Feeling blue? Blue phosphors for OLEDs (Review article)

10 October 2011 Research on organic light emitting diodes (OLEDs) has been revitalized, partly due to the debut of the OLED TV by SONY in 2008. While there is still plenty of room for improvement in efficiency, cost-effectiveness and longevity, it is timely…

10 October 2011

Research on organic light emitting diodes (OLEDs) has been revitalized, partly due to the debut of the OLED TV by SONY in 2008. While there is still plenty of room for improvement in efficiency, cost-effectiveness and longevity, it is timely to report on the advances of light emitting materials, the core of OLEDs, and their future perspectives. The focus of this account is primarily to chronicle the blue phosphors developed in our laboratory. Special attention is paid to the design strategy, synthetic novelty, and their OLED performance. The report also underscores the importance of the interplay between chemistry and photophysics en route to true-blue phosphors.

Hungshin Fua, Yi-Ming Chengb, Pi-Tai Choua,*, and Yun Chib,
*a Department of Chemistry, National Taiwan University, Taipei 10617, Taiwan
b Department of Chemistry, National Tsing Hua University, Hsinchu 30013, Taiwan
*E-mail: chop@ntu.edu.tw; ychi@mx.nthu.edu.tw

Organic light emitting diodes (OLEDs) have drawn great attention in the past two (plus) decades. Ever since the influential and  groundbreaking work by Tang and Van Slyke1 OLEDs have been  viewed as the next generation flat panel display (FPD) technology  as they offer several advantages for self-emitting displays, such  as a wide viewing angle (almost 180 º), a thin panel (< 2 mm),  light weight, a fast response time (microseconds and less), bright  emission, and high contrast. Moreover, they can be made on flexible  substrates, and are thus highly versatile. In reality, OLEDs have  already been incorporated into some commercial products, like  MP3 players, mobile phones, digital cameras, PDAs, etc. The SONY  XEL-1 was the world’s first commercial OLED TV, featuring a 3 mm  thick panel, as well as breathtaking image contrast, brightness, and  color. Later on, LG announced the debut of a 31” prototype and  planned to have a 55” TV on the production line in 2012, while  Samsung introduced a 42” prototype in 2011. In addition to its  main application in display technology, OLEDs are also used as a  lighting source, namely, as white-emitting OLEDs, or WOLEDs2 . For  lighting purposes, WOLEDs have to meet stringent requirements.  To compete with the fluorescent tube as a lighting source, a power  efficiency of > 70 lm/W is preferred and the lifetime must be at  least 10 000 h @ 1000 cd/m2, with a color rendering index (CRI)  greater than 80 %. For the OLED display, it is expected that the   turning point will be around 2014 when the yield rate of production is  expected to be  greater than 70 %. The consequent cost-effectiveness,  brightness, and portability of the OLED will represent a major threat to LCD products3.

The basic OLED structure consists of an organic active layer sandwiched between two electrodes; usually a reducing-metal cathode and a transparent oxide anode. Upon applying a voltage, the electrons and holes are injected from their respective electrodes, generating excitons, which then recombine at the designated organic layer and release energy as visible light. Amid development, OLED technology has been evolving from the initial single hetero-junction structure into a double hetero-junction structure, taking advantage of the separation of the charge transport and light emitting layers into two territories. The technology has further evolved into a multi-heterojunction architecture, which not only separates the two functions, but also strictly confines the carriers, resulting in a vast improvement in performance. Parallel to the structural evolution, key components with different functions in the OLED have also been substantially developed. These include electrodes, electron/hole transports, hosts, and emitters, among which the emitters are the key issue, drawing extensive academic and industrial interest. The driving force has been the development of suitable RGB emitters for both display and solidstate lighting sources.

To meet the above criteria, particularly the leap in efficiency
and hence an intense light output, phosphorescent emitters (or
phosphors) turn out to be indispensible. Theoretically, an internal
quantum efficiency (ηint
) as high as 100 % could be achieved for
phosphorescent OLEDs (PhOLED)4,5
, so these emitting materials would
be superior compared to their fluorescent counterparts, for which only
singlet excitons can be harvested, giving an upper efficiency limit of
25 %. Likewise, phosphors with third-row transition-metal elements
as the core become crucial for the fabrication of PhOLEDs6-9
. The
strong spin-orbit coupling effectively promotes singlet-to-triplet
intersystem crossing, and also enhances the subsequent radiative
transition, i.e., phosphorescence, the results of which facilitate strong
electroluminescence by harnessing both singlet and triplet excitons.
