Effective shielding of triplet energy transfer to conjugated polymer by its dense side chains from phosphor dopant for highly efficient electrophosphorescence

Article abstract of DOI:10.1021/ja076413i

To examine the quenching of a triplet exciton by low triplet energy (E _T) polymer hosts with different chain configurations for high E _T phosphor guests, the quenching rate constant measurements were carried out and analyzed by the standard Stern-Volmer equation. We found that an effective shielding of triplet energy transfer from a high E_T phosphor guest to a low E_T polymer host is possible upon introducing dense side chains to the polymer to block direct contact from the guest such that the possibility of Dexter energy transfer between them is reduced to a minimum. Together with energy level matching to allow charge trapping on the guest, high device efficiency can be achieved. The extent of shielding for the systems of phenylene-based conjugated structures from iridium complexes follows the sequence di-substituted (octoxyl chain) in the para position (dC ₈OPPP) is greater than monosubstituted (mC₈OPPP) and the PPPs with longer side chains are much higher than a phenylene tetramer (P4) with two short methyl groups. Further, capping the dialkoxyl-susbstituents with a carbazole (Cz) moiety (CzPPP) provides enhanced extent of shielding. Excellent device efficiency of 30 cd/A (8.25%) for a green electrophosphorescent device can be achieved with CzPPP as a host, which is higher than that of dC ₈OPPP as host (15 cd/A). The efficiency is higher than those of high E_T conjugated polymers, poly(3,6-carbazole) derivatives, as hosts (23 cd/A). This observation suggests a new route for molecular design of electroluminescent polymers as a host for a phosphorescent dopant.

Full text of DOI:10.1021/ja076413i

Published on Web 03/13/2008
Effective Shielding of Triplet Energy Transfer to Conjugated
Polymer by Its Dense Side Chains from Phosphor Dopant for
Highly Efficient Electrophosphorescence
Szu-Po Huang, Tzu-Hao Jen, Yen-Chun Chen, An-En Hsiao, Shu-Hui Yin,
Hsiang-Yun Chen, and Show-An Chen*
Chemical Engineering Department, National Tsing-Hua UniVersity,
Hsinchu, 30041, Taiwan, R.O.C.
Received August 26, 2007; E-mail: sachen@che.nthu.edu.tw
Abstract: To examine the quenching of a triplet exciton by low triplet energy (E
chain configurations for high E phosphor guests, the quenching rate constant measurements were carried
out and analyzed by the standard Stern-Volmer equation. We found that an effective shielding of triplet
energy transfer from a high E phosphor guest to a low E polymer host is possible upon introducing dense
T
) polymer hosts with different
T
T
T
side chains to the polymer to block direct contact from the guest such that the possibility of Dexter energy
transfer between them is reduced to a minimum. Together with energy level matching to allow charge
trapping on the guest, high device efficiency can be achieved. The extent of shielding for the systems of
phenylene-based conjugated structures from iridium complexes follows the sequence di-substituted (octoxyl
8 8
chain) in the para position (dC OPPP) is greater than monosubstituted (mC OPPP) and the PPPs with
longer side chains are much higher than a phenylene tetramer (P4) with two short methyl groups. Further,
capping the dialkoxyl-susbstituents with a carbazole (Cz) moiety (CzPPP) provides enhanced extent of
shielding. Excellent device efficiency of 30 cd/A (8.25%) for a green electrophosphorescent device can be
achieved with CzPPP as a host, which is higher than that of dC
8
OPPP as host (15 cd/A). The efficiency
is higher than those of high E conjugated polymers, poly(3,6-carbazole) derivatives, as hosts (23 cd/A).
T
This observation suggests a new route for molecular design of electroluminescent polymers as a host for
a phosphorescent dopant.
3
Introduction
are produced by F o¨ rster (long-range dipole-dipole induced)
energy transfer from the host or by charge trapping (CT)1
a,3h
In organic and polymeric light emitting diodes (LEDs), holes
and electrons injected from the anode and cathode can recom-
bine to yield singlet and triplet excitons at a population ratio of
on itself, and triplet excitons are generated by Dexter (short-
range electron exchange) transfer from the host or by CT on
3
itself. However, it is not the general case that triplet energy
transfer is via short-range Dexter transfer. It was found recently
that triplet energy transfer between two iridium (Ir) complex
1
1:3 according to quantum statistics and some experiments, to
2
which the ratio in polymer LED remains in debate. For a purely
fluorescent device, only singlet is emissive and most of the
energy is wasted by a nonemissive triplet. By doping with
heavy-metal complexes as guests in small molecules or polymers
4
molecules can also take place via long-range F o¨ rster transfer
and its transfer rate is dependent on intermolecular distance.4a
The reason is that large spin-orbital coupling associated with a
heavy metal (e.g., Ir) causes formally spin-forbidden S0 f T1
absorption to be allowed. Due to the larger extinction coefficient
3
as hosts, both singlet and triplet excitons can be harvested and
consequently the internal quantum efficiency is able to be
promoted toward 100%.
3d-f
The reason is that the heavy-metal
(
3) (a) Baldo, M. A.; O’Brien, D. F.; You, Y.; Shoustikov, A.; Sibley, S.;
Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151-154. (b) Lamansky,
S.; Djurovich, P.; Murphy, D.; Abdel-Razzaq, F.; Lee, H. E.; Adachi, C.;
Burrows, P. E.; Forrest, S. R.; Thompson, M. E. J. Am. Chem. Soc. 2001,
complexes can induce strong spin-orbital coupling and thus
promote an efficient intersystem crossing (ISC) from its singlet
excited state, eventually to the lowest-energy triplet excited state
of the guest followed by relaxation to the ground state via
phosphorescence at room temperature. In general, for such host/
guest systems, the singlet excitons of heavy-metal complexes
1
23, 4304-4312. (c) Sudhakar, M.; Djurovich, P. I.; Hogen-Esch, T. E.;
Thompson, M. E. J. Am. Chem. Soc. 2003, 125, 7796-7797. (d) Kawamura,
Y.; Goushi, K.; Brooks, J.; Brown, J. J.; Sasabe, H.; Adachi, C. Appl. Phys.
