Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Lamansky 2001

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

4304 J. Am. Chem. Soc.

2001, 123, 4304-4312

Highly Phosphorescent Bis-Cyclometalated Iridium Complexes:


Synthesis, Photophysical Characterization, and Use in Organic Light
Emitting Diodes
Sergey Lamansky,† Peter Djurovich,† Drew Murphy,† Feras Abdel-Razzaq,† Hae-Eun Lee,†
Chihaya Adachi,‡ Paul E. Burrows,‡,§ Stephen R. Forrest,*,‡ and Mark E. Thompson*,†
Contribution from the Department of Chemistry, UniVersity of Southern California,
Los Angeles, California 90089, and Department of Electrical Engineering, Princeton UniVersity,
Princeton, New Jersey 08544
ReceiVed October 16, 2000. ReVised Manuscript ReceiVed February 23, 2001

Abstract: The synthesis and photophysical study of a family of cyclometalated iridium(III) complexes are
reported. The iridium complexes have two cyclometalated (C∧N) ligands and a single monoanionic, bidentate
ancillary ligand (LX), i.e., C∧N2Ir(LX). The C∧N ligands can be any of a wide variety of organometallic
ligands. The LX ligands used for this study were all β-diketonates, with the major emphasis placed on
acetylacetonate (acac) complexes. The majority of the C∧N2Ir(acac) complexes phosphoresce with high quantum
efficiencies (solution quantum yields, 0.1-0.6), and microsecond lifetimes (e.g., 1-14 µs). The strongly allowed
phosphorescence in these complexes is the result of significant spin-orbit coupling of the Ir center. The lowest
energy (emissive) excited state in these C∧N2Ir(acac) complexes is a mixture of 3MLCT and 3(π-π*) states.
By choosing the appropriate C∧N ligand, C∧N2Ir(acac) complexes can be prepared which emit in any color
from green to red. Simple, systematic changes in the C∧N ligands, which lead to bathochromic shifts of the
free ligands, lead to similar bathochromic shifts in the Ir complexes of the same ligands, consistent with “C∧N2-
Ir”-centered emission. Three of the C∧N2Ir(acac) complexes were used as dopants for organic light emitting
diodes (OLEDs). The three Ir complexes, i.e., bis(2-phenylpyridinato-N,C2′)iridium(acetylacetonate) [ppy2Ir-
(acac)], bis(2-phenyl benzothiozolato-N,C2′)iridium(acetylacetonate) [bt2Ir(acac)], and bis(2-(2′-benzothienyl)-
pyridinato-N,C3′)iridium(acetylacetonate) [btp2Ir(acac)], were doped into the emissive region of multilayer,
vapor-deposited OLEDs. The ppy2Ir(acac)-, bt2Ir(acac)-, and btp2Ir(acac)-based OLEDs give green, yellow,
and red electroluminescence, respectively, with very similar current-voltage characteristics. The OLEDs give
high external quantum efficiencies, ranging from 6 to 12.3%, with the ppy2Ir(acac) giving the highest efficiency
(12.3%, 38 lm/W, >50 Cd/A). The btp2Ir(acac)-based device gives saturated red emission with a quantum
efficiency of 6.5% and a luminance efficiency of 2.2 lm/W. These C∧N2Ir(acac)-doped OLEDs show some of
the highest efficiencies reported for organic light emitting diodes. The high efficiencies result from efficient
trapping and radiative relaxation of the singlet and triplet excitons formed in the electroluminescent process.

Introduction of the 4d or 5d ion leads to efficient intersystem crossing of


the singlet excited states to the triplet manifold.3 The long
The photophysics of octahedral 4d6
and 5d6
complexes has lifetimes of these excited states are due to the triplet character
been studied extensively.1 These complexes, particularly those of this state. Mixing of the singlet and triplet excited states, via
prepared with Ru and Os, have been used in a variety of spin-orbit coupling, removes the spin-forbidden nature of the
photonic applications, including photocatalysis and photoelec- radiative relaxation of the triplet state, leading to high phos-
trochemistry.2 These d6 complexes are attractive in photochemi- phorescence efficiencies.
cal applications, because they generally have long-lived excited Researchers have directed their attention to the photochem-
states and high luminescence efficiencies, increasing the likeli- istry and photophysics of Ru2+ and Os2+ complexes, with the
hood of either energy or electron transfer occurring prior to majority of the effort being focused on metal-diimine com-
radiative or nonradiative relaxation. Strong spin-orbit coupling plexes, such as the bipyridine and phenanthroline complexes.4
* To whom correspondence should be addressed: (S.R.F.) forrest@ More recently, researchers have investigated the photophysics
ee.princeton.edu. (M.E.T.) met@usc.edu. (2) (a) Kalyanasundaran, K. Coord. Chem. ReV. 1982, 46, 159-244. (b)
† University of Southern California. Chin, K.-F.; Cheung, K.-K.; Yip, H.-K.; Mak, T. C. W.; Che, C. M. J.
‡ Princeton University. Chem. Soc., Dalton Trans. 1995, 4, 657-665. (c) Sonoyama, N.; Karasawa,
§ Current address: Pacific Northwest National Laboratory, K3-59 PO O.; Kaizu, Y. J. Chem. Soc., Faraday Trans. 1995, 91, 437-443. (d) Tan-
Box 999, Richland, WA 99352. Sien-Hee, L.; Mesmaeker, A. K-D. J. Chem. Soc., Dalton Trans. 1994,
(1) (a) Balzani, V.; Scandola, F. Supramolecular Photochemistry; Ellis 24, 3651-3658. (e) Kalyanasundaram, K.; Gratzel, M. Coord. Chem. ReV.
Horwood: Chichester, U.K., 1991. (b) Balzani, V.; Credi, A.; Scandola, F. 1998, 177, 347-414.
In Transition Metals in Supramolecular Chemistry; Fabbrizzi, L., Poggi, (3) (a) Baldo, M. A.; O’Brien, D. F.; You Y.; Shoustikov, A.; Sibley,
A., Eds.; Kluwer: Dordrecht, The Netherlands, 1994; p1. (c) Lehn, J.-M. S.; Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151-154. (b) Baldo,
Supramolecular ChemistrysConcepts and Properties; VCH: Weinheim, M A.; Lamansky, S.; Burrows, P. E.; Thompson, M. E.; Forrest, S. R. Appl.
Germany, 1995. (d) Bignozzi, C. A.; Schoonover, J. R.; Scandola, F. Prog. Phys. Lett. 1999, 75, 4-6. (c) Thompson, M. E.; Burrows, P. E.; Forrest,
Inorg. Chem. 1997, 44, 1. S. R. Cur. Opinion Solid State Mater. Sci. 1999, 4, 369.

