Red to blue emitting cationic iridium complexes with 2-phenyl-4-dimethylaminopyridine as the cyclometalating ligand: Synthesis, characterization and electroluminescent devices

Article abstract of DOI:10.1016/j.dyepig.2019.03.003

2-Phenyl-4-dimethylaminopyridine (dmappy) with high ³π-π^? triplet energy is, for the first time, employed as the cyclometalating ligand for constructing cationic iridium complexes with tunable emission properties. Three complexes, na

Full text of DOI:10.1016/j.dyepig.2019.03.003

Accepted Manuscript
Red to blue emitting cationic iridium complexes with 2-phenyl-4-
dimethylaminopyridine as the cyclometalating ligand: Synthesis, characterization and
electroluminescent devices
Xianwen Meng, Rubing Bai, Xiaoxiang Wang, FangFang Pan, Lei He
PII:
S0143-7208(19)30110-X
https://doi.org/10.1016/j.dyepig.2019.03.003
DYPI 7392
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 15 January 2019
Revised Date: 3 March 2019
Accepted Date: 3 March 2019
Please cite this article as: Meng X, Bai R, Wang X, Pan F, He L, Red to blue emitting cationic
iridium complexes with 2-phenyl-4-dimethylaminopyridine as the cyclometalating ligand: Synthesis,
characterization and electroluminescent devices, Dyes and Pigments (2019), doi: https://
doi.org/10.1016/j.dyepig.2019.03.003.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Red to Blue Emitting Cationic Iridium Complexes with
2-phenyl-4-dimethylaminopyridine as the Cyclometalating
Ligand: Synthesis, Characterization and Electroluminescent
Devices
Xianwen Meng,a,1 Rubing Bai,a,1 Xiaoxiang Wang,a FangFang Pan,a Lei He a*
a College of Chemistry, Key Laboratory of Pesticide & Chemical Biology of Ministry
of Education, Hubei International Scientific and Technological Cooperation Base of
Pesticide and Green Synthesis, International Joint Research Center for Intelligent
Biosensing Technology and Health, Central China Normal University, Wuhan
430079, P. R. China.
1 These authors contribute equally to this work.
E-mail: helei@mail.ccnu.edu.cn
Abstract
3
2-Phenyl-4-dimethylaminopyridine (dmappy) with high π-π* triplet energy is, for
the first time, employed as the cyclometalating ligand for constructing cationic
iridium complexes with tunable emission properties. Three complexes, namely,
[Ir(dmappy)2(bpy)]PF6
[Ir(dmappy)2(dma-pzpy)]PF6 (3), have been designed and synthesized (bpy = 2,
2′-bipyridine, pzpy 2-(1H-pyrazol-1-yl)pyridine and dma-pzpy
(1),
[Ir(dmappy)2(pzpy)]PF6
(2)
and
=
=
4-dimethylamino-2-(1H-pyrazol-1-yl)pyridine), and their photophysical and
electrochemical properties have been comprehensively investigated. By using
1

ACCEPTED MANUSCRIPT
ancillary ligands with increasing electron-richness (bpy → pzpy → dmapzpy),
complexes 1-3 emit red, green and blue light, respectively. Theoretical calculations
reveal that for complexes 1 and 2, the light emission stems predominantly from the
charge-transfer states (Ir/dmappy → bpy or pzpy), whereas for complex 3, the light
3
emission originates mainly from the dmappy-centered π-π* state. In the doped film,
fluorine-free complex 3 affords comparable emission color, luminescent efficiency
and excited-state lifetime to the typical blue-emitting, neutral iridium complex FIrpic
that uses fluorinated ligands. Single-layer, solution-processed organic light-emitting
diodes based on complex 3 affords blue electroluminescence (EL) peaked at 472 nm,
with the EL color competing with that from FIrpic and the peak current efficiency
(12.6 cd A−1) being among the highest reported for single-layer, solution-processed
blue OLEDs based on ionic iridium complexes.
Keywords: cationic iridium complex; color-tuning; blue; organic light-emitting diode
1. Introduction
Phosphorescent cationic iridium complexes have drawn growing research
interest in the past decade, due to their intrinsic ionic nature, high phosphorescent
efficiency, tunable emission color and rich redox properties. For these desirable
features, they have been widely employed as functional agents in biological imaging
[1,2], sensing [1,2], photocatalysis [3] and optoelectronic devices [4-7], to mention
but a few. In particular, they have been widely used as emitting materials in
light-emitting electrochemical cells (LECs) or as dopants in solution-processed
2

ACCEPTED MANUSCRIPT
organic light-emitting diodes (OLEDs) [4,6-9], which, as efficient emissive thin-film
devices, have emerged as promising candidates for next-generation solid-state lighting
or display applications.
Huge efforts have been devoted to tune the emission properties of cationic
iridium complexes, which is realized by control of both cyclometalating (CΛN) and
ancillary (NΛN) ligands [4,6,7,10]. The earliest reported phosphorescent cationic
iridium complex, [Ir(ppy)2(bpy)]PF6, with 2-phenylpyridine (Hppy) as the CΛN ligand
and 2,2′-bipyridine (bpy) as the NΛN ligand, emits orange-red light [11,12]. To shift
the light emission toward blue, the energy gap between the highest occupied
molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO)
needs to be widened [10,13,14]. To this end, various approaches have been adopted
[6,10]. For instance, electron-withdrawing groups (such as F, −CF3, −SF5) have been
attached at the phenyl ring of ppy to stabilize the HOMO [15-17] and
electron-donating groups (such as dimethylamino) have been attached at the pyridine
ring of bpy to destabilize the LUMO [18]; moreover, electron-rich NΛN ligands, such
as
2-(1H-pyrazo1-1-yl)pyridine
(pzpy)
and
4-dimethylamino-2-(1H-pyrazo1-1-yl)pyridine (dma-pzpy), have been developed to
destabilize significantly the LUMO [19-22]. For cationic iridium complexes, it has
been unveiled that along with the emission blue-shift, the character of the emitting
triplet state could change. For example, for orange-emitting [Ir(ppy)2(bpy)]PF6, the
emitting triplet state exhibits mixed metal-to-ligand charge-transfer (3MLCT) (Ir→
bpy)/ligand-to-ligand charge-transfer (3LLCT) (ppy→bpy) character [12,23], whereas
3