Lett. 2005, 86, 71104. (e) Adachi, C.; Baldo, M. A.; Thompson, M. E.;
Forrest, S. R. J. Appl. Phys. 2001, 90, 5048-5051. (f) Sun, Y. R.; Giebink,
N. C.; Kanno, H.; Ma, B. W.; Thompson, M. E.; Forrest, S. R. Nature
2
006, 440, 908-912. (g) Chen, F. C.; Chang, S. C.; He, G. F.; Pyo, S.;
(
1) (a) Baldo, M. A.; O’Brien, D. F.; Thompson, M. E.; Forrest, S. R. Phys.
ReV. B 1999, 60, 14422-14428. (b) Wilson, J. S.; Dhoot, A. S.; Seeley,
A.; Khan, M. S.; Kohler, A.; Friend, R. H. Nature 2001, 413, 828-831.
Yang, Y.; Kurotaki, M.; Kido, J. J. Polym. Sci., Part B: Polym. Phys.
2003, 41, 2681-2690. (h) Baldo, M. A.; Forrest, S. R. Phys. ReV. B 2000,
62, 10958-10966. (i) Turro, N. J. Modern Molecular Photochemistry;
University Science Books: Sausalito, CA, 1991; pp 311-316.
(
c) Segal, M.; Baldo, M. A.; Holmes, R. J.; Forrest, S. R.; Soos, Z. G.
Phys. ReV. B 2003, 68, 75211.
(4) (a) Namdas, E. B.; Ruseckas, A.; Samuel, I. D. W.; Lo, S. C.; Burn, P. L.
Appl. Phys. Lett. 2005, 86, 91104. (b) Wasserberg, D.; Meskers, S. C. J.;
Janssen, R. A. J. J. Phys. Chem. A 2007, 111, 1381-1388. (c) Rothe, C.;
King, S.; Monkman, A. Nat. Mater. 2006, 5, 463-466.
(
2) (a) Wohlgenannt, M.; Tandon, K.; Mazumdar, S.; Ramasesha, S.; Vardeny,
Z. V. Nature 2001, 409, 494-497. (b) Rothe, C.; King, S. M.; Monkman,
A. P. Phys. ReV. Lett. 2006, 97, 76602.
10.1021/ja076413i CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 4699-4707
9
4699

A R T I C L E S
Huang et al.
(ꢀ) of the S-T transition caused by this heavy atom effect, the
Scheme 1. Synthetic Route for Monomer 2 (dC8OPPP)
triplet energy transfer between two Ir-complex molecules can
take place via the long-range F o¨ rster mechanism. However, in
the case of Ir complexes doping into polymers (or small
molecules), the small ꢀ of the S-T transition of polymers (or
small molecules) without heavy metal pertubation, the triplet
energy transfer is mainly via the short-range Dexter mechanism
and negligible via the long-range F o¨ rster mechanism.
Scheme 2. Synthetic Route for Monomer 4 (CzPPP)
In order to confine triplet excitons on a phosphor guest, a
host material with a triplet energy level (ET) higher than that
of the phosphor guest is intuitively required as a significant
quench of triplet excitons by a low ET host for a high ET guests
can occur, as evidenced by the Stern-Volmer analysis for the
system of various Ir complexes ranging from a blue to red
emission with tris(9,9-dimethylfluorene) (F3) as the phosphor
quencher.3 High ET conjugated polymers such as poly(3,6-
c
5
carbazole) derivatives, P(3,6-Cz)’s with ET ≈ 2.6 eV, have been
prepared by limiting the conjugation length of the backbone to
biphenyl. Upon doping with a green emitting Ir complex (Ir-
2′
G, (bis(2-phenylpyridinato-N,C )iridium(acetylacetonate)), the
excellent performance with a maximum luminous efficiency (η)
of about 23 cd/A, equivalent to an external quantum efficiency
(
ηext) of about 6.6%, was obtained. The common blue-emitting
6
polymers, polyfluorenes (PFs) (ET ) 2.18 eV) and poly(p-
phenylene)s (PPPs) (ET ) 2.27 eV), are usually low in ET and
expected to be not suitable for use as hosts for high ET guests,
e.g., Ir-G. As stated above, since triplet energy transfer between
a conjugated polymer host and phosphor guest is dominated
Hence, the efficiency of the green electrophosphorescent devices
3
g,8b
9a,b
is still low for the use of low ET hosts (PFs
and PPPs ) as
5
compared to that with the high ET hosts, P(3,6-Cz)’s and
9
c-e
PVK
(the measured value ET ) 2.91 eV as shown in the
7
Supporting Information (SI), which is identical to that reported
via electron exchange (requiring a close contact within 15 Å),
some efforts8 have been attempted recently to reduce the back
triplet energy transfer from guest to host by increasing the
distance between their triplet centers. For the case with similar
a,b
in ref 10). More importantly, a design route for a low E polymer
T
as an effective host for a high ET guest still remains unclear.
