Energy transfer pathways in the carbazole functionalized β-diketonate europium complexes

Article abstract of DOI:10.1039/b705666d

Two novel Eu³⁺ complexes Eu(CDBM)₃·2H ₂O and Eu(CCDBM)₃·2H₂O have been synthesized (CDBM = 1-(4-(9-carbazol)phenyl)-3-phenyl-1,3-propanedione, CCDBM = 1-((4-(9-carbazol)methyl)phenyl)-3-phenyl-1,3-propanedione). Eu(CDBM) ₃·2H₂O showed strong internal ligand charge transfer (ILCT) fluorescence and sensitized emission of Eu³⁺. Due to the charge transfer character, the fluorescence band of CDBM shifted from 403 nm in cyclohexane to 559 nm in acetonitrile. Photophysical studies demonstrated that no energy was migrated from the ILCT excited state of the ligands to Eu³⁺, and that Eu³⁺ was sensitized by the triplet state which was localized in the 1,3-diphenyl-1,3-propanedione (DBM) part. The quantum efficiencies of Eu³⁺ in the three complexes are in the order Eu(DBM)₃·2H₂O > Eu(CCDBM)₃· 2H₂O > Eu(CDBM)₃·2H₂O in both solution and the solid state. The energy transfer pathways in the three Eu ³⁺ complexes were discussed in detail. Based on the systematic photophysical studies, a new guideline for the organo-lanthanide light emitting materials has been proposed: ILCT should be avoided during molecular modification. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique.

Full text of DOI:10.1039/b705666d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Energy transfer pathways in the carbazole functionalized b-diketonate
europium complexesw
a
a
a
b
a
Daobo Nie, Zhuqi Chen, Zuqiang Bian,* Jianqiang Zhou, Zhiwei Liu,
a
b
a
Fangfang Chen, Yongliang Zhao and Chunhui Huang*
Received (in Gainesville, FL, USA) 16th April 2007, Accepted 17th May 2007
First published as an Advance Article on the web 8th June 2007
DOI: 10.1039/b705666d
3+
Two novel Eu complexes Eu(CDBM)
3
Á 2H
2
O and Eu(CCDBM)
3
Á 2H
2
O have been synthesized
(
CDBM = 1-(4-(9-carbazol)phenyl)-3-phenyl-1,3-propanedione, CCDBM = 1-((4-(9-
carbazol)methyl)phenyl)-3-phenyl-1,3-propanedione). Eu(CDBM)
3
Á 2H
2
O showed strong internal
3
+
ligand charge transfer (ILCT) fluorescence and sensitized emission of Eu . Due to the charge
transfer character, the fluorescence band of CDBM shifted from 403 nm in cyclohexane to
5
59 nm in acetonitrile. Photophysical studies demonstrated that no energy was migrated from the
3
ILCT excited state of the ligands to Eu , and that Eu was sensitized by the triplet state which
+
3+
3
+
was localized in the 1,3-diphenyl-1,3-propanedione (DBM) part. The quantum efficiencies of Eu
in the three complexes are in the order Eu(DBM) O 4 Eu(CCDBM) Á 2H O 4
Á 2H
Eu(CDBM) O in both solution and the solid state. The energy transfer pathways in the
Á 2H
three Eu complexes were discussed in detail. Based on the systematic photophysical studies, a
3
2
3
2
3
2
3
+
new guideline for the organo-lanthanide light emitting materials has been proposed: ILCT should
be avoided during molecular modification.
Introduction
No matter which pathway the sensitization processes take,
the energy absorbed by the ligand should be transferred to the
lanthanide ions as much as possible to obtain a high quantum
yield. However, this factor is sometimes ignored in the design
of new organo-lanthanide materials. For instance, the carba-
zole functional group is usually incorporated into the
b-diketonate ligands to improve the hole transfer mobility of
1
Due to the forbidden character of the intra-4f transitions,
luminescent lanthanide ions exhibit long lifetime and high
color purity emissions, which have made them ideally suited
2
for various applications such as fluorescent probes, optical
3 4,5
signal amplifiers and organic light-emitting diodes. Unfor-
tunately, for the same reason, the absorption coefficients of
À1
12
lanthanide complexes. Because the carbazole functional
À1
6
lanthanide ions are very low (less than 10 M cm ). To
group is an electron donor and the carbonyl groups in
b-diketonates can be regarded as electron acceptors, this
modification may lead to the internal ligand charge transfer
overcome this problem, organic chromophores which have
4
À1 À1
cm ) are
much larger absorption coefficients (B10
M
usually coordinated to lanthanide ions to form lanthanide
complexes. In this case, the sensitization process of the lantha-
nide ions generally consists of three steps: the excitation of the
singlet excited state of the ligand, the subsequent intersystem
crossing to its triplet state and energy transfer from the triplet
(
ILCT) process. Unfortunately, most of the relevant papers
did not investigate the energy transfer pathways in these
modified complexes. Whether the energy can be transferred
from the ILCT excited states to the lanthanide ion still remains
an unsolved problem.
1
,6–8
state to the emission lanthanide ion.
On the other hand,
In order to investigate how these modifications influence the
sensitization process in a lanthanide complex, we synthesized
two new b-diketonate ligands by introducing a carbazole
moiety into a classic 1,3-diphenyl-1,3-propanedione ligand
direct energy transfer from the singlet excited state of the
9
ligand to the lanthanide ion has also been observed.