ACCEPTED MANUSCRIPT
for the blue-green-emitting [Ir(ppy)2(pzpy)]PF6, the emitting triplet state shows mixed
3MLCT (Ir→ppy)/3π-π* (ppy-centered) character [19,23]. For the latter complex, the
3
emission is mainly controlled by the π-π* state of the ppy CΛN ligand. Therefore,
3
CΛN ligands with high π-π* triplet energy are preferred for constructing cationic
iridium complexes with blue-shifted light emission.
It is noted that attaching electron-donating alkoxyl groups at the para position of
3
the pyridine ring in ppy can effectively enhance the π-π* triplet energy of the ligand
[22,24]. With a stronger electron-donating dimethylamino group at the para position
of the pyridine ring, 2-phenyl-4-dimethylaminopyridine (dmappy) possesses even
3
higher π-π* triplet energy and thus, is a suitable CΛN ligand for color-tuning toward
blue for cationic iridium complexes. Moreover, dmappy contains no fluorine
substitutions, which is beneficial for constructing blue-emitting iridium complexes,
because fluorine substitutions decrease the thermal and electrochemical stability of
the complexes [25,26]. To our knowledge, dmappy has so far not been examined as
the CΛN ligand for cationic iridium complexes, especially for blue-emitting ones.
Herein, we report three red to blue emitting cationic iridium complexes with
dmappy CΛN ligands, namely, [Ir(dmappy)2(bpy)]PF6 (1), [Ir(dmappy)2(pzpy)]PF6 (2)
and [Ir(dmappy)2(dma-pzpy)]PF6 (3) (Scheme 1). By using NΛN ligands with
increasing electron-richness (bpy → pzpy → dmapzpy), complexes 1-3 afford red,
green and blue light emission, respectively. Theoretical calculations reveal that from
complexes 1 and 2 to complex 3, the emitting triplet state changes from mixed
3
3MLCT (Ir→NΛN)/3LLCT (dmappy→NΛN) to mixed MLCT (Ir→dmappy)/3π-π*
4

ACCEPTED MANUSCRIPT
(dmappy-centered). In the lightly doped film, fluorine-free blue-emitting complex 3
affords comparable emission properties to the typical blue-emitting, neutral iridium
complex FIrpic that uses fluorinated ligands [27]. Single-layer, solution-processed
OLEDs using the complexes as dopants have been fabricated. The blue OLED based
on complex 3 affords comparable electroluminescence (EL) color to that based on
FIrpic and its efficiency (12.6 cd A−1) is among the highest reported for single-layer,
solution-processed blue OLEDs based on ionic iridium complexes [20,22,28-33].
2. Experimental
General experiments. All reactants and solvents, unless otherwise stated, were
1
purchased from commercial sources and used as received. H-NMR spectra were
recorded on a BRUKER 400/500 NMR spectrometer with Si(CH3)4 as the internal
standard. Mass spectrometry was conducted with a LTQ-ORBITRAP-ETD mass
spectrometer. Elemental analysis was determined on an EA3000 elemental analyzer
(Eurovector, Italy). Thermogravimetric analysis (TGA) was performed on a
simultaneous thermal analyzer (SDTQ650) with a temperature increasing rate of 10
oC/min under a nitrogen atmosphere. Absorption and photoluminescence (PL) spectra
were recorded with a UV-vis spectrophotometer (Agilent 8453) and a
fluorospectrophotometer (Jobin Yvon, FluoroMax-3), respectively. The PL transient
lifetimes were measured on a transient spectrofluorimeter (Edinburgh Instruments,
FLSP920) with a time-correlated single-photon counting technique. The
photoluminescent quantum yields (PLQYs) (excited at 380 nm) in degassed CH3CN
5

ACCEPTED MANUSCRIPT
were measured with Ru(bpy)3Cl2 (Φp = 0.028 in water) as the standard for complex 1
and quinine sulfate (Φp = 0.545 in 1 M H2SO4) as the standard for complexes 2 and 3.
The PLQYs in thin films were measured with an absolute PLQY measurement system
(Hamamatsu C11347) equipped with an integrating sphere. Cyclic voltammetry was
performed on a LK1100 voltammetric analyzer in CH2Cl2 solution (10-3 M) for
oxidation and in N,N-dimethylformamide (DMF) solution (10-3 M) for reduction at a
scan rate of 100 mV s-1, with a glass-carbon working electrode, a Ag wire reference
electrode and a Pt wire counter-electrode. Tetrabutylammonium hexafluorophosphate
(0.1 M) and ferrocene were used as the supporting electrolyte and the internal
standard, respectively. The solutions were degassed with argon before measurements.
Synthesis. The ligands pzpy and dma-pzpy were synthesized according to reported
procedures [19,22].
Synthesis of 4-chloro-2-phenylpyridine. Phenylboronic acid (3.0 g, 25 mmol),
2,4-dichloropyridine (3.7 g, 25 mmol), Pd(PPh3)4 (0.87 g, 0.75 mmol) and K2CO3
(10.4 g, 75 mmol) were dissolved in 1,2-dimethoxyethane/H2O (70/30 mL). The
mixture was stirred at 90 ºC under a nitrogen atmosphere for 18 h. After being cooled
to room temperature, the mixture was extracted with CH2Cl2 (3×50 mL). The organic
phase was washed with brine, dried over anhydrous Na2SO4 and purified by column
chromatography on silica gel with petroleum ether/CH2Cl2 (5:1) as the eluent,
1
yielding a light yellow oil (3.1 g, 16.9 mmol).Yield: 65%. H-NMR (CDCl3, 400
MHz, δ[ppm]): δ 8.59 (d, J = 5.3 Hz, 1H), 8.00-7.95 (m, 2H), 7.74 (d, J = 1.8 Hz, 1H),
6