This superiority has led to the continuous trend of shifting research
endeavors towards these heavy transition-metal based phosphors.
Among the three primary RGB colors, the synthetic protocols
and fabrication methods of green and red phosphors to meet the
necessary requirements have been well established10-14
. Conversely,
the design and fabrication of blue phosphors and ensuing devices is
still an ongoing challenge. From our viewpoint, three challenges stand
out regarding the development of blue phosphorescent emitters,
particularly the phosphors: (i) the chromaticity of the most available
blue phosphor is not a true-blue color, which gives a poorer color
gamut for applications in full color OLED displays; (ii) the emission
efficiency of blue phosphors, being inferior to green and red phosphors,
has impeded the development of white lighting, because lighting
demands a high power efficiency; (iii) the lifetime or long-term
stability of phosphors in OLEDs, which is a property relevant to their
phosphorescence efficiency and is influenced by their structural design
and device architecture.
The stability and efficiency of blue phosphors are, in many cases,
significantly poorer than their green and red counterparts. One major
deactivation mechanism should be ascribed to the fact that the
emissive state is in proximity to the metal-centered dπ
dσ*
state, so
that the phosphorescence is prone to quenching by the repulsive dd
state via contact with the potential energy surface (PES) with respect
to that of the ground state. As a result, there is a steep reduction of
the emission efficiency and likewise the photostability12
. Attaining
blue phosphors with high quantum yields (QY) and better stability
for OLEDs thus requires ingenious design and innovative synthetic
pathways. It suffices to say that both the device encapsulation and
adjustment of guest-host energy gaps play crucial roles; the latter is
often sophisticated due to the difficulty of matching the large triplet
energy gap15
.
The poor stability of blue phosphors not only causes a shorter
lifetime of the device, but perhaps more severely, it also induces poor
color stability for white lighting devices. Ideally, there should not be
a noticeable color change during operation. In reality, however, the
distinctive stability of each emitter induces significant ratiometric color
changes before a device reaches fatality. Although the color instability
could be mitigated by device design, e.g., by using different pixel sizes
according to their lifetime, practically, it is still of prime importance
for obtaining blue phosphorescence with high efficiency and long-term
stability.
Our aim in this account is to review progress on blue OLED phosphors;
its past and current status, as well as the future perspectives. Particular
attention will be paid to those new blue phosphors developed in our
lab, including their design strategy, synthetic assessment, and OLED
performance. A photograph of one such blue-emitting PhOLED is shown
in Fig. 1 to serve as an example. To be suitable for a general audience,
both the underlining fundamental and photophysical properties will be
discussed as succinctly as possible. This report also unveils why we “felt
blue” about research on blue phosphors, in particular, the early stage
true-blue phosphors, and how we partially circumvented the obstacles,
underscoring the importance of the interplay between chemistry and
Fig. 1 Photograph of a blue-emitting PhOLED fabricated by our collaborators
working at the Industrial Technology Research Institute of Taiwan.
MT1410p472_479.indd 473 21/09/2011 13:04:47REVIEW Feeling blue? Blue phosphors for OLEDs
4 7 4 OCTOBER 2011 | VOLUME 14 | NUMBER 10
photophysics. Nevertheless, complete success is yet to be achieved, but
perhaps we are not too far away from some pragmatic solutions.