,10
3
Recently, sensitization of Eu through charge transfer (CT)
+
1
1
excited states of the ligand has also been reported.
(
DBM) by different linkage (Scheme 1). In CDBM, the
carbazole fragment was directly linked to DBM, while in
CCDBM, the conjugation between carbazole fragment and
a
Beijing National Laboratory for Molecular Sciences, State Key
Laboratory of Rare Earth Materials Chemistry and Applications,
College of Chemistry and Molecular Engineering, Peking University,
Beijing, 100871, P. R. China. E-mail: chhuang@pku.edu.cn;
Fax: +86-10-62757156; Tel: +86-10-62757156
2
the DBM part was interrupted by a –CH – group to avoid the
internal ligand charge transfer process. Europium complexes
were prepared with these ligands because the triplet state
3
+
b
energy level of DBM matched the receiving state of Eu
College of Chemistry and Chemical Engineering, Inner Mongolia
University, Hohhot, 010021, P. R. China
very well. Photophysical studies of these complexes in both
3+
solutions and the solid state revealed that Eu could not be
w Electronic supplementary information (ESI) available: Absorption
spectra, emission spectra in the solid state, element analysis results of
the complexes, calculation details of the rate constants in Fig. 7. See
DOI: 10.1039/b705666d
sensitized by the ILCT excited states of CDBM. The sensitiza-
tion process still took the triplet pathway.
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1639–1646 | 1639

View Article Online
processing the lifetime data with the F900 reconvolution soft-
ware, a time resolution better than 1 ns for FLS920 and 20 ps
for Lifespec-Red can be achieved.
Phosphorescence spectra of the gadolinium complexes were
measured by mounting the samples into a cryostat (Oxford
Instruments, Optistat DN) which had been cooled by liquid
nitrogen to 77 K. A methanol–ethanol (1 : 4) mixture was
chosen as the solvent to form a transparent glass at low
temperatures.
The quantum yields of the complexes in solution were
determined using quinine sulfate in 0.05 M H
2
SO
The quantum yields of
the complexes in the solid state were measured with an
4
as a
1
4,15
f
standard (f = 0.546 at 25 1C).
1
6,17
integrating sphere
using a mercury lamp (365 nm) as the
excitation source and a photodiode as the detector. The
spectral response of the sphere and the photodiode were not
calibrated, so the results can only be used in comparison.
Scheme 1 Molecular structures of the three b-diketonate ligands.
Experimental
Computational details
Synthesis of the ligands
The calculations were carried out using the MOPAC6_RM1
1
8
package.
3+
The geometry optimization of the three
DBM was purchased from ACROS and used as received.
3
CDBM was synthesized according to our other paper. The
synthetic procedure of CCDBM was almost the same as that
Eu
complexes have been done employing a semiempirical
19,20
1
Sparkle/AM1 model.
of CDBM except that 4-fluorobenzonitrile was replaced by
1
4
-(bromomethyl)benzonitrile. See ESIw for the H NMR data
of CDBM and CCDBM.
Results and discussion
Molecular geometry
Synthesis of the complexes
3+
The optimized structures of the three Eu
complexes are
Eu(DBM)
3
Á 2H
2
O: To a 50 ml round-bottomed flask, DBM
shown in Fig. 1. Selected bond lengths around the central ion
are tabulated in Table 1. It has been reported that the accuracy
(
0.672 g, 3.0 mmol), NaOH (0.120 g, 3.0 mmol) were mixed in
3
+
1
0 ml ethanol and refluxed for 10 min, then it was added
Á 6H O (0.403 g,
.1 mmol). The mixture was refluxed for 2 h and then poured
of the optimizations of the Eu
complexes using Sparkle/
dropwise to 10 ml ethanol solution of EuCl
3
2
AM1 model is comparable with those using ab initio calcula-
2
1
1
tions. Because the crystal structures of the three complexes
are not available, we also optimized a Eu(DBM)
into water. The crude product was obtained by filtration and
purified by recrystallization in a mixture of THF–ethanol. All
the other complexes were synthesized by the same procedure.
The gadolinium complexes were synthesized to investigate the
triplets of the ligands. All the complexes were characterized by
element analysis (see ESIw for details).
3
Á phen com-
2
2
plex, in which the two water molecules are replaced by a
neutral phenanthroline ligand. From Table 1 we can see that
the discrepancies between the optimized and experimental
˚
structures of Eu(DBM)
3
Á phen are within 0.03 A, thus this
model is accurate enough to describe our systems.
The coordination environments of the three complexes are
all distorted square antiprisms. All the complexes belong
to the C1 point-group. From Eu(DBM) Á 2H O to Eu-
Photophysical measurements
Absorption spectra were recorded with Shimadzu UV 3100
spectrophotometer, and steady state fluorescence spectra were
taken with an Edinburgh Instruments FLS920 spectrometer.
All the fluorescence spectra were corrected for the instrument
response using a correction file provided by the manufacturer.
