New 3π-2spiro ladder-type phenylene materials: Synthesis, physicochemical properties and applications in OLEDs

Article abstract of DOI:10.1002/chem.200801428

New routes to ladder-type phenylene materials 1 and 2 are described. The oligomers 1 and 2, which possess a "3π-2spiro" architecture, have been synthesized by using extended diketone derivatives 3 and 10 as key intermediates. The physicochemical properties of the new blue-light emitter 2 were studied in detail and compared with those of the less-extended 1. Owing to the recent development of fluorenone derivatives and their corresponding more conjugated analogues as potential electron-transport materials in organic light-emitting diodes (OLEDs) and as n-type materials for photovoltaic applications, we also report herein the thermal, optical and electrochemical behavior of the key intermediates, diketones 3 and 10. Finally, the application of dispiro 2 as a new light-emitting material in OLEDs is reported.

Full text of DOI:10.1002/chem.200801428

DOI: 10.1002/chem.200801428
New 3p-2Spiro Ladder-Type Phenylene Materials:
Synthesis, Physicochemical Properties and Applications in OLEDs
Nicolas Cocherel,[a] Cyril Poriel,*[a] Joꢀlle Rault-Berthelot,*[a] Frꢁdꢁric Barriꢂre,[a]
Nathalie Audebrand,[a] Alexandra M. Z. Slawin,[b] and Laurence Vignau[c]
Abstract: New routes to ladder-type
phenylene materials 1 and 2 are de-
scribed. The oligomers 1 and 2, which
Owing to the recent development of
fluorenone derivatives and their corre-
sponding more conjugated analogues
as potential electron-transport materi-
als in organic light-emitting diodes
(OLEDs) and as n-type materials for
photovoltaic applications, we also
report herein the thermal, optical and
electrochemical behavior of the key in-
termediates, diketones 3 and 10. Final-
ly, the application of dispiro 2 as a new
light-emitting material in OLEDs is re-
ported.
possess
a
“3p-2spiro” architecture,
have been synthesized by using extend-
ed diketone derivatives 3 and 10 as key
intermediates. The physicochemical
properties of the new blue-light emitter
2 were studied in detail and compared
with those of the less-extended 1.
Keywords: ladder pentaphenylenes ·
luminescence · organic light-emit-
ting diodes · semiconductors · spiro
compounds
Introduction
430 nm for polyindenofluorenes and 450 nm for ladder-type
poly-p-phenylene. However, several groups have shown
that, in the solid state, the emission is unstable due to the
appearance of long-wavelength emission bands, which were
assigned to excimer formation[4,5] and to ketone defects.[6,7]
Among the numerous solutions that have been proposed to
suppress this long-wavelength emission, the introduction of
a rigid spiro linkage, such as spirobifluorene, into the poly-
mer/oligomer backbone has been extensively developed.
This spiro concept in electronics,[8,9] based on the connec-
tion, through an sp3-hybridized carbon atom, of two molecu-
lar p systems (the 2p-1spiro concept, Figure 1) with equal or
different functions (for example, emission, charge transport)
has been widely studied by Tour[10–13] and Salbeck[14–19] and
their co-workers. The use of such a linkage has been demon-
strated to strongly enhance the OLED properties, that is,
suppress excimer formation and ketonic defects, increase
the thermal and morphological stability, enhance the quan-
tum yield, improve the solubility and so forth.