The evolution of blue phosphors
A great deal of credit must be given to Thompson and his co-workers
for their seminal and elegant studies on the sky-blue phosphorescent
Ir(III) complex (FIrpic, see Fig. 2) and its application in PhOLEDs16
. Using
a carbazole based CBP as the host material, the power efficiency (ηp
)
and external quantum efficiency (ηext
) were reported to be 10.5 lm/W
and 5.7 %, respectively. Subsequently, the efficiency was found to be
closely related to the triplet state energy (E
T
) of the host material. When
doped in higher E
T
host materials like mCP and CDBP, the ηext
improved
to 7.8 % and 10.4 %, respectively17,18
. In 2005, Kido and co-workers
developed an FIrpic based OLED with a maximum ηp
of 37 lm/W and
ηext
of 19 % by using novel, high E
T
host materials such as 4CzPBP,
3DTAPBP, and mTPPP (see Fig. 2). The applications of these materials
are versatile, since they can provide highly balanced hole and electron
transport19
. They later produced another FIrpic-based OLED with a
maximum ηp
of over 37 lm/W and ηext
of 24 % by using a wide energy
gap electron transport layer20
. Although the efficiencies of FIrpic-based
OLEDs have been improved over the years, a few critical problems need
to be addressed. For example, it exhibits 1931 Commission Internationale
de L’Eclairage coordinates, CIE
x,y
, of (0.16, 0.29), which are still far from
the true-blue color (0.14, 0.08). Some endeavors were made to search for
better blue phosphors by modifying the ancillary chelate of FIrpic. These
include the second generation FIr621-23
, for which the picolinate ancillary
of FIrpic was replaced with a chelating pyrazolyl-borate, and FIrtaz
and FIrN4
24
, which were chelated by pyridyl triazolate and tetrazolate
ancillary, respectively. Other blue phosphors, such as phenylpyridinebased FCNIr25
, bipyridine-based Ir(dfpypy)3
26
, phenyltriazole-based
Ir(taz)3
27
and carbene-based mer-Ir(cn-pmic)15
, have also been reported.
Evidently, tuning phosphorescence to the true-blue region with ideal
CIE
x,y
coordinates of (0.14, 0.08) is highly challenging. The majority of
blue phosphors reported demonstrated inferior color chromaticity, with
the sum of CIE
x
+ CIE
y
being much greater than 0.3 or with a single CIE
y
coordinate being higher than 0.2521, 28
. To tackle this problem, we first
synthesized Os(II) complexes known as [Os(fppz)2
(CO)2
] (1)29,30
and
[Os(CO)3
(tfa)(fppz)] (2)31
, (see Fig. 3). The fppz chelate tends to exhibit
a much greater intra-ligand charge-transfer (ILCT) or ππ* energy gap
versus that of the sky-blue emitting 4,6-difluorophenyl pyridinato chelate
(dfppy). The electron-withdrawing CF3
group on fppz also stabilizes the
pyrazolate-centered highest occupied molecular orbital (HOMO) and
enlarges the respective ππ* energy level; thus, a true-blue phosphor could
be realized. As a result, the photophysical data (see Table 1) show better
chromaticity; unfortunately, the associated carbonyl ligands make them
electrochemically unstable. Moreover, the 2+ charged Os(II) cation is
limited by its ligand field strength; therefore, its stability and tunability
toward shorter wavelengths are restricted compared to the isoelectronic
Ir(III) complexes and hence, it is unsuitable as a true-blue OLED dopant.
To circumvent the inherent disadvantage of Os(II) complexes, we
switched to Ir(III) complexes, which provide stronger ligand fields due
to the 3+ charged metal core. Nevertheless, en route to the required
blue phosphor, some crucial factors must be optimized. Of prime
Fig. 2 The structures of FIrpic and related host and electron transporting materials.
Table 1 Photophysical data for Os(II) complexes (1) and (2)
Compound abs. λmax
(nm) PL λmax
(nm) Q.Y. τ (μs)
1 311 430, 457, 480 0.14 18.5
2 247, 292, 320 460 0.26 143
MT1410p472_479.indd 474 21/09/2011 13:04:49Feeling blue? Blue phosphors for OLEDs REVIEW
OCTOBER 2011 | VOLUME 14 | NUMBER 10 4 7 5
concern is increasing the contribution from the metal-to-ligand chargetransfer (MLCT) in the lowest-lying triplet manifold32,33
. The increase
of the MLCT contribution enhances the coupling of the orbital angular
momentum to the electron spin, such that the T1 → S0
transition
would have a large first order spin-orbit coupling term, which would
result in a substantial decrease of the radiative lifetime and hence
the possibility of increasing the phosphorescence QY34
. Moreover,
whilst enlarging the emission energy gap, care has to be taken to
avoid touching the metal-centered dd excited state, which may induce
radiationless decay or, more seriously, bond dissociation toward the
weakest metal-ligand site35
. Third, upon increasing the energy gap
towards blue, it becomes facile for the lowest-lying excited state,
which is mainly ILCT and MLCT in character, to mix with a thermally
accessible ligand-to-ligand charge-transfer (LLCT) transition involving
auxiliary chromophores. For clarity, Fig. 4 illustrates the electron
density flow of the MLCT, ILCT, and LLCT, using complex (3) as an
example (see Fig. 3). From the viewpoint of relaxation dynamics, LLCT
may possess a well bounded PES due to the π-electron delocalization
and hence the overall bond stabilization, which avoids intersection
and/or thermal population to the dd excited state, resulting in an
increased QY. Conversely, the associated vibration modes eligible
for quenching increase in LLCT, which may facilitate radiationless
deactivation and hence decrease QY. Therefore, the optimization of
various ligand structures and compositions is necessary before the
resulting transition metal complexes can maximize the blue emission
efficiency.