ACCEPTED MANUSCRIPT
7.52– 7.42 (m, 3H), 7.24-7.27 (m, 1H). HRMS (ESI, m/z): 190.0407 [M+H]+ (calc.
190.0418).
Synthesis of dmappy. 4-chloro-2-phenylpyridine (2.8 g, 15 mmol), dimethyllamine
hydrochloride (12.2 g, 150 mmol) and sodium hydroxide (6.0 g, 150 mmol) were
dissolved in DMF/H2O (50/25 mL). The mixture was stirred at room temperature for
30 minutes and then potassium carbonate (4.1 g, 30 mmol) was added. The mixture
º
was further stirred at 130 C for 10 h in a sealed tube. After being cooled to room
temperature, the mixture was extracted with CH2Cl2 (80 mL). The organic phase was
washed with brine, dried over anhydrous Na2SO4 and purified by column
chromatography on silica gel with CH2Cl2/CH3OH (50:1) as the eluent, yielding a
1
white solid (2.4 g, 12.1 mmol).Yield: 81%. H-NMR (CDCl3, 500 MHz, δ[ppm]):
8.37 (d, J = 6.4 Hz, 1H), 8.00 – 7.94 (m, 2H), 7.53 – 7.42 (m, 3H), 6.89 (d, J = 2.6 Hz,
1H), 6.56 (dd, J = 6.5, 2.5 Hz, 1H), 3.15 (s, 6H). HRMS (ESI, m/z): 199.1215 [M+H]+
(calc. 199.1230).
Synthesis of [Ir(dmappy)2Cl]2. The chloro-bridged iridium dimer [Ir(dmappy)2Cl]2
was prepared by refluxing IrCl3.nH2O and dmappy (2.5 eq.) in 2-ethoxyethanol/water
(3:1) under nitrogen[34,35]. The precipitate was collected, dried under vacuum and
used without further purification.
Synthesis of Ir[(dmappy)2(bpy)]PF6 (1). [Ir(dmappy)2Cl]2 (286 mg, 0.23 mmol) and
bpy (80 mg, 0.51 mmol) were dissolved in CH2Cl2/CH3OH (10/10 mL).The mixture
was refluxed overnight under a nitrogen atmosphere and then cooled to room
temperature. To the solution, NH4PF6 (375 mg, 2.3 mmol) was added. The mixture
7

ACCEPTED MANUSCRIPT
was stirred at room temperature for 2 h and the solvent was removed under vacuum.
The remaining solid was washed with water and dried under vacuum. The solid was
purified by column chromatography on silica gel (200-300 mesh) with
CH2Cl2/acetone as the eluent, yielding a red solid (123 mg, 0.14 mmol). Yield: 30%.
1H-NMR (DMSO-d6, 500 MHz, δ[ppm]): 8.84 (d, J = 8.2 Hz, 2H), 8.23 (td, J = 8.0,
1.4 Hz, 2H), 7.94 – 7.82 (m, 4H), 7.70−7.64 (m, 2H), 7.25 (d, J = 2.5 Hz, 2H), 6.98 –
6.92 (m, 4H), 6.86 (t, J=7.5 Hz, 2H), 6.42 – 6.38 (m, 2H), 6.34 (d, J = 7.4 Hz, 2H),
3.08 (s, 12H). HRMS (ESI, m/z): 743.2487 [M−PF6]+ (calc. 743.2469). Anal. found:
C, 48.59; H, 3.97; N, 9.41. Anal. calcd for C36H34F6IrN6P: C, 48.70; H, 3.86; N, 9.47.
Synthesis of Ir[(dmappy)2(pzpy)]PF6 (2). Complex 2 was synthesized with a similar
procedure to that for complex 1, except that bpy was replaced by pzpy, yielding a
1
yellow solid. Yield: 35%. H-NMR (DMSO-d6, 500 MHz, δ[ppm]): 9.26 (d, J = 3.0
Hz, 1H), 8.49 (d, J = 8.5 Hz, 1H), 8.30 (td, J=8.0, 1.5 Hz, 1H), 7.88 – 7.80 (m, 2H),
7.67 (dd, J = 5.5, 1.0 Hz, 1H), 7.54−7.49 (m, 1H), 7.28−7.23 (m, 2H), 7.22 (d, J = 2.0
Hz, 1H), 7.12 (d, J = 7.0 Hz, 1H), 7.09 (d, J = 7.0 Hz, 1H), 6.97 – 6.87 (m, 3H), 6.85
(td, J = 7.4, 1.3 Hz, 1H), 6.80 (td, J = 7.4, 1.3 Hz, 1H), 6.47 (dd, J = 7.0, 2.9 Hz, 1H),
6.44 (dd, J = 7.0, 2.9 Hz, 1H), 6.37−6.32 (m, 2H), 3.12−3.08 (m, 12H). HRMS (ESI,
m/z): 732.2385 [M−PF6]+ (calc. 732.2421). Anal. found: C, 46.42; H, 3.86; N, 11.10.
Anal. calcd for C34H33F6IrN7P: C, 46.57; H, 3.79; N, 11.18.
Synthesis of Ir[(dmappy)2(dma-pzpy)]PF6 (3). Complex 3 was synthesized with a
similar procedure to that for complex 1, except that bpy was replaced by dma-pzpy,
yielding a light yellow solid. Yield: 28%. 1H-NMR (CD3CN, 500 MHz, δ[ppm]): 8.68
8