10.1021/ja003693s CCC: $20.00 © 2001 American Chemical Society


Published on Web 04/13/2001
Phosphorescent Bis-Cyclometalated Iridium Complexes J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4305

of isoelectronic Rh3+ and Ir3+ complexes, with both diimine electron-hole recombination are trapped at the phosphor, where
and cyclometalated ligands, such as 2-phenylpyidinato-C2,N strong spin-orbit coupling leads to singlet-triplet state mixing
(ppy).5 The cyclometalated ligands are formally monoanionic and, hence, efficient phosphorescent emission at room temper-
and can thus be used to prepare neutral tris-ligand complexes, ature. Both singlet and triplet excited states can be trapped at
which are isoelectronic with the cationic trisdiimine complexes the phosphor. OLEDs prepared with these heavy metal com-
of Ru and Os, e.g., fac-M(ppy)3,6 fac-M(2-(R-thiopheneyl)- plexes are the most efficient OLEDs reported to date, with
pyridine)3 (fac ) facial).7 The d6 Ir complexes show intense internal quantum efficiencies exceeding 75% (photons/electrons)
phosphorescence at room temperature, while the Rh complexes (>15% external efficiency).12 Furthermore, OLEDs have been
give measurable emission only at low temperatures, consistent prepared with C∧N2Ir(LX) phosphor dopants, giving efficient
with stronger spin-orbit coupling of Ir relative to Rh. The green, yellow, or red emission. The external quantum efficien-
electronic transitions responsible for luminescence in these cies for these devices vary from 5% to nearly 20%. In this work,
complexes have been assigned to a mixture of metal-to-ligand we explore the photophysical and electroluminescent properties
charge-transfer (MLCT) and 3(π-π*) ligand states.8 We have of a series of Ir complexes used as efficient phosphorescent
recently found that highly emissive Ir complexes can be formed dopants in OLEDs. We demonstrate that, by optimizing the
with two cyclometalated ligands (abbreviated hereafter as C∧N) molecular structure of C∧N2Ir(LX) dopants and the energy-
and a single monoanionic, bidentate ancillary ligand (LX).9 The transfer process, exceedingly high external and power efficien-
emission colors from these complexes are strongly dependent cies can be obtained in the green to red spectral region.
on the choice of cyclometalating ligand, ranging from green to
red, with room-temperature lifetimes from 1 to 14 µs. The Experimental Section
photophysical properties of these C∧N2Ir(LX) complexes are Synthesis. All synthetic procedures involving IrCl3‚H2O and other
similar to those observed for the tris-cyclometalated complexes, Ir(III) species were carried out in inert gas atmosphere despite the air
and will be discussed below. stability of the compounds, the main concern being the oxidative
Heavy metal complexes, particularly those containing Pt and stability of intermediate complexes at the high temperatures used in
Ir, can serve as efficient phosphors in organic light emitting the reactions. NMR spectra were recorded on Bruker AMX 360- or
devices.10 In these devices, holes and electrons are injected into 500-MHz instruments. High-resolution mass spectrometry was carried
opposite surfaces of a planar multilayer organic thin film. The out by the mass specroscopy facility at the Frick Chemistry Laboratory,
holes and electrons migrate through the thin film, to a material Princeton University. Elemental analyses (C, H, N) were carried out
by standard combustion analysis by the Microanalysis Laboratory at
interface, where they recombine to form radiative excited states,
the University of Illinois, UrbanasChampagne.
or excitons. This electrically generated exciton can be either a
Cyclometalated Ir(III) µ-chloro-bridged dimers of a general formula
singlet or a triplet. Both theoretical predictions and experimental C∧N2Ir(µ-Cl)2IrC∧N2 were synthesized according to the Nonoyama
measurements give a singlet/triplet ratio for these excitons of 1 route, by refluxing IrCl3‚nH2O (Next Chimica) with 2-2.5 equiv of
to 3.11 Fluorescent materials typically used to fabricate organic cyclometalating ligand in a 3:1 mixture of 2-ethoxyethanol (Aldrich
light emitting diodes (OLEDs) do not give detectable triplet Sigma) and water.13
emission (i.e., phosphorescence), nor is there evidence for C∧N2Ir(acac), C∧N ) ppy, tpy, bzq, bt, Rbsn, and pq were prepared
significant intersystem crossing between the triplet and singlet as described previously.9
manifolds at room temperature. The singlet/triplet ratio thus Synthesis of (C∧N)2Ir(acac) Complexes. General Procedure. The
implies a limitation of 25% for the internal quantum efficiency chloro-bridged dimer complex (0.08 mmol), 0.2 mmol of acetyl acetone,
for OLEDs based on fluorescence. By doping OLEDs with and 85-90 mg of sodium carbonate were refluxed in an inert
heavy metal phosphors, we have shown that the singlet-triplet atmosphere in 2-ethoxyethanol for 12-15 h. After cooling to room
limitation can be eliminated.10 The excited states generated by temperature, a colored precipitate was filtered off and washed with
water, hexane, and ether. The crude product was flash chromatographed
(4) (a) Anderson, P. A.; Anderson, R. F.; Furue, M.; Junk, P. C.; Keene, on a silica column with dichloromethane mobile phase to yield ∼75-
F. R.; Patterson, B. T.; Yeomans, B. D. Inorg. Chem. 2000, 39, 2721- 90% of the pure C∧N2Ir(acac), after solvent evaporation and drying.
2728. (b) Li, C.; Hoffman, M. Z. Inorg. Chem. 1998, 37, 830-832. (c) thp2Ir(acac): Iridium(III) bis(2-(2′-thienyl)pyridinato-N,C3′) (acetyl
Berg-Brennan, C.; Subramanian, P.; Absi, M.; Stern, C.; Hupp, J. T. Inorg. acetonate) (yield 83%). 1H NMR (360 MHz, acetone-d6): δ, ppm 8.41
Chem. 1996, 35, 3719-3722. (d) Kawanishi, Y.; Kitamura, N.; Tazuke, S.
Inorg. Chem. 1989, 28, 2968-2975. (d, 2H, J 5.8 Hz), 7.79 (td, 2H, J 7.9, 1.6 Hz), 7.56 (d, 2H, J 7.9 Hz),
(5) (a) Balzani, V.; Juris, A.; Venturi, M.; Campagna, S.; Serroni, S. 7.22 (d, 2H, J 4.7 Hz), 7.11 (td, 2H, J 6.3, 1.6 Hz), 6.09 (d, 2H, J 4.7
Chem. ReV. 1996, 96, 759-834. (b) Shaw, J. R.; Sadler, G. S.; Wacholtz, Hz), 5.29 (s, 1H), 1.72 (s, 6H). Anal. Found C 45.33, H 3.00, N 4.81.
W. F.; Ryu, C. K.; Schmehl, R. H. New J. Chem. 1996, 20, 749. Calcd C 45.16, H 3.13, N 4.58.
(6) (a) Garces, F. O.; King, K. A.; Watts, R. J. Inorg. Chem. 1988, 27, btp2Ir(acac): Iridium(III) bis(2-(2′-benzothienyl)pyridinato-N,C3′)
3464-3471. (b) Dedian, K.; Djurovich, P. I.; Garces, F. O.; Carlson, G.;
Watts, R. J. Inorg. Chem. 1991, 30, 1685-1687. (c) Schmid, B.; Garces,
(acetylacetonate) (yield 72%). 1H NMR (360 MHz, acdtone-d6): δ,
F. O.; Watts, R. J. Inorg. Chem. 1994, 33, 9-14. (d) King, K. A.; Spellane, ppm 8.39 (d, 2H, J 5.9 Hz), 7.80 (t, 2H, J 7.9 Hz), 7.77 (d, 2H, J 8.0
P. J.; Watts, R. J. J. Am. Chem. Soc. 1985, 107, 1432-1433. Hz), 7.68 (d, 2H, J 8.0 Hz), 7.25 (t, 2H, J 7.0 Hz), 7.10 (t, 2H, J 7.1
(7) Colombo, M. G.; Brunold, T. C.; Riedener, T.; Güdel, H. U.; Förtsch, Hz), 6.82 (t, 2H, J 8.0 Hz), 6.40 (d, 2H, J 7.3 Hz), 5.70 (s, 1H), 1.90
M.; Bürgi, H.-B. Inorg. Chem. 1994, 33, 545-550. (s, 6H). Anal. Found C 52.51, H 3.29, N 4.01. Calcd C 52.30, H 3.26,
(8) (a) Wilde, A. P.; King, K. A.; Watts, R. J. J. Phys. Chem. 1991, 95, N 3.94.
629-634. (b) Sprouse, S.; King, K. A.; Spellane, P. J.; Watts, R. J. J. Am.
Chem. Soc. 1984, 106, 6647-6653. (c) Crosby, G. A. J. Chim. Phys. 1967, dpo2Ir(acac): Iridium(III) bis(2,4-diphenyloxazolato-1,3-N,C2′) (acetyl
64, 160. (d) Columbo, M. C.; Hauser, A.; Güdel, H. U. Top. Curr. Chem. acetonate) (yield 93%).1H NMR (360 MHz, CDCl3): δ, ppm 7.79 (d,
1994, 171, 143. 4H, J 7.4 Hz), 7.53 (d, 2H, J 7.9 Hz), 7.49 (m, 6H), 7.45 (s, 2H), 7.40
(9) Lamansky, S.; Djurovich, P.; Murphy, D.; Abdel-Razaq, F.; Kwong, (t, 2H, J 7.4 Hz), 6.84 (t, 2H, J 7.4 Hz), 6.76 (t, 2H, J 7.4 Hz), 6.62 (d,
R.; Tsyba, I.; Bortz, M.; Mui, B.; Bau, R.; Thompson, M. E. Inorg. Chem. 2H, J 7.9 Hz), 5.25 (s, 1H), 1.86 (s, 6H). Anal. Found C 56.35, H
2001, 40, 1704-1711.
(10) (a) Baldo, M. A.; O’Brien, D. F.; You, Y.; Shoustikov, A.; Sibley, 3.67, N 3.89. Calcd C 57.44, H 3.72, N. 3.83.
S.; Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151-154. (b) Baldo, C62Ir(acac): Iridium(III) bis(3-(2-benzothiazolyl)-7-(diethylamino)-
M A.; Lamansky, S.; Burrows, P. E.; Thompson, M. E.; Forrest, S. R. Appl. 2H-1-benzopyran-2-onato-N′,C4) (acetyl acetonate) (yield 59%). 1H
Phys. Lett. 1999, 75, 4-6. (c) Thompson, M. E.; Burrows, P. E.; Forrest,
S. R. Curr. Opin. Solid State Mater. Sci. 1999, 4, 369. (12) Adachi, C.; Baldo, M. A.; Forrest, S. R.; Thompson, M. E. Appl.
(11) Baldo, M. A.; O’Brien, D. F.; Thompson, M. E.; Forrest, S. R. Phys. Phys. Lett. 2000, 78, 1704.
ReV. B 1999, 60, 14422. (13) Nonoyama, M. Bull. Chem. Soc. Jpn. 1974, 47, 767-768.
4306 J. Am. Chem. Soc., Vol. 123, No. 18, 2001 Lamansky et al.

