder (421 mg, 43%). 1H NMR (CDCl3):
δH 1.52 (t, 3H, J ) 7.2 Hz), 4.54 (q, 2H, J ) 7.2 Hz), 7.81 (q, 2H,
J ) 4.5 Hz), 8.38 (d, 1H, J ) 8.7 Hz), 8.50 (dd, 1H, J ) 2.1, 8.7
Hz), 9.07 (d, 1H, J ) 1.5 Hz), 9.29 (t, 2H, J ) 2.7 Hz), 9.63 (d,
2H, J ) 8.1 Hz) ppm. Anal. Calcd for C21H15N4O2.5: C, 69.41; H,
4.16; N, 15.42. Found: C, 69.72; H, 4.17; N, 15.19.
11-Bromodipyrido[3,2-a:2′,3′-c]phenazine (2). 2 was prepared
in the same manner as 1, except that 4-bromo-1,2-benzenediamine
was used instead of ethyl 3,4-diaminobenzoate. The product was
obtained as a light brown solid (757 mg, 75%). 1H NMR (CDCl3):
δH 7.77 (q, 2H, J ) 4.5, 7.8 Hz), 7.95 (dd, 1H, J ) 2.1, 9 Hz),
8.16 (d, 1H, J ) 9 Hz), 8.49 (d, 1H, J ) 2.4 Hz), 9.26 (dd, 2H, J
) 1.8, 4.5 Hz), 9.52 (m, 2H) ppm. Anal. Calcd for C18H9N4Br: C,
59.85; H, 2.51; N, 15.52; Br, 22.12. Found: C, 60.01; H, 2.75; N,
15.25; Br, 22.05.
[Cu(PPh3)2(ethyl dipyrido[3,2-a:2′,3′-c]phenazine-11-car-
boxylate)]BF4‚H2O (3). Diethyl ether (30 mL) was purged with
N2 for 15 min, then [Cu(PPh3)4]BF4 (338 mg, 0.28 mmol) and 1
(100 mg, 0.28 mmol) were added, and the suspension was stirred
for 55 h under N2 at room temperature. The resulting yellow
Materials. Commercially available reagents were used as
received. Literature procedures were used to prepare the precursor
1,10-phenanthroline-5,6-dione,27 4-bromo-1,2-benzenediamine,28,29
cis-[Ru(bpy)2Cl2],30 and [Cu(PPh3)4]BF4‚H2O.31 Ligands were
prepared by the Schiff-base condensation of 1,10-phenanthroline-
5,6-dione with the appropriate diamino compound. All metal
complexes were synthesized by established literature techniques
with minor variations.18,31,32 Spectroscopic grade solvents were dried
and filtered through neutral aluminum oxide before being used for
electrochemical and spectroscopic measurements.
Instrumentation. Electronic spectra were recorded on a Varian
Cary 500 Scan UV-vis-NIR spectrophotometer. Emission spectra
were recorded on a Perkin-Elmer luminescence spectrometer LS
1
50 B. H NMR spectra were recorded on a Varian Unity Inova-
300 2-Channel FT 300 MHz spectrometer at 25 °C. Chemical shifts
are reported relative to residual solvent signals. Elemental analyses
were performed by the Campbell Microanalytical Laboratory,
University of Otago. FT-IR spectra were recorded in the solid state
(KBr disk) on a Perkin-Elmer Spectrum BX FT-IR System with
Spectrum v. 2.00. Melting points were recorded on a Kofler
Heizbank hot-stage apparatus. Cyclic voltammograms were re-
corded from nitrogen-purged 0.5 mM dichloromethane or aceto-
nitrile solutions with 0.1 M tetrabutylammonium perchlorate as the
supporting electrolyte. The electrochemical cell consisted of a
1-mm-diameter platinum working electrode embedded in a KeL-F
cylinder with a platinum auxiliary electrode and a saturated
potassium chloride calomel reference electrode. The cyclic volta-
mmograms were calibrated against the decamethylferrocene/deca-
methylferrocenium couple. The potential of the cell was controlled
by an EG&G PAR 273A potentiostat with model 270 software.
Emission lifetime measurements were made using a nanosecond-
pulsed excitation source (Spectra-Physics Quanta-Ray, Model DCR
(27) Smith, G. F.; Cagle, F. W., Jr. J. Org. Chem. 1947, 12, 781.
(28) Elder, J. W. J. Chem. Educ. 1994, 71, A142.
(29) Wilson, J. G.; Hunt, F. C. Aust. J. Chem. 1983, 36, 2317.
(30) Marmion, M. E.; Takeuchi, K. J. J. Am. Chem. Soc. 1988, 110, 1472.
(31) Rader, R. A.; McMillin, D. R.; Buckner, M. T.; Matthews, T. G.;
Casadonte, D. J.; Lengel, R. K.; Whittaker, S. B.; Darmon, L. M.;
Lytle, F. E. J. Am. Chem. Soc. 1981, 103, 5906.
(32) Browne, W. R.; O’Connor, C. M.; Hughes, H. P.; Hage, R.; Walter,
O.; Doering, M.; Gallagher, J. F.; Vos, J. G. J. Chem. Soc., Dalton
Trans. 2002, 4048.
(33) Hunt, J. W.; Thomas, J. K. Radiat. Res. 1967, 32, 149.
Inorganic Chemistry, Vol. 44, No. 10, 2005 3553

Lundin et al.
precipitate was isolated by filtration and recrystallized from
methanol to give a bright yellow powder (157 mg, 54%). Anal.
Calcd for C57H46N4BO3F4P2Cu: C, 65.37; H, 4.43; N, 5.35; P, 5.91.
Found: C, 65.24; H, 4.25; N, 5.35; P, 6.08.
[Cu(PPh3)2(11-bromodipyrido[3,2-a:2′,3′-c]phenazine)]BF4 (4).
This complex was prepared in the same manner as 3 but using
ligand 2 instead of 1, yielding a bright yellow powder (226 mg,
78%). Anal. Calcd for C54H39N4BF4P2CuBr: C, 62.60; H, 3.79; N,
5.41; P, 5.98; Br, 7.71. Found: C, 62.30; H, 3.87; N, 5.37; P, 5.78;
Br, 7.54.