Here, we demonstrate that an effective shielding of triplet
energy transfer from a high ET phosphor guest to a low ET
polymer host is possible upon introducing dense side chains to
the polymer to block direct contact from the guest such that
the possibility of Dexter energy transfer between them is reduced
to a minimum. Together with energy levels matching to allow
charge trapping on the guest, high device efficiency can be
achieved. The system investigated is dialkoxyl-substituted PPP
ET values for the polymer host and iridium guest, the device
ηext can be improved from 1.1% to 2.0%8a by grafting red-
emitting (btp)2Ir(acac) (ET ) 2.0 eV) with the spacer, -(CH2)8-,
as the side chain8 on poly(9,9′-dioctyl-fluorene) (PFO, ET )
a,c
2
.18 eV) as compared to that without the spacer. Alternatively,
introducing a bulky tert-butyl group as a side group on each
ring in the ligands of green-emitting Ir(ppy)3 (ET ) 2.4 eV)
8
b
(dC OPPP) (E ) 2.31 eV) as the host and Ir-G (E ) 2.41
has also been found to promote device ηext from 0.1 to 0.4%
8
T
T
eV) as the guest, which gives a high device η of 15 cd/A
4.12%). With further capping of the dialkoxyl-susbstituents with
when used as a dopant for poly(9,9′-spirobifluorene) (ET ) 2.18
eV) relative to the system with the same dopant but without
such modification, which reflects a reduced triplet energy
transfer back to the host. However, the chance for Ir-complex
molecules to contact with main chains for both cases still
remains high, since the Ir complexes on a side chain can still
be able to be in contact with the polymer backbone nearby and
the distance from the hydrogen atom of the tert-butyl group to
the center Ir atom is within 10 Å (within the effective distance
for Dexter energy transfer). In addition, trace host emission
under electrical excitation can still be observed in both cases.
(
a carbazole (Cz) moiety (CzPPP), the device η is further
promoted to 30 cd/A (8.25%). The further promoted efficiency
can be attributed to better chemical compatibility between the
host and guest provided by the Cz. This observation suggests a
new route for molecular design of electroluminescent polymers
as a host for a phosphorescent dopant.
Experimental Section
Synthesis of Monomers and Polymers. The polymers (mC
dC OPPP, CzPPP, PFO, and CzPF) were synthesized by a Yamamoto
polymerization method, and mC
8
OPPP,
8
(
5) (a) van Dijken, A.; Bastiaansen, J.; Kiggen, N. M. M.; Langeveld, B. M.
W.; Rothe, C.; Monkman, A.; Bach, I.; Stossel, P.; Brunner, K. J. Am.
Chem. Soc. 2004, 126, 7718-7727. (b) Chen, Y. C.; Huang, G. S.; Hsiao,
C. C.; Chen, S. A. J. Am. Chem. Soc. 2006, 128, 8549-8558.
6) Hertel, D.; Setayesh, S.; Nothofer, H. G.; Scherf, U.; Mullen, K.; Bassler,
H. AdV. Mater. 2001, 13, 65-70.
11
8c
8
OPPP and PFs syntheses were
(9) (a) Zhu, W. G.; Mo, Y. Q.; Yuan, M.; Yang, W.; Cao, Y. Appl. Phys. Lett.
2002, 80, 2045-2047. (b) Zhu, W. G.; Liu, C. Z.; Su, L. J.; Yang, W.;
Yuan, M.; Cao, Y. J. Mater. Chem. 2003, 13, 50-55. (c) Suzuki, M.;
Tokito, S.; Sato, F.; Igarashi, T.; Kondo, K.; Koyama, T.; Yamaguchi, T.
Appl. Phys. Lett. 2005, 86, 103507. (d) Kim, T. H.; Yoo, D. H.; Park, J.
H.; Park, O. O.; Yu, J. W.; Kim, J. K. Appl. Phys. Lett. 2005, 86, 171108.
(e) Choulis, S. A.; Mathai, M. K.; Choong, V. E.; So, F. Appl. Phys. Lett.
2006, 88, 203502.
(10) (a) Burkhart, R. D.; Chakraborty, D. K. J. Phys. Chem. 1990, 94, 4143-
4147. (b) Pina, J.; de Melo, J. S.; Burrows, H. D.; Monkman, A. P.;
Navaratnam, S. Chem. Phys. Lett. 2004, 400, 441-445.
(
(
(
7) Turro, N. J. Modern Molecular Photochemistry; University Science
Books: Sausalito, CA, 1991; pp 328-330.
8) (a) Evans, N. R.; Devi, L. S.; Mak, C. S. K.; Watkins, S. E.; Pascu, S. I.;
Kohler, A.; Friend, R. H.; Williams, C. K.; Holmes, A. B. J. Am. Chem.
Soc. 2006, 128, 6647-6656. (b) King, S. M.; Al-Attar, H. A.; Evans, R.
J.; Congreve, A.; Beeby, A.; Monkman, A. P. AdV. Funct. Mater. 2006,
1
6, 1043-1050. (c) Chen, X. W.; Liao, J. L.; Liang, Y. M.; Ahmed, M.
O.; Tseng, H. E.; Chen, S. A. J. Am. Chem. Soc. 2003, 125, 636-637.
4
700 J. AM. CHEM. SOC. VOL. 130, NO. 14, 2008
9

Shielding of Triplet Energy Transfer to Conjugated Polymer
A R T I C L E S
Figure 1. Chemical structures of F3, P4, PFO, CzPF, mC8OPPP, dC8OPPP, and CzPPP as hosts (quenchers in Stern-Volmer analysis) and Ir-G, FIrpic,
and Ir-B as guests.
carried out according to our previous reports. The reaction routes for
monomers are given in Schemes 1 and 2. The detailed synthetic
procedures for the monomers and polymers are described in the SI.
Instrumentation. Ultraviolet-visible (UV-vis), photoluminescence
and CzPPP are shown in Figure 1, and their UV-vis absorption
and photoluminescence (PL) spectra in THF solutions and as
thin solid films along with the absorption spectrum of carbazole
in THF are shown in Figure 2. The absorption maxima of dC8-
OPPP and mC8OPPP in THF are 337 and 335 nm, respectively.
However, the absorption peaks of CzPPP in THF at 261, 294,
(PL), photoexcitation (PLE), electroluminescence (EL), phosphores-
cence, gel permeation chromatography (GPC), scanning probe micros-
copy (SPM), the film thickness monitor and power supplier as well as
luminance meter for measurements of device performance are also
described in detail in the SI.