Phenylene-based polymers are one of the most important
classes of blue-emitting materials used in organic light-emit-
ting diodes (OLEDs).[1] Three classes of polyphenylene that
have been of particular interest are polyfluorenes, polyinde-
nofluorenes and ladder-type poly-p-phenylene.[2,3] These
polymers show blue to blue-green emission in solution. As
the chain rigidity increases with an increasing number of
fused planar rings, a bathochromic shift of the emission
wavelength is observed, that is, 410 nm for polyfluorenes,
[a] N. Cocherel, Dr. C. Poriel, Dr. J. Rault-Berthelot, Dr. F. Barriꢁre,
Dr. N. Audebrand
Universitꢂ de Rennes 1
UMR CNRS 6226 “Sciences Chimiques de Rennes”
Bat 10C, Campus de Beaulieu, 35042 Rennes Cedex (France)
Fax : (+33)223-236732
E-mail: cyril.poriel@univ-rennes1.fr
For the last 20 years Mꢀllen and Scherf and co-workers
have designed many ladder-type oligomers and polymers for
various optoelectronic applications, which has led to the sig-
nificant enhancement in the performance of organic devi-
ces.[2,3,20–25] However, to the best of our knowledge, the syn-
thesis of such oligomers incorporating spiro linkages has
only been reported in a few papers.[26,27] In this context, we
recently designed new blue luminescent dispiro building
blocks, such as 1, called dispirofluorene-indenofluorene
joelle.rault-berthelot@univ-rennes1.fr
[b] Prof. Dr. A. M. Z. Slawin
School of Chemistry, University of St Andrews
North Haugh, St Andrews
Fife, KY16 9ST (UK)
[c] Dr. L. Vignau
Universitꢂ de Bordeaux, IMS/UMR CNRS 5218-site
ENSCPB, 16 Avenue Pey-Berland, 33607 Pessac Cedex (France)
Supporting Information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801428.
11328
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

FULL PAPER
Results and Discussion
Synthesis: 9,9’-Spirobifluorene
can be considered to be the
fusion of two fluorene units
through a shared spiro carbon
and is readily obtained in a
two-step synthesis from 2-halo-
genobiphenyl and fluorenone
through
a metal–halogen ex-
change reaction.[10,37–40] As
1
can be considered to be the
fusion of two 9,9’-spirobifluor-
ene moieties through a shared
phenyl ring, our previous retro-
Figure 1. General molecular structure of spiro-linked molecules (from Salbeck and co-workers[8]): the 2p-
1spiro concept (left). Molecular design adopted in this work: the 3p-2spiro concept (right).
synthetic analysis of
based on a double lithium–hal-
ogen exchange reaction
(Scheme 1, route a) and in-
1 was
(DSF-IF), for use in OLED applications.[28–30] The “3p-
2spiro” architecture of these molecules (Figure 1) preserves
the p conjugation of the indenofluorene core, that is, p-
system-1, which plays a prominent role in their electronic
and electrochemical properties. The introduction of the
spiro-linked fluorene rings, that is, p-system-2 units, gives
the molecules a highly rigid structure. Contrary to 2p-1spiro
compounds,[8] 3p-2spiro compounds have only been reported
on a few occasions.[27,31,32] For example, Bo and co-workers
recently reported monodisperse ladder-type oligomers with
two, three, four and seven spiro bridges using an elegant ap-
proach involving the oxidation of the fluorene C-9 carbon
atom as a key step.[26]
On the other hand, aromatic diketones such as fluorenone
derivatives and their corresponding more conjugated ana-
logues are good electron-accepting materials and have re-
cently received attention as electron-transport materials
(ETMs) in OLEDs and as n-type materials for photovoltaic
applications.[21,33–35] For example, new indenofluorenone-
and bisindenofluorenone-based molecular building blocks
have recently been synthesized and appear to be a promis-
ing family for n-channel semiconducting polymers.[36]
volved 2,2’’-diiodo-1,1’;4’,1’’-terphenyl (2,2’’-DITP) as the
key starting material.[29] Despite the efficiency of this route,
it was, however, evident that a similar synthetic approach
might not be used to prepare 2 because the preparation of
the key diiodinated intermediate appeared to be highly
complicated (Scheme 1, route c). Moreover, we have recent-
ly shown that, when the fluorenone core is substituted in the
2- and 7-positions, route a leads to positional isomers of
DSF-IFs[30] and is therefore inadequate for the synthesis of
2. Stimulated by the work of Mꢀllen and Scherf and their
co-workers, we thus decided to investigate a new route to
3p-2spiro molecules such as 1 (Scheme 1, route b), which
may also be used for the synthesis of the more extended oli-
gomer 2 (Scheme 1, route d). These two approaches were
both based on a key coupling reaction between a diketone
derivative, that is, 3 or 10, and 2-halogenobiphenyl. A simi-
lar approach has recently been developed by Vak et al. for
the synthesis of a polyindenofluorene bridged by two spiro
anthracene fragments.[27]
Starting from the commercially available 2,5-dibromo-1,4-
dimethylbenzene, the diketone 3 was prepared on the multi-
gram scale in a three-step synthesis as previously report-
ed.[29,41] As shown in Scheme 2, the lithium–iodine exchange
of 2-iodobiphenyl[42] 4 with n-butyllithium at a low tempera-
ture followed by quenching with the diketone 3 afforded the
dialcohol 5, which was not purified at this stage and was im-
mediately used in an intramolecular cyclisation reaction
under acidic conditions, that is, AcOH/HCl, and led to com-
pound 1 in a yield of 26% (not optimiszed). This new route,
performed in an overall yield of 18%,[43] is short, efficient
and highly adaptable to the preparation of various DSF-IF
derivatives substituted either on the p-system-1 (substitution
of the diketone 3) or on the p-system-2 (substitution of the
iodobiphenyl 4).