3
2
(
CDBM) Á 2H O and Eu(CCDBM) Á 2H O, the incorpora-
3
2
3
2
tion of the carbazole group does not change the coordination
3
+
environment of Eu significantly. The geometries of the first
coordination sphere in the three complexes remain almost the
same. This result indicates that the structure difference among
3
+
Lifetimes of Eu
were measured with Edinburgh Instru-
3
+
the three Eu
complexes is not responsible for the big
ments FLS920 based on the time correlated single photon
counting technology. A microsecond flash lamp was used as
the excitation source and the signals were detected with a
photomultiplier (Hamamatsu R955). Lifetimes of the ligands
longer than 1 ns were detected with the same equipment using
a hydrogen filled nanosecond flash lamp as the excitation
source. A picosecond lifetime spectrometer Edinburgh Instru-
ments Lifespec-Red was used to measure lifetimes shorter than
3
+
discrepancy of the Eu emission (see the following text).
Absorption
The absorption spectra of the complexes are very similar to
that of the corresponding ligands, indicating that the singlet
excited states of the ligands are not significantly affected by the
3
+
6
complexation to the Eu
ion (see Fig. S1 in ESIw). DBM
1
ns. In this equipment, a diode laser (operating at 372 nm,
shows a broad and structureless band centered at 350 nm.
CDBM and CCDBM both have a sharp and structured band
around 292 nm. We assigned this band to the locally excited
pulse duration 69 ps) controlled by a picosecond light pulser
(
Hamamatsu PLP-10) was used as excitation source. By
1
640 | New J. Chem., 2007, 31, 1639–1646
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007

View Article Online
In the region from 300 to 400 nm, the absorption spectrum
of CCDBM is very similar to that of DBM, except that a
narrow and sharp band arises at 330 nm. Compared with
CDBM, the disappearance of the broad band at 380 nm and
the existence of the narrow band at 330 nm indicate that only
the CT excited state related to carbonyl group 1 exists in
CCDBM. One possible explanation for this phenomenon is
that in CCDBM, the distance between carbazole fragment and
carbonyl group 2 gets too long to form a CT excited state.
Excitation
The excitation spectra together with the absorption spectra of
3+
the three Eu complexes are depicted in Fig. 2. The excita-
tion spectra of Eu(DBM) Á 2H O and Eu(CCDBM) Á 2H O
3
2
3
2
5
7
monitored at 611 nm (corresponding to D - F transition
0
2
3
+
of Eu ) closely resemble the corresponding absorption spec-
tra (top and bottom panel in Fig. 2). In contrast, the excitation
spectrum of Eu(CDBM)
3
Á 2H
2
O monitored at 611 nm lost
some features around 340 nm comparing with the absorption
spectrum and the excitation spectrum monitored at the ILCT
emission maximum (see the arrows in Fig. 2). This indicates
that in Eu(DBM)
3
Á 2H
ions are excited via the whole b-diketonate ligands,
while the situation in Eu(CDBM) O is somewhat differ-
Á 2H
ent due to the different conjugation system.
2
O and Eu(CCDBM)
3
Á 2H
2
O, the
3
+
Eu
3
2
3+
Sensitized Eu emission
Upon excitation near the absorption maximum of the ligands,
the three complexes all showed characteristic narrow band
3
+
5
emissions of Eu corresponding to the D
7
0 J
- F (J = 0–4)
transitions (Fig. 3). The emission bands at 580 and 650 nm are
7
5
very weak because their corresponding transitions D
0
- F0,3
2
3
are forbidden both in magnetic and electric dipole schemes.
The intensity of the emission band at 592 nm is relatively
strong and independent of the coordination environment
5
because the corresponding transition D
7
0
- F
1
is a magnetic
5
7
transition; on the contrary, the D - F transition is an
0
2
induced electric dipole transition and its corresponding intense
emission at B615 nm is very sensitive to the coordination
environment.
3
+
2
3
Fig. 1 Optimized geometries of the three Eu complexes using the
Sparkle/AM1 model. Hydrogen atoms are omitted for clarity.
The energy transfer pathways
1
3
state of carbazol moiety. Compared with DBM, the broad
and strong absorption band of CDBM is red shifted to
CDBM is highly fluorescent, DBM is non-fluorescent and
CCDBM showed very weak fluorescence in all solvents (not
3
+
3
80 nm. Around 330 nm, CDBM shows a small and narrow
shown). After coordinating with Eu , the ligand fluorescence
band shapes did not change very much.
shoulder. We assigned these two absorption bands to the
charge transfer (CT) excited states corresponding to electron
transfer from carbazole moiety to carbonyl group 2 and 1 (see
At first glance to the fluorescence spectra of Eu
(CDBM) Á 2H O, one may note that the intensity of the
3
2
1
3
3+
Eu
Scheme 1), respectively.
emission varies as the ligand emission band shifts
3+
˚
Table 1 Selected bond lengths (A) around the central Eu ion
a
1
Eu–O
Eu–O
2
Eu–H
2
O
C
1
–O
1
C
3
–O
2
C
1
–C
2
2 3
C –C
Eu(DBM)
Eu(CDBM)
Eu(CCDBM)
3
Á 2H
2
O
2.377
2.377
2.380
2.380
2.351
b
2.375
2.376
2.380
2.385
2.383
2.399
2.398
2.398
—
1.290
1.290
1.291
1.288
1.267
1.291
1.291
1.290
1.289
1.271
1.410
1.409
1.407
1.409
1.407
1.409
1.410
1.410
1.409
1.401
3
2
Á 2H O
3
Á 2H
2
O
Eu(DBM)
Eu(DBM)
a
3
Á phen
b
3
Á phen
—
See Fig. 1 for atom numbering. Crystal structure from ref. 22.
