Neutral Ru
II-Based Emitting Materials: A Prototypical Study on Factors
Governing Radiationless Transition in Phosphorescent Metal
Complexes
†Elise Y. Li, Yi-Ming Cheng, Cheng-Chih Hsu, Pi-Tai Chou,* and Gene-Hsiang Lee Department of Chemistry and Instrumentation Center, National Taiwan UniVersity, Taipei 106, Taiwan
I-Hui Lin and Yun Chi*
Department of Chemistry, National Tsing Hua UniVersity, Hsinchu 300, Taiwan Chao-Shiuan Liu
Department of Chemistry, SooChow UniVersity, Taipei 111, Taiwan Received January 13, 2006
In addition to the metal-centered dd transition that is widely accepted as a dominant radiationless decay channel, other factors may also play important roles in governing the loss of phosphorescence efficiency for heavy-transition-metal complexes. To conduct our investigation, we synthesized two dicarbonylruthenium complexes with formulas [Ru(CO)2(BQ)2] (1) and [Ru(CO)2(DBQ)2] (2), for which the cyclometalated ligands BQ and DBQ denote benzo-[h]quinoline and dibenzo[f,h]quinoxaline, respectively. Replacing one CO ligand with a P donor ligand such as PPh2Me and PPhMe2caused one cyclometalated ligand to undergo a 180°rotation around the central metal atom, giving highly luminous metal complexes [Ru(CO)L(BQ)2] and [Ru(CO)L(DBQ)2], where L)PPh2Me and PPhMe2 (3−6), with emission peaksλmaxin the range of 571−656 nm measured in the fluid state at room temperature. It is notable that the S0−T1 energy gap for both 1 and 2 is much higher than that of 3−6, but the corresponding phosphorescent spectral intensity is much weaker. Using these cyclometalated Ru metal complexes as a prototype, our experimental results and theoretical analysis draw attention to the fact that, for complexes 1 and 2, the weaker spin−orbit coupling present within these molecules reduces the T1−S0interaction, from which the thermally activated radiationless deactivation may take place. This, in combination with the much smaller3MLCT contribution than that observed in 3−6, rationalizes the lack of room-temperature emission for complexes 1 and 2.
1. Introduction
One of the important research subjects toward organic light emitting diodes (OLEDs) is the development of phospho-rescent materials that emit all three primary colors for full-color displays. This approach leads to the essentiality of preparing phosphors incorporating second- and third-row transition-metal complexes.1 The associated strong
spin-orbit coupling in heavy metals would promote singlet-to-triplet intersystem crossing as well as enhance the subsequent radiative transition from the triplet to the ground state, giving
good phosphorescence efficiency for these metal complexes. Among the phosphorescent complexes, green-emitting com-plexes2have been known for years and were fabricated as
OLED components with∼100% internal quantum efficiency, while the red-emitting complexes are also accessible through judicious choices of chelate chromophores to lower the energy gap and extend the triplet-state lifetime3as well as
†Dedicated to Prof. John R. Shapley on the occasion of his 60th birthday. * To whom correspondence should be addressed. E-mail: [email protected] (P.-T.C.), [email protected] (Y.C.).
(1) (a) Holder, E.; Langeveld, B. M. W.; Schubert, U. S. AdV. Mater. 2005, 17, 1109. (b) Hwang, F.-M.; Chen, H.-Y.; Chen, P.-S.; Liu, C.-S.; Chi, Y.; Shu, C.-F.; Wu, F.-I.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H. Inorg. Chem. 2005, 44, 1344. (c) Tung, Y.-L.; Lee, S.-W.; Chi, Y.; Chen, L.-S.; Shu, C.-F.; Wu, F.-I.; Carty, A. J.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H. AdV. Mater. 2005, 17, 1059. (d) Chou, P.-T.; Chi, Y. Eur. J. Inorg. Chem. 2006, 3319 (DOI 10.1002/ejic.200600364).
Inorg. Chem. 2006, 45, 8041
−
8051
10.1021/ic060066g CCC: $33.50 © 2006 American Chemical Society Inorganic Chemistry, Vol. 45, No. 20, 2006
8041
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
to circumvent their intrinsic obstacle, namely, the rapid nonradiative deactivation predicted by the energy gap law.4
Presently, researchers have turned their attention to the preparation of the remaining blue-emitting phosphorescent complexes.5This task, however, is even more difficult to
achieve than those of the other two cases. One major challenge lies in the selection of suitable chelate ligands that are able to form complexes with sufficiently large ligand-centered ππ* transition energies and/or metal-to-ligand
charge-transfer (MLCT) energies. Such an approach might inevitably raise the ligand-centered transition (or MLCT) to a region very close to or even higher than the metal-centered dd states (or ligand-field, LF, states), such that a very efficient radiationless decay pathway may take place through a shallow potential energy surface or a possible T1-S0
inter-section due to the weakness of the metal-ligand bonds. To circumvent this obstacle, a few attempts have been made through the use of strong field ancillary ligands such as CO or cyanide with an aim to increase the dd transition.6
This, in combination with the incorporation of third-row metal elements, further strengthens the metal-ligand bond-ing.7More recently, a series of blue-emitting pyridyl azolate
osmium carbonyl complexes8as well as iridium complexes
with 2,4-difluorophenylpyridyl, pyrazolyl, and even N-heterocyclic carbene ligands have been reported.9 These
exquisite works demonstrate the feasibility of achieving a
saturated blue color or even near-UV phosphorescence.10
Despite this perspective, however, almost all of these blue phosphorescent complexes are inevitably subject to a pro-nounced decrease in the luminance efficiency at room temperature.
Because both strong-field metal elements and ligands are selected in assembling this class of metal complexes, it is reasonable to expect that the metal-centered dd state will be inaccessible from the lowest emissive triplet state. Thus, the dominant radiationless deactivation generalized by a quench-ing mechanism incorporatquench-ing dd transition may be ground-less.6,11As such, the call for alternative, convincing
expla-nations to account for this ubiquitous observation is urgent. Bearing this challenge in mind, we then made an assiduous effort to design and synthesize a new series of RuIIcomplexes
possessing cyclometalated chromophores and other strong-field ligands, aimed at probing the radiationless pathways in correlation with their chemical structures.12It is notable
that RuIIcomplexes are well suited for this approach mainly
because of their relatively small ligand field and weaker metal-ligand bonding. This, in combination with its less heavy atom effect and hence the weaker spin-orbit coupling, leads us to believe that, under the same ligand configuration, the induction of radiationless transition in RuIIcomplexes is
expected to be more drastic than that of the third-row metal congeners. Accordingly, it becomes more plausible to explore the undermining factors causing such intriguing phenomena, i.e., loss of the emission intensity, from fundamental aspects. As an equally important issue, once the radiationless channels are inhibited, the prevailing of RuIIcomplexes to the
third-row transition-metal complexes toward OLED application are apparently due to the lower cost and greater abundance of Ru.