[Ru(bpy)2(ethyl dipyrido[3,2-a:2′,3′-c]phenazine-11-carboxy-
late)](PF6)2 (5). cis-[Ru(bpy)2Cl2] (104 mg, 0.2 mmol) and 1 (100
mg, 0.28 mmol) were dissolved in ethanol/H2O (1:1, 10 mL), and
the solution was refluxed overnight. The solvent was removed on
the rotary evaporator, the residue was dissolved in water (20 mL),
and the solution was filtered to remove unreacted ligand. Saturated
aqueous NaPF6 (2 mL) was added dropwise to the stirred filtrate,
and the resulting orange precipitate was isolated by filtration;
washed with water (2 × 10 mL), methanol (1 × 10 mL), and ether
(1 × 10 mL); and air-dried. The red-brown solid was suspended in
boiling methanol (20 mL) for 1 h, resulting in a bright orange
powder, which was isolated by filtration to give 5 (203 mg, 94%).
1H NMR (CDCl3): δH 1.52 (t, 3H, J ) 7.5 Hz), 4.54 (q, 2H, J )
7.2, 14.4 Hz), 7.83 (dd, 2H, J ) 4.5, 8.1 Hz), 8.42 (d, 1H, J ) 8.7
Hz), 8.52 (dd, 1H, J ) 1.2, 9 Hz), 9.10 (d, 1H, J ) 2.1 Hz), 9.31
(dd, 2H, J ) 2.4, 3.6 Hz), 9.68 (dd, 2H, J ) 0.9, 9 Hz) ppm. Anal.
Calcd for C41H30N8O2F12P2Ru: C, 46.55; H, 2.86; N, 10.60; P, 5.76.
Found: C, 46.77; H, 2.90; N, 10.76; P, 5.57.
corrected for Lorentz and polarization effects using SAINT.34 The
structures were solved by direct methods using SIR-9735 running
within the WinGX package,36 with the resulting Fourier map
revealing the location of all non-hydrogen atoms. A weighted full-
matrix refinement on F2 was carried out using SHELXL-97,37 with
all non-hydrogen atoms being refined anisotropically for 7 and 8.
Hydrogen atoms were included in calculated positions and were
refined as riding atoms with individual (or group, if appropriate)
isotropic displacement parameters. A possible disorder of C(30) in
7 was evidenced by large thermal parameters, but this was not
investigated further. The bromine atom in 8 was found to be
disordered over two sites, and this was modeled successfully using
a 50:50 occupancy. This, of course, means that the entire ligand
must be disordered over two sites, and such disorder is perhaps
suggested by the elongated ellipsoids of the bromine-substituted
ring. However, all attempts to model the disorder of the entire ligand
resulted in numerous atoms displaying nonpositive definite thermal
parameters, and thus, this disorder was not examined further.
Computational Studies. The Gaussian 03W package38 was used
to carry out geometry optimizations and frequency calculations for
compounds 1-8. The geometry optimizations, vibrational frequen-
cies, and their IR and Raman intensities were calculated using DFT
(B3LYP/6-31G(d)). Rhenium(I) and ruthenium(II) centers were
represented using the effective core potential and associated
LANL2DZ basis set. The optimized structure was used to calculate
vibrational frequencies. Importantly, none of the frequency calcula-
tions generated imaginary frequencies; this is consistent with an
energy minimum for the optimized geometry.
[Ru(bpy)2(11-bromodipyrido[3,2-a:2′,3′-c]phenazine)]-
(PF6)2.H2O (6). This complex was prepared in the same manner
as 5 but using ligand 2 instead of 1, giving the product as an orange
powder (193 mg, 89%). Anal. Calcd for C38H27N8OF12P2BrRu: C,
42.16; H, 2.51; N, 10.35; Br, 7.38. Found: C, 42.23; H, 2.46; N,
10.20; Br, 7.52.
[Re(CO)3Cl(ethyl dipyrido[3,2-a:2′,3′-c]phenazine-11-carbox-
ylate)] (7). Ethanol (40 mL) was degassed with N2 for 15 min;
then, [Re(CO)5Cl] (50 mg, 0.14 mmol) and 1 (49 mg, 0.14 mmol)
were added, and the solution was refluxed for 5 h. The solution
was cooled on ice, and the resulting orange precipitate was filtered
(74 mg, 81%). Crystals suitable for X-ray diffraction were grown
by slow evaporation of a dichloromethane solution. 1H NMR
(CDCl3): δH 1.54 (t, 3H, J ) 6.9 Hz), 4.56 (q, 2H, J ) 6.6 Hz),
8.03 (t, 2H, J ) 5.7 Hz), 8.39 (d, 1H, J ) 8.7 Hz), 8.58 (d, 1H, J
) 8.7 Hz), 9.06 (s, 1H), 9.46 (d, 2H, J ) 3.6 Hz), 9.77 (d, 2H, J
) 7.8 Hz) ppm. Anal. Calcd for C24H14N4O5ClRe: C, 43.67; H,
2.14; N, 8.49; Cl, 5.37. Found: C, 43.91; H, 2.02; N, 8.58; Cl,
5.33.
[Re(CO)3Cl(11-bromodipyrido[3,2-a:2′,3′-c]phenazine)] (8).
The complex was synthesized in the same manner as 7 but using
ligand 2 instead of 1. The desired product was obtained as a yellow
powder (64 mg, 69%). Crystals suitable for X-ray diffraction were
grown by slow evaporation of a dichloromethane solution. 1H NMR
(CDCl3): δH 8.03 (s, 2H, J ) 2.7 Hz), 8.10 (dd, 1H, J ) 2.1, 7.2
Hz), 8.30 (d, 1H, J ) 9 Hz), 8.63 (d, 1H, J ) 2.1 Hz), 9.47(dt, 2H,
J ) 0.9, 5.4 Hz), 9.80 (m, 2H) ppm. Anal. Calcd for C21H9N4-
BrClO3Re: C, 37.82; H, 1.36; N, 8.40; Br, 11.98; Cl, 5.32. Found:
C, 37.96; H, 1.45; N, 8.23; Br, 12.12; Cl, 5.39.
Solid-State Structures. Crystals of complexes 7 and 8 were
grown from a concentrated dichloromethane solution, and X-ray
crystallographic data were collected at -97 or -109 °C, respec-
tively, on a Siemens SMART system using graphite monochromated
Mo KR radiation with exposures over 0.3°. No appreciable decay
of the crystals was detected during data collection. Data were
Electroluminescence Studies. OLEDs were fabricated by depo-
sition onto an indium tin oxide (ITO) glass substrate. The ITO (Rs
) 100 Ω/square) was patterned by etching with 37% HCl/FeCl3
and cleaned using acetone and Opti-Clean (Datronix) optical
cleaning polymer.