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1639–1646 | 1641

View Article Online
Fig. 2 The excitation and absorption spectra of the three complexes
in CH Cl . The solid lines are the excitation spectra monitored at
11 nm; The dotted line in the middle panel is the excitation spectrum
monitored at 511 nm, which is the emission maximum of CDBM in
CH Cl
2
2
6
Fig. 3 The corrected emission spectra of the three complexes re-
À5
corded in solution. Upper panel: emission spectrum of 10
Eu(DBM) O in CH Cl
Á 2H
M
À5
2
2
.
Á 2H
2
; middle panel: emission spectra of 10
3
2
2
M Eu(CDBM)
3
2
O in various solvents; bottom panel: emission
À5
spectra of 10 M Eu(CCDBM)
acetonitrile.
3
2 2 2
Á 2H O in cyclohexane, CH Cl and
(
Fig. 3, middle panel). Since sensitization of lanthanide ion
2
4
through the singlet pathway has been proposed theoretically
9
,10
and observed experimentally,
it is reasonable to assume
that the ILCT excited state of CDBM may be involved in the
listed in Table 2. The fluorescence properties of the free ligand
3
sensitization process of Eu . Generally speaking, two me-
+
are also presented to study the heavy atom effect of the
3+
3
chanisms are invoked in the sensitization process of Ln ion:
+
lanthanide ions. Gd
can not receive energy from most
2
Dexter (exchange) mechanism and Fo
5
¨
According to these two mechanisms, the en-
rster (dipole–dipole)
organic ligands because the energy level of its lowest excited
1
,26
À1 7,9
3+
mechanism.
state is too high (about 32 000 cm ). Thus Gd complex
can be used as a model to study the ligand fluorescence in the
presence of heavy atom effect but in the absence of energy
transfer. If energy transferred from the ILCT excited states of
ergy-transfer rate is proportional to the overlap of the ligand
emission spectrum and the lanthanide absorption spec-
1
,24
trum.
It is very probable that in some solvents, the energy
3
+
level of the ILCT excited state of CDBM may perfectly match
3
CDBM to Eu in certain solvents, quantum yield reduction
+
the receiving states of Eu and efficient energy transfer may
occur. In order to expose the energy transfer pathways, we
and lifetime shortening should be observed for CDBM in the
3
+
3+
Eu
complex compared with those in the Gd
complex.
studied the photophysical properties of Eu(CDBM)
3
Á 2H
2
O
However, both quantum yields and lifetimes of CDBM are
almost the same in both complexes in all solvents (Table 2).
and Gd(CDBM) O in several solvents. The results are
3
Á 2H
2
3+
3+
Table 2 Photophysical properties of CDBM as a free ligand and in Gd , Eu complexes in various solvents
t/ns
Q
s
À1
Solvent
Band position /1000 cm
Free ligand
Gd
Eu
Free ligand
Gd
Eu
a
b
Acetonitrile
Acetone
CH Cl
2 2
17.89
19.19
19.65
20.70
21.23
22.83
23.26
24.81
3.88
3.89
5.22
5.49
3.17
1.95
0.17
0.14
0.03
3.93
5.33
5.55
3.13
1.89
0.17
0.12
0.03
0.036
0.014
0.043
0.069
0.094
0.051
0.007
0.010
0.0006
0.013
0.041
0.062
0.094
0.054
0.006
0.009
0.0008
5.21
5.54
3.19
2.04
0.18
0.13
0.04
0.073
0.154
0.131
0.074
0.015
0.013
0.001
CHCl
THF
3
Diethyl ether
Toluene
Cyclohexane
a
Error Æ2%. Error Æ5%.
b
1
642 | New J. Chem., 2007, 31, 1639–1646
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007

View Article Online
Fig. 5 Phosphorescence spectra of Gd(DBM)
3
Á 2H
2
O (dashed line),
Fig.
Gd(CDBM)
line) at 77 K in methanol–ethanol (1 : 4) glass.
4
The emission spectra of CDBM (solid line),
Gd(CDBM)
line).
3
Á 2H
2
O (solid line) and Gd(CCDBM)
3
Á 2H
2
O (dotted
3
Á 2H O (dashed line) and Eu(CDBM) Á 2H O (dotted
2
3
2
3
+
These data imply that Eu can not be sensitized by the ILCT
excited state of CDBM. Another striking result is that the
lifetimes of the free ligand remained all the same upon
Gd(CCDBM)
Gd(CDBM)
3
Á 2H
2
O, the results together with that of
Á 2H
O are illustrated in Fig. 5. One can find
2
3
from Fig. 5 that Gd(DBM)
Á 2H
O displays weak fluorescence
3
2
3
+
3+
coordination with Gd
and Eu . That is to say, there is
band around 420 nm (lifetime 1 ns) and strong phosphores-
cence band around 500 nm (1.8 ms). Both bands show fine
structures and closely resembled each other in band shape. The
no detectable heavy atom effect for the ILCT excited state of
CDBM.