2. Experimental Section
General Information and Materials. Elemental analyses and
mass spectroscopy (operating in fast atom bombardment, FAB, mode) were carried out at the NSC Regional Instrument Centre at National Chiao Tung University, Hsinchu, Taiwan. 1H and 13C
NMR spectra were recorded on a Varian Mercury 400 or an Inova 500-MHz instrument; chemical shifts are quoted with respect to internal standard Me4Si. Details of the fabrication and
characteriza-tion of electroluminescent devices are as reported previously.1bAll
synthetic manipulations were performed under a N2 atmosphere,
while solvents were used as received. Benzo[h]quinoline (BQ) was purchased from TCI Japan, while dibenzo[f,h]quinoxaline (DBQ) was prepared from condensation of phenanthrene-9,10-dione with ethylenediamine.13The RuIImetal complex [Ru(CO)
2(BQ)2] (1) was (2) (a) Hua, F.; Kinayyigit, S.; Cable, J. R.; Castellano, F. N. Inorg. Chem.
2005, 44, 471. (b) Huang, W.-S.; Lin, J. T.; Chien, C.-H.; Tao, Y.-T.; Sun, S.-S.; Wen, Y.-S. Chem. Mater. 2004, 16, 2480. (c) Tokito, S.; Iijima, T.; Tsuzuki, T.; Sato, F. Appl. Phys. Lett. 2003, 83, 2459. (d) Lo, S.-C.; Namdas, E. B.; Burn, P. L.; Samuel, I. D. W.
Macro-molecules 2003, 36, 9721. (e) Adachi, C.; Baldo, M. A.; Forrest, S.
R.; Thompson, M. E. Appl. Phys. Lett. 2000, 77, 904.
(3) (a) Tung, Y.-L.; Lee, S.-W.; Chi, Y.; Tao, Y.-T.; Chien, C.-H.; Cheng, Y.-M.; Chou, P.-T.; Peng, S.-M.; Liu, C.-S. J. Mater. Chem. 2005,
15, 460. (b) Anthopoulos, T. D.; Frampton, M. J.; Namdas, E. B.;
Burn, P. L.; Samuel, I. D. W. AdV. Mater. 2004, 16, 557. (c) Tsuboyama, A.; Iwawaki, H.; Furugori, M.; Mukaide, T.; Kamatani, J.; Igawa, S.; Moriyama, T.; Miura, S.; Takiguchi, T.; Okada, S.; Hoshino, M.; Ueno, K. J. Am. Chem. Soc. 2003, 125, 12971. (d) Adachi, C.; Baldo, M. A.; Forrest, S. R.; Lamansky, S.; Thompson, M. E.; Kwong, R. C. Appl. Phys. Lett. 2001, 78, 1622.
(4) (a) Kober, E. M.; Caspar, J. V.; Lumpkin, R. S.; Meyer, T. J. J. Phys.
Chem. 1986, 90, 3722. (b) Perkins, T. A.; Pourreau, D. B.; Netzel, T.
L.; Schanze, K. S. J. Phys. Chem. 1989, 93, 4511.
(5) (a) Yeh, S.-J.; Wu, M.-F.; Chen, C.-T.; Song, Y.-H.; Chi, Y.; Ho, M.-H.; Hsu, S.-F.; Chen, C.-H. AdV. Mater. 2005, 17, 285. (b) Ren, X.; Li, J.; Holmes, R. J.; Djurovich, P. I.; Forrest, S. R.; Thompson, M. E. Chem. Mater. 2004, 16, 4743. (c) Karatsu, T.; Nakamura, T.; Yagai, S.; Kitamura, A.; Yamaguchi, K.; Matsushima, Y.; Iwata, T.; Hori, Y.; Hagiwara, T. Chem. Lett. 2003, 32, 886. (d) Holmes, R. J.; D’Andrade, B. W.; Forrest, S. R.; Ren, X.; Li, J.; Thompson, M. E.
Appl. Phys. Lett. 2003, 83, 3818. (e) Tanaka, I.; Tabata, Y.; Tokito,
S. Chem. Phys. Lett. 2004, 400, 86.
(6) (a) Anderson, P. A.; Keene, F. R.; Meyer, T. J.; Moss, J. A.; Strouse, G. F.; Treadway, J. A. J. Chem. Soc., Dalton Trans. 2002, 3820. (b) Koike, K.; Okoshi, N.; Hori, H.; Takeuchi, K.; Ishitani, O.; Tsubaki, H.; Clark, I. P.; George, M. W.; Johnson, F. P. A.; Turner, J. J. J.
Am. Chem. Soc. 2002, 124, 11448.
(7) (a) Lee, C.-L.; Das, R. R.; Kim, J.-J. Chem. Mater. 2004, 16, 4642. (b) Nazeeruddin, M. K.; Humphry-Baker, R.; Berner, D.; Rivier, S.; Zuppiroli, L.; Graetzel, M. J. Am. Chem. Soc. 2003, 125, 8790. (8) (a) Yu, J.-K.; Hu, Y.-H.; Cheng, Y.-M.; Chou, P.-T.; Peng, S.-M.;
Lee, G.-H.; Carty, A. J.; Tung, Y.-L.; Lee, S.-W.; Chi, Y.; Liu, C.-S.
Chem.sEur. J. 2004, 10, 6255. (b) Wu, P.-C.; Yu, J.-K.; Song,
Y.-H.; Chi, Y.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H. Organometallics 2003, 22, 4938.
(9) Sajoto, T.; Djurovich, P. I.; Tamayo, A.; Yousufuddin, M.; Bau, R.; Thompson, M. E.; Holmes, R. J.; Forrest, S. R. Inorg. Chem. 2005,
44, 7992.
(10) (a) Li, J.; Djurovich, P. I.; Alleyne, B. D.; Tsyba, I.; Ho, N. N.; Bau, R.; Thompson, M. E. Polyhedron 2004, 23, 419. (b) Yang, C.-H.; Li, S.-W.; Chi, Y.; Cheng, Y.-M.; Yeh, Y.-S.; Chou, P.-T.; Lee, G.-H.; Wang, C.-H.; Shu, C.-F. Inorg. Chem. 2005, 44, 7770. (c) Li, J.; Djurovich, P. I.; Alleyne, B. D.; Yousufuddin, M.; Ho, N. N.; Thomas, J. C.; Peters, J. C.; Bau, R.; Thompson, M. E. Inorg. Chem. 2005, 44, 1713.
(11) Zalis, S.; Farrell, I. R.; Vlcek, A. J. Am. Chem. Soc. 2003, 125, 4580. (12) Tung, Y.-L.; Chen, L.-S.; Chi, Y.; Chou, P.-T.; Cheng, Y.-M.; Li, E. Y.; Lee, G.-H.; Shu, C.-F.; Wu, F.-I.; Carty, A. J. AdV. Funct. Mater. 2006, 16, 1615.
(13) Steel, P. J.; Caygill, G. B. J. Organomet. Chem. 1990, 395, 359.
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
prepared using Ru3(CO)12and BQ in a mixture of
1,2-dimethoxy-ethane and octane following literature procedures.14
Cyclic voltammetry (CV) measurements were performed using a BAS 100 B/W electrochemical analyzer. The oxidation and reduction measurements were recorded, respectively, in anhydrous CH2Cl2and anhydrous THF solutions containing 0.1 M TBAPF6
as the supporting electrolyte, at a scan rate of 100 mV s-1. The potentials were measured against an Ag/Ag+ (0.01 M AgNO3)
reference electrode with the ferrocene/ferrocenium couple as the internal standard.