3
32, and 345 nm show similar features to those of Cz in THF,
causing identification of the absorption (at 332 or 345 nm) for
the polymer backbone or pendant Cz to be difficult. The absorp-
tion peaks of PPPs in thin solid films (Figure 2b) show a red-
shift by about 2 nm as compared with their corresponding peaks
in solutions except for dC8OPPP (4 nm red-shift). The PL
spectra of all the PPPs in THF show the same emission maxima
at 399 nm regardless of the difference in side groups. However,
the emission maxima in thin solid films show red-shifts to 400,
Device Fabrication and General Instrumentation. Indium-tin
oxide (ITO) glass plates were treated with oxygen plasma and then
1
2
coated with a CFx thin film as a hole injection layer by plasma
polymerization of CF H in a chamber of parallel electrode type equipped
3
with a 13.56 MHz radio frequency (rf) power generator. Its thickness
is about 2 nm as estimated using an ellipsometric spectrometer (from
Jobin Yvon). On top of it, a thin layer (emitting polymer layer, ca.100
nm) of emitting polymer was spin-cast from its solution (10 mg/mL)
in THF. The 1,3,5-tri(phenyl-2-benzimidazolyl)-benzene (TPBI) layer
4
11, and 442 nm for CzPPP, mC8OPPP, and dC8OPPP, res-
pectively, as compared with the corresponding peaks in solutions
all at 399 nm). The large red-shifts of emission maxima for
mC8OPPP and dC8OPPP could be attributed to a presence of
(30 nm), which was used as a hole/exciton blocking layer, was grown
(
by thermal evaporation in a vacuum thermal evaporator through a
-
6
shadow mask at a vacuum of 2 × 10 Torr. Finally, thin layers of
CsF (about 1 nm) and then calcium (about 5 nm) covered with a layer
of aluminum as the cathode for the bipolar device was deposited in a
vacuum thermal evaporator through a shadow mask at a pressure of
1
3
aggregates which emit a lower energy light or an increase of
conjugation length in films relative to that in solutions.
Figure 3a shows the UV-vis absorption spectrum of Ir-G
and PL emission spectra of PFs and PPPs in THF solutions.
The F o¨ rster energy transfer from the hosts to guest is expected
to occur due to the good spectral overlap between the absorption
of metal-to-ligand charge-transfer bands (410-460 nm) of Ir-
-
6
2
less than 2 × 10 Torr. The active area of the diode is about 10 mm .
A surface profiler (Tencor P-10) was used for determining polymer
film thickness.
Results and Discussion
3
b
G
and the emission spectra of the polymers. However, the
Photophysical Properties (UV-vis Absorption, Photolu-
minescence). The chemical structures of mC8OPPP, dC8OPPP,
PL spectrum of a 8 wt % Ir-G doped PFO film does not show
an Ir-G emission (peak maximum at 518 nm, Figure 3b) at
all, resulting from triplet excitons in Ir-G being significantly
quenched by the lower ET PFO. The extent of quenching is
reduced as CzPF is used as the host as reflected by a weak
(
11) Chen, S. A.; Chao, C. I. Synth. Met. 1996, 79, 93-96.
3g
(
12) (a) Hsiao, C. C.; Chang, C. H.; Jen, T. H.; Hung, M. C.; Chen, S. A. Appl.
Phys. Lett. 2006, 88, 33512. (b) Hsiao, C. C.; Chang, C. H.; Lu, H. H.;
Chen, S. A. Org. Electron. 2007, 8, 343-348. (c) Hung, L. S.; Zheng, L.
R.; Mason, M. G. Appl. Phys. Lett. 2001, 78, 673-675. (d) Tong, S. W.;
Lee, C. S.; Lifshitz, Y.; Gao, D. Q.; Lee, S. T. Appl. Phys. Lett. 2004, 84,
emission from Ir-G relative to the emission from the host. Both
4
032-4034. (e) Tang, J. X.; Li, Y. Q.; Hung, L. S.; Lee, C. S. Appl. Phys.
(13) (a) Peng, K. Y.; Chen, S. A.; Fann, W. S. J. Am. Chem. Soc. 2001, 123,
11388-11397. (b) Peng, K. Y.; Chen, S. A.; Fann, W. S.; Chen, S. H.; Su,
A. C. J. Phys. Chem. B 2005, 109, 9368-9373.
Lett. 2004, 84, 73-75. (f) Tang, J. X.; Li, Y. Q.; Zheng, L. R.; Hung, L.
S. J. Appl. Phys. 2004, 95, 4397-4403.
J. AM. CHEM. SOC.
9
VOL. 130, NO. 14, 2008 4701

A R T I C L E S
Huang et al.
doped mC8OPPP are observed. However, as shown in Figure
3
c, the relative host/guest intensities between Ir-G doped PFs
as well as PPPs exhibit similar trends in Figure 3b. Hence, we
can rule out the doping level effect (wt % or mol %) that might
lead to a variation in relative host/guest intensities between Ir-G
and host in the range of doping level investigated. But the other
possible factors can also affect the spectra in the solid state,
such as a shielding effect of triplet energy transfer to conjugated
polymers by configuration of their side chains, ET of host, and
3
g,9d,14
chemical compatibility between host and guest,
be discussed later.
as will
Shielding Effect Provided by Side Chains: Quantitative
Measurements via Stern-Volmer Analysis for Quenching
Rate Constants. In order to examine more specifically the
quenching of triplet excitons in phosphor dopants by low ET
polymer hosts with different chain configurations, F3, P4, PFs,
and PPPs as triplet quenchers for Ir-G, Firpic, and Ir-B (their
chemical strucutres are shown in Figure 1) were measured and
3
c
analyzed by the standard Stern-Volmer equation (see SI for
detail). The quenching analysis involves two parts: (a) a
fluorene-based series containing F3 with a short side chain
(methyl) and PFs with long side chains and (b) a phenylene-
based series containing P4 with a short side chain (methyl) and
PPPs with long side chains. The results of the Stern-Volmer
quenching analysis along with the ET’s of the materials
investigated are shown in Table 1. Due to the variations in ET’s
of the materials (Table 1 and SI), the ET difference between
guest and host (4 ET ) ET,guest - ET,host) is plotted versus the
quenching rate constant (Kqsv) as shown in Figure 4, though
the Kqsv is influenced not only by 4ET but also by several other
8
a,b
parameters, like steric hindrance.