Based on the above, we have synthesized heteroleptic isomeric
iridium complexes [Ir(dfppy)(fppz)
2
] (3) and (4)36
. As demonstrated
by the computational analysis (density functional theory, DFT), the
lowest lying state for both complexes (3) and (4) comprises MLCT and
ILCT mixed with a substantial LLCT component. As a result, (3) and
(4) exhibit phosphorescence maxima at 450 and 480 nm with QYs as
high as 50 % and 15 %, respectively, supporting the positive effect of
LLCT suppressing the radiationless pathways. Note that the higher QY
in (3) also reflects its larger MLCT percentage (26.6 %) versus that of
(4) (16.9 %) and hence a radiative decay rate of 1.4 × 105
s
-1
(3) versus
3.2 × 104
s
-1
(4), verifying the above first criterion. Accordingly, an
OLED device (see Fig. 5) prepared with complex (3) demonstrated CIE
coordinates of (0.16, 0.18), ηext
of 8.5 %, and a peak ηp
of 8.5 lmW-1
.
While the delocalization concept mentioned above is beneficial
for the emission QY, the π-delocalization among distinctive chelates
for LLCT, per se, may reduce the energy gap, which has adverse
Fig. 3 Structural drawings of assorted complexes, as described in the text.
Fig. 4 The MLCT, ILCT, and LLCT orbital transitions of complex (3). The orbital
contour in purple corresponds to the HOMO, while the contour in yellow
represents the second orbital above the LUMO (i.e., LUMO+2). Note that the
iso-density of the orbital contours was set to 0.05.
MT1410p472_479.indd 475 21/09/2011 13:04:50REVIEW Feeling blue? Blue phosphors for OLEDs
4 7 6 OCTOBER 2011 | VOLUME 14 | NUMBER 10
consequences for reaching true-blue emission. We then proceeded to
design a class of Ir(III) complexes, for which the (lowest lying) emitting
state was restrained to a single chromophore. Experimentally, we were
keen to exploit fppz or an analogous chelating triazolate (c.f., fptz) as
the key chromophore, thanks to their high energy gaps and versatility
in chemical modification. Using the aforementioned Ir(III) complexes
(3) and (4) as the prototypes, a way to reduce the involvement of
the auxiliary dfppy ligand is to cut off the π-conjugation between the
2,4-difluorophenyl and pyridyl moieties. Thus, we proceeded to study
the so-called nonconjugated chelate, in which the two designated
moieties are strategically linked by a methylene group to reduce the
cross-talk. Accordingly, a blue phosphorescent Ir(III) complex (5) with
nonconjugated N-benzylpyrazole (dfb-pz) ligands was synthesized37
.
Table 2 lists the comparison between compound (5) with dual dfb-pz
chelates and its counterpart (6) that possesses N-phenylpyrazole
chelates (dfpz)38
. Clearly, the higher emission QY for (5), versus
that of (6), proves the concept of non-conjugating auxiliary ligands,
which is also verified by DFT calculation, confirming that only the
ILCT of 2-pyridyl triazolate fragment (fptz) in (5) is involved in
the emitting state, while the lowest lying state for (6) comprises a
substantial LLCT contribution38
. The low emission QY for complex (6)
(c.f., 5) is consistent with the studies by Thompson and co-workers
on the temperature dependence of Ir(III) complexes based on the
N-phenylpyrazole chelate (ppz)39
. It was found that Ir(ppz)3
is nonemissive at room temperature due to the small activation energy to a
non-emissive dd excited state.