Table 1. Photophysical Data for C∧N2Ir(acac) Complexesa


C∧N emission lifetime (µs) quantum
ligand absorbance λ (log ) λmax (nm) 298 K efficiency
bo 262 (4.7), 269 (4.6), 298 (4.6), 310 (4.5), 343 (4.0), 525 1.1 0.25
383 (3.8), 425 (3.7), 462 (3.6), 510 (2.6)
bon 266 (4.8), 298 (4.7), 326 (4.6), 360 (4.4), 410 (4.1), 586 1.3 0.11
458 (4.0), 491 (4.0)
bzq9 260 (4.6), 360 (3.9), 470 (3.3), 500 (3.2) 548 4.5 0.27
R-bsn 274 (4.7), 300 (4.6), 345 (4.5), 427 (4.0), 476 (4.0), 506 (3.9) 606 1.8 0.22
β-bsn 328 (4.7), 348 (4.6), 420 (3.8), 496 (3.5) 594 2.2 0.16
btth 286 (4.4), 327 (4.4), 405 (4.0), 437 (4.0), 478 (3.9) 593 3.6 0.21
bt9 269 (4.6), 313 (4.4), 327 (4.5), 408 (3.8), 447 (3.8), 557 1.8 0.26
493 (3.4), 540 (3.0)
btp 286 (4.5), 340 (4.1), 355 (3.9), 495 (3.7) 612 5.8 0.21
C6 444 (4.8), 472 (4.8) 585 14 0.6
op 258 (4.4), 294 (4.2), 336 (4.1), 372 (3.8), 456 (3.5) 520 2.3 0.14
dpo 297 (4.8), 369 (4.0), 420 (3.9), 443 (3.9) 550 3.0 0.1
ppy9 260 (4.5), 345 (3.8), 412 (3.4), 460 (3.3), 497 (3.0) 516 1.6 0.34
pq9 268 (5.0), 349 (4.4), 433 (3.9), 467 (3.9), 553 (3.6) 597 2 0.1
thp 302 (4.4), 336 (4.1), 387 (3.8), 453 (3.5) 562 5.3 0.12
tpy9 270 (4.5), 370 (3.7), 410 (3.5), 460 (3.4), 495 (3.0) 512 3.1 0.31
a All of the data were collected for 2-methyltetrahydrofuran solutions.