Steady-state absorption and emission spectra were recorded by a Hitachi U-3310 spectrophotometer and an Edinburgh FS920 fluorimeter, respectively. Emission quantum yields were measured at excitation wavelength λexc ) 480 nm in CH2Cl2 at room
temperature. 4-(Dicyanomethylene)-2-methyl-6-[p-(dimethylamino)-styryl]-4H-pyran (DCM,Φr) 0.44) in a methanol solution was
used as the reference, and the equation
was used to calculate the emission quantum yields, whereΦsand Φrare the quantum yields of the unknown and reference samples,
η is the refractive index of the solvent, Arand Asare the absorbance
of the reference and the unknown samples at the excitation wavelength, and Isand Irare the integrated areas under the emission
spectra of interest, respectively. For the phosphorescence lifetime measurements in the microsecond region, a third harmonic of an Nd:YAG laser of 355 nm was used as the excitation source. For this approach, emission decay was detected with a photomultiplier tube and averaged over 500 shots using an oscilloscope, and laser energy was reduced to e 1 mJ pulse-1to prevent possible photo-chemical decomposition. For the nanosecond lifetime measure-ments, the fundamental train of pulses from a Ti-sapphire oscillator (82 MHz, Spectra Physics) was used to produce second harmonics (375-425 nm) as an excitation light source. The signal was detected by a time-correlated single-photon-counting system (Edinburgh OB 900-L).
Synthesis of [Ru(CO)2(DBQ)2] (2). A mixture of Ru3(CO)12
(111 mg, 017 mmol), dibenzo[f,h]quinoline (250 mg, 1.09 mmol), 1,2-dimethoxyethane (2 mL), and octane (15 mL) was heated at reflux for 12 h, after which time a dark-brown precipitate had deposited from the brown solution. The solvent was then removed using a rotary evaporator, and the residue was treated with a mix-ture of 1 mL of diethylamine and 3 mL of methanol to remove the probable side product [Ru4H4(CO)12]. The remaining solid was
collected by filtration and washed with methanol (3 × 2 mL) and hexane (3× 3 mL) three times. The expected DBQ complex
2 was obtained as a yellow-brown solid (83 mg, 0.34 mmol,
80%).
Spectral Data of 2. MS (FAB,102Ru): observed m/z
[assign-ment] 616 [M+]. IR (CH2Cl2): ν(CO) 2015 (s), 1948 (s) cm-1.1H
NMR (400 MHz, CD2Cl2): δ 9.04 (d, 2H, J ) 7.6 Hz), 8.73 (d,
2H, J ) 8.4 Hz), 8.55 (d, 2H, J ) 7.2 Hz), 8.45 (d, 2H, J ) 8.0 Hz), 8.37 (d, 2H, J ) 2.8 Hz), 7.90-7.81 (m, 4H), 7.77 (d, 2H, J ) 2.8 Hz), 7.70 (t, 2H, J ) 7.6 Hz). Anal. Calcd for C34H18N4O2
-Ru: C, 66.34; H, 2.95; N, 9.10. Found: C, 66.37; H, 3.01; N, 8.95.
Preparation of [Ru(CO)(BQ)2(PPh2Me)] (3). A 50-mL reaction flask was first charged with 1 (300 mg, 0.58 mmol) and 20 mL of
anhydrous diethylene glycol monoethyl ether (DGME) and then immersed into an oil bath maintained at∼130°C. Freshly sublimed Me3NO (109 mg, 1.46 mmol) dissolved in 12 mL of DGME was
dropwise added over a period of 5 min, followed by the addition of PPh2Me (500µL, 2.63 mmol). The resulting mixture was then
stirred at 160°C for 24 h. Finally, the solvent was evaporated under vacuum and the residue placed into 30 mL of CH2Cl2. The solution
was washed with distilled water (150 mL× 2), dried over MgSO4,
and evaporated to dryness. The residue was purified by silica gel column chromatography using ethyl acetate/hexane (1:5) as the eluent. Recrystallization was conducted from a mixture of CH2Cl2
and methanol at room temperature, giving yellow crystalline solids (238 mg, 0.348 mmol) in 60% yield. Complex [Ru(CO)(BQ)2
-(PPhMe2)] (4) was prepared in 80% yield using similar procedures.
Spectral Data for 3. MS (FAB,102Ru): observed m/z
[assign-ment] 686 [M+]. IR (CH2Cl2): ν(CO) 1909 (s) cm-1. 1H NMR (500 MHz, acetone-d6): δ 9.01 (dd, 1H, J ) 5.0 and 1.0 Hz), 8.52 (dd, 1H, J ) 1.5 and 1.0 Hz), 8.30-8.29 (m, 1H), 7.85 (d, 1H, J ) 7.8 Hz), 7.78-7.67 (m, 5H), 7.56-7.40 (m, 7H), 7.27 (d, 1H, J ) 7.5 Hz) 7.02-6.81 (m, 6H), 6.70 (dd, 1H, J ) 8.0 and 5.5 Hz), 6.64-6.60 (m, 2H), 1.32 (d, 3H, J ) 6.5 Hz).31P NMR (202 MHz,
acetone-d6): δ 18.35 (s). Anal. Calcd for C40H29N2OPRu: C, 70.06;
H, 4.26; N, 4.09. Found: C, 69.75; H, 4.57; N, 4.01.
Spectral Data for 4. MS (FAB,102Ru): observed m/z
[assign-ment] 624 [M+]. IR (CH2Cl2): ν(CO) 1909 (s) cm-1. 1H NMR (400 MHz, CD2Cl2): δ 8.42 (dt, 1H, J ) 5.6 and 1.6 Hz), 8.35 (d, 1H, J ) 5.2 Hz), 8.26 (dd, 1H, J ) 6.6 and 1.6 Hz), 7.91 (d, 1H, J ) 8.0 Hz), 7.82 (d, 1H, J ) 8.8 Hz), 7.00 (d, 1H, J ) 8.8 Hz), 7.64-7.54 (m, 3H), 7.47 (d, 1H, J ) 8.8 Hz), 7.30-7.23 (m, 7H), 7.13 (dt, 1H, J ) 4.8 and 1.6 Hz), 6.98 (td, 1H, J ) 7.4 and 1.2 Hz), 6.95-6.83 (m, 2H), 1.56 (d, 3H, J ) 7.2 Hz), 0.61 (d, 3H, J ) 6.4 Hz).31P NMR (202 MHz, CDCl 3): δ 18.32 (s). Anal. Calcd
for C35H27N2OPRu: C, 67.41; H, 4.36; N, 4.49. Found: C, 67.52;
H, 4.46; N, 4.74.
Preparation of [Ru(CO)(DBQ)2(PPh2Me)] (5) and
[Ru(CO)-(DBQ)2(PPhMe2)] (6). The synthesis procedures were essentially identical with those described for 3, using similar ratios of 1, freshly sublimed Me3NO, and the phosphine ligand PPh2Me or PPhMe2.
Dark-red 5 and 6 were obtained from a mixture of CH2Cl2 and
methanol at room temperature. Yield: 68-70%.
Spectral Data for 5. MS (FAB,102Ru): observed m/z
[assign-ment] 788 [M+]. IR (CH2Cl2): ν(CO) 1928 (s) cm-1. 1H NMR (400 MHz, CD2Cl2): δ 9.22 (dd, 1H, J ) 7.8 and 1.6 Hz), 8.91 (d, 1H, J ) 8.0 Hz), 8.77-8.75 (m, 2H), 8.69 (d, 1H, J ) 8.0 Hz), 8.52 (d, 1H, J ) 7.6 Hz), 8.41 (d, 1H, J ) 7.6 Hz), 8.23 (d, 1H, J ) 8.0 Hz), 7.98 (d, 1H, J ) 8.0 Hz), 7.86-7.44 (m, 11H), 7.10-7.06 (m, 2H), 6.91-6.90 (m, 1H), 6.76-6.69 (m, 3H), 6.56-6.52 (m, 2H), 1.41 (d, 3H, J ) 7.5 Hz).31P NMR (202 MHz, CDCl 3):
δ 18.26 (s). Anal. Calcd for C46H31N4OPRu: C, 70.13; H, 3.97;
N, 7.11. Found: C, 69.86; H, 3.71; N, 6.41.