For fluorene-based F3 and PFs as quenchers for Ir-G, the
9
8
Kqsv’s of F3, PFO, and CzPF are 1.14 × 10 , 2.06 × 10 , and
Figure 2. UV-vis absorption and normalized PL emission spectra of PPPs
8
-1
1
.76 × 10 (Ms) (Table 1), respectively. The Kqsv’s of PFO
(a) in THF solution and (b) as thin solid films. The absorption spectrum of
and CzPF are lower than that of F3 by 1 order of magnitude,
though the ET of F3 is significantly higher than those of PFs
by 0.13 eV. This would indicate that doping the Ir complex
carbazole is also provided in (a).
cases reveal the predominant host emission (the corresponding
mol % of Ir-G doping for PFO and CzPF based on their repeat
units are 5 mol % and 10 mol %, respectively). Surprisingly,
the predominant Ir-G emissions are observed for the cases with
PPPs as hosts also at the 8 wt % doping level (the corresponding
mol % of Ir-G doping for mC8OPPP, dC8OPPP, and CzPPP
based on their repeat units are 2.7 mol %, 4.4 mol %, and 9.5
mol %, respectively), in which no host emission for CzPPP, a
weak host emission at 442 nm for dC8OPPP, and a moderate
host emission at 411 nm for mC8OPPP are observed even though
the PPPs have a lower ET (ET of CzPPP ) 2.39 eV, ET of dC8-
OPPP ) 2.31 eV, and ET of mC8OPPP ) 2.3 eV) than Ir-G
into PFO instead of grafting onto the PF backbone spaced by
8a
-(CH2)8- can also reduce triplet energy transfer to the host
and that side chains do provide a significant shielding effect
for quenching.
For phenylene-based P4 and PPPs as quenchers, the Kqsv of
P4 for Ir-G is lower than those of PPPs by a factor of 2-15
(Table 1) due to the fact that the triplet energy transfer from P4
to Ir-G is energetically unfavorable (4 ET < 0) leading to the
lowest Kqsv. Interestingly, the Kqsv’s of PPPs for Ir-B are lower
than those of the corresponding PPPs for FIrpic by a factor of
1
.3-2.1 even though the energy gaps are larger by about 0.1
(2.41 eV).
eV. It appears that the Kqsv is also influenced by the steric
One may argue that Ir-G emission in CzPF may be due to
hindrance provided by the larger substituents at the Ir-core of
relative high mol % (10%) Ir-G doping concentration as
compared to that of PFO (5%). Therefore, the emission spectra
of PFs doped with 10 mol % Ir-G and PPPs doped with 4.4
mol % are also provided in Figure 3c to clarify this point. The
dopant concentrations of 10 mol % Ir-G for PFs are equal to
those of 16 and 8 wt % for PFO and CzPF, respectively, and
those of 4.4 mol % for PPPs are equal to those of 13, 8, and
8
b
Ir-B. In general, as 4 ET > 0, the Kqsv’s of P4 and PPPs for
the same Ir complex follow the sequence P4 > mC8OPPP >
dC8OPPP > CzPPP even though the ET of P4 is higher than
(14) Noh, Y. Y.; Lee, C. L.; Kim, J. J.; Yase, K. J. Chem. Phys. 2003, 118,
2
853-2864.
(
15) (a) Adachi, C.; Kwong, R. C.; Djurovich, P.; Adamovich, V.; Baldo, M.
A.; Thompson, M. E.; Forrest, S. R. Appl. Phys. Lett. 2001, 79, 2082-
2084. (b) Holmes, R. J.; D’Andrade, B. W.; Forrest, S. R.; Ren, X.; Li, J.;
3
.7 wt % for mC8OPPP, dC8OPPP, and CzPPP, respectively.
Thompson, M. E. Appl. Phys. Lett. 2003, 83, 3818-3820. (c) Chi, C. Y.;
Im, C.; Wegner, G. J. Chem. Phys. 2006, 124, 24907. (d) Allen, B. D.;
Benniston, A. C.; Harriman, A.; Llarena, I.; Sams, C. A. J. Phys. Chem. A
As compared with Figure 3b (8 wt % Ir-G doped polymers),
only a very weak CzPPP emission for 3.7 wt % Ir-G doped
CzPPP and a drop of mC8OPPP emission for 13 wt % Ir-G
2
007, 111, 2641-2649. (e) Lafolet, F.; Welter, S.; Popovic, Z.; De Cola,
L. J. Mater. Chem. 2005, 15, 2820-2828.
4702 J. AM. CHEM. SOC.
9
VOL. 130, NO. 14, 2008

Shielding of Triplet Energy Transfer to Conjugated Polymer
A R T I C L E S
Figure 3. (a) UV-vis absorption spectrum of Ir-G and normalized PL emission spectra of PFs and PPPs in THF soulutions. (b) Normalized PL spectra
of PFs and PPPs as thin films doped with 8 wt % Ir-G and (c) normalized PL spectra of PFs as thin film doped with 10 mol % Ir-G and PPPs as thin film
doped with 4.4 mol % Ir-G, respectively. The inserted table on the right side in Figure 3b is the wt % and corresponding mol % Ir-complex doping
concentrations, and that in Figure 3c is the mol % and corresponding wt % Ir-complex doping concentrations. The mol % is calculated based on their repeat
unit of the polymer (PFs are excited at 388 nm, and PPPs are excited at 340 nm).
that of all PPPs (Table 1 and Figure 4). This reflects that the
shielding provided by the side chains could overcompensate for
the difference in ET, and the extent of shielding by the
disubstitution is significantly higher than that of the monosub-
stitution. The reason is that the two side chains of dC8OPPP
are schematically nearly orthogonal16 with those of its neighbor-
ing repeat unit; thus the available space is so crowded for an
Ir-G molecule to insert into, and the chance for the aromatic
rings of the ligands to be in contact with the main chain is only
possible at a specific molecular orientation (Figure S5). For mC8-
OPPP, there is only one side chain per aromatic ring, leading
to a sufficiently large space available for an Ir-G molecule to
be accommodated at any orientation. Furthermore, Kqsv’s
(
16) (a) Park, K. C.; Dodd, L. R.; Levon, K.; Kwei, T. K. Macromolecules
996, 29, 7149-7154. (b) Pasco, S. T.; Baker, G. L. Synth. Met. 1997, 84,
275-276.