A solution of N,N′-diphenyl-N,N′-di(3-methylphenyl)benzidine
(TPD) (Aldrich) in 1,1,2-trichloroethane (10 mg/mL) was spin
coated onto ITO as a hole transport layer. The emitting layer
containing poly(vinylcarbazole) (PVK; 10 mg/mL), TPD (1 mg/
mL), 2-[1,1′-biphenyl]-4-yl-5-[4-(1,1-dimethylethyl)phenyl]-1,3,4-
oxadiazole (from Aldrich; 4 mg/mL), and the dopant (0.5 mg/mL)
in 1,1,2-trichloroethane was spin coated over the hole transport
layer. Some washing away of the hole transport layer is expected.
(34) SAINT, version 4; Siemens Analytical X-ray Systems: Madison, WI,
1996.
(35) Altomere, A.; Burla, M. C.; Camilla, M.; Cascarano, G. L.; Giaco-
vazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna,
R. J. Appl. Crystallogr. 1999, 32, 115.
(36) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
(37) Sheldrick, G. M. SHELXL-97; University of Go¨ttingen: Go¨ttingen,
Germany, 1997; Program for the Refinement of Crystal Structures
section.
(38) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. J. A.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone,
V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G.
A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev,
O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P.
Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas,
O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J.
B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision
B.04; Gaussian, Inc.: Pittsburgh, PA, 2003.
3554 Inorganic Chemistry, Vol. 44, No. 10, 2005

Functionalized Dipyridophenazine Complexes
Table 1. Crystallographic Data for 7 and 8
7
8
formula
fw
C24H14ClN4O5Re
660.04
C44H24Br2Cl2N8O8Re2
1395.83
space group
Z
P1h
2
C2/c
4
cryst syst
a (Å)
b (Å)
triclinic
6.403(5)
10.388(5)
16.976(5)
84.087(5)
84.161(5)
79.369(5)
1099.9(11)
1.993
-97
0.710 69
5.692
0.0536
0.0932
monoclinic
19.506(5)
18.043(5)
13.320(5)
c (Å)
R (deg)
â (deg)
γ (deg)
V, Å3
114.936(5)
4251(2)
2.181
-109
0.710 69
7.762
0.0298
0.0827
D calcd (g cm-3
temp (°C)
λ, (Å)
)
µ (mm-1
R1
)
wR2
Table 2. Selected Bond Lengths (Å) and Angles (deg) for 7 and 8
7
Re(1)-N(1)
2.170(6)
2.185(5)
2.4683(19) Re(1)-C(3)
75.6(2)
83.12(16)
97.2(3)
172.6(3)
95.8(2)
Re(1)-C(1)
Re(1)-C(2)
1.921(8)
1.927(8)
1.890(8)
94.1(2)
93.3(2)
91.1(2)
Re(1)-N(2)
Re(1)-Cl(1)
N(1)-Re(1)-N(2)
N(1)-Re(1)-Cl(1)
N(1)-Re(1)-C(1)
N(1)-Re(1)-C(2)
N(1)-Re(1)-C(3)
N(2)-Re(1)-Cl(1)
N(2)-Re(1)-C(1)
N(2)-Re(1)-C(2)
N(2)-Re(1)-C(3)
Cl(1)-Re(1)-C(1)
Cl(1)-Re(1)-C(2)
Cl(1)-Re(1)-C(3) 177.1(2)
C(1)-Re(1)-C(2)
C(1)-Re(1)-C(3)
C(2)-Re(1)-C(3)
87.5(3)
89.4(3)
90.0(3)
83.12(16)
172.3(3)
99.4(3)
8
Re(1)-N(1)
2.171(4)
2.168(4)
2.4648(15) Re(1)-C(300)
75.21(15)
80.51(11)
Re(1)-C(100)
Re(1)-C(200)
1.911(5)
1.902(6)
1.900(6)
175.1(2)
96.14(16)
Re(1)-N(2)
Figure 2. (A) (i) ORTEP view of 7 with atom labeling scheme. (ii)
Alternative view of 7 emphasizing distortion of the dppz ligand. (B) ORTEP
view of 8 showing the disorder of the bromine atom over two sites (50:50
occupancy). Ellipsoids are drawn at the 30% probability level in all cases.
Re(1)-Cl(1)
N(1)-Re(1)-N(2)
N(1)-Re(1)-Cl(1)
N(1)-Re(1)-C(100) 171.51(17)
N(1)-Re(1)-C(200) 96.0(2)
N(1)-Re(1)-C(300) 101.4(2)
N(2)-Re(1)-Cl(1)
N(2)-Re(1)-C(100) 96.76(18)
N(2)-Re(1)-C(200) 94.4(2)
N(2)-Re(1)-C(300)
Cl(1)-Re(1)-C(100)
Cl(1)-Re(1)-C(200) 176.46(19)
Cl(1)-Re(1)-C(300) 91.8(2)
C(100)-Re(1)-C(200) 87.3(2)
C(100)-Re(1)-C(300) 86.4(2)
C(200)-Re(1)-C(300) 89.4(3)
Re(1) is distorted octahedral because of a N(1)-Re(1)-N(2)
bond angle of 75.6(2)°.
84.26(11)
A comparison of the solid-state structure of 7 with the
structure obtained from DFT calculations shows close
agreement between the calculated and experimentally ob-
served bond lengths (Table 3). The difference between bond
lengths within the crystal and calculated structures of 7 is
less than 0.02 Å. This contrasts with a comparison of
calculations performed on the dppz ligand using B3LYP/6-
31G(d)41 and crystallographic data for [Ru(bpy)2(11,12-
Thermal evaporation at a pressure of 0.01 mTorr was used to
deposit a layer of 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline
to a thickness of approximately 6 nm, and tris(8-hydroxyquinoli-
nate)aluminum (Alq3) was deposited to a thickness of 20 nm. This
was followed by deposition of the aluminum cathode, at a thickness
of 200 nm. The thickness of the layers was monitored using a
Taylor-Hobson Talystep mechanical probe.
Emission spectra were obtained on a Perkin-Elmer LS-50B
fluorescence spectrometer. A luminance meter (Topcon BM-9) was
used to measure the brightness of the devices, and the current was
monitored with a Fluke multimeter.