ACCEPTED MANUSCRIPT
(d, J = 3 Hz, 1H), 7.74−7.68 (m, 2H), 7.28 (d, J = 7.0 Hz, 1H), 7.21 (d, J = 6.5 Hz,
1H), 7.18−7.11(m, 4H), 7.03 (d, J = 2.5 Hz, 1H), 6.96−6.84 (m, 2H), 6.86−6.77 (m,
2H), 6.69−6.66 (m, 1H), 6.50 (dd, J = 7.0, 2.5Hz, 1H), 6.43−6.31(m, 4H), 3.13−3.09
(m, 18H). HRMS (ESI, m/z): 775.2809 [M−PF6]+ (calc. 775.2843). Anal. found: C,
46.89; H, 4.24; N, 12.11. Anal. calcd for C36H38F6IrN8P: C, 47.00; H, 4.16; N, 12.18.
X-ray crystallography. The single-crystal X-ray experiments were conducted on a
Rigaku CCD Saturn 724＋deffractometer equipped with graphite monochromatized
Mo Kα radiation. Direct phase determination led to the positions of all non-hydrogen
atoms which were subjected to anisotropic refinement. All hydrogen atoms were
produced theoretically and rode on their parent atoms in the final refinement. All
computations were conducted using the SHELXTL program. Single crystal structures
of complexes 1 and 2 were deposited to Cambridge Crystallographic Data Centre
(CCDC) with deposition numbers of 1887889 and 1887890, respectively.
Theoretical calculations. For calculations on the ground and excited electronic states,
density functional theory (DFT) at the B3LYP level was employed [36,37]. “Double-ξ”
quality basis sets were used for C, H and N (6-31G**) and the Ir (LANL2DZ). An
effective core potential (ECP) replaces the inner core electrons of iridium, leaving the
outer core (5s)2(5p)6 electrons and the (5d)6 valence electrons of Ir(III). The
geometries of singlet ground states (S0) were fully optimized with the C1 symmetry.
The lowest-lying ten triplets were calculated at the optimized S0 geometries with the
time-dependent DFT (TD-DFT) approach. The geometries of the triplet excited states
were fully optimized at the spin-unrestricted B3LYP level with a spin multiplicity of 3.
9

ACCEPTED MANUSCRIPT
All the calculations were performed in the presence of solvent (CH3CN). Solvent
effects were considered within the self-consistent reaction field (SCRF) theory
adopting the polarized continuum (PCM) model [38]. All calculations were conducted
with Gaussian 09 software package [39]. The geometries of selected triplet states were
directly compared with the S0 geometries by using the VMD program [40]. The root
mean square displacement/deviation (RMSD) values between the two geometries
were given by the VMD program.
Fabrication and characterizations of OLEDs. Indium-tin-oxide (ITO) substrates
with a sheet resistance of 15 Ω/□ were sufficiently cleaned in detergent and then in
deionized water, and treated with UV-zone before use. The PEDOT: PSS layers were
deposited onto the ITO substrates in air and baked at 200 °C for 10 minutes. The
substrates coated with PEDOT: PSS were transferred into a nitrogen-filled glove box,
in which the emissive layers were spin coated on top of PEDOT:PSS from
1,2-dichloroethane solution. The films were baked at 80 °C for 30 minutes and then
transferred into a vacuum chamber, where LiF and Al were evaporated at 0.1−0.2 Ǻ/s
and 5−8 Ǻ/s, respectively. The devices were characterized with a Keithley 4200
semiconductor characterization system. The EL spectra were recorded with a Photo
Research PR705 spectrophotometer under ambient conditions.
3. Results and discussion
3.1. Synthesis and structural characterizations
10

ACCEPTED MANUSCRIPT
Cl
N
Cl
B(OH)2
NH HCl
N
N
Pd(PPh3)4
K2CO3
N
+
NaOH/K2CO3
Cl
dmappy
+
N
N
N
N
N
N
N
N
N
Cl
Cl
N
Ir
_
NH4PF6
(PF6)
+
Ir
Ir
N
2
2
N
N
N
N
N
N
N
=
N
N
N
N
N
Complex 1 Complex 2
Complex 3
Scheme 1. Synthesis of complexes 1-3.
Scheme 1 depicts the synthetic routes to dmappy and complexes 1-3. Dmappy was
previously synthesized by a Pd(0)-catalyzed Suzuki coupling reaction between
phenylboronic acid and 2-iodo-4-dimethylaminopyridine with just a moderate yield of
46%; moreover, the synthesis of the intermediate 2-iodo-4-dimethylaminopyridine
required a complex procedure including the use of n-butyllithium [35]. In this work,
dmappy was readily synthesized by a nucleophilic substitution reaction between
4-chloro-2-phenylpyridine and dimethylamine with a high yield of 81% (Scheme 1).
The intermediate 4-chloro-2-phenylpyridine was conveniently prepared by a Suzuki
coupling reaction between phenylboronic acid and 2,4-dichloropyridine (Scheme 1).
11

ACCEPTED MANUSCRIPT
Complexes 1-3 were prepared by cleaving the chloro-bridged iridium dimer,
[Ir(dmappy)2Cl]2, with the NΛN ligand, followed by an anion-exchanging reaction
−
from Cl− to PF6 . Chemical structures of all the complexes were verified by NMR
spectroscopy, high-resolution mass spectrometry and elemental analysis (see
Experimental). Thermal stability of complexes 1-3 was characterized by
thermogravimetric analysis (Fig. S1), which shows that complexes 1-3 exhibit high
decomposition temperature (5% weight loss) of 380, 330 and 305 oC, respectively.
Figure 1. Single crystal structures of complexes 1 and 2. Thermal ellipsoids were
−
drawn at 30% probability. PF6 counter anions, solvent molecules and all hydrogen
atoms were omitted for clarity.
Single crystals of complexes 1 and 2 were grown from solution and characterized
12

ACCEPTED MANUSCRIPT
by X-ray experiments. Fig. 1 depicts the determined single-crystal structures.
Pertinent bond lengths and ligand bite angles are presented in Table S1 in the
supporting information. The geometries of the iridium centers are distorted
octahedrons with the two dmappy ligands adopting C, C-cis and N, N-trans
configurations. The Ir−C, Ir−N bond lengths and the ligand bite angles fall within
those found in analogous complexes [19,22]. It is noted that the C−N bonds
(1.32−1.36 Å) connecting the dimethylamino groups and the pyridine rings in the
dmappy ligands resemble the typical C=N double bond (1.35 Å), which indicates that
the nitrogen atoms of the dimethylamino groups are in conjugation with the pyridine
rings, i.e., they donate their lone-pair electrons to the pyridine rings. For this reason,
the dmappy ligand bear higher 3π-π* triplet energy than the ppy ligand [22,24].
3.2. Photophysical properties
Figure 2. Absorption and PL spectra of complexes 1-3 in CH3CN solution.
13