NMR (360 MHz, CDCl3): δ, ppm 7.86 (d, 2H, J 8.0 Hz), 7.58 (d, 2H, (CBP) host and a phosphorescent dopant, with both deposition rates
J 8.0 Hz), 7.29 (t, 2H, J 8.0 Hz), 7.21 (t, 2H, J 7.4 Hz), 6.29 (s, 2H), being controlled with two independent quartz crystal oscillators. ppy2-
6.05 (d, 2H, J 9.7 Hz), 5.83 (d, 2H, J 9.1 Hz), 5.27 (s, 1H), 3.20 (m, Ir(acac), bt2Ir(acac), and btp2Ir(acac) with a dopant concentration of
8H), 1.68 (s, 6H), 1.03 (t, 12H, J 7.4 Hz). Anal. Found C 52.45, H ∼7% were utilized to promote short-range Dexter transfer of the triplet
4.33, N 5.33. Calcd C 54.58, H 4.17, N 5.66. excitons without concentration quenching.1 Next, a 10-nm-thick 2,9-
bon2Ir(acac): Iridium(III) bis(2-(1-naphthyl)benzooxazolato-N,C2′) dimethyl-4,7-diphenyl-1,10-phenanthroline (BCP) as a hole and exciton
(acetyl acetonate) (yield 70%). 1H NMR (360 MHz, CDCl3): δ, ppm blocking layer (HBL) and 40-nm-thick tris(8-hydroxyquinoline)-
1.82 (s,6 H), 5.24 (s, 1H), 6.70 (d, 2H, J 8.5 Hz), 7.07 (d, 2H, J 8.2 aluminum (Alq3) as an electron transport layer were deposited on the
Hz),7.28 (t, 2H, J 7.5 Hz), 7.41(t, 2H, J 7.5 Hz), 7.47 (d, 2H, J 8.2 emitter layer. Finally, a shadow mask with 1-mm-diameter openings
Hz), 7.55 (m, 6H), 7.80 (d, 2H, J 7.50 Hz), 8.89 (d, 2H, J 8.5 Hz). was used to define the cathode consisting of a 100-nm-thick Mg-Ag
High-resolution MS: calculated M+ 780.1600; observed M+ 780.1592. layer, with a 20-nm-thick Ag cap. Current density-voltage-luminance
βbsn2Ir(acac): Iridium(III) bis(2-(2-naphthyl)benzothiazolato-N,C2′) (J-V-L) measurements were obtained using a semiconductor param-
(acetyl acetonate). (yield 70%). 1H NMR (360 MHz, CD2Cl2): δ, ppm eter analyzer and a calibrated silicon photodiode.
8.25 (s, 2 H), 8.16 (d, 2 H, J 8 Hz), 8.05 (d, 2 H, J 8 Hz), 7.71 (t, 2
H, J 4.5 Hz), 7.52 (m, 4 H), 7.18 (m, 6 H), 6.74 (s, 2 H), 5.24 (s, 1 H), Results and Discussion
1.79 (s, 6 H). Anal. Found C 56.68, H 3.22, N 3.59. Calcd C 56.56, H
3.51, N 3.03. Synthesis and Characterization of C∧N2Ir(LX) Complexes.
op2Ir(acac): Iridium(III) bis(2-phenyl oxazolinato-N,C2′(acetyl ac- C∧N2Ir(LX) complexes have been prepared, with several dif-
etonate) (yield 70%). 1H NMR (360 MHz, CD2Cl2): δ, ppm 7.26 (d,
ferent C∧N and LX ligands, Figure 1 and Table 1. The synthetic
2H, J 7.8 Hz), 6.86 (m, 4 H), 6.8 (t, 2H, J 6.8 Hz), 5.22 (s, 1 H), 4.98
(dd, 2H, J 8.6, 8.3 Hz), 4.87 (dd, 2H, J 8.6, 8.3 Hz), 4.04 (dd, 2H, J procedure used to prepare these complexes involves two steps.9
12.7, 7.7 Hz), 3.82 (dd, 2H, J 12.7, 7.7 Hz), 1.80 (s, 6H). High- In the first step, IrCl3‚nH2O is reacted with an excess of the
resolution MS: calculated M+ 584.1287; observed M+ 584.1295. desired C∧N ligand to give a chloride-bridged dimer, i.e., C∧N2-
Optical Measurements. Absorption spectra were recorded on AVIV Ir(µ-Cl)2IrC∧N2, eq 1. The NMR spectra of these complexes
model 14DS UV-visible-IR spectrophotometer (re-engineered Cary
14) and corrected for background due to solvent absorption. Emission
2IrCl3‚nH2O + 2C∧NH f
spectra were recorded on PTI QuantaMaster model C-60SE spectro-
fluorometer with 928 PMT detector and corrected for detector sensitivity C∧N2Ir(µ-Cl)2IrC∧N2 + 2HCl + 2Cl (1)
inhomogeneity. Emission quantum yields were determined using fac-
Ir(ppy)3 as a reference.14 Emission lifetimes were obtained by expo-
C∧N2Ir(µ-Cl)2IrC∧N2 + 2LXH f 2C∧N2Ir(LX) + 2HCl
nential fit of emission decay15 curves recorded on a PTI TimeMaster (2)
model C-60SE spectrofluorometer.
OLED Fabrication and Testing. The OLED structure employed
in this study is shown in the inset of Figure 6. Organic layers were are consistent with the heterocyclic rings of the C∧N ligands
fabricated by high-vacuum (10-6 Torr) thermal evaporation onto a glass being in a trans disposition, as shown in Figure 1b. The chloride-
substrate precoated with an indium-tin-oxide (ITO) layer with a sheet bridged dimers can be readily converted to emissive, monomeric
resistance of 20 Ω/0. Prior to use, the ITO surface was ultrasonicated complexes by replacing the bridging chlorides with bidentate,
in a detergent solution followed by a deionized water rinse, dipped monoanionic β-diketonate ligands (LX), eq 2. These reactions
into acetone, trichloroethylene, and 2-propanol, and then degreased in give C∧N2Ir(LX) with a yield of typically >80%.
2-propanol vapor. After degreasing, the substrate was oxidized and
X-ray crystallographic studies have been carried out for two
cleaned in a UV-ozone chamber before it was loaded into an evaporator.
A 50-nm-thick film of 4,4′-bis[N-(naphthyl-N-phenylamino)biphenyl
C∧N2Ir(LX) complexes, i.e., (ppy)2Ir(acac) and (tpy)2Ir(acac)
(R-NPD) served as the hole transport layer (HTL). The light emitting (acac ) acetylacetonate).9 The Ir in both of these C∧N2Ir(acac)
layer was prepared by coevaporating a 4,4′-N,N′-dicarbazolebiphenyl complexes is octahedrally coordinated by the three chelating
ligands, with the pyridyl groups in a trans disposition, as shown
(14) King, K. A.; Spellane, P. J.; Watts, R. J. J. Am. Chem. Soc. 1985, schematically in Figure 1b. The coordination geometries of the
107, 1432-1433.
(15) O’Connor, D. V.; Phillips, D. Time-Correlated Single Photon “C∧N2Ir” fragment in both the tpy and ppy iridium acac
Counting; Academic Press: London, 1984; p 287. complexes are the same as that reported for the chloride-bridged
Phosphorescent Bis-Cyclometalated Iridium Complexes J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4307