In this context, we have proposed new routes to 3p-2spiro
molecules using extended diketone derivatives as key inter-
mediates. We have thus investigated a short and efficient
new route to 1 starting from indenoACHTUNTRGNEUNG[1,2-b]fluorene-6,12-
dione (3) (Scheme 1, route b). This approach was extended
to the synthesis of the dispirofluorene ladder pentapheny-
lene 2 starting from the corresponding extended diketone 10
(Scheme 1, route d). Herein, we first report our synthetic in-
vestigations towards these new oligomers as potential emit-
ters for OLED applications. The thermal, structural, electro-
chemical, theoretical and optical properties of these oligo-
mers and their corresponding diketone intermediates are
then discussed. Finally, we present the behavior of oligomer
2 as an emissive layer within an OLED.
In the light of these results, we decided to follow the
same strategy to prepare the ladder-type pentaphenylene 2
(Scheme 3). In this context, the diketone 10 was the key in-
termediate. Suzuki coupling of the fluorene-2-boronate
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11329

C. Poriel, J. Rault-Berthelot et al.
Scheme 1. Retrosynthetic approaches to dispiro 1 and 2 (X=Br, I; R=C8H17).
7[44–46] and the dibromoterephthalate 6[47] generated the key
intermediate 8 (see the Supporting Information for full de-
tails of the syntheses), previously reported with a methyl
ester group by Mꢀllen and co-workers.[21] In the solid state
(see the molecular structure of 8 from single-crystal X-ray
diffraction data in the Supporting Information), the two flu-
orene rings of 8 lie on each side of the central phenyl ring
with angles of around 50.88. The shortest distance between
the carbon atom of the carbonyl group and the correspond-
ing adjacent carbon atom in the fluorene ring is 3.1 ꢄ. Un-
fortunately, acidic treatment of 8, as previously described
for similar structures, that is, in concentrated H2SO4 at
1658C,[21,36] in AcOH/HBr at 1108C[41] or in CH3SO3H at
658C,[48] failed to produce 10. Saponification of the diester 8
under basic conditions (NaOH/EtOH at reflux) readily pro-
vided the corresponding terephthalic acid 9, which was con-
verted into the corresponding acid dichloride (oxalyl chlo-
ride, DMF, RT). Finally, Lewis acid promoted intramolecu-
lar Friedel–Crafts acylation
(TiCl4, 08C)[49] of the acid di-
chloride readily took place to
give the diketone 10 in a yield
of 91% over the two steps.
With this key building block in
hand, the last step required its
coupling to the 2-metallobi-
phenyl. Note that this reaction
appears to be difficult to per-
form and is highly sensitive to
the reaction conditions, includ-
ing magnesium versus lithium,
Scheme 2. Synthesis of DSF-IF 1.
iodobiphenyl versus bromobi-
11330
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

3p-2Spiro Ladder-Type Phenylene Materials
FULL PAPER
Scheme 3. Syntheses of 10, 13 and 2 (R=C8H17, Ph2Br=2-bromobiphenyl).
phenyl, THF versus diethyl ether, the nature of the base and
the time of formation.
After several attempts, the dialcohol 11 was finally ob-
tained in a yield of 31%[50] and cyclized under mild condi-
tions, that is, BF3·OEt2 in dichloromethane, to provide the
expected 3p-2spiro pentaphenylene 2.[51]
of the biphenyl moieties due to a sterically hindered envi-
ronment. The monoketone 13 appeared to be an interesting
model compound to study the electronic properties of 3p-
2spiro derivatives. It is also a key intermediate for the syn-
thesis of mixed 3p-2spiro molecules with two different p-
system-2 moieties. Indeed, with an accurate molecular
design, that is, electron-donating and/or -withdrawing sub-
stituents, these mixed structures may present highly interest-
ing charge balance properties. Note that compounds 10, 13
and 2 exhibit good solubility in common organic solvents,
such as CHCl3, CH2Cl2, THF and DMF.