ACCEPTED MANUSCRIPT
Table 1. Photophysical and electrochemical data of complexes 1-3.
Emission at room temperature
Absorption
PL at Electrochemi-
cal data e
CH3CN b
2 wt.% in PMMA c
77 k
λ [nm]
(ε[×104
[nm]
d
PL
Φem
kr, knr
PL
Φem
kr, knr
Eox
[V]
Ered
[V]
[nm] (τ[µs]) [105 s-1]
[nm] (τ[µs]) [105 s-1]
M-1 cm-1]) a
1
2
3
486 (0.09)
405 (0.48)
383 (0.97)
270 (6.15)
450 (0.08)
407 (0.35)
382(1.03)
269 (6.50)
446 (0.04)
402 (0.40)
380 (1.08)
266 (7.76)
620
0.03
1.8, 57
1.8, 14
3.6, 87
582
491
0.50
6.6, 6.6
3.8, 2.0
4.8, 1.4
563
0.61 −1.85
(0.17)
(0.76)
526
0.12
0.65
453, 0.59 −2.32 ir
483
(0.65)
(1.72)
464,
489
0.04
465,
489
0.76
459, 0.53 −2.62 ir
491
(0.11)
(1.59)
c
a In CH3CN solution (10-5 M). b In degassed CH3CN solution (10-5 M). Films coated
d
e
on quartz substrates. In CH3CN glass at 77 K. Potentials were recorded versus
Fc+/Fc; the superscript “ir” denotes irreversible reduction processes.
Fig. 2 depicts the absorption and PL spectra of complexes 1-3 in CH3CN solution.
Detailed photophysical characteristics were summarized in Table 1. As shown in Fig.
2, complexes 1-3 exhibit strong absorption bands in the ultra-violet region (below 350
1
nm), which arise from the spin-allowed ligand-centered π-π* transitions, and
relatively weak absorption bands (above 350 nm) that extend to the visible region,
which can be tentatively assigned to 1MLCT/3MLCT, 1LLCT/3LLCT and
3
ligand-centered π-π* transitions [11,41,42]. The spin-forbidden triplet transitions
become partially permitted owing to the strong spin-orbital coupling endowed by the
heavy iridium atoms.
In degassed CH3CN solution, complexes 1-3 emit red, green and blue light with
14

ACCEPTED MANUSCRIPT
the emission peaks at 620, 526 and 464 nm, respectively, with their emission spectra
spanning nearly the entire visible spectrum (Fig. 2). As shown in Fig. 2, complexes 1
and 2 afford featureless and broad PL spectra, which indicates that their light emission
comes predominantly from charge-transfer (CT) states, whereas complex 3 affords a
structured and narrowed PL spectrum, which indicates that its light emission
originates mainly from a ligand-centered 3π-π* state. Apparently, the blue-shift of light
emission is accompanied with the change of the character of emitting triplet state
3
3
from dominant CT to ligand-centered π-π*. The underlying mechanism for this
character change will be discussed in theoretical calculations.
The PL spectra of complexes 1-3 in the frozen CH3CN glass at 77 K were
measured, as shown in Fig. S2 in the supporting information. At 77 K, complex 1
exhibits an almost featureless PL spectrum that is largely blue-shifted by about 60 nm
compared to the room-temperature PL spectrum, which agrees with the dominant 3CT
character assigned to the light emission of complex 1 [43]. At 77 K. complex 2
affords a structured PL spectrum peaked at 483 nm, which indicates the development
3
of ligand-centered π-π* character in the low-temperature emission, in contrast with
3
the dominant CT character in the room-temperature emission. It is thus suggested
that the 3CT state and the ligand-centered 3π-π* state in complex 2 lie close in energy,
the order of which depend on the local environment (such as temperature) [44]. At 77
K, complex 3 exhibits a highly structured PL spectrum that is slightly blue-shifted by
about 5 nm compared to the room temperature PL spectrum, which agrees with the
dominant ligand-centered 3π-π* character assigned to the light emission of complex 3.
15

ACCEPTED MANUSCRIPT
In the degased CH3CN solution, complexes 1-3 afford PLQYs of 0.03, 0.12 and
0.04, respectively, and excited-state lifetimes of 0.17, 0.65 and 0.11 µs, respectively
(Table 1). According to these PLQY and excited-state lifetime values, the radiative (kr)
and non-radiative (knr) decay rates in the degassed CH3CN solution were calculated
and summarized in Table 1. Complex 1 exhibits a normal kr of 1.8×105 s−1 but quite
a large knr of 5.7×106 s−1. The large knr of red-emitting complex 1 can be explained
by the energy gap law which predicts that a narrower energy gap leads to a larger knr
[45-47]. Complex 2 exhibits kr and knr of 1.8×105 s−1 and 1.4×106 s−1, respectively.
Compared to complex 1, green-emitting complex 2 exhibits a largely decreased knr,
which agrees with the energy gap law. However, compared to typical green-emitting
iridium complexes which show knr values in the order of 105 s−1, complex 2 still
exhibits a relatively large knr. Blue-emitting complex 3 exhibits a kr of 3.6×105 s−1
and a knr of 8.7×106 s−1 that is the largest among the series of complexes, which
seems to disobey the energy gap law. The relatively large knr of complex 2 and the
abnormally large knr of complex 3 indicate that there exist efficient non-radiative
deactivation channels for the emitting triplet states of complexes 2 and 3 in the
solution.
Emission properties of complexes 1-3 in the
2
wt.
%
doped
poly(methylmethacrylate) (PMMA) films were measured, with the PL spectra
presented in Fig. S3 in the supporting information and the photophysical data
summarized in Table 1. In the PMMA film, complex 1 affords orange-red light
emission with a featureless PL spectrum centered at 582 nm, a PLQY of 0.50 and an
16