Figure 1. Cyclometalating (C∧N) and ancillary (LX) ligands used to prepare C∧N2Ir(LX) complexes, (a). The abbreviations used throughout the
paper for each ligand are listed below the “C∧N2Ir” or “Ir(LX)” fragment. The coordination geometries of the chloride-bridged dimer and C∧N2-
Ir(LX) complexes are shown in (b).

dimers and C∧N2Ir(bpy)+ complexes of the same C∧N ligands.16 6000 M-1 cm-1, respectively. The π-π* absorption bands for
The C∧N2Ir(LX) complexes are stable in air and can be sublimed these complexes fall in the ultraviolet and closely resemble the
in a vacuum without decomposition. spectra of the free C∧N ligands. Both 1MLCT and 3MLCT bands
Photophysical Properties of C∧N2Ir(LX) Complexes. In are typically observed for these complexes. The high degree of
order for the C∧N2Ir(LX) complexes to be useful as phosphors spin-orbit coupling is evident in comparing the oscillator
in organic light emitting diodes, strong spin-orbit coupling must strengths for the two MLCT bands. For example, the singlet
be present to efficiently mix the singlet and triplet excited states. and triplet MLCT bands for ppy2Ir(acac) fall at 410 and 460
Clear evidence for significant mixing of the singlet and triplet nm, respectively (see Figure 2 and Table 1), with less than a
excited states is seen in both the absorption and emission spectra factor of 2 difference in their extinction coefficients. Strong
of these complexes. All of the C∧N2Ir(LX) complexes show spin-orbit coupling on Ir gives the formally spin-forbidden 3-
intense absorption from C∧N ligand π-π* and MLCT transi- MLCT an intensity comparable to the allowed 1MLCT. The
tions. The absorption spectra for the ppy, bzq, thp, and btp energies of these singlet and triplet MLCT absorptions are very
complexes are shown in Figure 2. The extinction coefficients similar to those reported for ppy2Ir(bpy)+,17 ppy2Ir(H2O)2+, 18
for these bands are in the ranges expected for π-π* ligand-
centered and MLCT bands, about 10 000-35 000 and 2000- (17) (a) Ohsawa, Y.; Sprouse, S.; King, K. A.; DeArmond, M. K.; Hanck,
K. W.; Watts, R. J. J. Phys. Chem. 1987, 91, 1047-1058. (b) Colombo,
(16) Carlson, G. A.; Djurovich, P. I.; Watts, R. J. Inorg. Chem. 1993, M. G.; Hauser, A.; Gudel, H. U. Inorg. Chem. 1993, 32, 3088-3092.
32, 4483-4484. (18) Schmid, B.; Garces, F. O.; Watts, R. J. Inorg. Chem. 1994, 33, 9-14.
4308 J. Am. Chem. Soc., Vol. 123, No. 18, 2001 Lamansky et al.

Figure 2. Absorption and luminescence spectra of C∧N2Ir(acac)


complexes. (top) Absorption spectra for C∧N ) ppy, bzq, thp, and btp
complexes. The spectra have been offset for clarity. All four complexes
have extinction coefficients of ∼0.0 at 550 nm. (bottom) Photolumi- Figure 3. Solution photoluminescence, excitation, and absorption
nescence spectra for the same complexes. spectra of C62Ir(acac) and C6 (top) and dpo2Ir(acac) and dpo (bottom).

and ppy2Ir(µ-Cl)2Irppy2.19 The ppy, bzq, thp, and btp complexes and no observable phosphorescence. Further, C6 and dpo have
have very similar 3MLCT energies, with λmax values ranging the requisite structures to make them suitable for cyclometalation
from 440 to 490 nm (top inset of Figure 2 and Table 1). The reactions with Ir and have been used to make C∧N2Ir(acac)
similarity of MLCT energies for these complexes is not complexes. Coordination to Ir shifts the emission maximum of
surprising, since all four complexes have MLCT states involving C6 from 500 nm in the free dye to 570 nm for C62Ir(acac), as
very similar (pyridyl) acceptors. shown in Figure 3, providing striking evidence for intersystem
1. Correlation of Absorption and Emission Bands. In crossing induced by the proximity of the heavy Ir atom. This Ir
addition to 3MLCT absorption bands with high oscillator complex has a high quantum yield for emission of 0.6 and the
strengths, strong spin-orbit coupling leads to efficient phos- lifetime is 14 µs. The excitation spectrum for C62Ir(acac) is
phorescence in the majority of the C∧N2Ir(LX) complexes very similar to that of pure C6 and gives a good match to the
reported here. The room-temperature (solution) quantum yields absorption spectrum of the Ir complex. A more marked red shift
of these complexes range from 0.1 to 0.6 (Table 1), and their is observed on metalation of dpo, whose λmax for emission shifts
luminescent lifetimes fall between 1 and 14 µs, consistent with from 365 nm for fluorescence from the free dye to 550 nm for
emission from a triplet excited state. The positions of the phosphorescence from dpo2Ir(acac). The excitation spectrum
maximums in the excitation spectra for these complexes are matches the absorption spectrum of dpo2Ir(acac), with a
very similar to those in their absorption spectra. Pumping either maximum efficiency at an energy close to the MLCT band. The
ligand-based or MLCT transitions efficiently gives the same quantum efficiency and lifetime for the dpo Ir complex are 0.1
phosphorescent excited state, as illustrated for C62Ir(acac) and and 3 µs, as expected for Ir-promoted phosphorescence from
dpo2Ir(acac) in Figure 3. the dpo ligand. The quantum yields for these phosphorescent
Note that C6 and dpo are common fluorescent laser dyes. transitions are high, especially C62Ir(acac), which has a quantum
Solutions of C6 give green fluorescence, while solutions of dpo yield nearly as high as those of the fluorescent transition C6
give UV/violet fluorescence.20 Both of these laser dyes give alone. In both C6- and dpo-based complexes, the Ir center
high quantum efficiencies for fluorescence at room temperature facilitates intersystem crossing into the triplet levels of the laser
(19) Sprouse, S.; King, K. A.; Spellane, P. J.; Watts, R. J. J. Am. Chem. (20) Brackman, U. Lambdachrome Laser Dyes, 2nd ed.; Lambda Physik
Soc. 1984, 106, 6647-6653. Ink.: Göttingen, Germany, 1997.
Phosphorescent Bis-Cyclometalated Iridium Complexes J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4309