It is also noteworthy that the ILCT fluorescence quantum
yields of CDBM decreased dramatically after coordination
triplet state energy level of Gd(DBM)
À1
3
Á 2H
2
O was determined
to be 20 700 cm from the 0–0 transition of the phosphores-
cence band. This result agrees well with that reported in
3
+
3+
with Gd and Eu (Table 2). We will explain this confusing
phenomenon in the following text.
2
literature. In the spectrum of Gd(CDBM)
8
3
2
Á 2H O, the ILCT
If the sensitization process does not take the singlet path-
way, it must have taken the triplet pathway. Low-temperature
fluorescence becomes dominate and the intensity of the phos-
phorescence band gets relatively weak. However, despite the
dissimilarity in relative intensity, the phosphorescence band
spectra of free CDBM, Gd(CDBM)
3
Á 2H
2
O and Eu
(
CDBM) O have been measured to prove this point of
3
Á 2H
2
shape and the 0–0 transition band position of
À1
view. The results are shown in Fig. 4. The three profiles all
showed strong emission bands at B423 nm, which was con-
firmed to be a fluorescence band by lifetime measurements
Gd(CDBM)
3
Á 2H
2
O (20 450 cm ) are almost the same as
that of Gd(DBM)
3
2
Á 2H O. This result suggests that the in-
corporation of the carbazole functional group into DBM only
causes the ILCT fluorescence. It does not affect the triplet
energy level of DBM at all. In fact, the population of the
triplet states gets smaller because most of the energy has been
released in the form of ILCT fluorescence. Generally speaking,
the triplet state energy level of the ligands will decrease if the
conjugation system is enlarged. This result indicates that the
conjugation system of CDBM only has been slightly enlarged
compared with that of DBM. The optimized molecular struc-
(
about 2 ns). In the spectrum of Gd(CDBM)
3
Á 2H
2
O, a
moderately strong emission band emerged at B489 nm. Life-
time measurements proved it to be a phosphorescent band
(
lifetime 1.6 ms). In the case of Eu(CDBM) Á 2H O, this
3
2
phosphorescence band almost completely disappeared and
3
the characteristic narrow emission bands of Eu
+
appeared
(
Fig. 4). This result demonstrated that the sensitization process
+
3
of Eu still took the triplet pathway. As a consequence, the
Dexter type energy transfer should be the dominant mechan-
ture of Eu(CDBM)
Á 2H
O also supported this conclusion. As
3
2
3
+ 24,27
.
ism in the sensitization process of Eu
can be seen from Fig. 1, the carbazole moiety and the DBM
part are not coplanar. The carbazole fragment is twisted about
501 with respect to the phenyl ring. As a result, the conjugation
system of CDBM remains almost the same as that of DBM.
However, there is still one confusing problem left: where
does the phosphorescence band come from? In the first part of
this section, we have already concluded that the coordinating
lanthanide ions have little heavy atom effects on the ILCT
excited states of CDBM. Obviously the ILCT excited state is
not the origin of the phosphorescence band. We noticed that
2
Since the D–p–A system has been interrupted by a –CH –
group, the ILCT fluorescence of CCDBM has been greatly
suppressed and the phosphorescence spectrum closely resem-
bles that of DBM (Fig. 5). This result further supports our
conclusion.
À1
the triplet state energy level of CDBM (20 450 cm ) lies very
À1 28
close to that of DBM (20 520 cm ). It is very likely that the
triplet excited state of CDBM still localizes in the DBM
fragment. To confirm this assumption, we recorded the
As already discussed above, the introducing of carbazole
functional group to DBM depopulates the triplet excited
3
+
phosphorescence spectra of Gd(DBM)
3
2
Á 2H O
and
states, through which the Eu
is sensitized. Based on this
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1639–1646 | 1643

View Article Online
3+
(%) of Eu
Table 3 The overall quantum yields Q
complexes in various solvents
s
in the three
spectra are very similar, thus the spectra overlaps are nearly
identical, too. For these two reasons, Fet are also excluded for
the possible explanation. In the case of Eu(DBM) Á 2H O, the
3
2
Solvents
DBM–Eu
CDBM–Eu
CCDBM–Eu
F
isc
should be close to unity because the phosphorescence is
Acetonitrile
0.63
0.31
1.94
0.10
0.08
0.07
0.47
0.15
0.29
dominant in the low-temperature emission spectrum of
3
2 2
CH Cl
3
Gd(DBM) Á 2H O (Fig. 5).
However, the Fisc of Eu
2
O is far from unity because the intersystem
Cyclohexane
3
2
(
CDBM)
3
Á 2H
crossing efficiency for the ILCT excited state is very low. This
can be clearly seen from the low-temperature emission spec-
conclusion, we further supposed that the ILCT fluorescence
which was caused by incorporation of carbazole group into
trum of Gd(CDBM)
3
Á 2H
2
O, in which the ILCT fluorescence
3
+
DBM impaired the sensitization efficiency of Eu
Eu(CDBM) O system. The overall quantum yields of
Á 2H
Eu in the three complexes are listed in Table 3. We can see
in the
dominates (Fig. 5). Thus, we concluded that the incorporation
3
2
of the carbazole group reduced the Fisc of CDBM and further
3
+
3+
impaired the whole sensitization process of Eu
in the
3
+
from Table 3 that the overall quantum efficiency of Eu are
Eu(CDBM) Á 2H O system. Although the phosphorescence
3
2
in the order Eu(DBM) Á 2H O 4 Eu(CCDBM) Á 2H O 4
3
2
3
2
spectrum of Gd(CCDBM) Á 2H O is very close to that of
3
2
Eu(CDBM)
3
2
Á 2H O in all solvents.