ACCEPTED MANUSCRIPT
excited-state lifetime of 0.76 µs; complex 2 emits green-blue light with an almost
featureless PL spectrum centered at 491 nm, a PLQY of 0.65 and an excited-state
lifetime of 1.72 µs; complex 3 emits blue light with a structured PL spectrum peaked
at 465 nm, a PLQY of 0.76 and an excited-state lifetime of 1.59 µs. The spectrum
profiles indicate that in the lightly doped films, the emission of complexes 1 and 2 and
3
the emission of complex 3 originate predominantly from the CT states and the
ligand-centered 3π-π* state, respectively, similar to those observed in the solution. It is
worth noting that in the doped film, fluorine-free blue-emitting complex 3 exhibits
comparable emission peak, PLQY and excited-state lifetime to the typical
blue-emitting neutral iridium complex FIrpic that uses fluorinated CΛN ligands [27].
In the doped PMMA film, FIrpic affords an emission peak at 468 nm with a PLQY of
0.89 and an excited state lifetime 1.7 µs [48].
According to the PLQY and excited-state lifetime values, the kr and knr values in
the doped PMMA films were calculated and summarized in Table 1. Compared to
those in the solution, the knr values in the PMMA films are significantly reduced,
owing to the suppressed non-radiative deactivations in the solid rigid films. On the
other hand, the kr values are remarkably enhanced compared to those in the solution
3
(Table 1). In general, for blue light emission with dominant ligand-centered π-π*
character, the kr value is usually small because of the reduced 3MLCT character in the
emitting triplet state [49]. However, complex 3 still exhibits a relatively large kr of 4.8
×105 s-1. This result agrees with the previous work on blue-emitting complexes using
dma-pzpy NΛN ligand, in which it was revealed that the highly electron-rich
17

ACCEPTED MANUSCRIPT
dma-pzpy ligand is capable of enhancing the kr values of blue-emitting complexes[22].
Owing to the suppressed knr and enhanced kr, complex 3 exhibits relatively high
PLQY and short excited-state lifetime in the doped film, which benefits its application
as dopants in OLEDs.
3.3. Electrochemical properties
Figure 3. Cyclic voltammograms of complexes 1-3 in solution.
Electrochemical properties of complexes 1-3 were measured by cyclic
voltammetry in solution. Fig. 3 depicts the cyclic voltammograms. Complexes 1-3 all
show reversible oxidation processes, with oxidation potentials at 0.61, 0.59 and 0.53 V,
respectively. With these oxidation potentials, the HOMO levels of complexes 1-3
were calculated to be −5.41, −5.39 and −5.33 eV, respectively. Complexes 1-3 show
similar HOMO levels (within 0.08 eV), although the trend is that the HOMO level is
slightly destabilized from complex 1, to 2 and to 3.
18

ACCEPTED MANUSCRIPT
As shown in Fig. 3, complex 1 shows a reversible reduction process, with the
reduction potential at −1.85 V, whereas complexes 2 and 3 exhibit irreversible
reduction processes, with the peak reduction potentials at −2.32 and −2.62 V,
respectively. From these reduction potentials, the LUMO levels of complexes 1-3
were calculated to be −2.95, −2.48 and −2.18 eV, respectively. From complex 1, to 2
and to 3, the LUMO level is stepwisely destabilized by a large extent (0.30−0.47 eV)
[19,22]. As a result, the energy gap is largely enlarged from complex 1, to 2 and to 3,
which consists with the gradual blue-shift of light emission experimentally observed.
3.4. Theoretical calculations
Figure 4. Optimized S0 geometries and surface distributions (|Ψ| = 0.025) and energy
levels of HOMO (lower images) and LUMO (upper images) orbitals calculated for
complexes 1-3.
19

ACCEPTED MANUSCRIPT
To gain deep insight into the photophysical and electrochemical properties of the
complexes, theoretical calculations on the complexes were conducted (see
Experimental). Fig. 4 depicts the optimized geometries of the ground states (S0) and
surface distributions and energy levels of the HOMO and LUMO orbitals. Those of
other molecular orbitals were presented in Table S2 in the supporting information.
The geometrical parameters of the optimized S0 states were summarized in Table S3
in the supporting information. The calculated bond lengths and ligand bite angles
agree well with those observed in the single-crystal structures (Tables S1 and S3).
As shown in Fig. 4, the HOMO orbitals reside predominantly on the iridium ions
and the phenyl rings of the dmappy ligands, while the LUMO orbitals distribute
almost exclusively over the NΛN ligands. From complex 1, to 2 and to 3, the HOMO
level exhibits only a slight change (within 0.08 eV) but the LUMO level is gradually
destabilized by a large extent of about 0.4 eV, which agrees with the electrochemical
measurements. Because the LUMO resides on the NΛN ligand, increasing the
electron-richness of the NΛN ligand from bpy, to pzpy and to dmapzpy results in the
stepwise destabilization of the LUMO level, which agrees with previous studies on
complexes using electron-rich NΛN ligands [19,22].
As shown in Table S2, for complexes 1-3, the HOMO-1 and HOMO-2 reside on
the pyridine rings and dimethylamino groups of the dmappy ligands and the iridium
ions, while the HOMO-3 distributes predominantly over the entire dmappy ligands.
For complexes 1 and 2, the LUMO+1 resides mainly over the NΛN ligands, while the
LUMO+2 and LUMO+3 distribute mainly over the dmappy ligands. For complex 3,
20