Note that the much lower QY for (6), to a certain extent, reflects
the negative effect of LLCT compared to complexes (3) and (4) (see
above). Thus, the counteracting effect of LLCT on the QY seems to vary
case by case, the trend of which is unfortunately still unpredictable at
this stage11,12
. Nevertheless, via an ingenious chemical design, if the
LLCT is intentionally placed in the higher excited state near to the dd
state, LLCT with a steeper PES should be able to by-pass/cross the dd
repulsive PES. In this case, the bond dissociation occurring in the higher
excited states can be mitigated, increasing the emission efficiency of
the blue phosphor. Together with a single emitting chromophore, i.e.,
ILCT (mixed with MLCT) that may also be easily chemically modified, it
is thus possible to produce a blue phosphor possessing a true-blue hue,
high emission yield, and great stability. A conceptual design strategy
for such a system is depicted in Fig. 6.
Fig. 5 Architecture of the OLED using (3) as the dopant and molecular drawings of the electron or hole transporting material. Reproduced with permission from36
.
Copyright Wiley 2007.
Table 2 Comparison of photophysical properties between complexes (5) and (6)
Compound abs. λmax
(nm) PL λmax
(nm) Q.Y. (Φ) τobs
(ns) kr (×105
s
-1
)
[Ir(dfb-pz)2
(fptz)] (5) 261, 368 437 0.1 100 10
[Ir(dfpz)2
(fptz)] (6) 300, 349 457 0.0046 8.6 5.4
MT1410p472_479.indd 476 21/09/2011 13:04:51Feeling blue? Blue phosphors for OLEDs REVIEW
OCTOBER 2011 | VOLUME 14 | NUMBER 10 4 7 7
Bearing this concept in mind, we reported modified blue phosphors
employing benzyldiphenylphosphine and its difluorinated analogue
as the ancillary chelate40
. As for the strategy, cyclometalation of
benzyldiphenylphosphine can produce the required anionic chelate.
Moreover, both the PPh2
unit and the non-conjugated benzyl group
exert the greatest ligand field stabilization energy as well as the
largest ππ* energy gap among all established chelates41
. A deep-blue
phosphor, possessing a superior emission QY, was thus anticipated.
In this approach, a heteroleptic complex [Ir(dfppy)2
(dfbdp)] (7) was
successfully isolated, which showed an improved efficiency with
respect to its chloride precursor, [Ir(dfppy)2
(dfbdpH)Cl], as well as two
blue-shifted emission maxima at 457 and 480 nm versus those of
standard FIrpic (470 and 490 nm). An OLED device fabricated with (7)
demonstrated a low turn-on voltage (defined as the voltage obtained at
1 cd m-2
) of 4.6 V. The peak ηext
, ηl
, and ηp
are approximately 10.24 %,
15.95 cdA-1
, and 10.07 lmW-1
, respectively. The CIE coordinates
calculated from the resulting spectrum are (0.16, 0.20), which is more
blue-shifted compared to the majority of dfppy based phosphors.
A parallel study resulted in another class of Ir(III) complexes
with either dfppy or chelating azolate chromophores, plus one nonconjugated phosphine chelate42
The non-conjugated phosphine
chelate not only greatly restricted its participation in the lowest-lying
electronic transition but also enhanced the coordination strength.
These two factors led to authentic blue phosphorescence as well as
suppressed nonradiative deactivation, thus improving the emission
efficiency. Thus, the blue-emitting complexes [Ir(dfpbpy)2
(P^N)] (8) and
[Ir(fppz)2
(P^N)] (9), were used in the fabrication of OLED devices. Of
particular interest was the (9)-doped OLEDs, which exhibited remarkable
peak ηext
, ηl
, and ηp
values of 6.9 %, 8.1 cdA-1
, and 4.9 lmW-1
, and with
a true-blue chromaticity CIE
x,y
= (0.16, 0.15).