Spectral Data for 6. MS (FAB,102Ru): observed m/z
[assign-ment] 726 [M+]. IR (CH2Cl2): ν(CO) 1924 (s) cm-1. 1H NMR (500 MHz, CD2Cl2): δ 9.21 (dd, 1H, J ) 7.8 and 1.0 Hz), 8.99 (dd, 1H, J ) 7.8 and 1.0 Hz), 8.86 (d, 1H, J ) 2.4 Hz), 8.81 (d, 1H, J ) 8.3 Hz), 8.70 (d, 1H, J ) 3.0 Hz), 8.60 (d, 1H, J ) 8.5 Hz), 8.45 (d, 1H, J ) 7.0 Hz), 8.32 (d, 1H, J ) 8.0 Hz), 8.25 (d, 1H, J ) 2.5 Hz), 8.00 (d, 1H, J ) 7.5 Hz), 7.86-7.67 (m, 5H), 7.49 (dd, 1H, J ) 2.5 and 1.0 Hz), 7.22-6.89 (m, 7H), 1.76 (d, 3H, J ) 7.5 Hz), 1.03 (d, 3H, J ) 7.5 Hz).31P NMR (202 MHz,
CD2Cl2): δ 18.22 (s). Anal. Calcd for C41H29N4OPRu: C, 67.85;
H, 4.03; N, 7.72. Found: C, 68.17; H, 4.36; N, 7.94.
X-ray Structural Measurements. Single-crystal X-ray analysis
was measured on a Bruker SMART Apex CCD diffractometer using (14) (a) Patrick, J. M.; White, A. H.; Bruce, M. I.; Beatson, M. J.; Black,
D. S. C.; Deacon, G. B.; Thomas, N. C. J. Chem. Soc., Dalton Trans. 1983, 2121. (b) Bruce, M. I.; Liddell, M. J.; Pain, G. N. Inorg. Synth. 1989, 26, 171. Φs) Φr
(
ηs2ArIs ηr 2 AsIr)
Neutral Ru -Based Emitting Materials
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
µ(Mo KR) radiation (λ ) 0.710 73 Å). The data collection was executed using the SMART program. Cell refinement and data reduction were made by the SAINT program. The structure was determined using the SHELXTL/PC program and refined using full-matrix least squares. All non-H atoms were refined anisotropically, whereas H atoms were placed at the calculated positions and included in the final stage of refinements with fixed parameters. The crystallographic refinement parameters of 3 are summarized in Table 1, and the selective bond distances and angles are listed in Table 2.
Theoretical Approach. Time-dependent density functional
theory (TDDFT)15calculations using the B3LYP16functional were
performed based on the structures obtained from single-crystal X-ray diffraction data. A “double-ζ” quality basis set consisting of Hay and Wadt’s effective core potentials (ECPs; LANL2DZ)17 was
employed for Ru atoms and the 6-31G* basis set18for H, C, N,
and P atoms. A relativistic ECP replaced the inner-core electrons of RuII, leaving the outer-core (4s24p6) electrons and the 4d6valence
electrons. Typically, the lowest 10 triplet and 10 singlet roots of
the nonhermitian eigenvalue equations were obtained to determine the vertical excitation energies. Oscillator strengths were deduced from the dipole transition matrix elements (for singlet states only). The excited-state TDDFT calculations were carried out using Gaussian03, as described in our previous publications.19 3. Results
Synthesis and Characterization. It has been reported that the direct reaction of BQ with Ru3(CO)12in a mixture of
1,2-dimethoxyethane and octane yielded a cyclometalated RuII complex 1, which contains two mutually orthogonal
C,N-chelating BQ ligands (see Scheme 1).14 The X-ray
structural analysis of 1 revealed that the central Ru atom has approximate octahedral coordination, for which the two mutually cis CO ligands lie trans to the N atoms of the cyclometalated BQ ligands and the cyclometalated C atoms are located at the opposite dispositions.
Moreover, treatment of a more conjugated heteroaromatic DBQ with Ru3(CO)12 under similar conditions led to the
isolation of a second derivative complex 2. The structure of 2 was readily characterized by NMR analyses, the results of which revealed several multiplets betweenδ 9.94 and 7.70
due to the aromatic proton resonances, while its final structural identification was achieved by the observation of two sharp IRν(CO) stretching bands at 2015 and 1948 cm-1, attributed to the cis-oriented CO ligands. To the best of our understanding, complex 2 is one of the few known examples to contain cyclometalated DBQ ligands. Well-known ex-amples include the recently reported orange-emitting complex [Ir(DBQ)2(acac)], for which the N atoms in two DBQ ligand
chelates are located at the mutual trans disposition.20In sharp
contrast, however, RuII complexes 1 and 2 possess a cis
orientation between the two DBQ N atoms.
For preparation of the luminescent complexes (vide infra), phosphine-substituted RuIIderivatives 3-6 were synthesized
in two consecutive steps, employing an excess of decarbo-nylation reagent Me3NO, followed by addition of the
phos-phine ligand. The resulting mixture was then stirred at 160 °C for 24 h, and the product was isolated by routine chromatography and recrystallization. These metal complexes were found to be highly soluble in most organic solvents such as acetone and CHCl3and have been characterized using
various spectroscopic methods including FAB MS, IR, and
1H and31P NMR (see the Experimental Section). Importantly,
these complexes showed only one sharp IRν(CO) stretching
signal in the range 1928-1909 cm-1 and are in good agreement with the retention of a single carbonyl ligand. It is also notable that the amount of Me3NO added in the initial
reaction mixture is sufficient to remove both CO ligands in the parent complexes 1 and 2; this observation implies that the second, remaining CO ligand is essentially inert to the Me3NO reagent.
Figure 1 depicts the ORTEP diagram of 3, showing octa-hedral arrangement around the RuIImetal center. However,
to our surprise, the BQ ligand orientation is distinctive from
(15) (a) Jamorski, C.; Casida, M. E.; Salahub, D. R. J. Chem. Phys. 1996, 104, 5134. (b) Petersilka, M.; Grossmann, U. J.; Gross, E. K. U. Phys. ReV. Lett. 1996, 76, 1212. (c) Bauernschmitt, R.; Ahlrichs, R.; Hennrich, F. H.; Kappes, M. M. J. Am. Chem. Soc. 1998, 120, 5052. (d) Casida, M. E. J. Chem. Phys. 1998, 108, 4439. (e) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109, 8218.
(16) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (b) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(17) (a) Hay, P. J.; Wadt, R. W. J. Chem. Phys. 1985, 82, 270. (b) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284. (c) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(18) Hariharan, P. C.; Pople, J. A. Mol. Phys. 1974, 27, 209.
(19) Yu, J.-K.; Cheng, Y.-M.; Hu, Y.-H.; Chou, P.-T.; Chen, Y.-L.; Lee, S.-W.; Chi, Y. J. Phys. Chem. B 2004, 108, 19908.