1
J. AM. CHEM. SOC.
9
VOL. 130, NO. 14, 2008 4703

A R T I C L E S
Huang et al.
Table 1. Data for the Stern-Volmer Quenching Analysis along
a
with the Triplet Energies of the Materials Investigated
∆E
T
)
iridium complex
triplet energy)
quencher
E
T,guest
−
E
T,host
K
qsv (M- s-1)
b
1
(
(triplet energy)
(eV)
9
Ir-G
F3 (ET ) 2.31 eV)f
0.10
0.23
0.23
-0.03
0.11
1.14 × 10
2.06 × 10
1.76 × 10
6.60 × 10
1.00 × 10
3.75 × 10
1.44 × 10
2.74 × 10
1.83 × 10
9.24 × 10
3.48 × 10
1.92 × 10
(
ET ) 2.41 eV)c PFO (ET ) 2.18 eV)
CzPF (ET ) 2.18 eV)
8
8
5
7
6
6
g
P4 (ET ) 2.44 eV)
mC8OPPP (ET ) 2.3 eV)
dC8OPPP (ET ) 2.31 eV)
CzPPP (ET ) 2.39 eV)
0.10
0.02
9
8
7
7
7
Flrpic
F3
0.31
0.18
0.32
0.31
0.23
(
ET ) 2.62 eV)d P4
mC8OPPP
dC8OPPP
CzPPP
8
7
7
6
Ir-B
P4
0.28
0.42
0.41
0.33
2.88 × 10
6.49 × 10
2.71 × 10
9.22 × 10
(
ET ) 2.72 eV)e mC8OPPP
dC8OPPP
CzPPP
a
The maximum of the first vibronic transition (T1ν)0 f S0ν)0) of the
phosphorescence spectrum is taken as ET (see the section “delay lumines-
b
cence” in SI). The Stern-Volmer equation τ0/τ ) τ0Kqsv[Q] + 1 was used,
where τ and τ0 are lifetimes of the phosphor with and without the quencher,
Kqsv is the Stern-Volmer quenching rate constant, and [Q] is the
c
d
e
concentration of the quencher. Reference 3b. Reference 15a. Reference
1
f
g
5b. Reference 15c. Reference 15d and e.
Figure 5. Phase-contrast images of the eight samples: (a) mC8OPPP, (c)
dC8OPPP and (e) CzPPP, and (b) mC8OPPP, (d) dC8OPPP, (f) CzPPP, (g)
PFO and (h) CzPF doped with 8 wt % Ir-G, as measured using a scanning
probe microscope (from Digital Instrument Nanoscope III).
of PFO and CzPF are located far below the F3 line indicating
a possible shielding effect provided by the side chains. In group
Figure 4. Triplet energy difference between guest and host (4 ET ) ET,guest
ET,host) versus quenching rate constant Kqsv. The Kqsv data of F3 for various
Ir complexes are taken from ref 3c (see SI for details). The solid lines are
for a visual guide.
-
(b), the three polymers exhibit the following sequence of
shielding effect as can be seen from the sequential decending
levels: CzPPP > dC8OPPP > mC8OPPP .
become even smaller by the incorporation of Cz on the end of
each side chain for CzPPP and CzPF relative to their analogues
dC8OPPP and PFO, respectively. This is mainly due to the plane
configuration of Cz providing more steric hindrance than the
line configuration of an alkoxyl or alkyl group. In addition, the
ET of CzPPP being higher than that of dC8OPPP should also
be responsible for the lower Kqsv in CzPPP.
The triplet energy transfer from the phosphor to conjgated
polymers is mainly governed by their difference in ET and
shielding effect provided by side chains in the polymer and/or
ligands in the phosphor. The plot of Kqsv versus 4ET given in
Figure 4 provides an overall picture of the correlation. The linear
line given for each quencher (polymer) does not mean an
existence of a linear relationship; it is merely given for a visual
guide. There are two groups of lines: (a) nonshielding in F3
and P4 and (b) shielding in PPPs. In group (a), the P4 line is
underneath the F3 line, which reflects less efficiency in
quenching resulting from the higher ET of the former. The Kqsv’s
One further question to be clarified is “Is molecular diffusion
in solution a rate-limiting step for the energy transfer process?”
If the energy transfer process is molecular diffusion-controlled,
the experimental value of Kqsv should essentially be invariant
for quenchers of widely varying structures; i.e., a mechanical,
not a molecular structure, feature is rate-limiting. However,
in our Stern-Volmer experiment, for example, the quenching
rate constants of the Ir-B system with mC8OPPP is larger than
that with dC8OPPP by a factor of 2.4 resulting from the
molecular structure difference. Thus the energy transfer is not
a molecular diffusion-controlled process.