An additional strategy to achieve efficient blue emission was to use
high-field-strength ligands such as NHC carbenes15,32
. In this contribution,
we have reported the use of benzyl carbene chelate and allowed it
to react with [IrCl3
(tht)3
] in synthesizing a series of Ir(III) complexes,
[(fbmb)2
Ir(bptz)] (10) and [(dfbmb)2
Ir(fptz)] (11)43
. Complexes (10)
and (11) exhibited blue emission λmax
at 460 and 458 nm, and QYs of
0.22 and 0.73, respectively. Complex (11) was used as the dopant for
the fabrication of blue PhOLEDs due to its superior emission quantum
efficiency. The respective device configuration and materials used were:
ITO/α-NPD (30 nm)/TCTA (20 nm)/CzSi (3 nm)/CzSi: 6 % (11) (3 nm)/
UGH2 (2 nm)/BCP (50 nm)/CsCO3
(2 nm)/Al (150 nm). Figs. 7a and
7b show the electroluminescence (EL) spectrum and the corresponding
CIE color coordinates, while Figs. 7c and 7d depict the current-voltageluminance (I-V-L) characteristics and the EL efficiencies of the device. The
EL spectrum exhibits two distinct emission peaks at 434 and 460 nm. The
CIE
x,y
coordinates were (0.16, 0.13). The respective peak ηext
and ηp
were
6.0 % and 4.0 lmW-1
, but dropped to about 2.7 % and 0.9 lmW-1
at a
higher current density. Despite the common efficiency roll-offs, their color
chromaticities were far better than other PhOLEDs fabricated from wellknown blue phosphors such as FIrpic17
, FIr621
, FIrtaz24
, and FIrN424
.
Finally, we move on to the recent discovery of heteroleptic Ir(III)
complexes with a tripodal, facially coordinated phosphite (or phosphonite)
Fig. 6 Concept for obtaining blue phosphors with a high quantum yield and photostability.
MT1410p472_479.indd 477 21/09/2011 13:04:52REVIEW Feeling blue? Blue phosphors for OLEDs
4 7 8 OCTOBER 2011 | VOLUME 14 | NUMBER 10
(denoted as P^C2
) to serve as the ancillary ligand44
. In addition to
fulfilling the criteria discussed above, of both non-conjugation and a
strong ligand field, the tridentate ligand offers great binding stability.
As a result, highly efficient blue phosphorescence is attained with
good OLED performance. The synthetic protocol of this class of Ir(III)
complexes, (12) and (13), is depicted in Fig. 8. Complex (13) was chosen
for the fabrication of an OLED device due to its high solid state emission
QY of 0.97. The device configuration and materials being used were:
ITO/TAPC (30 nm)/TCTA (10 nm)/CzSi (3 nm)/CzSi: 4 % (13) (25nm)/
UGH2: 4 % (13) (2nm)/TmPyPB (50nm)/LiF (0.8nm)/Al (150 nm). The
wide-gap host CzSi and UGH2, which have a triplet energy gap of 3.02
eV and 3.08 eV, respectively, were employed for optimal efficiency. In
addition, good confinement of the excitons and carriers was realized
by using double emitting layers and double buffer layers to balance the
charge transport and to move the exciton-formation zone away from
the adjacent carrier-transport layers. The device demonstrated a turnon voltage of 4.1 V, and peak ηext
, ηl
, and ηp
values of 11.0 %, 22.3
cdA-1
, and 16.7 lmW-1
, respectively.
Concluding remarks
Despite the intensive progress on blue phosphors that has been made, the
design of a highly efficient blue phosphor based on a third-row transition
metal is still a challenge today. We have demonstrated that pushing the
emission gap towards the authentic blue region requires not only ingenious
molecular design but also consideration of the subtle interference between
various low-lying electronic states. The chelates chosen for pursuing the
Fig. 7 (a) EL spectrum, (b) CIE chromaticity coordinates, (c) I-V-L characteristics (J = current density, L = brightness), and (d) external quantum efficiency (ηext
)
versus L for the device containing dopant (11). Reproduced with permission from43
. © Wiley 2008.
Fig. 8 Reaction protocol that provided the blue-emitting Ir(III) complexes with
the tripodal ancillary ligand.