3+
Gd(DBM) Á 2H O, the overall quantum efficiency of Eu
3
2
Now it is clear that the introducing of carbazole functional
3
in Eu(CCDBM) Á 2H O is still lower than that in Eu
+
group impaired the whole sensitization process of Eu in the
Eu(CDBM) O system, but we still do not know exactly
3
2
(
DBM)
3
Á 2H
2
O (Table 3). This relatively small decrease of
3
Á 2H
2
sensitization efficiency may be due to the weak ILCT fluores-
cence of CCDBM and the locally excited states of carbazole
moiety (around 350 nm, not shown in Fig. 5).
which steps are disturbed in the overall process. The overall
efficiency of the lanthanide emission depends on three factors:
F = F F F
isc et Ln
(1)
Now we are able to discuss the unsolved problem in the
third paragraph of this sub-section: why does the ILCT
fluorescence of CDBM decrease after coordinating with
where Fisc is the singlet to triplet intersystem crossing effi-
ciency, Fet is the energy transfer quantum yield and FLn is the
3+
3+
Gd and Eu ? Because CDBM is a typical Donor–p–
Acceptor molecule containing a carbazole moiety, the ILCT
7
,29
intrinsic luminescence quantum yield.
3
The intrinsic lumi-
+
nescence quantum yield of Eu (FLn) can be determined by
eqn (2):
fluorescence of CDBM can be described with a ‘‘twisted
34
intramolecular charge transfer (TICT)’’ model. According
to this model, the locally excited state S
was formed immediately after excitation (Scheme 2, left side).
Then the molecule undergoes a charge transfer reaction (k ) to
1
(non-CT character)
F
Ln = tobs/t
r
(2)
3
+
in which tobs is the observed lifetime of Eu and t stands for
the radiative lifetime of Eu . The radiative lifetime t
r
a
3
+
5
r 0
of D
form a TICT state (SCT), in which the carbazole moiety is
perpendicular to the benzene ring. The two states can establish
state can be calculate according to the procedures in litera-
3
0,31
3+
ture.
The calculated t and FLn of Eu in various solvents
r
equilibrium and decay to the ground state independently
35,36
are listed in Table 4. It can be seen from Table 4 that the
differences between the intrinsic quantum yields are relatively
small. This can not be responsible for the big discrepancy in
the overall quantum yields. According to the Dexter mechan-
ism, the energy transfer quantum yield Fet is determined by
two factors: one is the distance between the ligand and the
lanthanide ion; the other one is the spectra overlap of the
ligand phosphorescence spectrum and the absorption spec-
trum of the lanthanide ion. As discussed above, the geometry
(
Scheme 2, left side).
However, after coordinating with
the lanthanide ions, the intersystem crossing rate (k ) of S
isc
1
has been greatly enhanced due to the heavy atom effect. As a
result, k not only has to compete with k , but also has to
1
a
compete with kisc (Scheme 2, right side). That is to say, the
heavy atom effect of the lanthanide ions indirectly reduced the
quantum yield of the ILCT fluorescence.
It is instructive to compare our results with that of Kim
et al., who have observed a different phenomenon in their
of the central Eu(DBM)
3
Á 2H
2
O part in the three complexes
11
3+
experiments. The authors concluded that the Eu ion could
be sensitized by the CT excited states of the 4-naphthalen-
remains almost the same. Therefore, the distances between the
ligands and the lanthanide ions can be regarded as the same
because the outer carbazole moieties are not involved in the
sensitization process. On the other hand, the phosphorescence
1
-ylbenzoic acid (NAC-1) and 4-[4-(4-methoxyphenyl)-
naphthalen-1-yl]benzoic acid (NAC-2) ligands. The discre-
pancy between their observation and ours might be due to
the following facts: (i) the charge transfer character of NAC-1
and NAC-2 is not as strong as CDBM. When the solvent was
r
Table 4 The observed lifetime tobs (in ms), radiative lifetime t (in ms)
and the intrinsic quantum yields of Eu in various solvents
3
+
DBM–Eu CDBM–Eu CCDBM–Eu
20 a
D
Solvent
n
t
obs
t
r
F
Ln
t
obs
t
r
F
Ln
t
obs
t
r
F
Ln
Acetonitrile 1.3441 0.20 1.50 0.13 0.17 1.23 0.14 0.16 1.52 0.11
CH Cl 1.4242 0.21 2.10 0.10 0.16 1.41 0.11 0.20 1.48 0.14
Cyclohexane 1.4262 0.24 2.12 0.11 0.16 1.65 0.10 0.20 2.09 0.10
2
2
a
Refractive index at 20 1C, taken from ref. 32.