dye and efficient phosphorescence from these low-energy triplet


states. Efficient, ligand-based phosphorescence is common in
C∧N2Ir(LX) complexes, making the design of new phosphors
straightforward, as described below.
Although the ppy, bzq, thp, and btp complexes have similar
MLCT absorption energies, their emission spectra differ mark-
edly, Figure 2 (bottom). The ppy and bzq complexes exhibit a
small Stokes shift between the 3MLCT absorption and emission
bands, while the thp and btp complexes give larger shifts. The
energy differences between λmax for the 3MLCT absorption and
emission spectra of ppy2Ir(acac) and bzq2Ir(acac) are 55 and
58 nm, respectively (∼2200 cm-1), wheras significantly larger
shifts are observed for the thp and btp complexes (105 and 120
nm, respectively, ∼4100 cm-1). It has been shown that the
emission spectra of fac-Irppy3 and fac-Irthp3 complexes result
from mixtures of MLCT and 3(π-π*) transitions.21 fac-Irppy3
emits from an excited state that is predominantly due to MLCT,
while fac-Irthp3 emits from a largely ligand-based 3(π-π*)
excited state. The 3(π-π*) level for phenylpyridine (ppy) has Figure 4. Solution photoluminesccence spectra of bo2Ir(acac), bt2Ir-
an energy of 460 nm,22 which puts it at a sufficiently high energy (acac), bon2Ir(acac) and Rbsn2Ir(acac). The structures of the individual
such that the 3MLCT becomes the lowest energy excited state. C∧N ligands are shown above the corresponding spectrum.
In contrast, the 3(π-π*) transition for thienylpyridine (thp) is
at λmax ) 564 nm, which is at an energy below the 3MLCT All four complexes show vibronic fine structure in their
energy, so a predominantly ligand-based state is the lowest emission spectra, as expected for ligand-based transitions. The
energy excited state in fac-Irthp3. The Stokes shifts observed highest energy emission is observed for the benzoxazole (bo)
for the C∧N2Ir(acac) complexes support similar assignments complex. Substitution of S for O in a chromophore (bo f bt)
here. Emission from a predominantly 3MLCT state would be leads to a 30-nm red shift, due the higher polarizibility and
expected to have a small Stokes shift between the 3MLCT basicity of sulfur relative to oxygen,26 in this ligand-based
absorption and emission bands, as seen for ppy and bzq excited state. Increasing the size of the ligand π system is
complexes. Emission from a predominantly ligand-based excited expected to bathochromically shift electronic transitions, as is
state, however, should give a large Stokes shift between the observed in converting a phenyl group to a naphthyl group (bo
3MLCT absorption and emission bands, as observed for the thp f bon), which leads to a 60-nm red shift. The effects of the
and btp complexes. naphthyl and sulfur substitutions are nearly additive, leading to
The line shapes of the phosphorescence spectra of these an 80-nm red shift when comparing bo to Rbsn complexes.
complexes also support the hypothesis that the ppy and bzq Unfortunately, we have been unable to detect phosphorescence
complexes emit primarily from a MLCT state, while the other spectra from the free ligands, even at temperatures below 10
complexes emit from 3(π-π*) C∧N states. The emission spectra K, so we cannot directly correlate the phosphorescence spectra
for thp2Ir(acac) closely resemble the phosphorescence spectrum of the Ir complexes with those of the 3(π-π*) of the free
reported for the free organic ligand,7,17,23 supporting the assign- ligands. The luminescence energies of the four Ir complexes
ment of the lowest energy excited state to be predominantly do follow the same trend seen in the fluorescence spectra of
ligand-based.24 Vibronic fine structure is clearly observed for the free ligands, however, which have λmax values of 350, 360,
the thp and btp ligands and is absent for the ppy and bzq 380 and 400 nm, respectively.
complexes. Emission bands from MLCT states are generally We also attempted to decrease the size of the ligand π system,
broad and featureless, while 3(π-π*) states typically give highly in the hope of achieving a shift form green to blue emission.
structured emission.25 Significant vibronic fine structure is also The first approach we chose was to replace the metalated phenyl
observed in the PL spectra of both the C6 and dpo complexes, group of the ppy ligand with a smaller group, such as a vinyl
consistent with emission from ligand-based excited states in or cycloalkene (e.g., C∧N ) 2-vinylpyridine, vinylbenzoxazole,
these complexes as well. and 1-cyclohexenylbenzoxazole). Unfortunately, we have been
2. C∧N Ligand Tuning of Phosphorescence. Further unable to cyclometalate ligands in which the phenyl has been
evidence for significant ligand character in the emission spectra replaced with olefinic groups. We have succeeded in decreasing
can be seen in a series of C∧N2Ir(acac) complexes prepared with the size of the π system in the benzoxazole portion of the ligand,
bo, bt, bon, and Rbsn ligands. The photoluminescence spectra however, i.e., op (2-(1-phenyl)oxazole, Figure 1). In op, the
of these complexes show a pronounced red shift, Figure 4. phenylene group of the bo benzoxazole group has been replaced
with a vinylene group. The emission spectrum of op2Ir(acac)
(21) (a) Colombo, M. G.; Brunold, T. C.; Riedener, T.; Güdel, H. U.; has its λmax at 520 nm, only 5-nm blue shifted from bo2Ir(acac),
Förtsch, M.; Bürgi, H.-B. Inorg. Chem. 1994, 33, 545-550. (b) Colombo, Table 1. The π* state in both of these complexes is presumably
M. G.; Hauser, A.; Gudel, H. U. Inorg. Chem. 1993, 32, 3088-3092.
(22) Sarkar, A.; Sankar, C. J. Lumin. 1995, 65, 163-168. localized in the oxazole group and does not extend significantly
(23) Colombo, M. G.; Hauser, A.; Güdel, H. U. Top. Curr. Chem. 1994, into the added phenyl ring of the benzoxazole group, leading
171, 143-171 (Electronic and Vibronic Spectra of Transition Metal to only a modest shift in the phosphorescence energy on removal
Complexes I).
(24) Sandrini, D.; Maestri, M., Ciano, M.; Balzani, V.; Lueoend, R.; of the phenyl ring.
Deuschel-Cornioley, C.; Chassot, L.; von Zelewsky, A. Gaz. Chim. Ital. Two different isomers of the naphthylbenzothiazole (Rbsn
1988, 118, 661. and βbsn) have been used to prepare cyclometalated Ir
(25) Sandrini, D.; Maestri, M., Ciano, M.; Balzani, V.; Deuschel-
Cornioley, C.; von Zelewsky, A.; Jolliet, P. HelV. Chim. Acta 1988, 71, (26) Zollinger, H. Color Chemistry: Syntheses, Properties and Applica-
1053. Balton, C. B.; Murtaza, Z.; Shaver, R. J.; Rillema, D. P. Inorg. Chem. tions of Organic Dyes and Pigments, 2nd ed.; VSH: Weinheim, Germany,
1992, 31, 3230. 1991.
4310 J. Am. Chem. Soc., Vol. 123, No. 18, 2001 Lamansky et al.