(fppz)H = 3-(trifluoromethyl)-5-(2-pyridyl) pyrazole
tfa = trifluoroacetate
dfbdpH = 4,6-difluorobenzyl diphenylphosphine
(dfpbpy)H = 2-(4,6-difluorophenyl-4-tert-butylpyridine)
(P^N)H = 5-(diphenylphosphinomethyl)-3-trifluoromethylpyrazole)
tht = tetrahydrothiophene
fbmbH = 1-(4-fluorobenzyl)-3-methylbenzimidazolium
dfbmbH = 1-(2,4,-difluorobenzyl)-3-methylbenzimidazolium)
ITO = indium tin oxide
TAPC = di[4-N,N-ditolylamino]phenyl)cyclohexane
TCTA = 4,4’,4”-tris(carbazol-9-yl)-triphenylamine
CzSi = 9-(4-tert-butylphenyl)-3,6-bis(triphenylsilyl)-9H-carbazole
UGH2 = p-bis(triphenylsilyl)benzene
TmPyPB = 1,3,5-tri[(3-pyridyl)-phen-3-yl]benzene)
MT1410p472_479.indd 478 21/09/2011 13:04:53Feeling blue? Blue phosphors for OLEDs REVIEW
OCTOBER 2011 | VOLUME 14 | NUMBER 10 4 7 9
blue phosphor include: (i) C^N chelates such as 2,4-difluorophenylpyridine,
(ii) N^N chelates like 2-pyridyl azolate, (iii) C^C chelates comprising both
C-H site and neutral NHC carbene fragment, (iv) C^P chelates with both
benzylic C-H site and neutral phosphine donor, and (v) N^P chelates with
an azolic N-H group11
. These cyclometalating chelates, in theory, may
afford a series of Ir(III) based phosphors with photophysical properties
tuned specifically for the fabrication of PhOLEDs.
We have systematically pushed the chromaticity of phosphors closer
to the true-blue region. However, even if they reach the authentic
blue color, the close proximity between the metal-centered dd and
the emissive states may inevitably lead to a lower emission efficiency
and photostability. These inferiorities make us feel somewhat “blue”
about attaining a true-blue hue, though some proposed mechanisms
such as increasing the ligand-field strength or placing LLCT near the dd
state may partly improve the performance. An alternative solution for
achieving more stable OLEDs may rely on the sky-blue phosphor, for
which the excited state is lower in energy which would thus strengthen
the photostability. Moreover, by selecting better host/charge transport
materials and optimizing the OLED device structure, we believe that these
sky-blue phosphors, together with orange ones, have latent potential for
making all phosphorescent WOLEDs.
Nevertheless, white light generation using multiple chromophores,
namely RGB triads or dual colors, such as sky-blue plus orange (or blue
plus yellow), would still suffer from relative-stability issues due to the
intrinsically different lifespans of different materials. In theory, a single
system emitting white light could ultimately overcome this hurdle. In
the case of fluorescence emitters, it has been proposed that a single
system utilizing both normal Franck-Condon and exciplex emissions may
accomplish this goal45-48
. A similar strategy, however, is not attainable
for the Ir(III) complexes because of the octahedral configuration. Since
the heavy atom effect and hence the spin-orbit coupling is empirically
inversely proportional to r6
, where r is the distance between the emissive
chromophore and core Ir(III) element, one approach is to de-emphasize
the spin-orbit coupling by elongating their spatial separation49,50
. If the
S1
-Tm (m ≥ 1) intersystem crossing has a low efficiency for a designated
Ir(III) complex and the fluorescence yield is virtually 100 % upon optical
excitation, the electrical pumping (in OLED applications) may then give
25 % and 75 % population in the S1
and T1
states, respectively. As
opposed to the null phosphorescence upon optical excitation, this 75 %
T1
population may exhibit phosphorescence due to the heavy atom effect,
producing dual emission (fluorescence + phosphorescence). On the basis
of proper chemical derivation, white light generation is not impossible,
per se. Of course, such a system may be limited by the maximal 25 %
fluorescence yield, considering white light generation. Another possibility
may count on the design of transition-metal complexes that show
fast isomerization at the triplet manifolds. What if the isomerization is
adiabatic, meaning the reaction is along the triplet PES, with a certain
barrier, so that the system exhibits dual phosphorescence covering a
gamut of the visible spectrum? We believe that such a system should be
both theoretically and synthetically feasible. Certainly, more ingenious
and original concepts should be available, and we hope this review will
stimulate and encourage the OLED community.