Scheme 2
1
644 | New J. Chem., 2007, 31, 1639–1646
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007

View Article Online
state. Because the ILCT excited states are not involved in the
sensitization process both in the solid state and solutions, it is
3
+
expected that the quantum efficiency of the three Eu com-
plexes in the solid state should present the same trend as that
in solutions. We measured the quantum efficiencies of the
3
+
three Eu
are 0.30, 0.09 and 0.18% for Eu(DBM)
CDBM) and Eu(CCDBM) Á 2H O, respectively.
Á 2H
These data agree very well with those obtained in solutions
especially that in CH Cl ). This result strongly suggests that
the sensitization process of the three Eu
complexes in the solid state and the results
3
Á 2H O, Eu-
2
(
3
2
O
3
2
(
2
2
3
+
complexes in the
solid state should be similar to that in solution.
Conclusions
Photophysical properties of the b-diketonate–europium com-
plex Eu(CDBM) Á 2H O have been thoroughly studied in
3
2
Fig. 6 Schematic representation of the energy transfer pathways in
7
À1
2 T
O system: kflu E10 s , k E600 s , ket 4
3
À1
several solvents. It is revealed that the ILCT fluorescence of
the ligand CDBM is not involved in the sensitization process
the Eu(CDBM)
7
3
Á 2H
À1
À1
1
0 s , kobs = 6.3 Â 10 s . See ESIw for the calculation details of the
3
+
3+
rate constants.
of Eu
through the triplet excited state which localizes in the DBM
part. In another europium complex Eu(CCDBM) O, in
in this complex. The Eu
ion is still sensitized
3
Á 2H
2
changed from cyclohexane to acetonitrile, NAC-1 and NAC-2
showed a red shift of 30 and 40 nm, respectively, whereas
CDBM showed a red shift of 156 nm. (ii) NAC-1 and NAC-2
are benzoic acids and CDBM is a b-diketone. These two kinds
of ligands are very different in acidity, formation constants
and so on. We also want to emphasize that our conclusion is
suitable for the b-diketonate system which has strong CT
character.
which the conjugation system between the carbazole moiety
and the DBM part has been interrupted, decreased ILCT
3
+
fluorescence and increased Eu
emissions have been ob-
served. This supported our conclusion in another way. The
3
+
quantum efficiencies of the three Eu complexes in the solid
state have been observed to show the same trend as that in
solutions. This proved that our conclusion is still valid in the
solid state. All these observations give an important implica-
tion for the design of organic–lanthanide light-emitting mate-
rials. That is, one should avoid ILCT fluorescence when one
applies chemical modifications to the ligand to get better
performances of the materials.
To sum up, the ILCT excited states in Eu(CDBM)
3
Á 2H
2
O
can neither be transformed into triplet excited states nor
3
sensitize Eu ion directly. However, the energy of the ILCT
+
3
states can be partially transferred to the Eu indirectly. The
+
3
+
Eu
which localizes in the DBM part. The energy transfer pro-
ion is still sensitized through the triplet excited state
cesses in Eu(CDBM)
Fig. 6.
3
2
Á 2H O are schematically represented in
Acknowledgements
We thank the National Basic Research Program
(
2006CB601103), and the NNSFC (20021101, 20423005,
0471004, 90401028, 50372002, 20671006) for financial
Solid state
2
How the chemical modifications of the ligands affect the
sensitization process of lanthanide complexes in the solid state
is an essential issue in many practical applications (e.g.
OLEDs). In this regard, we measured the emission spectra
support.
References
3
+
and the quantum yields of the three Eu
solid state.
complexes in the
¨
1 J.-C. G. Bunzli and C. Piguet, Chem. Soc. Rev., 2005, 34, 1048.
F. J. Steemers, W. Verboom, D. N. Reinhoudt, E. B. van der Tol
2
and J. W. Verhoeven, J. Am. Chem. Soc., 1995, 117, 9408.
L. H. Slooff, A. Polman, S. I. Klink, G. A. Hebbink, L. Grave, F.
C. J. M. van Veggel, D. N. Reinhoudt and J. W. Hofstraat, Opt.
Mater., 2000, 14, 101.
A distinct difference of the emission spectra between the
solid state (Fig. S2 in ESIw) and solution (Fig. 3) is that there is
no observable ligand emission in the solid state emission
spectra. It is possible that the ILCT emission energy may have
3
4
5
H. Xin, F. Y. Li, M. Shi, Z. Q. Bian and C. H. Huang, J. Am.
Chem. Soc., 2003, 125, 7166.
H. Xin, M. Shi, X. C. Gao, Y. Y. Huang, Z. L. Gong, D. B. Nie,
H. Cao, Z. Q. Bian, F. Y. Li and C. H. Huang, J. Phys. Chem. B,
3
+
been transferred to the Eu ion or that the ILCT fluorescence
may be quenched due to the intermolecular interactions (p–p
stacking, hydrogen bonds, etc.) in the solid state. If it is the
energy transfer that lead to the disappearance of the strong
2
004, 108, 10796.
6
S. I. Klink, G. A. Hebbink, L. Grave, P. G. B. Oude Alink, F. C. J.
M. van Veggel and M. H. V. Werts, J. Phys. Chem. A, 2002, 106,
3681.
ILCT emission bands in Eu(CDBM)
able to observe ILCT fluorescence for Gd(CDBM)
However, no fluorescence can be detected for the three Gd
3
Á 2H
2
O, we should be
Á 2H O.