identical to those of the same complexes prepared with an acac


LX ligand. All of these complexes emit from states within the
“C∧N2Ir” fragment as discussed above. When the LX ligand is
bza, emission from the ppy, bzq, and thp complexes are largely
quenched (φphos < 0.01). In contrast, the btp complex (btp2Ir-
(bza)) gives a quantum efficiency of 0.1 and an emission
spectrum identical to that of btp2Ir(acac). For the C∧N2Ir(dbm)
complexes, all four C∧N ligands give very weak phosphores-
cence (φphos < 0.01). The triplet levels of these three β-dike-
tonates fall in the order tmd > bza > dbm. Apparently, only
the ligand with the lowest triplet energy, i.e., C∧N ) btp, has
a triplet level lower than dbm. Thus, only the btp complex gives
emission from an excited state on the “C∧N2Ir” fragment for
bza complexes. The other three complexes have their excited
states localized predominantly on the on the bza ligand, leading
Figure 5. Solution phosphorescence efficiencies of C∧N2Ir(LX) as a to weak phosphorescence. All four C∧N ligands give “(C∧N)2-
function of the LX ligand for acac, tmb, bza, and dbm (see Figure 1). Ir” fragments with triplet levels higher than dbm, so the emission
results from a dbm-based excited state, giving rise to very
complexes. The only chemical difference between these com- inefficient phosphorescence for all four C∧N ligands.
plexes is the substitution at the naphthyl groups. The emission 4. OLEDs Prepared with C∧N2Ir(LX) Complexes. Using
efficiencies, lifetimes, and emission spectral line shapes of the the phosphors discussed above it is possible to prepare efficient
two complexes are similar. The principal difference in the OLEDs, which emit in a variety of colors. The OLED structure
photophysical properties of Rbsn and βbsn complexes are blue used is the same one previously developed for fac-Ir(ppy)3-based
shifts in the phosphorescence and absorption spectra of the βbsn OLEDs, which gave green electroluminescence (EL), at an
relative to the Rbsn complex (e.g., emission λmax ) 593 and external efficiency of 9% (ηext, photons/electrons).30 In the C∧N2-
606 nm, respectively). Both phosphorescence spectra show Ir(acac)-based devices, the Ir phosphor was doped into the
vibronic fine structure, consistent with ligand-based emission. emissive layer of the OLED, at a concentraion of 7 wt %. Figure
Unfortunately, we have been unable to observe phosphorescence 6 (top) shows external quantum efficiency (ηext) and power
from frozen solutions of the Rbsn and βbsn free ligands, so we efficiency as functions of current density for OLEDs with a
cannot determine if the blue shift for the β-isomer is due the ppy2Ir(acac) dopant. A maximum external quantum efficiency
electronic structure of the two ligands. On the basis of the (ηext) of 12.3 ( 0.3% and a power efficiency of 38 ( 3 lm/W31
photophysics of related naphthyl complexes, the βbsn blue shift were obtained at a current density 0.01 mA/cm2. The device
is most likely not related to the Ir substitution. Similar blue showed a gradual decrease in ηext with increasing current
shifts are observed in other acceptor substituted naphthalene density, which is attributed to increasing triplet-triplet an-
derivatives, substituted in the 1 (R)- and 2 (β)-positions. For nihilation of the phosphor-bound excitons.10-12 A maximum
example, the phosphorescence energy of 2-nitronaphthalene is optical output of 32 500 cd/m2 was obtained at J ) 2.4 A/cm2.
blue shifted relative to that of 1-nitronaphthalene by 16 nm.27
OLEDs prepared with btp2Ir(acac) gave a peak quantum
3. C∧N vs LX Centered Emission. Three fac-IrC∧N3
efficiency of 6.6% and a power efficiency of 2.2 lm/W (at 1
complexes have been reported, i.e., fac-Ir(ppy)3,28 fac-Ir(bzq)3,9
mA/cm2). The external quantum efficiency and power efficiency
and fac-Ir(thp)3.29 All three tris-chelate complexes have emission
as functions of current density for btp2Ir(acac) based OLEDs
spectra that are nearly identical to the C∧N2Ir(acac) complexes
are shown in Figure 6 (bottom). These efficiency values are
with the same ligands.9 The phosphorescence quantum yields
the highest values reported for a red emissive OLED. Table 2
and lifetimes for the two classes of complex for a given ligand
summarizes the EL performance with ppy2Ir(acac), bt2Ir(acac),
are also the same. The similarity of the two classes of complexes
and btp2Ir(acac) doped OLEDs. These devices give high
is consistent with emission predominantly from the “C∧N2Ir”
performance with EL emission colors in green, yellow, and red,
fragment in C∧N2Ir(acac) complexes.
respectively. The device structures and layer thickness used here
For all of the complexes discussed thus far, the triplet levels
were identical to those reported for fac-Ir(ppy)3-based OLEDs.10(b)
of the LX ligand (acac) lie well above the energies of the C∧N
They have not been optimized for either low-voltage or high-
ligand and MLCT excited states. Thus, the luminescence is
efficiency operation, so the quantum and power efficiency values
dominated by C∧N and MLCT transitions, leading to efficient
are lower limits. For example, lowering the doping concentration
phosphorescence. If the triplet-state energy of the LX ligand is
of bt2Ir(acac) from 7 to 4% increases the maximum ηext from
lower in energy than the C∧N 3(π-π*) or 3MLCT, however, a
9.5 to 11.9%, Figure 6 (middle). Data are also shown for an
triplet LX level will be the lowest energy excited state. A switch
OLED in which the bt2Ir(acac)-doped CBP film has been
from “C∧N2Ir” to LX-based emission can be seen in a series of
replaced with a pure bt2Ir(acac) luminescent film. Increasing
complexes prepared with different β-diketonates (i.e., acac, tmd,
the concentration of bt2Ir(acac) to 100% lowers the device
bza, and dbm in Figure 1). The phosphorescence quantum
efficiency to 1.5%, due to enhanced self-quenching in the neat
efficiencies for these complexes are provided in the bar graph
bt2Ir(acac) film. The device performances and electrical proper-
of Figure 5. The emission spectra and phosphorescence ef-
ties of the three C∧N2Ir(LX)-based OLEDs are very similar,
ficiencies for the tmd complexes of ppy, bzq, thp, and btp are
which is consistent with similar mechanisms for exciton
(27) Rusakowicz, R.; Testa, A. C. Spectochim. Acta, Part A 1971, 27, formation and trapping at the phosphorescent centers. The lower
787-92.
(28) (a) Garces, F. O.; King, K. A.; Watts, R. J. Inorg. Chem. 1988, 27, (30) Baldo, M. A.; Lamansky, S.; Burrows, P. E.; Thompson, M. E.;
3464-3471. (b) Dedian, K.; Djurovich, P. I.; Garces, F. O.; Carlson, G.; Forrest, S. R. Appl. Phys. Lett. 1999, 75, 4-6.
Watts, R. J. Inorg. Chem. 1991, 30, 1685-1687. (31) The lumen is a unit of optical power that is weighted by the human
(29) Colombo, M. G.; Brunold, T. C.; Riedener, T.; Güdel, H. U.; Förtsch, eye response. For the EL spectrum of ppy2Ir(acac), there are roughly 500
M.; Bürgi, H.-B. Inorg. Chem. 1994, 33, 545-550. lumens/optical watt.
Phosphorescent Bis-Cyclometalated Iridium Complexes J. Am. Chem. Soc., Vol. 123, No. 18, 2001 4311