42
dimethyldipyrido[3,2-a:2′,3′-c]pyridine)](ClO4)2 and [Pt-
(dppz)(2,4,6-trimethylphenyl)2]‚toluene,16 which showed a
large number of bond lengths differing from the calculated
values by more than 0.03 Å.
The dppz ligand in the complexes [Re(I)(dppz)(4-
methylpyridine)(CO)3](OTf)4 and [Re(benzo[i]dipyrido[3,2-
a:2′,3,-c]phenazine)(CO)3(pyridine)](OTf)43 is essentially
Results and Discussion
Structural Studies. X-ray data for 7 and 8 are presented
in Table 1, and selected bond lengths and angles are shown
in Table 2. ORTEP diagrams of the two structures are
presented in Figure 2. A labeled diagram of 1 is shown in
Figure 1. The rhenium-ligand bond lengths of 7 are
comparable to those observed in the similar complexes fac-
[Re(1,10-phenanthroline)(CO)3Cl]39 and fac-[Re(dipyrido-
[2,3-a:3′,2′-c]phenazine)(CO)3Cl].40 The geometry about
(39) Haddad, S. F.; Marshall, J. A.; Crosby, G.; Twamley, B. Acta
Crystallogr., Sect. E 2002, E58, m559.
(40) Polson, M. I. J.; Howell, S. L.; Flood, A. H.; Burrell, A. K.; Blackman,
A. G.; Gordon, K. C. Polyhedron 2004, 23, 1427.
(41) Matthewson, B. J.; Flood, A.; Polson, M. I. J.; Armstrong, C.; Phillips,
D. L.; Gordon, K. C. Bull. Chem. Soc. Jpn. 2002, 75, 933.
(42) Komatsuzaki, N.; Katoh, R.; Himeda, Y.; Sugihara, H.; Arakawa, H.;
Kasuga, K. J. Chem. Soc., Dalton Trans. 2000, 3053.
(43) Yam, V. W.-W.; Lo, K. K.-W.; Cheung, K.-K.; Kong, R. Y.-C. J.
Chem. Soc., Chem. Commun. 1995, 1191.
Inorganic Chemistry, Vol. 44, No. 10, 2005 3555

Lundin et al.
Table 3. Selected Calculated and Observed Bond Lengths of Ligands and Complexes
compound;
bond
5;
7;
7;
crystal
6;
8;
crystal
8;
6-31G(d),
9;
crystal
1;
3;
6-31G(d),
6-31G(d),
2;
4;
6-31G(d),
parametera 6-31G(d)b 6-31G(d) LANL2DZ LANL2DZ structure 6-31G(d) 6-31G(d) LANL2DZ structure LANL2DZ structure
r2
r2′
r3
r4
r5
r6
r7
r8
r9
1.383
1.383
1.404
1.328
1.347
1.415
1.476
1.462
1.440
1.353
1.434
0.1
1.427
1.494
1.349
1.447
1.517
1.386
1.386
1.399
1.338
1.359
1.407
1.458
1.463
1.434
1.351
1.441
0.1
1.428
1.501
1.342
1.454
1.516
1.386
1.386
1.400
1.340
1.365
1.408
1.449
1.463
1.432
1.349
1.446
0.1
1.429
1.504
1.338
1.458
1.515
1.386
1.386
1.398
1.338
1.359
1.407
1.453
1.463
1.438
1.352
1.437
0.1
1.427
1.496
1.346
1.449
1.517
1.367
1.352
1.389
1.327
1.367
1.396
1.448
1.461
1.429
1.350
1.415
0.8
1.420
1.482
1.337
1.442
1.402
1.383
1.383
1.404
1.328
1.347
1.415
1.476
1.462
1.440
1.352
1.434
0.0
1.386
1.386
1.399
1.338
1.359
1.407
1.458
1.463
1.433
1.351
1.441
0.1
1.386
1.386
1.400
1.340
1.365
1.408
1.449
1.463
1.431
1.350
1.445
0.1
1.352
1.365
1.398
1.323
1.347
1.394
1.442
1.445
1.426
1.348
1.408
0.6
1.385
1.385
1.398
1.338
1.358
1.407
1.452
1.463
1.438
1.351
1.436
0.0
1.370
1.337
1.356
1.376
1.356
1.383
1.424
1.448
1.421
1.336
1.468
2.6
r11
r12
∆
r
r15
r16
r18
r19
r20
φ1
1.421
1.909
1.425
1.897
1.428
1.890
1.405
1.834
1.423
1.903
1.396
119.1
119.1
119.1
118.9
119.3
119.1
119.1
119.1
119.4
119.0
118.9
φ1′
φ4
φ4′
φ5
φ5′
φ9
φ9′
119.1
118.2
118.2
122.3
122.3
119.6
119.5
0.1
119.1
118.1
118.1
122.6
122.6
119.8
119.7
0.1
119.1
118.4
118.4
122.4
122.4
119.8
119.8
0.1
119.0
118.5
118.6
122.3
122.3
119.8
119.7
0.1
120.7
118.5
118.3
122.1
122.8
119.2
120.6
0.6
119.1
118.1
118.1
122.3
122.3
119.5
119.6
0.3
119.1
118.1
118.1
122.6
122.6
119.7
119.8
0.2
119.1
118.5
118.4
122.4
122.4
119.8
119.9
0.2
118.8
119.0
117.5
122.2
123.2
120.2
119.4
0.6
119.0
118.6
118.6
122.3
122.3
119.7
119.8
0.3
119.8
116.9
120.3
123.6
121.2
118.8
120.5
1.2
∆
φ
a Bond lengths (r) in Å, average difference in bond lengths about C2V axis (∆r) in pm; bond angles (φ) in degrees, average difference in bond angles about
C2V axis (∆φ) in degrees. b Basis sets used in geometry calculations.
planar, showing mean deviations of 0.019 Å (maximum
deviation ) 0.1174 Å) and 0.0084 Å (maximum deviation
) 0.1104 Å), respectively, from the least-squares plane of
N4, N5, C11, and C12 (Figure 1). However, the dppz ligand
in 7 was found to be significantly distorted (Figure 2) with
deviations from the least-squares plane of the atoms N(1),
N(2), C(24), and C(25) of up to 0.2537 Å for N(4) and a
mean deviation of 0.0578 Å. Ligand distortion in dppz-based
complexes has been previously observed in crystal structures
of [Ru(dppz)(2,9-dimethyl-1,10-phenanthroline)2](PF6)2,44 [Cu-
(dppz)(tris(3-phenylpyrazolyl)borate)](ClO4),45 and [Os4-
(dppz)(µ-H)4(CO)10].46 Similar distortion has been observed
for other polypyridyl ligands in rhenium(I) complexes such
as [Re(dipyrido[2,3-a:3′,2′-c]phenazine)(CO)3Cl].40 For this
complex, it was suggested that a destabilizing steric interac-
tion between equatorial carbonyls and protons ortho to the
nitrogen donors caused distortion of the polypyridyl ligand.