ACCEPTED MANUSCRIPT
the LUMO+1 and LUMO+2 reside predominantly over the dmappy ligands, while the
LUMO+3 resides mainly on the pyridine ring of the dma-pzpy ligand.
Table 2. Selected triplet states of complexes 1-3 calculated from the TDDFT
approach.
State
ET
Dominant Excitation
Character
b
(eV) a
2.39
2.65
2.69
1
2
T1
T2
T3
H →L (97%)
3MLCT (Ir→bpy)/3LLCT (dmappy→bpy)
3MLCT (Ir→bpy)/3LLCT (dmappy→bpy)
3MLCT (Ir→bpy)/3LLCT (dmappy→bpy)
H-1 →L (84%)
H-2 →L (90%)
T1
T2
2.83
2.94
H →L (97%)
3MLCT (Ir→pzpy)/3LLCT (dmappy→pzpy)
3MLCT (Ir→dmappy)/3π-π* (dmappy-centered)
H →L+2 (44%),
H-3 →L+3 (18%)
H →L+1 (13%)
H →L+3 (40%),
H-3 →L+2 (23%)
H →L+1 (59%),
H-3 →L+2 (15%)
H →L+2 (45%),
H-3 →L+1 (23%)
H →L (74%),
T3
T1
T2
T3
2.99
2.92
2.98
3.12
3MLCT (Ir→dmappy)/3π-π* (dmappy-centered)
3MLCT (Ir→dmappy)/3π-π* (dmappy-centered)
3MLCT (Ir→dmappy)/3π-π* (dmappy-centered)
3MLCT(Ir→dmapzpy)/3LLCT (dmappy→dmapzpy)
b
3
H-1 →L (13%)
a
Excitation energies calculated for the triplet states. H and L denote HOMO and
LUMO, respectively; data in parentheses are the contributions of corresponding
excitations; only excitations with contributions larger than 10% are listed.
To reveal the triplet-state character, TDDFT calculations were performed on the
optimized S0 geometries. Table 2 summarizes the characteristics of the first three
triplet states of complexes 1-3. The character of the triplet states were concluded by
analyzing the dominant transitions and the surface distributions of the molecular
orbitals involved in the transitions. For complex 1, the T1, T2 and T3 states all exhibit
3
3
mixed MLCT (Ir → bpy)/3LLCT (dmappy → bpy) character, i.e., pure CT
21

ACCEPTED MANUSCRIPT
3
character. For complex 2, the T1 state exhibits mixed MLCT (Ir → pzpy)/3LLCT
3
(dmappy → pzpy) character, i.e., pure CT character, whereas the T2 and T3 states
3
exhibit mixed MLCT (Ir → dmappy)/3π-π* (dmappy-centered) character. For
complex 3, the T1 and T2 states exhibit mixed 3MLCT (Ir → dmappy)/3π-π* (dmappy
centered) character, whereas the T3 state shows mixed 3MLCT (Ir →
dma-pzpy)/3LLCT (dmappy → dma-pzpy) character, i.e., pure 3CT character.
From Table 2, it can be observed that from complex 1, to 2 and to 3, the 3MLCT
(Ir → NΛN)/3LLCT (dmappy → NΛN) state, i.e., the pure 3CT state, is destabilized in
a stepwise manner (2.39→2.83→3.12 eV) because of the gradual destabilization of
3
the LUMO residing on the NΛN ligand, whereas the MLCT (Ir → dmappy)/3π-π*
(dmappy-centered) state (around 2.94 eV) remains almost unchanged. The gradual
3
destabilization of the CT state finally leads to the change of the energy level order
3
between the two states (Table 2). For complexes 1 and 2, the CT states (Ir/dmappy
→ NΛN) are the lowest-lying triple states that account for the light emission, while for
3
complex 3, the MLCT (Ir → dmappy)/3π-π* (dmappy-centered) state becomes the
lowest-lying triplet state from which the light emission occurs. The assignments of the
emitting triplet states consist with the character of light emission concluded from the
photophysical characterizations (vide supra). It is worth noting that for complex 3, the
3CT state (T3) and the 3MLCT (Ir → dmappy)/3π-π* (dmappy-centered) state (T1/T2)
lie close in energy (within 0.20 eV), which indicates that thermal population from the
latter to the former is likely to occur [50-53].
22

ACCEPTED MANUSCRIPT
Figure 5. Optimized geometries and unpaired-electron spin-density contours (0.002 e
bohr-3) calculated for a) T1 of complex 1, b) T1 of complex 2, c) T1 of complex 3 and
d) T3 of complex 3.
Table 3. Triplet energies (in eV) calculated for complexes 1-3.
complex
T1
T3
−
3MC
3.06
3.07
3.06
∆E(3MC−T1)
0.83
1
2
3
2.23
2.69
2.80
−
0.38
2.96
0.26
To gain further insight into the character of triplet states, they were optimized with
the spin-unrestricted DFT method (see Experimental). Fig. 5 depicts the geometries
and spin-density distributions of the optimized triplet states. The geometrical
parameters of the optimized triplet states were presented in Table S3 in the supporting
information. As shown in Fig. 5, for the T1 states of complexes 1 and 2 and the T3
state of complex 3, the spin-densities distribute over the iridium ions, the phenyl rings
of dmappy ligands and the entire NΛN ligands, which match very well with the the
3
topologies of HOMO and LUMO orbitals (Fig. 4) and agree with the MLCT (Ir→
NΛN)/3LLCT (dmappy→NΛN) character assigned to these states (Table 2). For the T1
state of complex 3, the spin-density distributes over the iridium ion and one dmappy
23

ACCEPTED MANUSCRIPT
3
ligand, which consists with the MLCT (Ir→dmappy)/3π-π* (dmappy-centered)
character assigned to this state (Table 2).
Table 3 summarizes the triplet energies of the optimized triplet states. From
complex 1, to 2 and to 3, the triplet energy of the T1 state (the emitting state) increases
gradually from 2.23 to 2.69 and to 2.80 eV, which accords with the gradually
blue-shifted light emission experimentally observed. For complex 3, the optimized T1
and T3 states still lie close in energy (within 0.16 eV), which further indicates that
thermal population from the T1 to the T3 states could occur at ambient conditions
[50-54].
Structural distortion between the excited and ground states usually results in a
large temperature-independent non-radiative decay of the excited-state [45,55-58].
The geometries of the optimized triplet states were thus compared with those of S0
states (see Experimental), as shown in Fig. S4. For complex 1, the T1 state exhibits a
small structural distortion compared to the S0 state, with a small RMSD value of
0.102 (Fig. S4a). However, because of its narrow energy gap, complex 1 still shows
quite a large non-radiative decay in the solution (Table 1). For complex 2, the T1 state
shows a considerably large structural distortion with respect to the S0 state (Fig. S4b),
with a relatively large RMSD value of 0.179, mainly because of the twist of the pzpy
ligand on the T1 state, as observed in similar complexes with pzpy-type NΛN ligands
[53,54]. For this relatively large structural distortion, complex 2 should show a
relatively large non-radiative decay in the solution, which was indeed observed in the
photophysical measurement (Table 1). For complex 3, the T1 state exhibits a small
24