(20) Duan, J.-P.; Sun, P.-P.; Cheng, C.-H. AdV. Mater. 2003, 15, 224. Table 1. X-ray Structural Data of Complex 3
empirical formula C40H29N2OPRu
mol wt 685.69
cryst syst monoclinic
space group P21/c T, K 150(1) a, Å 9.5823(5) b, Å 17.7987(10) c, Å 18.2986(10) β, deg 96.148(1) V, Å3 3102.9(3) Z 4 Dc, g cm-3 1.468 F(000) 1400 µ(Mo KR), mm-1 0.593 cryst size, mm 0.30× 0.25 × 0.20 h, k, l ranges -11 < h < 12, -23 < k < 23, -23 < l < 23 reflns collected 25 478
indep reflns 7127 [R(int) ) 0.0375]
data/restraints/parameters 7127/0/407
GOF on F2 1.110
R1, wR2 with I > 2σ(I) 0.0348, 0.0840 D map, max/min, e/Å-3 0.510/-0.384
Table 2. Selected Bond Distances (Å) and Angles (deg) for Complex 3
Ru-P(1) 2.3789(6) Ru-C(1) 1.833(3) Ru-C(2) 2.079(2) Ru-C(15) 2.041(2) Ru-N(1) 2.208(2) Ru-N(2) 2.151(2) C(1)-O(1) 1.157(3) ∠P(1)-Ru-C(2) 176.09(6) ∠N(1)-Ru-C(15) 165.80(8) ∠N(2)-Ru-C(1) 170.34(9) ∠N(1)-Ru-C(2) 78.58(8) ∠N(2)-Ru-C(15) 80.10(8) ∠Ru-C(1)-O(1) 176.1(2)
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
that of its parent complex 1, namely, one cyclometalated C atom is shifted to a new position trans to the PPh2Me
phosphine ligand, while the second C atom resides trans to the N atom of the second BQ ligand. This unique structural behavior demonstrates that one BQ ligand undergoes a 180° rotation around the RuIIcenter during the course of phosphine
substitution. A similar coordination arrangement was ob-served from the X-ray structural determination of a PPh3
derivative prepared using UV irradiation.21
For the metric parameters, in 3, the Ru-N(1) distance [2.208(2) Å] of the first BQ ligand is obviously longer than the respective Ru-N(2) distance [2.151(2) Å] and the Ru-N distances observed in 1 (2.148 and 2.161 Å), showing a large labilization effect attributed to its trans Ru-C(15)σ bond
versus that of the trans CO ligand exerted to the second BQ
ligand. Apparently, the trans competition through metal-C
σ bonding is more pronounced in this class of complexes,
although the CO ligand was among the best π-acceptor
ligands but has failed to impose a sufficient amount of trans effect compared to the metal-Cσ bond that exerted to the
first BQ ligand. Moreover, the Ru-C distance of the second BQ ligand [Ru-C(15) ) 2.041(2) Å] is shorter than that of the other Ru-C bond [Ru-C(2) ) 2.079(2) Å], which is, in turn, much shorter than those of the mutually trans-oriented Ru-C bonds of its parent complex 1 [2.12-2.13(1) Å].12
Again, this increase in the Ru-C bond distances could be attributed to the bonding competition exerted by their trans ligands, for which the metal-ligand bond strengths follow the order of, i.e., cyclometalated C atom (C) > phosphine (P) > pyridine (N).
Electrochemistry. The redox potentials of the RuII
com-plexes were determined from cyclic voltammograms, and the data are summarized in Table 3. It is believed that the oxidation occurred mainly at the metal site, with minor contributions from the cyclometalated chelate and other ancillary ligands. For 3 and 4 possessing BQ ligands, an oxidation potential at 0.13 V was acquired in CH2Cl2, while
DBQ complexes 5 and 6 exhibited a higher oxidation potential at 0.35 V due to the presence of the quinoxaline fragment, which reduced the electron donation to the metal atom with the presence of an additional N atom. Moreover, variation of phosphine has caused almost no change to the oxidation potential. Finally, the corresponding parents 1 and 2 give an irreversible oxidation half-wave at 0.74 and 0.95 V, which are in good agreement with their electron-deficient nature induced by the carbonyl ligands.
As for the reduction behavior, the dicarbonyl complexes 1 and 2 give three closely spaced, reversible reduction peaks in the THF solution, a result of two one-electron reductions at each of the cyclometalated ligands as well as the possible reduction at the metal site to give RuIspecies. It is possible
that this metal reduction is coupled with the [Ru(CO)2]
fragment, allowing an easy delocalization of electron density
(21) Zhang, Q.-F.; Cheung, K.-M.; Williams, I. D.; Leung, W.-H. Eur. J.
Inorg. Chem. 2005, 4780.
Scheme 1
Figure 1. ORTEP diagram of 3 with thermal ellipsoids shown at the 50% probability level.
Neutral Ru -Based Emitting Materials
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
to the CO ligands. However, no further attempt was made to support this hypothesis. On the other hand, the phosphine-substituted complexes 3-6 exhibit only two reversible reduction signals between the narrow ranges of -2.57 to -2.92 and -2.07 to -2.32 V for the pair of BQ and DBQ complexes, respectively. This finding led us to propose that the observed reduction was strongly associated with both BQ and DBQ groups, while the ancillary phosphine ligand has very little influence on the reduction potential.
Photophysical Measurements. The photophysical proper-ties of these RuIIcomplexes can be systematically varied by
modifying the cyclometalated ligand chromophores and the ancillary ligands; the respective data are listed in Table 3. It is notable that the lowest absorption of their parent BQ and DBQ complexes 1 and 2 appear atλmax∼395 and 420 nm,
respectively, which are tentatively assigned to the ligand-centeredππ*, mixed with small amounts of MLCT
transi-tions. The slightly higher extinction coefficient and occur-rence of multiple peak maxima in the higher energy region for the DBQ complex 2 are obviously caused by the extended
π conjugation of the DBQ ligands.
Moreover, the respective PPh2Me- and PPhMe2-substituted
BQ complexes showed the occurrence of red-shifted, less intense transitions atλmax456 nm for 3 and 458 nm for 4, which are tentatively assigned to the transition incorporat-ing a state mixincorporat-ing among sincorporat-inglet and triplet metal-ligand charge transfer (1MLCT and3MLCT) and, to a certain extent,
the intraligand 3ππ transitions. The close energetics and
absorptivity between 1MLCT and 3MLCT bands suggest
that the3MLCT transition, induced by the spin-orbit
cou-pling and the proximal energy levels with respect to1MLCT,
is greatly enhanced and becomes partially allowed.22
Following the same principle, the MLCT transitions of both complexes 5 (506 nm) and 6 (513 nm) can be assigned, for which the more electron-rich PPhMe2-substituted 6 has
induced a slightly further bathochromic shift compared with
the slightly electron-deficient RuII metal core in 5. Further
support for these assignments is given in the Discussion section.
Although they have the largest S0f S1 absorption gap
among complexes 1-6, to our surprise, both 1 (Φ∼ 0) and
2 (Φ e 5.0 × 10-4) are nearly nonemissive in a
room-temperature CH2Cl2solution, as well as showing significant
temperature-dependent emissive behavior. For example, upon cooling of the solution from 298 to 203 K, the emission yield gradually increased from null to 8× 10-4for 1. The emission possesses a well-resolved vibronic progression with a 0-0 transition peak at∼ 485 nm. Both 1 and 2 revealed decent emission signals in the 77 K solid CH2Cl2matrix, and data
are depicted in Figure 2 and Table 3.