This coupling reaction can also be stopped at the monoal-
cohol 12 (Scheme 1), which can then be cyclized to give the
monospiro monoketone derivative 13. An extra biphenyl
linkage was introduced by using the same conditions (see
above) and led to the final dispiro 2 via the corresponding
alcohol 14 (not isolated). The structure of 2 was confirmed
by HMBC spectroscopy. Indeed, we recently showed that
spiro carbon atoms are powerful probes for detecting neigh-
boring hydrogen atoms through long-range shift correla-
tions.[30] As expected, three cross-peaks were detected for
the spiro carbon atoms, which correspond to three bond cor-
relations (3J) with the three neighboring hydrogen atoms (in
CD2Cl2: d=7.20 (s), 6.98 (s), 6.88 ppm (d); see the figure in
the Supporting Information). Note that the hydrogen atoms
of the biphenyl linkages in alcohols 11 and 12 appear as
broad signals in the 1H NMR spectra (see the Supporting In-
formation). This feature was attributed to the slow rotation
Physicochemical properties of DSF-IF derivatives
Structural properties: Single crystals of 2 were obtained from
a solution in CDCl3 and analyzed by X-ray diffraction to
elucidate its molecular structure and packing characteristics.
The structure of 2 is presented in Figure 2. To the best of
our knowledge, the characterisation of this kind of long
ladder-type p-phenylene derivative by X-ray crystallography
has only been reported once by Wong et al. in 2006.[53] Com-
pound 2 presents a linear antarafacial geometry[32] and its p-
system-1 (19.4 ꢄ long) does not adopt a strict coplanar con-
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11331

C. Poriel, J. Rault-Berthelot et al.
Figure 3. Crystal-packing diagram of 2 along the b stacking axis. The
alkyl chains and the hydrogen atoms have been omitted for clarity.
Figure 2. Views of the molecular structure of 2 from single-crystal X-ray
diffraction data (hydrogen atoms have been omitted for clarity).
Table 1. Electrochemical data for 2, 3, 10 and 13 (vs. SCE).
onset
ox
onset
red
E1ox E2ox
E
E1red
[V]
E2red
[V]
E
HOMO LUMO DEEl
[V] [V] [V]
[V]
[eV][a]
[eV][b]
[eV][c]
figuration as two distortions exist on both sides. The dihe-
dral angles between the plane of the central phenyl ring 3
and those of side rings 1 and 5 are 8.48 (see the labelling of
the phenyl rings in Figure 2). These distortions are larger
than those reported by Wong et al.,[53] that is, 2.58, for an
analogous tetra p-tolyl-substituted ladder-type oligo-penta-
phenylene. The angle between the plane of the central
phenyl ring 3 of p-system-1 and that of p-system-2 (through
the central cyclopentane ring) is 86.28 (Figure 2).
As indicated in the crystal packing diagram of 2 shown in
Figure 3, in which the alkyl chains have been omitted for
clarity, the two p-system-2 moieties lie on either side of the
p-system-1 plane. The closest C–C distance between two ter-
minal phenyl rings of two neighboring molecules of the p-
system-1 moieties was estimated to be 3.3 ꢄ (Figure 3, plain
circle). The closest C–C distance between two p-system-2
moieties of two neighboring molecules was also determined
to be 3.3 ꢄ (Figure 3, dashed line). Despite these two short
distances, these packing diagrams indicate the lack of effec-
tive interchromophore p–p interactions in the solid state.
2
3
1.09 1.66 0.96
–
–
ꢀ2.3 ꢀ5.36
ꢀ2.1
3.26
2.5
2.19
2.21
1.70
–
1.65 ꢀ0.98 ꢀ1.38 ꢀ0.85 ꢀ6.05
ꢀ3.55
ꢀ3.48
ꢀ3.25
10 1.40 1.68 1.27 ꢀ1.07 ꢀ1.52 ꢀ0.92 ꢀ5.67
13 1.23 1.56 1.06 ꢀ1.35 ꢀ1.15 ꢀ5.46
–
onset
[a] Calculated from the onset oxidation potential E . [b] Calculated
from the onset reduction potential
jHOMOꢀLUMOj from the redox data.
ox
onset
E
.