ACCEPTED MANUSCRIPT
structural distortion compared to the S0 state (Fig. S4c), with a small RMSD value of
0.084; however, the T3 state shows a large structural distortion compared to the S0
state, with a large RMSD value of 0.258, mainly because of the remarkable twist of
the dma-pzpy ligand (Fig. S4d) [53,54]. Therefore, for complex 3, the T3 state should
be dominated by non-radiative decay. On the other hand, the T3 state could be
thermally accessed by the T1 state; as a result, the T1 state could undergo additional
non-radiative decay through thermal population to the T3 state [53,54], which partially
explains the abnormally large non-radiative decay of complex 3 in the solution.
The metal-centered states (3MC) resulting from the d-d transitions in the iridium
ions are well known non-radiative dark states in iridium complexes [49,59]. For
3
complexes 1-3, their MC states were optimized and the optimized geometries and
3
spin-density distributions were presented in Fig. S5. On the MC states, the Ir−N
bonds between iridium and dmappy ligands are remarkably elongated by about 0.4 Å,
which indicates bond ruptures and the de-coordination of the pyridine rings. The
spin-densities distribute mainly on the iridium ions, consistent with the metal-centered
3
character. As shown in Table 3, the MC state lies 0.83, 0.38 and 0.26 eV higher in
energy than the T1 state for complex 1, 2 and 3, respectively. For complex 1, the much
3
higher-lying MC state should hardly affect the T1 state and thus makes little
contribution to the overall non-radiative decay. For complex 2, non-radiative decay
3
through thermal population to the MC state cannot be precluded in the solution. For
complex 3, it is very likely that non-radiative decay occurs through thermal
3
population to the MC state which lies moderately higher in energy than the T1 state.
25

ACCEPTED MANUSCRIPT
Therefore, for complex 3, thermal populations to the higher-lying non-radiative T3
3
and MC states represent important non-radiative decay channels for the T1 state,
which lead to the abnormally large non-radiative decay observed in the solution
(Table 1). It is worth noting that in the doped film, these non-radiative decay channels
3
should be remarkably suppressed because of the destabilization of the CT state (T3)
[41,42] and the suppression of geometry relaxation for the triplet states in the rigid
matrix. As a result, in the doped PMMA film, complex 3 affords a remarkably
suppressed non-radiative decay and a high PLQY (Table 1).
3.5. Electroluminescence devices
Figure 6. EL spectra of the OLEDs based on complexes 2 and 3.
26

ACCEPTED MANUSCRIPT
Figure 7. Current-density and brightness versus voltage curves of the OLEDs based
on complexes 2 and 3.
Table 4. Electrical characteristics of the OLEDs based on complexes 2 and 3.
Complex
Von
Bmax
η c, p
EQEp
(%) d
EL λ
(nm)
CIE
(x,y)
(V) a
(cd m2) b
(cd A-1) c
2
3
6.7
13600
16.2
6.5
506
(0.27, 0.50)
6.8
7000
12.6
6.3
472, 488 (0.19, 0.35)
b
c
a Voltage required to reach 1 cd m−2. The maximum brightness. The peak current
efficiency. d The peak external quantum efficiency.
Complex 1 shows limited solubility in organic solvents with low or medium
polarity, which hinders its use in solution-processed OLEDs. Complex 2 and complex
3 show good solubility in organic solvents with medium polarity and also high
PLQYs in the doped films (Table 1). They were thus used as dopants in single-layer,
solution-processed OLEDs. The device structure is ITO/PEDOT: PSS (40 nm)/PVK:
30 wt.% OXD-7: 10 wt.% complex (90 nm) /LiF (1 nm)/Al (100 nm), where poly(N-
vinylcarbazole) (PVK) was used as the hole-transporting host and
27

ACCEPTED MANUSCRIPT
1,3-bis(5-(4-(tert-butyl)phenyl)-1,3,4-oxadiazol-2-yl)benzene (OXD-7) was doped
into PVK to facilitate the electron-transport. The film morphologies of the emissive
layers (PVK: OXD-7: complex) were characterized and the root-mean square
roughness values for the (PVK: OXD-7: complex 2) and (PVK: OXD-7: complex 3)
films (5×5 µm2) are 0.75 and 0.72 nm, respectively, which indicates that the emissive
layers have smooth surfaces. These smooth film surfaces agree with a previous study
on similar (PVK:OXD-7:complex) films [60].
Fig. 6 and Fig. 7 depict the EL spectra and the current density and brightness
versus voltage curves of the OLEDs, respectively. Detailed electrical characteristics
were summarized in Table 4. The OLED based on complex 2 affords green-blue EL
centered at 506 nm, with CIE (Commission Internationale de L’Eclairage) coordinates
of (0.27, 0.50). The OLED based on complex 3 affords sky-blue EL peaked at 472
and 488 nm, with CIE (Commission Internationale de L’Eclairage) coordinates of
(0.19, 0.35), which is comparable to that from the OLED based on FIrpic (ELpeak
=
474 nm, CIE=(0.17, 0.34)) [60]. The PL spectra of the light emitting layers show
residual light emission around 400 nm originating from the PVK:OXD-7 host (Fig.
S6), which indicates incomplete energy transfer from the host to the complex under
photo-excitation. However, the EL spectra show no such residual light emission from
the host, which indicates that in addition to the energy transfer, charge-trapping
followed by direct exciton formation on the complex occurs under electrical
excitation [61,62]. The energy level diagrams of the OLEDs reveal that complexes 2
and 3 behave as strong hole trappers due to their shallow HOMO levels (Fig. S7),
28