In sharp contrast to their parent complexes, i.e., 1 and 2, complexes 3-6 showed strong to weak luminescence in a room-temperature, degassed CH2Cl2 solution with peak
wavelengths located at 571, 575, 655, and 656 nm, respec-tively (see Figure 2). Owing to the rapid quenching of luminescence in the aerated solution for, e.g., complex 3 (not shown here), the assignment of the emission mainly origi-nating from the triplet manifold is unambiguous. Moreover, for complexes 3-6, the partial overlap between the emission
(22) (a) Kavitha, J.; Chang, S.-Y.; Chi, Y.; Yu, J.-K.; Hu, Y.-H.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H.; Tao, Y.-T.; Chien, C.-H.; Carty, A. J.
AdV. Funct. Mater. 2005, 15, 223. (b) Chang, S.-Y.; Kavitha, J.; Li,
S.-W.; Hsu, C.-S.; Chi, Y.; Yeh, Y.-S.; Chou, P.-T.; Lee, G.-H.; Carty, A. J.; Tao, Y.-T.; Chien, C.-H. Inorg. Chem. 2006, 45, 137.
Table 3. Photophysical and Electrochemical Properties for Complexes 1-6 in Degassed CH2Cl2at Room Temperature λmax abs/nm (× 10-3) λ max em/nm Φ k obs(s-1) kr(s-1) E1/2 ox E 1/2 red 1 320 (7.2), 381 (3.8), 395 (4.2) -[485, 521, 563, 617 (sh)]a -[4.1 × 10 2]a 0.74 [irr] -2.41, -2.64, -2.77 2 332 (14), 376 (7), 421 (3.8) 562 [526]a ∼5.0 × 10-4 3.7× 106[7.9× 103]a ∼1.9 × 103 0.95 [irr] -1.91, -2.11, -2.34 3 310 (14), 370 (7.4), 456 (2.4) 571 [553]a(539)b 0.24 1.4× 105[3.8× 103]a (2.5× 107)b 3.3× 104 0.13 -2.57, -2.92 4 312 (18), 378 (9.2), 458 (3) 575 [556]a(542)b 0.18 1.2× 105[5.5× 103]a (2.0× 107)b 2.1× 104 0.13 -2.59, -2.86 5 344 (13), 359 (12), 387 (8), 506 (1.4) 655 [632]a(642)b 0.008 4.2× 106[1.7× 104]a (1.4× 107)b 3.3× 104 0.35 -2.07, -2.32 6 343 (11), 362 (16), 387 (11), 513 (1.8) 656 [633]a(645)b 0.004 7.1× 106[1.8× 104]a 2.9× 104 0.35 -2.09, -2.32 aData in square brackets are measured in a CH
2Cl2matrix at 77 K, and “sh” denotes “shoulder”.bData in parentheses are measured in a thin solid film at room temperature.
Figure 2. UV-vis absorption and emission spectra of complexes 1-6 in a CH2Cl2solution, for which the emissions of 1 and 2 were taken in a 77 K matrix, while those of 3-6 were recorded at room temperature.
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
onset and the lowest-energy absorption bands, in combination with a broad, structureless spectral profile, leads us to conclude that the luminescence originates primarily from the
3MLCT state.23In comparison to the BQ complex 3, its DBQ
counterpart 5 exhibits a∼84-nm bathochromic shift in λmax,
the result of which can qualitatively be rationalized by an increase ofπ conjugation and the incorporation of a C4N2
hexagon for the DBQ heterocyclic ligands.24In comparison
to 3, possessing PPh2Me, a very slightly bathochromic shift
of only ∼4 nm was observed for the PPhMe2-substituted
derivative 4. A similar small red shift was resolved between 5 (PPh2Me) and 6 (PPhMe2; see Table 3). This is apparently
caused by the decrease of the π-accepting strength of
PPhMe2versus the PPh2Me ligand. However, the variation
seems significantly lower than those expected based on the electronic properties of phosphine ligands.25One possible
explanation is that the apparent variation viewed from peak maxima is much reduced because of the broad, structureless spectral profiles. It is also noteworthy that the phosphores-cence quantum yields for 5 (8.0× 10-3) and 6 (4.0× 10-3)
(23) Vlcek, A., Jr. Coord. Chem. ReV. 1998, 177, 219.
(24) Song, Y.-H.; Yeh, S.-J.; Chen, C.-T.; Chi, Y.; Liu, C.-S.; Yu, J.-K.; Hu, Y.-H.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H. AdV. Funct. Mater. 2004, 14, 1221.
Figure 3. Selected frontier orbitals of 1 involved in the lower-lying transitions.
Neutral Ru -Based Emitting Materials
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
in CH2Cl2 are significantly lower than those of
yellow-emitting complexes 3 (0.24) and 4 (0.18) (see Table 2). For 3-6, the difference could be qualitatively rationalized using an empirical energy gap law.4Upon a decrease in the energy
gap, rapid quenching may take place via the T1-S0
inter-system crossing through coupling of certain high-frequency vibration motions and subsequent solvent collisional deac-tivation in the fluid solution. In sharp contrast, although the T1-S0 energy gap for both 1 and 2 is larger than that of 3-6, the corresponding phosphorescence is much weaker. Certainly, the results for 1 and 2 are opposite to the empirical energy gap law and must be associated with other mech-anisms of deactivation. Details regarding types of specific modes involved in the radiationless deactivation are elabo-rated in the Discussion section.
4. Discussion
In light of the search for blue phosphorescent materials for fabrication of OLEDs, tuning of the emission color becomes an urgent task, the feasibility of which may be made by systematic variation of the ligand chromophores as well as the ancillary ligands. However, a fundamental question is promptly raised regarding the inferior factors that might induce the rapid radiationless decay processes, giving rise to low quantum efficiency for most of the authentic blue phosphors documented in the literature.8-10Apparently, the
room-temperature nonemissive complex 1 (with a 485-nm emission peak at 77 K) falls in this category. Following the initial population to the low-lying3MLCT or3ππ* excited
states, one widely accepted radiationless deactivation channel may be ascribed to their rapid crossing to the adjacent or even lower metal-centered 3dd state. Because of its
anti-bonding character, the potential energy surface of the dd state, theoretically, is expected to be shallow, and it may intercept with the potential energy surface associated with the ground state and, in an extreme case, may even undergo bond dissociation. The net results should cause rapid energy dissipation through metal-ligand bond stretching. This, in combination with the forbidden nature of the S0 f 3dd transition, further signifies the importance of the3dd state
in manipulating the radiationless pathways.
To gain detailed insights into this fundamental issue, theoretical approaches (TDDFT, see the Experimental Sec-tion) on the photophysical properties were performed for the RuIIcomplexes studied. Figure 3 depicts the features of the
selected occupied and unoccupied frontier orbitals mainly involved in the lower-lying transitions for complex 1, while the descriptions and energy gaps of each transition are listed in Table 4. As shown in Table 4, the lowest singlet S0f S1
transition of 391 nm is in good agreement with the ππ*
transition of 395 nm obtained from its absorption spectra. Likewise, the estimated S0f T1transition occurring at 471
nm is consistent with the resolved 485 nm emission in the 77 K CH2Cl2matrix.