2
7 S. I. Klink, L. Grave, D. N. Reinhoudt, F. C. J. M. van Veggel, M.
H. V. Werts, F. A. J. Geurts and J. W. Hofstraat, J. Phys. Chem.
A, 2000, 104, 5457.
3
3
+
complexes in the solid state. This leads to the conclusion that
the ILCT fluorescence must have been quenched in the solid
8
K. Binnemans, Handbook on The Physics and Chemistry of Rare
Earths, Elsevier, Amsterdam, 2005, vol. 35, ch. 225.
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1639–1646 | 1645

View Article Online
9
G. A. Hebbink, S. I. Klink, L. Grave, P. G. B. Oude Alink and F.
C. J. M. van Veggel, ChemPhysChem, 2002, 3, 1014.
24 F. R. Goncalves e Silva, O. L. Malta, C. Reinhard, H.-U. Gu
C. Piguet, J. E. Moser and J.-C. G. Bunzli, J. Phys. Chem. A, 2002,
106, 1670.
25 D. L. Dexter, J. Chem. Phys., 1953, 21, 836.
26 G. F. de Sa, O. L. Malta, C. de Mello Donega
¨
del,
¨
1
0 C. Yang, L. M. Fu, Y. Wang, J. P. Zhang, W. T. Wong, X. C. Ai,
Y. F. Qiao, B. S. Zou and L. L. Gui, Angew. Chem., Int. Ed., 2004,
3, 5010.
1 H. K. Kim, N. S. Baek and H. K. Kim, ChemPhysChem, 2006, 7,
13.
4
´
´
, A. M. Simas, R. L.
1
1
1
Longo, P. A. Santa-Cruz and E. F. da Silva, Jr, Coord. Chem. Rev.,
2000, 196, 165.
2
2 M. R. Robinson, M. B. O’Regan and G. C. Bazan, Chem.
Commun., 2000, 1645.
3 D. B. Nie, Z. Q. Bian, A. C. Yu, Z. Q. Chen, Z. W. Liu and C. H.
Huang, J. Phys. Chem. A, submitted.
27 G. A. Hebbink, L. Grave, L. A. Woldering, D. N. Reinhoudt and
F. C. J. M. van Veggel, J. Phys. Chem. A, 2003, 107, 2483.
28 G. A. Crosby, R. E. Whan and R. M. Alire, J. Chem. Phys., 1961,
34, 743.
1
1
1
4 W. H. Melhuish, J. Phys. Chem., 1961, 65, 299.
5 G. A. Crosby and J. N. Demas, J. Phys. Chem., 1971, 75, 991.
6 N. C. Greenham, I. D. W. Samuel, G. R. Hayes, R. T. Phillips, Y.
A. R. R. Kessener, S. C. Moratti, A. B. Holmes and R. H. Friend,
Chem. Phys. Lett., 1995, 241, 89.
7 L. S. Rohwer and J. E. Martin, J. Lumin., 2005, 115, 77.
8 O. F. Ricardo, MOPAC_6RM1, 2006.
9 R. O. Freire, G. B. Rocha and A. M. Simas, Inorg. Chem., 2005,
29 L. M. Ying, A. C. Yu, X. S. Zhao, Q. Li, D. J. Zhou, C. H.
Huang, S. Umetani and M. Matasai, J. Phys. Chem., 1996,
100, 18387.
30 A. Beeby, L. M. Bushby, D. Maffeo and J. A. G. Williams,
J. Chem. Soc., Dalton Trans., 2002, 48.
31 C. Gorrler-Walrand and K. Binnemans, Handbook on the Physics
¨
and Chemistry of Rare Earths, Elsevier, Amsterdam, 1998, vol. 25.
32 C. Reichardt, Solvent and Solvent Effects in Organic Chemistry,
Wiley-VCH, Weinheim, 3rd edn, 2003.
1
1
1
44, 3299.
2
0 M. d. C. Fernandez-Fernandez, R. Bastida, A. Macias, P. Perez-
Lourido, C. Platas-Iglesias and L. Valencia, Inorg. Chem., 2006,
33 We assumed that the ISC efficiency is similar at low and room
3+
temperatures because the Eu efficiencies are of the same order at
77 K as those detected at room temperature.
45, 4484.
2
2
2
1 R. O. Freire, G. B. Rocha, R. Q. Albuquerque and M. S. Alfredo,
J. Lumin., 2005, 111, 81.
2 M. O. Ahmed, J.-L. Liao, X. Chen, S.-A. Chen and J. H. Kaldis,
Acta Crystallogr., Sect. E, 2003, 59, m29.
3 M. H. V. Werts, R. T. F. Jukes and J. W. Verhoeven, Phys. Chem.
Chem. Phys., 2002, 4, 1542.
34 Z. R. Grabowski, K. Rotkiewicz and W. Rettig, Chem. Rev., 2003,
103, 3899.
35 T. Yoshihara, S. I. Druzhinin, A. Demeter, N. Kocher, D. Stalke
and K. A. Zachariasse, J. Phys. Chem. A, 2005, 109, 1497.
36 K. Dahl, R. Biswas, N. Ito and M. Maroncelli, J. Phys. Chem. B,
2005, 109, 1563.
1
646 | New J. Chem., 2007, 31, 1639–1646
This journal is ꢀc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007