[c] Calculated as DEEl
=
red
first oxidation E1ox is reversible (Figure 4c) and occurs in a
potential range similar to that observed for analogous copla-
nar penta-p-phenylene oligomers reported by Wong,[53]
Bo[26] and Mꢀllen[25] and their co-workers. This process has
been assigned to the formation of a radical cation delocal-
ized throughout the p-system-1. The potential difference be-
tween the first oxidation potential E1ox of 1[29] (1.47 V) and 2
(1.09 V) is around 0.38 V, which clearly shows the extension
of conjugation of the p-system-1 in 2. The second oxidation
wave (Figure 4b) is multi-electronic and the intensity ratio
of E1ox/E2ox, measured in the first scan, is around 1:4. This
process was assigned to the second oxidation of the p-
system-1, which leads to a dication, as has also been ob-
served for other coplanar penta-p-phenylene deriva-
tives,[25,26,53] concomitant with the oxidation of the two spiro-
linked p-system-2 moieties. Indeed, this second wave Eo2x
(1.66 V) corresponds well with the primary oxidation waves
of fluorene and 9,9’-spirobifluorene centered at 1.62 and
1.69 V, respectively.[29] This behavior led us to conclude that
the radical cation is highly stable and strongly delocalized
throughout the p-system-1 and that there is no or very little
charge transfer between p-system-1 and the two p-system-2
Electrochemical properties: The redox properties of the dis-
piro derivative 2 were investigated by cyclic voltammetry
(CV). Diketones 3 and 10 and monoketone 13, key inter-
mediates in our synthetic approaches, were also electro-
chemically studied because they may be interesting as elec-
tron-accepting materials.[21,36,54] Electrochemical data are
summarized in Table 1.
The oxidation of 2 recorded in CH2Cl2 between 0.3 and
2.25 V (Figure 4a) occurs in three steps, the maxima of
which are at Eo1x (1.09 V), Eo2x (1.66 V) and Eo3x (1.95 V). The
11332
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

3p-2Spiro Ladder-Type Phenylene Materials
FULL PAPER
Figure 4. CV traces of 2 (2ꢅ10ꢀ3 m) in CH2Cl2 (0.2m Bu4NPF6). Working
electrode: platinum disk, diameter 1 mm; sweep rate: 100 mVsꢀ1. The
scanning potential window was 0.3–2.25 V (two cycles) in (a), 0.3–1.52 V
in (b) and 0.3–1.2 V in (c) (one cycle). The current scale S is equal to
4 mA in (a) and to 0.8 mA in (b) and (c).
Figure 5. CV traces of 3 (1.13ꢅ10ꢀ2 m), 10 (1.1ꢅ10ꢀ3 m) and 13 (10ꢀ3 m) in
CH2Cl2 (0.2m Bu4NPF6). Working electrode: platinum disk, diameter
1 mm; sweep rate 100 mVsꢀ1. The scanning potential window was 0.0–
1.8 V and back to ꢀ1.75 V for 3, ꢀ1.8 V for 10 and ꢀ1.95 V for 13. The
current scale S is equal to 0.4 mA for 3 and 0.8 mA for 10 and 13.
moieties. For 1, the second oxidation maximum E2ox is shifted
to 1.95 V due to charge transfer from the p-system-1 radical
cation to the neutral p-system-2. Consequently the p-
system-2 is more difficult to oxidise. Recurrent cycles in a
potential range that includes the two E1ox and Eo2x waves of 2
do not lead to any electrodeposition, contrary to what was
observed for 1. The absence of electropolymerisation of 2
might be explained by the high stability of the radical cation
delocalized throughout the p-system-1 and by the sterically
hindered environment of the p-system-2 moieties (see the
crystal structure determined from the X-ray diffraction data
in Figure 3). The reduction of 2, close to the dichlorome-
thane electrolyte discharge, was detected as a threshold at
ꢀ2.3 V.
teresting electron-accepting properties. The value of E1red for
10 (ꢀ1.07 V) is consistent with the first reduction potential
of the analogous diketone, slightly more extended, described
by Mꢀllen and co-workers (ꢀ1 V).[21] Relative to the less ex-
tended diketone 3 (E1red =ꢀ0.98 V), the value of E1red for 10
is shifted to more negative values. This shift may be attribut-
ed to a less efficient electron-withdrawing effect of the two
carbonyl groups on a more extended aromatic backbone. As
expected, the presence of only one carbonyl group in com-
pound 13 leads to a negative shift of the first reduction po-
tential compared with that of 10.