3
i
Improvement in Chemical Compatibility between Host
and Guest. Figure 5 shows the scanning probe microscopy
(SPM) phase contrast images of pristine polymer films, mC8-
OPPP, dC8OPPP, and CzPPP before and after doping with 8
wt % Ir-G and of PFO and CzPF also doping with 8 wt %
Ir-G. Phase separation occurs in the Ir-G doped polymers,
and the extents of phase separation for the cases with CzPPP
4704 J. AM. CHEM. SOC.
9
VOL. 130, NO. 14, 2008

Shielding of Triplet Energy Transfer to Conjugated Polymer
A R T I C L E S
The EL spectra of dC8OPPP and CzPPP doped with Ir-G
8 wt % for the former and 4-12 wt % for the latter) are all
(
identical to that of the Ir-G emission (peak maximum at 518
nm) (Figure 8). Apparently, singlet excitons formed in the host
transfer to the guest, and triplet excitons in the guest must
originate from charge trapping in addition to the F o¨ rster energy
transfer as to be revealed below. As can be seen in Figure 7,
the HOMO and LUMO levels of Ir-G lay in between those of
the main chain; thus an occurrence of charge trapping upon the
electric field excitation can be expected. In addition, the trace
host emission in the PL spectrum for dC8OPPP doped with 8
wt % Ir-G (Figure 3b) dissapears in the corresponding EL
spectrum, and the current density of CzPPP is higher than those
of CzPPP doped with Ir-G (Figure 9a). These observations
show that both energy transfer and charge trapping are the
mechanisms responsible for the EL performances. The occur-
rence of charge trapping is very important for such a host/guest
system to obtain high performance. Without this condition,
triplet excitons generated directly in the guest will be small in
population, and those produced by Dexter energy transfer from
the host are significantly suppressed since the opportunity for
energy transfer is very low due to shielding by side chains.
Besides, it has been demonstrated in recent work4c that a
resonant and remote intermolecular heavy-atom effect that
increases the intersystem crossing rate in the polymer host was
found to occur in the Ir-SC4 (a substituted derivative of Ir-
(ppy)3 with increased solubility) doped polyspirobifluorene
system. The overlap between the emission spectrum of polyspiro-
bifluorene and MLCT absorption bands of Ir-SC4 is an
important factor giving rise to this remote heavy-atom effect.
In addition, the wavefunction overlap of delocalized pi electron
orbitals between the polymer and aromatic ligands of the
complex is an another important factor. This mechanism can
lead to an increase in the number of host triplet excitons,
Figure 6. Wide-angle X-ray diffraction patterns of PPPs.
and CzPF are significantly less than those with mC8OPPP, dC8-
OPPP, and PFO. These results indicate that the Cz moiety serves
as a compatibilizer between the dopant and polymer. Conse-
quently, CzPPP and CzPF as hosts are expected to allow
efficient use of Ir-G (by improved dispersion) upon photoex-
citation or electric field excitation.3
g,9d,15
WAXD patterns for PPPs are shown in Figure 6. CzPPP is
featureless (amorphous and no side chain alignment), while mC8-
OPPP is also essentially amorphous but with a very small extent
of side chain alignment, as reflected in the presence of a small
hump in the 2θ range 3°-5°. For dC8OPPP, the side chain
alignment is very pronounced as reflected in the presence of an
intense and sharp diffraction peak at 3.42°. These results indicate
that the Cz moiety prevents an order alignment of the side chain
for CzPPP and thus reduces the extent of phase separation of
Ir-G from the CzPPP matrix, which is in agreement with the
results of AFM studies.
indepentent of host E ; quenching up to 95% of all singlet
T
Efficient Green Electrophosphorescence from Ir-G Doped
Low ET Polymers with Dense Side Chains. Figure 7 shows
the band diagram of the materials investigated in which the
excitons and competing effectively with F o¨ rster energy transfer
are observed. Consequently, devices of FIrpic and Ir-B doped
PPPs are expcted to be poor in performance due to the lack of
energy levels matching for hosts and guests (Figure 7). However,
the charges (holes and electrons) transporting from host to guest
would not be shielded by the dense side chains due to the field
dependence of charge transport and lower energy levels of Ir-
G. The distance of the fully extended side chain -O-(CH2)10-
Cz is estimated to be about 20 Å (from the oxygen atom to the
hydrogen atom bound to carbazole unit at 3 or 6 positon), which
is within the charge hopping distance (21 Å), as was observed
energy levels of CFx,1 Ir complexes,
2a,b
3b,15a,17
and TPBI are
5b
taken from previous reports. The HOMO levels of PPPs were
measured by cyclic voltammograms (see SI), and the corre-
sponding LUMO levels are deduced from optical band gaps
and the HOMO levels. For electroluminescence (EL) studies,
dC8OPPP and CzPPP are chosen as the hosts and Ir-G is chosen
as the guest for the bipolar devices (ITO/CFx/polymer: dopant/
TPBI/CsF/Ca/Al). The device performaces are shown in Figures
18
8
-10, and their characteristic values are listed in Table 2.
in the molecularly doped polymers (such as 5% triphenylamine
Figure 7. Energy level diagram of the materials used in the devices.
J. AM. CHEM. SOC.
9
VOL. 130, NO. 14, 2008 4705

A R T I C L E S
Huang et al.
Figure 10. Current efficiency and brightness versus current density for 8
wt % Ir-G doped CzPPP devices.
Figure 8. EL spectra of thin films of CzPPP with different Ir-G doping
levels and EL spectrum of dC8OPPP doped with 8 wt % Ir-G.
Table 2. Performances of the Devicesa
dopantb
concn
turn-on
voltage
(V)
host
max. efficiency
(cd/A) [
max. brightness
polymer
(wt %)
η
η
ext, %]
B
max (cd/m2)
CzPPP
4 (Ir-G)
13
13
14
16
10 [2.8]
30 [8.25]
7 [1.87]
15 [4.12]
2600
6127
4663
1500
8
(Ir-G)
1
2 (Ir-G)
8 (Ir-G)
dC8OPPP
a
ITO/CFx (2 nm)/emitting polymer layer (100 nm)/TPBI (30 nm)/CsF
b
2′
(
1 nm)/Ca (5 nm)/Al. Ir-G: (bis(2-phenylpyridinato-N,C )iridium(ace-
tylacetonate).