It is unlikely that this is the case for 7, as the C(1)-H(19)
distance of 2.812 Å and the C(2)-H(10) distance of 2.936
Å are similar to the distances for the complexes [Re(I)(dppz)-
(4-methylpyridine)(CO)3](OTf) (2.724 and 2.950 Å)4 and
[Re(benzo[i]dipyrido[3,2-a:2′,3,-c]phenazine)(CO)3(pyridine)]-
(OTf) (2.919 and 2.801 Å)43 in which the dppz ligands are
approximately planar. The distortion of the dppz ligand in 7
may instead be a combination of crystal packing effects and
π-π stacking interactions between the dppz ligands (Figure
3). The optimized geometries of 7 and 8 as calculated by
density functional theory display a planar aromatic backbone
of dppz because crystal packing forces are not considered
in calculations performed on single molecules.
The geometry about the rhenium atom in 8 is similar to
that of 7. The N(1)-Re(1)-N(2) bond angle of 75.21(15)°
again leads to a distorted octahedral geometry. The Re-C,
Re-N, and Re-Cl bond lengths of 8 are comparable to those
of 7 and other similar complexes. The differences in
corresponding bond lengths between the solid-state and
calculated geometries of 8 are similar to the differences
observed for 7. Disagreements between observed and cal-
culated bond lengths of 8 are generally less than 0.02 Å.
Complex 8 also shows distortion of the backbone of the
dppz-based ligand, with deviations from the least-squares
plane of N(1), N(2), C(15), and C(16) of up to 0.1877 Å for
N(3) and a mean deviation of 0.0536 Å. The distortion of
the dppz-based ligand in 8 allows the formation of isomers
in the solid state because it causes the two possible
orientations of the ligand to become nonequivalent. Refine-
ment of the occupancies of these two isomers shows them
to be present in an essentially 1:1 ratio (0.50063:0.49937).
The isomers are not resolved in solution by 1H NMR
spectroscopy, implying that the ligand distortion observed
in the solid-state structure is either not present or not fixed
in solution. The distances of the ortho hydrogen atoms from
the nearest carbonyl carbon are 2.772 Å [C(100) and H(10)]
and 2.970 Å [C(300) and H(1)], similar to that observed for
the crystal structure of 7, suggesting that the ligand distortion
again does not arise because of a steric interaction between
the carbonyl ligands and the dppz-based ligand. Complex 8
shows strong π-π stacking interactions, which may account
(44) Liu, J.-G.; Zhang, Q.-L.; Shi, X.-F.; Ji, L.-N. Inorg. Chem. 2001, 40,
5045.
(45) Dhar, S.; Reddy, P. A. N.; Nethaji, M.; Mahadevan, S.; Saha, M. K.;
Chakravarty, A. R. Inorg. Chem. 2002, 41, 3469.
(46) Choi, Y.-Y.; Wong, W.-T. J. Organomet. Chem. 1999, 573, 189.
3556 Inorganic Chemistry, Vol. 44, No. 10, 2005

Functionalized Dipyridophenazine Complexes
Table 4. Electronic Absorption Data for Ligands and Complexes
absorption λmax/nma
(ꢀ × 10-4/mol-1 L cm-1
π* r π
)
mixed
MLCT
1
276(3.02)
348(0.53)
356(0.60)
366(0.82)
374(0.67)
386(0.81)
279(2.83)
329(0.70)
366(0.61)
385(0.68)
280(9.15)
334(1.84)
362(1.35)
282(6.03)
330(1.00)
366(1.20)
283(12.1)
362(1.88)
284(8.42)
364(1.25)
375(1.55)
281(14.0)
323(2.88)
283(4.01)
321(1.23)
367(1.11)
2
3b
4b
381(1.19)
385(1.39)
5
6
442(1.75)
441(1.31)
7
8
363(2.44)
384(1.05)
a Spectra were recorded in dichloromethane unless otherwise stated.
b Recorded in acetonitrile.
than 0.01 Å along the phenazine backbone and ester
functionality. Binding to a copper(I) center produces moder-
ate (0.005-0.010 Å) bond length changes along the phena-
zine backbone and ester bonds. These findings are consistent
with the weak effect the {Cu(PPh3)2}+ fragment is known
to have on the structure of bound ligands.17
Figure 3. π-π stacking interactions of 7 (A) and 8 (B) showing centroid-
centroid distances (Å). Hydrogen atoms and solvent molecules have been
omitted for clarity.
Electronic Absorption Spectra. The UV-visible spectral
data of 1-8 are presented in Table 4. The ligands both show
typical absorption spectra for 11-substituted dppz ligands18
with an intense transition at approximately 280 nm and a
series of weaker bands in the 360 nm region. On the basis
of the similarity of the spectra of 1 and 2 to those of other
ligands of this type, the bands at 280 and 360 nm are assigned
as LC transitions. Complexation of 1 or 2 to a metal center
results in a red shift of the LC transition. The LC transitions
for 2 and its complexes occur at slightly lower energy than
for 1 and its corresponding complexes in all cases. Because
substitution of the ligand affects the position of the LC
transition, an orbital involved in this transition is phz-based.
The complexes absorb weakly in the visible region, with
patterns similar to those of other complexes of the type.14
The ruthenium complexes show MLCT bands at around 440
nm, due to the known chromophore Ru(II) f bpy,47 whereas
the copper and rhenium complexes have their higher-energy
MLCT transitions overlapped by the more-intense LC
transition. The MLCT shoulder of 7 extends further into the
visible region than for 8.
for the distortion of the dppz skeleton (Figure 3). The ligands
π-stack in an alternating fashion with an average centroid-
centroid distance of 3.6489 Å.