The diketone 3[35] presents an ill-defined wave with a max-
imum E1ox centered at 1.70 V. The diketone 10 and the mono-
ketone 13 show two reversible oxidation waves with maxima
at 1.40/1.23 V (E1ox) and 1.68/1.56 V (Eo2x), respectively
(Figure 5). Compared with 2, the presence of one (for 13)
and two (for 10) electron-withdrawing carbonyl groups
causes their first oxidation potentials to be shifted to more
anodic values by around 0.17 and 0.31 V, respectively. The
oxidation of 10 and 13 leads to the formation of radical cat-
ions and dications that are stable on the CV timescale, as
shown by the successive reversible mono-electronic oxida-
tion waves. Moreover, a comparison of E1ox for 3 (1.70 V)
and 10 (1.40 V) shows the extended p delocalisation of the
phenylene ladder backbone from three to five phenyl units
(in 3 and 10, respectively). Note that under the same experi-
mental conditions, 9-fluorenone presents a first reversible
oxidation wave with a maximum centered at 2 V (vs. SCE),
see the Supporting Information for details. In the cathodic
range, the diketones 10 and 3 present two reversible reduc-
tion waves leading to stable radical anions, which reveals in-
Theoretical calculations[55] are consistent with the electro-
chemical measurements, pointing to p-system-1-based pri-
mary oxidation and reduction in 2, as shown by the nature
of the frontier molecular orbitals (Figure 6).[56] Indeed, the
HOMO and LUMO show p-system-1 character. This was
also observed in the case of 1.[29] The two pairs of quasi-de-
generate LUMO+1 and LUMO+2 show p-system-2 charac-
ter. The HOMOꢀ1 has mixed character, whereas the
HOMOꢀ2 has mainly p-system-2 character. The electron af-
finity (LUMO energy) and ionisation potential (HOMO
energy) have been estimated from the redox data
(Table 1).[57,58] The estimated HOMO and LUMO energies
for 2 are ꢀ5.36 and ꢀ2.1 eV, respectively, determined by
using an SCE energy level of 4.4 V relative to vacuum. Rel-
ative to 1 (HOMO: ꢀ5.76 eV; LUMO: ꢀ2.17 eV, DEEl =
3.59 eV),[29] the LUMO of 2 has the same energy. However,
the HOMO of 2 is more strongly affected due to the extend-
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11333

C. Poriel, J. Rault-Berthelot et al.
tion bands have also been de-
scribed previously in the litera-
ture for fluorenone[63] (assigned
to p–p* transitions) and analo-
gous diketones of 3 (assigned to
p–p*
and
n–p*
transi-
tions).[21,36,61]
In the case of 10, a similar
strong absorption band is ob-
served, redshifted to 326 nm,
which clearly shows the extend-
ed conjugation. This strong ab-
sorption band also presents a
shoulder at
a lower wave-
length.[21] The monoketone-
monospiro 13 presents a similar
band at 322 nm, but with a sig-
nificantly lower intensity, be-
cause 13 possesses only one car-
bonyl group. Owing to the pres-
Figure 6. Sketch of frontier molecular orbitals for a simplified model of 2 (methyl instead of octyl groups)
from Gaussian 03 B3LYP/6-31G* calculations.[56, 59,60]
ed conjugation of the p-system-1 and lies 0.4 eV higher than
the HOMO of 1. Consequently, for OLED application, the
energy level of the HOMO in 2 would be better suited to
hole injection from the ITO/poly(3,4-ethylenedioxythio-
phene) (PEDOT) anode (see Figure 12). Thus, and com-
pared with 1, dispiro 2 possesses a smaller band gap, esti-
mated at 3.26 eV.