Giving another example of the PPh2Me-substituted
com-plex 3 (see Figure 4 and Table 5), the lowest singlet state (S1) with an energy gap of 457 nm consists of a highest
occupied molecular orbital (HOMO) f lowest unoccupied molecular orbital (LUMO) transition, while the lowest triplet state (T1), with a lower gap of 507 nm, mainly involves
HOMO f LUMO+1 transition, for which the electron densities of the HOMO are located on the C6hexagon of
one BQ ligand trans to the CO ligand and the central Ru atom, whereas those of the LUMO+1 are primarily distrib-uted on the same BQ ligand. The calculated lowest excited singlet and triplet states at 457 and 507 nm, respectively, are also qualitatively in agreement with those (the absorption peak maximum at 456 nm and the phosphorescence onset
of∼510 nm) observed experimentally. We believe that the
small deviation of theoretical prediction from experimental results is mainly due to its limitation in predicting the min-ute nuclear motion during the electronic transitions as well as the negligence of the solvation energy in the current theoretical approach. Qualitative consistency between theo-retical approaches and experimental results is also seen in the rest of the metal complexes, for which the theoretical results of the DBQ complex 5 are revealed in Figure 5 and Table 6. As for a general trend, the S0-S1and S0-T1energy
gaps for complexes filled with dual ancillary CO ligands are much larger than those anchored by one CO and one phosphine, showing the expected hypsochromic shift imposed by the strongerπ-accepting CO ligands.
Theoretical results provide valuable clues to the explora-tion of the correlaexplora-tion between each electronic transiexplora-tion and its corresponding frontier orbital contribution. A similar trend was also resolved for these complexes, in which the S0f T1 transition is dominated by either RuII f BQ (for 1, 3, and 4) or RuII
f DBQ (for 2, 5, and 6) MLCT in combin-ation withππ* transitions within BQ or DBQ ligands. This (25) (a) Vogler, A.; Kunkely, H. Coord. Chem. ReV. 2002, 230, 243. (b)
Tung, Y.-L.; Wu, P.-C.; Liu, C.-S.; Chi, Y.; Yu, J.-K.; Hu, Y.-H.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H.; Tao, Y.; Carty, A. J.; Shu, C.-F.; Wu, F.-I. Organometallics 2004, 23, 3745. (c) Tsubaki, H.; Tohyama, S.; Koike, K.; Saitoh, H.; Ishitani, O. Dalton Trans. 2005, 385.
Table 4. Calculated Energy Levels of the Lower-Lying Transitions of 1
assignment [nm] E [eV] f T1 HOMO f LUMO (+33%) 471.4 2.63 ∼ 0 HOMO-1 f LUMO (+20%) HOMO f LUMO+2 (+9%) HOMO-3 f LUMO (9%) HOMO-1 f LUMO+2 (+7%) HOMO f LUMO+1 (7%) HOMO-4 f LUMO (6%) HOMO-1f LUMO+1 (6%) T2 HOMO f LUMO+1 (+15%) 452.8 2.74 ∼0 HOMO-1 f LUMO+1 (13%) HOMO-3 f LUMO+1 (12%) HOMO-1 f LUMO (10%) HOMO-1 f LUMO+3 (8%) HOMO f LUMO (+7%) HOMO-4 f LUMO+1 (+7%) HOMO f LUMO+3 (+6%) HOMO-2 f LUMO+1 (6%) Singlet States S1 HOMO f LUMO (+75%) 391.4 3.17 0.0083 HOMO-1 f LUMO (13%) HOMO-2 f LUMO (+8%) S2 HOMO-1 f LUMO (+57% 382.0 3.25 0.0149 HOMO f LUMO+1 (+17%) HOMO f LUMO (+11%) HOMO-2 f LUMO (6%)
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
characteristic is somewhat similar to those of Ir(ppy)3 and
its derivatives, which were also attributed to a3ππ* manifold,
which, to a great extent, was mixed with the 3MLCT
character.26 Upon careful examination of the associated
frontier orbitals involved in each transition of our complexes, to our surprise, the calculated lowest-lying T1state and even
the S0f T2or higher energy transitions have no contribution
from the metal-centered dπdσ* and/or ligand-to-metal πdσ* states. Instead, the unoccupied orbitals with dσ*character are ascribed to LUMO+4 for 1 (Figure 3) and even higher for other complexes. For the case of 1, the dπdσ*state is higher in energy than the T1state by 1.5 eV. This gap is so large
that, even under the excited-state geometry relaxation, a chance of the intersection of the potential energy surface between dπdσ*and T1states seems to be rather slim and the
associated dπdσ*triplet state is strictly thermally inaccessible. These results simply discard a generally accepted mechanism
incorporating the 3d
πdσ* (or 3πdσ*) state being responsible for the lack of room-temperature emission for dicarbonyl complexes 1 and 2.
On the other hand, the typical lowest-energy transition in the triplet manifold involves a great extent ofππ* and MLCT
mixing. Because MLCT incorporates a dπf π* transition, its key role in enhancing the spin-orbit coupling is obvious; namely, the more3MLCT contribution, the greater the
spin-orbit coupling, and hence the faster radiative rate of the
3MLCT f S
0transition.10,27We thus carefully examined the
percentage of3MLCT contribution to the T
1state. As shown
in Table 7, it appears to us that the percentages of MLCT contribution in the T1state of 1 (10.2%) and 2 (12.4%) are
significantly lower than those of 3-6 (> 40%). The results can be rationalized by the greaterπ-accepting properties of
CO than those of PPh2Me and PPhMe2 ligands, such that
complexes 1 and 2 with dual CO ligands render a much reduced electron density in the dπ orbital, giving a lesser amount of the MLCT contribution.
Further support of the above viewpoint is given by two additional observations. First, from the steady-state approach, the observation of vibronic progression of phosphorescence for 1 in 77 K CH2Cl2 provides indirect evidence that
phosphorescence should contain an excess ofππ* character.
Conversely, 3-6 showed a broad, diffusive phosphorescence even at 77 K, strongly indicating their dominance of the
3MLCT character. Moreover, as for the time-resolved
meas-(26) Hay, P. J. J. Phys. Chem. A 2002, 106, 1634.
(27) Yutaka, T.; Obara, S.; Ogawa, S.; Nozaki, K.; Ikeda, N.; Ohno, T.; Ishii, Y.; Sakai, K.; Haga, M. Inorg. Chem. 2005, 44, 4737. Figure 4. Selected frontier orbitals of 3 involved in the lower-lying transitions.
Table 5. Calculated Energy Levels of the Lower-Lying Transitions of 3
assignment [nm] E [eV] f T1 HOMO f LUMO+1 (+88%) 506.9 2.45 ∼0 T2 HOMO-1 f LUMO (+66%) 483.6 2.56 ∼0 HOMO-2 f LUMO (+15%) HOMO-4 f LUMO (+8%) HOMO-1 f LUMO+3 (+7%) Singlet States S1 HOMO f LUMO (+96%) 457.3 2.71 0.0002 S2 HOMO f LUMO+1 (+89%) 441.9 2.81 0.0217
Neutral Ru -Based Emitting Materials
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
urement, although the radiative decay time of 1 could not be deduced because of the nearly nonemissive properties at ambient temperature, the decay time of 2400µs measured
at 77 K is far longer than the 30-50 µs deduced for
complexes 3-6. A rather similar long radiative decay time of 540 µs was also deduced for complex 2. For 1 and 2,
according to the exceedingly longer radiative decay time, its dominant 3ππ character in the T1 state is obvious,
consistent with the theoretical approach. The emission quantum yieldΦ is defined as Φ ) kr/kobs) kr/(kr+ knr), in
which subscripts r, nr, and obs denote radiative, nonradiative, and observed decay components, respectively. Thus, for complexes possessing a small krvalue, such as 1 and 2, the knr value should quite effectively influence the emission
quantum yield and hence the emission intensity. Experimen-tally, this is apparently true and can qualitatively explain the nonemissive (weak-emissive) property for 1 and 2.