Calculated Structural Data. The calculated structures
offer some insight into the effects of metal bonding and
ligand conformation. Rotation about bond r16 allows the
possibility of two conformational isomers for 1 that lie at
local energy minima. It is possible to model this isomeriza-
tion using computational methods. B3LYP/6-31G(d) calcula-
tions on each conformer of 7 showed the trans conformer,
in which r14 is trans to r17, to be 1.7 kJ mol-1 lower in energy
than the cis conformer. Because these two rotational con-
formers lie close in energy, structures and spectra were
calculated for the trans conformer, although the obtained
crystal structure of 7 showed the presence of the cis
conformer only.
The calculated structural parameters for 1-8 are compared
with the crystallographic data in Table 3. A comparison of
calculated optimized structures for 1-8 shows that the ligand
structure is calculated to be sensitive to different metal
centers. Binding of 1 or 2 to a rhenium(I) center produces
calculated bond length changes relative to the free ligand of
more than 0.01 Å in the phenanthroline section of the ligand.
However, binding to a ruthenium(II) center generates pre-
dicted bond length changes relative to the free ligand of more
Electrochemical Studies. Electrochemical data for com-
pounds 1-8 are presented in Table 5. Substitution of dppz
with an ester group or bromine substituent results in reduction
(47) Hartshorn, R. M.; Barton, J. K. J. Am. Chem. Soc. 1992, 114, 5919.
Inorganic Chemistry, Vol. 44, No. 10, 2005 3557

Lundin et al.
Table 5. Electrochemical Data for Ligands and Complexes
E0/Va
the emission wavelength. Complex 3 was emissive at two
wavelengths, one similar to that of the free ligand and the
other at a longer wavelength.
E0/Va
compound
reduction
compound
reduction
The ruthenium complexes, 5 and 6, displayed the shortest
excited-state lifetimes. These were similar, which suggests
that substitution of the 11-position of dppz has little effect
on the dynamics of these excited states. The emission
maxima are shifted by 30 nm on going from 5 to 6.
It is typical for the MLCT excited states of rhenium
complexes to show dynamic behavior dominated by the
energy-gap law.49 The absorption of light by 7 extends to
lower energy than 8, implying a lower-lying MLCT state
for 7. The electrochemical data show that 7 is 50 mV easier
to reduce than 8, and the fact that the excited-state lifetime
of 7 is shorter than that of 8 is consistent with these findings.
The copper complexes display by far the most interesting
excited-state properties. The excited-state lifetimes of the
complexes in dichloromethane are comparatively long but
completely quenched if the solvent is changed to methanol.
This marked solvent dependence suggests that the excited
states of 3 and 4 are MLCT in nature, as copper(I) complexes
that emit from MLCT excited states are strongly solvent-
dependent.17 For copper(I) complexes, the main influence
on the excited-state lifetime is its ability to form a solvent-
based exciplex with increased steric protection, thereby
prolonging the lifetime.50 The long lifetimes of 3 and 4
suggest that the solvent-based exciplexes are relatively stable
for an excited state in a nonprotic solvent such as dichlo-
romethane. Although this result is surprising, it is consistent
with electrochemical data.
Vibrational Spectra. The efficacy of the calculations for
those complexes for which structural data are not available
may be tested by using the calculations to predict other
observables of the complexes. The vibrational spectra offer
an opportunity to compare calculated and experimental
parameters. Experimental and calculated Raman spectra for
7 are presented in Table 7, along with experimental and
calculated IR spectra and assignments of vibrational modes.
A comparison of the calculated and experimental spectra
shows good correlation. Bands in the IR and Raman spectra
between 1000 and 1650 cm-1, with a relative intensity (RI)
of at least 25% of the most intense band, have an absolute
mean deviation of less than 6 cm-1 for compounds 1-8.
The carbonyl band is not as well-predicted by the calculations
as the other modes;51 however, it is easily identifiable as a
strong band at about 1700 cm-1 (mode 124 for 7: calcd,
1756 cm-1; exptl, 1708 cm-1). Figure 4 shows experimental
and calculated IR and Raman spectra for 6. A comparison
of the experimental and calculated spectra for 6 shows good
agreement, with an absolute mean deviation for bands with
RI 25% of 5 cm-1. The FT-IR spectrum shows five strong
peaks between 1400 and 1500 cm-1, which are reproduced
1
2
3
4
-1.07
-1.11
-0.85
-0.95
5
6
7
8
-0.77
-0.86
-0.82
-0.87
a Reduction potentials of compounds 1, 2, 7, and 8 were recorded in
dichloromethane with the decamethylferrocene/decamethylferrocenium
couple as a reference. Compounds 3-6 had their reduction potentials
recorded in acetonitrile. 0.1 M tetrabutylammonium perchlorate was used
as the supporting electrolyte in all cases.
Table 6. Fluorescent Maxima and Excited-State Lifetimes of Ligands
and Complexes
excited-state absorption
compound
λ
max/nma
τ/µs
1
2
446
453
443, 605
650
622
652
432
456
not determined
not determined
1.06
3b
4b
5
2.25
0.27
0.24
1.01
6
7
8
1.57
a Fluorescence and lifetime measurements were performed in dichlo-
romethane unless otherwise stated. b Recorded in acetonitrile.
occurring at less-negative potentials than for unsubstituted
dppz (-1.60 V vs Fc+/Fc).14 Coordination of the ligands to
a metal center also causes reductions to occur at less negative
potentials. It is generally found that, for metal polypyridyl
complexes, the stabilization of the first ligand-based reduction
potential (E°/[first reduction of the free ligand] - E°/[first
reduction of the metal complex]) is greater for Ru(II) and
Re(I) than for Cu(I).18 This is found to occur for complexes
of 1 and 2. This is a surprising result as unsubstituted dppz
complexes show very little change in reduction potential as
a function of the chelating metal center.14 This has been
rationalized in terms of the reducing electron for dppz
complexes causing structural distortions over the phenazine
portion of the dppz ligand and not affecting the phenanthro-
line section. In terms of a MO picture of the reduction
process, the reducing electron may be thought of as occupy-
ing a redox MO,48 which has considerable phenazine
character. The tuning of the reduction potential with 11-
position substitution and metal binding does not fit within
the confines of the Kaim MO model12-16 and suggests that
the reduction is not partitioned in the fashion inferred from
data on unsubstituted dppz.