REFERENCES
1. Tang, C. W., and VanSlyke, S. A., Appl Phys Lett (1987) 51, 913.
2. Zhou, G., et al., J Photochem Photobio C (2010) 11, 133.
3. Tsai, J. -H., and Chen, L. -R., Wealth Magazine (2011) 373, 66.
4. Kawamura, Y., et al., Appl Phys Lett (2005) 86, 071104.
5. Adachi, C., et al., J Appl Phys Lett (2001) 90, 5048.
6. Baldo, M. A., et al., Nature (1998) 395, 151.
7. Holder, E., et al, Adv Mater (2005) 17, 1109.
8. Gong, X., et al., Appl Phys Lett (2002) 81, 3711.
9. Wong, W. -Y., and Ho, C. -L., J Mater Chem (2009) 19, 4457.
10. Chou, P. -T., and Chi, Y., Chem-Eur J (2007) 13, 380.
11. Chi, Y., and Chou, P. -T., Chem Soc Rev (2010) 39, 638.
12. Chou, P. -T. et al., Coord Chem Rev (2011), doi:10.1016/j.ccr.2010.12.013
13. Wong, W. -Y., and Ho, C. -L., Coord Chem Rev (2009) 253, 1709.
14. Zhou, G., et al., Chem Asian J (2011) 6, 1706.
15. Haneder, S., et al., Adv Mater (2008) 20, 3325.
16. Adachi, C., et al., Appl Phys Lett (2001) 79, 2082.
17. Holmes, R. J., et al., Appl Phys Lett (2003) 82, 2422.
18. Tokito, S., et al., Appl Phys Lett (2003) 83, 569.
19. Sasabe, H., and Kido, J., Chem Mater (2011) 23, 621.
20. Sasabe, H., et al., J Chem Mater (2008) 20 5951.
21. Holmes, R. J., et al., Appl Phys Lett (2003) 82, 3818.
22. Ren, X., et al., Chem Mater (2004) 16, 4773.
23. Li, J., et al., Polyhedron (2004) 23, 419.
24. Yeh, S. J., et al., Adv Mater (2005) 17, 285.
25. Jeon, S. O., et al., Adv Funct Mater (2009) 19, 3644.
26. Lee, S. J., et al., Inorg Chem (2009) 48, 1030.
27. Lo, S. -C., et al., Chem Mater (2006) 18, 5119.
28. Zheng, Y., et al., Appl Phys Lett (2008) 92, 223301.
29. Wu, P. -C., et al., Organometallics (2003) 22, 4938.
30. Yu, J. -K., et al., Chem-Eur J (2004) 10, 6255.
31. Cheng, Y. -M., et al., Inorg Chem (2007) 46, 10276.
32. Sajoto, T., et al., Inorg Chem (2005) 44, 7992.
33. Li, J., et al., Inorg Chem (2005) 44, 1713.
34. Hsu, C. -W., et al., J Am Chem Soc (2011) 133, 12085.
35. Yeh, Y. -S., et al., ChemPhysChem (2006) 7, 2294.
36. Yang, C. -H., et al., Angew Chem Int Ed (2007) 46, 2418.
37. Song, Y. -H., et al., Chem-Eur J (2008) 14, 5423.
38. Yang, C. -H., et al., Inorg Chem (2005) 44, 7770.
39. Sajoto, T., et al., J Am Chem Soc (2009) 131, 9813.
40. Hung, J. -Y., et al., Dalton Trans (2009) 38, 6472.
41. Chiu, C. -L., et al., Adv Mater (2009) 21, 2221.
42. Chiu, Y. -C., et al., ACS Appl Mater Inter (2009) 2, 433.
43. Chang, C. -F., et al., Angew Chem Int Ed (2008) 47, 4524.
44. Lin, C. -H., et al., Angew Chem Int Ed (2011) 50, 3182.
45. Zhou, G., et al., Chem Commun (2009) 45, 3574.
46. Zhou, G., et al., J Mater Chem (2010) 20, 7472.
47. Williams, E. L., et al., Adv Mater (2007) 19, 197.
48. Kalinowski, J., et al., Adv Mater (2007) 19, 4000.
49. Chen, Y. -L., et al., ChemPhysChem (2005) 6, 2012.
50. Kozhevnikov, D. N., et al., Inorg Chem (2011) 50, 3804.
MT1410p472_479.indd 479 21/09/2011 13:04:54