Table 2. OLED Performance of ppy2Ir(acac)-, bt2Ir(acac)-, and


btp2Ir(acac)-Based OLEDsa
dopant phosphor
ppy2Ir(acac) bt2Ir(acac) btp2Ir(acac)
EL color green yellow red
peak wavelength (nm) 525 565 617
CIE-x 0.31 0.51 0.68
CIE-y 0.64 0.49 0.33
luminance 3.300 2500 470
at 10 mA/cm2(cd/m2)
drive voltage (V) 9.3 9.5 11.6
at 10 mA/cm2
ext quantum effic (%)
at 1 mA/cm2 10.0 9.7 6.6
at 10 mA/cm2 7.6 8.3 6.0
at 100 mA/cm2 5.4 5.5 4.6
power effic (lm/W) 18 11 2.2
at 1 mA/cm2
a Device structure, ITO/R-NPD (500 Å)/doped CBP (300 Å)/BCP

(100 Å)/Alq3 (400 Å)/Mg-Ag. All three devices were prepared with
the dopant at a 7% loading in the CBP layer.

efficiency of the btp 2Ir(acac)-based OLED relative to the ppy2-


Ir(acac) and bt2Ir(acac) may be due to a lower phosphorescence
efficiency of the btp complex relative to the ppy and bt
complexes.
The EL spectra of the C∧N2Ir(acac)-based devices match those
of the same phosphors in a dilute solution. Thus, all EL emission
originates from the triplet excited states of the phosphors. The
Commission Internationale de L’Eclairage (CIE) coordinates for
the three OLEDs are shown in Figure 7 (left). The CIE system
is the standard for evaluating color quality for visual applica-
tions.32 The btp2Ir(acac)-doped OLED gives a saturated red
emission that has CIE coordinates close to the National
Television Standards Committee (NTSC) recommended red for
a cathode ray tube (CRT). Green emission from ppy2Ir(acac) is
very similar to Ir(ppy)3-based OLEDs and a common fluores-
cence-based green OLED (Coumarin6:Alq3). The three com-
plexes used to fabricate OLEDs were chosen to be representative
of the family of C∧N2Ir(LX) phosphors. Figure 7 (right) shows
the CIE coordinates of the solution phosphorescent spectra of
all of the C∧N2Ir(acac) complexes reported here. All of the C∧N2-
Ir(acac) complexes are expected to give OLEDs with efficiencies
similar to those reported for the ppy, bt, and btp complexes.
We expect that the EL and photoluminescence spectra of these
phosphors are similar, so the coordinates of the phosphorescence
spectra shown in Figure 7 (right) will likely be the same for
OLEDs prepared with these phosphors. The C∧N2Ir(acac) are
clearly a broad and widely tunable class of OLED phosphors.

Summary
It has been reported that phosphorescence in iridium tris-
cyclometalated complexes comes from a mixture of ligand
centered, 3(π-π*), and 3MLCT excited states.7,33 While phos-
phorescence results from a mixture of the two triplet states (and
the 1MLCT via spin-orbit coupling), the emissive excited state
is predominantly 3(π-π) or 3MLCT, depending on the energies
of the two excited states. The same situation is present in the
C∧N2Ir(LX) complexes. ppy2Ir(acac), tpy2Ir(acac), and bzq2Ir-
Figure 6. External quantum and power efficiencies of OLEDs using
ppy2Ir(acac):CBP (top), bt2Ir(acac):CBP and bt2Ir(acac) (middle), and
(acac) have high-energy 3(π-π)* states and, thus, give unstruc-
btp2Ir(acac):CBP (bottom) emissive layers. The inset to the top plot tured, predominantly 3MLCT emission. The other complexes
shows the OLED structure used in this study. The EL spectrum of each reported here give phosphorescence spectra with a reasonable
device is shown as an inset on the relevant plots. R-NPD ) 4,4′-bis-
(32) Whitaker, J., Electronic Displays: Techlnology, Design, and Ap-
[N-(naphthyl-N-phenylamino)biphenyl, CBP ) 4,4′-N,N′-dicarbazolyl- plications; McGraw-Hill: New York, 1994; p 92.
biphenyl, BCP ) 2,9-dimethyl-4,7- diphenyl-1,10-phenanthroline (ba- (33) Colombo, M. G.; Hauser, A.; Gudel, H. U. Inorg. Chem. 1993, 32,
thocuproine), and Alq3 ) tris(8-hydroxyquinoline)aluminum. 3088-3092.
4312 J. Am. Chem. Soc., Vol. 123, No. 18, 2001 Lamansky et al.

Figure 7. The Commission Internationale de L′Eclairage (CIE) chromaticity coordinates of OLEDs and phosphorescence spectra of C∧N2Ir(LX)
complexes. The CIE coordinates for OLEDs with ppy2Ir(acac):CBP, bt2Ir(acac):CBP, and btp2Ir(acac)Ir:CBP are shown relative to the fluorescence-
based devices, coumarine6:Alq3 and DCJT:Alq3 on the left. The CIE coordinates of the phosphorescence spectra of many of the C∧N2Ir(LX)
complexes prepared here are shown to the right. The NTSC standard coordinates for the red, green, and blue subpixels of a CRT are at the corners
of the black triangle.

degree of vibronic fine structure and significant Stokes shifts, can result in a reduction of operating voltage and increased
consistent with predominantly 3(π-π)* C∧N-based emission. quantum efficiency.
By changing the C∧N ligands in cyclometalated Ir complexes,
Acknowledgment. The authors thank Universal Display
we demonstrated green to red electrophosphorescence with high
Corp., the Defense Advanced Research Projects Agency, and
ηext. High phosphorescence efficiencies and lifetimes less than
the National Science Foundation for financial support of this
10 µs resulted in record high-performance OLEDs operating
work.
from the green to the red. We note that the device structure has
the potential for further optimization. For example, the use of Supporting Information Available: Synthetic and spec-
Li-based cathodes,12,34 optimization of dopant concentration and troscopic data for all of the C∧N2Ir(LX) complexes where LX
thickness of organic layers, and other device configurations11 ) tmd, bza, and dbm reported in this paper. This material is
available free of charge via the Internet at http://pubs.acs.org.
(34) Hung, L. S.; Tang, C. W.; Mason, M. G. Appl. Phys. Lett. 1997,
70, 152. JA003693S

You might also like