Fluorescence and Lifetime Studies. The fluorescence and
lifetime measurements are presented in Table 6. The free
ligands displayed fluorescence at around 450 nm, values
consistent with other 11-substituted dppz ligands.18 All of
the complexes are fluorescent, and their behavior may be
divided into three categories. Complexes 7 and 8 are
fluorescent at a similar wavelength to that of the free ligand,
suggesting that the emissions from 7 and 8 are LC in nature.
Complexes 4, 5, and 6 displayed fluorescent maxima at lower
energy, suggesting that the presence of a metal ion can affect
(49) Caspar, J. V.; Meyer, T. J. J. Phys. Chem. 1983, 87, 952.
(50) McMillin, D. R.; Gamache, R. E., Jr; Kirchoff, J. R. Copper
Coordination Chemistry: Biochemical and Inorganic PerspectiVes;
Adenine Press: New York, 1983.
(51) This is, in part, due to the greater anharmonicity of the CO modes
versus the polypyridyl ligand modes for which the scaling factor is
optimized
(48) De Armond, M. K.; Myrick, M. L. Acc. Chem. Res. 1989, 22, 364.
3558 Inorganic Chemistry, Vol. 44, No. 10, 2005

Functionalized Dipyridophenazine Complexes
Table 7. Calculated and Experimental Frequencies (cm-1) and Relative
IR and Raman Intensities for Selected Normal Modes of 7 and 8
in both the experimental and calculated IR spectra. The FT-
Raman spectrum has a band and shoulder at 1310 cm-1,
which appears in the calculated Raman spectrum as two
bands. A broad band at 1604 cm-1 may conceal a shoulder
at 1616 cm-1. Table 7 shows the experimental and calculated
FT-IR and FT-Raman spectral data of 8 with assigned
vibrational modes. A comparison of the calculated and
experimental spectra of 8 shows good correlation, with an
absolute mean deviation for bands with RI 25% of 5 cm-1.
In all DFT calculations on 3-8, an unusually large intensity
was predicted for a vibrational mode due to the phz stretch
at about 1410 cm-1. The increased intensity of this band is
not reflected in the experimental data or the calculated IR
spectrum. In the Raman spectrum of 6, the predicted band
at 1406 cm-1 shows an intensity relative to the 1310 cm-1
band of 4:1; experimentally, this value is closer to 0.8:1
(Figure 4). This finding is not without precedent; in a study
of substituted terthiophene compounds, Casado et al.52 found
that the strongest predicted band at 1393 cm-1 was almost 5
times more intense than that observed in the experimental
data.
Electro-Optical Properties. The development of elec-
troluminescent (EL) systems based on conducting polymers
and inorganic components has led to much research into
materials for display and lighting technologies.53-55 The
design of such materials can lead to improved efficiency,
color tuning, and an increased lifetime of EL displays. The
incorporation of metal polypyridyl complexes into electrolu-
minescent systems has resulted in interesting developments
for the technology.
ν/cm-1
ν/cm-1
modea
major bonds involvedb
calcd (int.)c exptl
(int.)c
7
92 r11 - r1 - r1′
1193 (9, 3)
1182 (0, 6)
93 r5 + r8′ + r10′ - r8 - r11 - r18 1216 (39, 2) 1209 (41, 4)
102 r7 - r4 - r4′
1325 (2, 27) 1316 (4, 77)
1353 (23, 20) 1347 (0, 18)
1356 (18, 1) 1358 (50, 17)
1410 (4, 100) 1409 (55, 93)
1418 (8, 7)
1444 (7, 7)
1479 (2, 3)
1490 (1, 1)
1552 (1, 8)
1585 (0, 0)
103 r9 - r8 - r8′ - r12
104 r10 + r10′ + r12 - r9
106 r12 + r14 + r14′ - r13 - r13
107 r20
1416 (71, 100)
1445 (38, 80)
1472 (0, 1)
1494 (6, 22)
1536 (1, 26)
1574 (9, 1)
109 r6 + r6′ - r7
111 r5 + r8 + r13′ - r1 - r7
114 r20
117 r10 + r11′ - r10′ - r11
119 r6 + r6′ - r2 - r2′
120 r2 + r2′ - r1 - r1′
1597 (1, 12) 1594 (0, 59)
8
78 r7 + r11 - r1 - r1′
79 r5 + r8′ - r5′ - r8′
82 r5 + r5′ - r7
85 r5 + r8 - r2 - r4
87 r8 + r8′ - r10 - r10′
88 r9 - r12
89 r12 + r14 + r14′ - r13 - r13′
90 r6′ + r10 - r6 - r13
91 r6 + r6′ - r7
1193 (5,2)
1217 (3,0)
1282 (3,6)
1316 (5,3)
1351 (4,9)
1183 (9,3)
1212 (9,2)
1279 (0,2)
1316 (0,53)
1344 (0,7)
1352 (100,7) 1357 (74,7)
1405 (6, 100) 1405 (0,100)
1418 (35,1)
1443 (3,9)
1454 (29,7)
1495 (58,3)
1547 (4,0)
1584 (0,1)
1597 (4,9)
1413 (78,0)
1446 (48,49)
1463 (34,0)
1491 (100,9)
1538 (0,14)
1577 (0,2)
92 r1′ + r11 - r8 - r13
95 r8 + r15 - r3′ - r7 - r13′
96 + r14 + r14′ - r12 - r15
98 r6 + r6′ - r2 - r2′
99 r1 + r1′ + r4′ + r7 - r3′
1595 (45,44)
1617 (12,7)
102 r6 + r13′ + r14 - r6′ - r13 - r14′ 1621 (30,2)
a Mode numbers for 7 and 8. b Main bond stretches that are involved in
the mode of vibration. c Relative intensities for IR and Raman bands,
respectively, normalized so that the most intense band in the reported spectral
region is equal to 100.