maximum luminous efficiency of 30 cd/A (corresponding to
external quantum efficiency 8.25%) (Table 2). The efficiency
of 30 cd/A is higher than those of high ET polymers as hosts,5
but it drops to half (15 cd/A, 4.12%) under the higher current
2
density 40 mA/cm (Figure 10) which is lower than the
5
a
efficiency reported by van Dijken et al. where the efficiency
of 23 cd/A is stable over a large current range up to 46 mA/
2
cm . The 8 wt % doping level gives optimal performance in
the range 4-12 wt % as can be seen from the data in Table 2
which were calculated from the curves in Figure 9. For 8 wt %
Ir-G doped dC8OPPP, the performances give a maximum
2
brightness of 1500 cd/m and maximum luminous efficiency
of 15 cd/A (4.12%). As can be seen in Figure 7, despite the
uses of CFx and CsF/Ca/Al for reducing the injection barriers,
the turn-on voltages are still high (13-16 V) and the brightness
is only modest for both devices. [The work function of CFx
(5.3-5.7 eV) deposited on the ITO surface can be finely tuned
by using different rf power (20-50 W) during the plasma
1
2b
polymerization, and the CFx with a work function of 5.3 eV
and thickness 2 nm was obtained by using 50 W rf power. The
use of a multilayer CsF/Ca/Al cathode was found to be effective
for electron injection since CsF can be reduced to form low
Figure 9. (a) Current density and (b) brightness versus voltage for the
devices of Cz-PPP with or without Ir-G doping.
19
in poly(styrene)18c). Consequently, charge hopping from the host
to the guest is expected to take place.
work function cesium (2.2 eV) by the Ca layer. ] This is mainly
9
a,b,20
due to the low charge mobilities of PPPs
and thus low
current densities (Figure 9a). The promoted efficiency of the
The performances of the bipolar device with 8 wt % Ir-G
2
doped CzPPP give a maximum brightness of 6127 cd/m and
(
19) (a) Greczynski, G.; Fahlman, M.; Salaneck, W. R. J. Chem. Phys. 2001,
1
14, 8628-8636. (b) Fung, M. K.; Lai, S. L.; Bao, S. N.; Lee, C. S.; Lee,
(
17) D’Andrade, B. W.; Brooks, J.; Adamovich, V.; Thompson, M. E.; Forrest,
S. R. AdV. Mater. 2002, 14, 1032-1036.
S. T.; Wu, W. W.; Inbasekaran, M.; O’Brien, J. J. J. Vac. Sci. Technol. A
2002, 20, 911-918. (c) Brown, T. M.; Friend, R. H.; Millard, I. S.; Lacey,
D. J.; Butler, T.; Burroughes, J. H.; Cacialli, F. J. Appl. Phys. 2003, 93,
6159-6172.
(18) (a) Borsenberger, P. M. J. Appl. Phys. 1992, 72, 5283-5287. (b)
Borsenberger, P. M.; Magin, E. H.; Fitzgerald, J. J. J. Phys. Chem. 1993,
9
7, 9213-9216. (c) Borsenberger, P. M.; Oregan, M. B. Chem. Phys. 1995,
00, 257-263. (d) Borsenberger, P. M.; Magin, E. H.; Oregan, M. B.;
(20) (a) Yang, Y.; Pei, Q.; Heeger, A. J. J. Appl. Phys. 1996, 79, 934-939. (b)
Huang, J. S.; Zhang, H. F.; Tian, W. J.; Hou, J. Y.; Ma, Y. G.; Shen, J. C.;
Liu, S. Y. Synth. Met. 1997, 87, 105-108. (c) Ma, D. G.; Hummelgen, I.
A.; Li, R. W. C.; Gruber, J. J. Phys. D: Appl. Phys. 2000, 33, 1376-
1379.
2
Sinicropi, J. A. J. Polym. Sci., Part B: Polym. Phys. 1996, 34, 317-323.
(
e) Borsenberger, P. M.; Magin, E. H.; Sinicropi, J. A.; Lin, L. B. Jpn. J.
Appl. Phys. Part 1 1998, 37, 166-170.
4
706 J. AM. CHEM. SOC. VOL. 130, NO. 14, 2008
9

Shielding of Triplet Energy Transfer to Conjugated Polymer
A R T I C L E S
CzPPP-device over the dC8OPPP-device can be attributed to
the better chemical compatibility between the host and guest
provided by the incorporation of Cz (Figure 5). In addition,
incorporation of 8 wt % Ir-G into the polymers leading to
reduction of charge injection barriers (Figure 7) and promoted
charge balance within emissive layer could also be the reason
between host and guest. If the HOMO and LUMO levels of
the phosphor dopant are intercalated in those of the host polymer
so that charge trapping can occur, the host/guest system is able
to emit light solely from the guest efficiently even though the
ET of the host is lower than that of the guest. This observation
suggests that the ET of polymer host is not necessary to be higher
than that of phosphor guest for efficient electrophosphorescence.
Such a finding provides more freedom for molecular design of
electroluminescent polymers as hosts for phosphor dopants.
2
1
for the improved performance in Ir-G doped CzPPP.
Conclusions and Significance
In summary, we found that the dialkoxyl side chains provide
effective shielding for the backbone in PPPs such that energy
transfer of a triplet exciton from a phosphor dopant (Ir complex)
is sigificantly reduced or almost completely prevented. The
shielding effect is more obvious when the ET of the host is far
lower than that of the guest. Moreover, incorporation of Cz on
the end of the alkoxyl side chain can further hinder the back
triplet energy transfer and enhance chemical compatibility
Acknowledgment. We thank the Ministry of Education
Project 91E-FA04-2-4A) and the National Science Council
Project 95E-2752-E007-005 and 008-PAE) for financial sup-
(
(
port.
Supporting Information Available: Instrumental, synthesis,
device fabrication, optical measurement, Stern-Volmer analysis,
AFM, and WXRD (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
(
21) (a) Al Attar, H. A.; Monkman, A. P. AdV. Funct. Mater. 2006, 16, 2231-
2
242. (b) Matsusue, N.; Ikame, S.; Suzuki, Y.; Naito, H. Appl. Phys. Lett.
2004, 85, 4046-4048. (c) Matsusue, N.; Suzuki, Y.; Naito, H. Jpn. J. Appl.
Phys. Part 1 2005, 44, 3691-3694.
JA076413I
J. AM. CHEM. SOC.
9
VOL. 130, NO. 14, 2008 4707