Nevertheless, from the quantitative viewpoint, the deduced radiationless decay rates of 1 [g 4.2 × 106s-1, assuming a
detection limit ofΦ∼ 10-4in our current system and 2400
µs (at 77 K) for the radiative decay time] and 2 (∼ 3.7 ×
106s-1) are larger than those deduced from 3 (1.04 × 105
s-1) and 4 (9.6× 104s-1), which possess the same BQ and
DBQ ligands. For rationalization, our first attempt is the weakening of the Ru-CO bonds upon excitation if the net result of MLCT is to induce an electron transfer to the BQ or DBQ chelates trans to the CO ligands. However, thorough frontier orbital analysis clearly indicates that the Ru-CO back-π-bonding character is present only in the very stable
occupied molecular orbital of HOMO-5 (Figure 3) and even lower ones (not shown here). Thus, it is very unlikely that the strength of such a strong Ru-CO bond, upon excitation, could be drastically weakened; consequently, the associated Ru-CO stretching modes should not act as a main radia-tionless deactivation pathway. In fact, strong to moderate emission has been reported in numerous carbonyl-containing third-row transition-metal complexes, in which MLCT transition occurs at the ligand chromophore trans to the ancillary CO ligands.28
Alternatively, in our opinion, it is more plausible to correlate the rapid radiationless deactivation in 1 and 2 with respect to the increase of the vibrational modes channeling into the radiationless deactivation pathways. As depicted in Tables 4-6, it is obvious that, despite a very simple contribution, i.e., HOMO f LUMO+1, to the T1state in 3
and 5, a much more complicated frontier orbital contribution was observed in 1, incorporating HOMO to HOMO-4 and LUMO to LUMO+2 (see Table 4). Careful analyses indicated that the associated frontier orbitals spread out to both BQ ligands simultaneously rather than to only one BQ site trans to the unique CO ligand for 3 and 4. Multiple contributions were observed for T1 in complex 2 (see the
Supporting Information). The results can be tentatively rationalized by the two BQ and DBQ chelate ligands being subjected to identical coordinating environments in both 1 and 2. In contrast, a simplified T1state was also noted in 5
and 6. The radiationless decay rates were deduced to be 4.12× 106and 7.12× 106s-1, both more than 1 order of
magnitude larger than those of 3 and 4, and can be rationalized by the greater degree of vibrational freedom in combination with radiationless transition governed by the empirical energy gap law.4 Accordingly, despite the same
order of magnitude in radiative decay time with respect to 3 and 4, greatly inferior phosphorescence quantum yields were observed for 5 and 6.
As for the thermal activation on the nonradiative deactiva-tion, we have performed a temperature-dependent study on
(28) (a) Van Slageren, J.; Stufkens, D. J. Inorg. Chem. 2001, 40, 277. (b) Yam, V. W.-W. Chem. Commun. 2001, 789. (c) Cheng, Y.-M.; Yeh, Y.-S.; Ho, M.-L.; Chou, P.-T.; Chen, P.-S.; Chi, Y. Inorg. Chem. 2005,
44, 4594. (d) Chen, Y.-L.; Lee, S.-W.; Chi, Y.; Hwang, K.-C.; Kumar,
S. B.; Hu, Y.-H.; Cheng, Y.-M.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H.; Yeh, S.-J.; Chen, C.-T. Inorg. Chem. 2005, 44, 4287.
Figure 5. Selected frontier orbitals of 5 involved in the lower-lying transitions.
Table 6. Calculated Energy Levels of the Lower-Lying Transitions of 5
assignment [nm] E [eV] f T1 HOMO f LUMO+1 (+98%) 539.9 2.30 ∼0 T2 HOMO f LUMO (+88%) 507.6 2.44 ∼0 HOMO-1 f LUMO (7%) Singlet States S1 HOMO f LUMO (+96%) 505.0 2.46 0.0002 S2 HOMO f LUMO+1 (+90%) 485.4 2.55 0.0166 Table 7. Percentage of the MLCT State Contributing to the Lowest Electronic Transition in Singlet and Triplet Manifolds
complex
state 1 2 3 4 5 6
S1 11.0 13.1 44.1 45.4 45.3 43.9
T1 10.1 12.4 40.5 46.4 46.2 48.3
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
complex 1. Because of its lack of emission at ambient temperature, the experiment was performed in a temperature range of 220-150 K. As a result, the plot of the logarithm of knr(T) vs 1/T, in which knr(T) ∼ kobs - kr, rendered a
sufficiently straight line (see Figure 6). As a result, a∆Ea
value of 2.78 kcal mol-1 and a preexponential factor of 1.7× 108s-1 were deduced. The results indicate that the
nonradiative decay rate is dominated by rather low frequency motions, possibly involving the geometry distortion motions coupled with the T1-S0 intersystem crossing. Although
specific modes inducing radiationless transition, at this stage, are pending resolution, the overlap of multiple frontier orbitals in the case of 1 (vide supra) should statistically increase the number of active vibrational modes involved in the radiationless deactivation.
5. Conclusion
In conclusion, a series of new RuII complexes bearing
cyclometalated BQ and DBQ ligands have been designed and synthesized with an aim to investigate the associated photoluminescence properties. Our results solidify the cor-relation between MLCT and radiative lifetime and hence the efficiency of radiationless deactivation. Complexes 1 and 2, possessing much less 3MLCT contribution because of the
electron deficiency in the dπorbital, render a great extent of the3ππ* character for T1with an exceedingly long radiative
lifetime. This, in combination with perhaps the increase of the deactivating modes, gives rise to an essentially nonemis-sive (very weak emission) property. Under the premise of this proposed mechanism, one may be able to analyze the lower lying electronic transitions and the corresponding frontier orbital analyses, providing a guideline to predict not only the phosphorescence properties such as peak wavelength and transition properties but also the quantum efficiency in a qualitative manner. We thus believe that the results presented here may turn out to be of great importance in the design of luminescent materials incorporating both second-and third-row transition-metal elements.
Acknowledgment. The authors are grateful for financial support from the National Center for High-Performance Computing, The Ministry of Economy, and the National Sci-ence Council of Taiwan. We also thank Prof. Ching-Fong Shu for helpful advice on the CV measurements.
Supporting Information Available: X-ray crystallographic data
file (CIF) of complex 3 and the calculated energy levels and associated frontier orbitals of the DFT calculation on complexes
1-6. This material is available free of charge via the Internet at
http://pubs.acs.org.
IC060066G
Figure 6. Plot of the logarithm of knr(T) vs 1/T in the range of 220-150 K (see the text for the definition of knr).
Neutral Ru -Based Emitting Materials
Downloaded by NATIONAL TAIWAN UNIV on July 27, 2009