Phosphorescent (triplet) emitters such as Ru(II) and Re(I)
polypyridyl complexes have an advantage over fluorescent
(singlet) emitters,56 as only 25% of the excitons formed in
the EL process are singlets.57 This means that fluorescent
emitters have an absolute efficiency limit of 25%. Incorpo-
rating some Ru(II) complexes into organic polymers also
increases charge mobility in the polymer film because the
Ru(II) complex acts as a charge carrier.58 Das et al.59 studied
the electrophosphorescence of Ir(Me-ppy)3, a triplet emitter
(Me-ppy ) 3-methyl-2-phenylpyridine). Methyl substitution
at the 3-position influences the extent of delocalization
between the pyridyl and phenyl rings. This allows fine tuning
of the energy levels of ppy MOs, which gives a color-tuning
capability. Studies of electroluminescent systems containing
fac-[Re(CO)3Cl(pp)] and [Ru(bpy)2(pp)]2+ complexes, where
(pp) is a polypyridyl ligand with triphenylamine and oxa-
(52) Casado, J.; Pappenfus, T. M.; Mann, K. R.; Orti, E.; Viruela, P. M.;
Milian, B.; Hernandez, V.; Navarrete, J. T. L. ChemPhysChem 2004,
5, 529.
(53) Yu, J.-S. K.; Chen, W.-C.; Yu, C.-H. J. Phys. Chem. A 2003, 107,
4268.
(54) Lee, C.-L.; Lee, K. B.; Kim, J.-J. Appl. Phys. Lett. 2000, 77, 2280.
(55) Drolet, N.; Beaupre, S.; Morin, J.-F.; Tao, Y.; Leclerc, M. J. Opt. A:
Pure Appl. Opt. 2002, 4, S252.
(56) Lamansky, S.; Kwong, R. C.; Nugent, M.; Djurovich, P. I.; Thompson,
M. E. Org. Electron. 2001, 2, 53.
(57) Baldo, M. A.; O’Brien, D. F.; Thompson, M. E.; Forrest, S. R. Phys.
ReV. B: Condens. Matter Mater. Phys. 1999, 60, 14422.
(58) Chan, W. K.; Ng, P. K.; Gong, X.; Hou, S. J. Mater. Chem. 1999, 9,
2103.
(59) Das, R. R.; Lee, C.-L.; Noh, Y.-Y.; Kim, J.-J. Opt. Mater. (Amsterdam,
Neth.) 2003, 21, 143.
Figure 4. Vibrational spectra for 6: (a) FT-IR spectrum (KBr disk), (b)
calculated IR spectrum, (c) FT-Raman spectrum, and (d) calculated Raman
spectrum.
in the calculated IR spectrum in the same region. The 1488
cm-1 band is predicted at 1494 cm-1 and also appears in
the Raman spectrum. A feature at 1354 cm-1 also appears
Inorganic Chemistry, Vol. 44, No. 10, 2005 3559

Lundin et al.
Conclusion
The electronic absorption spectra of these complexes are
similar to those observed in other dppz systems. The lowest
energy transition in the copper(I) and rhenium(I) complexes
is the MLCT, and this appears as a shoulder against the LC
band. The spectra of the Ru(II) complexes differs in that
they show Ru f bpy transitions at 440 nm. These MLCT
transitions involving the ancillary ligands obscure dppz-based
transitions.
The electrochemical behavior of these complexes is
interesting. The presence of metal moieties significantly
stabilizes the first reduction potential, by up to 300 mV. It
is also found that the differing substitution, which is at the
phz end of the ligand, results in up to 100 mV changes in
the reduction potential. These two findings show that the
electronic nature of the these substituted dppz ligands differ
from that of the unsubstituted dppz. It suggests that, for these
complexes, reduction is affecting the structure of the entire
ligand.
Figure 5. Emission profile for EL devices containing complex 7 at 3 wt
% (solid line), 0.7 wt % (long dashed line), 0.07 wt % (short dashed line),
and 0.007 wt % (dotted line).
The fluorescence spectra for these compounds reveal the
presence of a number of excited states. The rhenium(I)
complexes are dominated by LC emissions, whereas the
ruthenium complexes show MLCT emissions; the copper(I)
complexes show both types of emission behavior. In addition,
the copper(I) complexes show extremely long excited-state
lifetimes that appear to be MLCT in origin. Such long-lived
copper(I) MLCT excited states normally require substituents
about the 3,6 positions to inhibit exciplex formation and
excited-state deactivation.61
Figure 6. Electro-optical characteristics of electroluminescent devices
containing complex 7 at 3 wt % (diamond), 0.7 wt % (circle), 0.07 wt %
(triangle), and 0.007 wt % (square): (a) luminance-voltage plot, (b)
electroluminescence efficiency plot.
Vibrational spectra of the complexes modeled by DFT
display an absolute mean deviation between predicted and
observed spectral bands of 6 cm-1. The molecular orbitals
provided by DFT calculations point to ligand-based LUMOs
that are delocalized across the entire ligand structure.
Although these findings contrast with those for unsubstituted
dppz in which phen- and phz-based MOs are calculated, they
are, nonetheless, consistent with the electrochemical studies.
diazole units,60 have shown that these materials have a higher
charge-carrier mobility compared to other oxadiazoles and
triarylamines.
Figure 5 shows the emission profiles of 7 doped into an
EL device at different doping levels. The lowest doping level
shows a blue-green emission from the PVK polymer11 and
Alq3 layer. However, at doping levels as low as 0.07 wt %,
the emission becomes dominated by a feature at 650 nm.
This is consistent with emission from the MLCT state of 7;
the electroluminescence differs from the solution phase
photoluminescence that originates from a ligand-centered
state at 432 nm.
Electroluminescent devices may be fashioned from these
complexes, and in the case of 7, emission from an MLCT
state is observed even at very low doping levels (<0.1% by
wt) for the complex. The electroluminescent properties of
these complexes will be the subject of further investigation.
Acknowledgment. We thank the MacDiarmid Institute
for Advanced Materials and Nanotechnology for providing
funding for this work.
The electro-optical properties of the EL devices are shown
in Figure 6. The addition of 7 at higher doping levels
impaired the efficiency and light output of the device. These
results suggest that the optimal doping level for this dye is
close to 0.07 wt %. The devices had luminance-current
efficiencies approaching 0.1 Cd A-1 at low current densities
(Figure 6).
Supporting Information Available: Crystallographic data in
CIF format. This material is available free of charge via the Internet
at http://pubs.acs.org.
IC050179K
(60) Gong, X.; Ng, P. K.; Chan, W. K. AdV. Mater. (Weinheim, Ger.) 1998,
10, 1337.
(61) Cuttell, D. G.; Kuang, S.-M.; Fanwick, P. E.; McMillin, D. R.; Walton,
R. A. J. Am. Chem. Soc. 2002, 124, 6.
3560 Inorganic Chemistry, Vol. 44, No. 10, 2005