Owing to the electron-withdrawing effects of their carbon-
yl groups, 3, 10 and 13 each possess a low-lying LUMO. As
observed by Mꢀllen and co-workers,[21] we note that the
LUMO energies of 10 and 3 fit well with metal cathodes
such as magnesium (work function: ꢀ3.7 eV), which means
that the injection of electrons should be favored in such de-
vices. This latter point is of great interest as it may prevent
the use of the highly reactive calcium cathode. To date there
are still only a few ETMs that efficiently function as hole-
Figure 7. UV/Vis spectra of 1 (orange), 2 (blue), 3 (red), 10 (black) and
13 (green) in CH2Cl2 (10ꢀ5 m). Inset: a) The 400/650 nm portion; b) UV/
Vis spectrum of ter-9,9’-dibutylfluorene in solution in CH2Cl2 (10ꢀ5 m).
blocking materials as well. For example, tris(8-
hydroxyquinolinolato)aluminium(III) (Alq3, HOMO:
AHCTUNGTRENNUNG
ꢀ5.6 eV; LUMO: ꢀ3.1 eV), a well-known ETM, does not
work efficiently as a hole blocker.[1] In this context, the dike-
tone 3 appears to be interesting because its HOMO level is
very low (ꢀ6.05 eV), which gives this molecule hole-block-
ing properties. However, the low solubility of 3 in common
organic solvents is still an important issue that must be
solved before practical application.[61]
Table 2. UV/Vis data and decomposition temperatures Td for 1–3, 10 and
13.
lAbs [nm][a]
Td [8C][b]
2
230, 253, 263, 277, 312, 323, 339, 355, 373, 394
231, 254, 300, 311, 330, 337, 345
227, 262, 279, 289, 335, 485
228, 309, 322, 365, 385, 474
228, 262, 326, 363, 381, 542
365
355
265
345
385
1[c]
3
13
10
Optical properties: The normalized solution UV/Vis absorp-
tion spectra of ladder-type oligomers 1 and 2 and their cor-
responding ketones 3, 10 and 13 are presented in Figure 7
and the absorption data in Table 2. Diketone 3 possesses a
strong absorption band with a maximum at 289 nm and a
shoulder at a lower wavelength. It also possesses weaker ab-
sorption bands at 335 and 485 nm.[35] The main absorption
band for 3 (289 nm) is blueshifted by 45 nm from the inden-
ofluorene main absorption band (334 nm).[62] Three absorp-
[a] Measured in CH2Cl2 (10ꢀ5 m). [b] Decomposition temperatures Td cor-
responding to 5% loss, determined by thermogravimetric analysis
(TGA). [c] From ref. [29].
ence of two chromophores, that is, pentaphenylene and aryl
ketone, 10 and 13 also present absorption bands typical of a
pentaphenylene ladder-type oligomer[22,26] at 363/381 nm and
11334
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

3p-2Spiro Ladder-Type Phenylene Materials
FULL PAPER
365/385 nm, respectively. Very weak absorption bands were
also observed at lower energy, around 474 nm for 13 and
542 nm for 10, which correspond to symmetry forbidden n–
p* transitions of the carbonyl function, as recently assigned
by Mꢀllen and co-workers (Figure 7, inset b).[33,54] At this
stage, it should be stressed that the interpretation of the in-
trinsic optical properties of fluorenone incorporating poly-
mer or oligomer still remains controversial.[64] The UV/Vis
absorption spectrum of 2 presents a well-defined vibronic
structure with two main absorptions at 373 and 394 nm, also
observed in 1 (330 and 345 nm)[29] but bathochromically
shifted by around 43/49 nm (optical band gap:[65] 3.07 eV for
2 vs. 3.51 eV for 1). As this absorption arises from the cen-
tral p-system-1, this clearly indicates that the degree of p
conjugation has been effectively extended. The absorption
maximum of 2 was also redshifted by around 40 nm relative
to those of terfluorene derivatives (for example, ter(9,9’-dio-
ctylfluorene)s:
352 nm;[66,67]
ter(9,9’-dibutylfluorene)s:
354 nm (Figure 7, inset b);[68] ter(9,9’-diarylfluorene)s: 353–
356 nm).[69] Moreover, and in contrast to the case of 2, it is
known that these terfluorene absorption bands are struc-
tureless due to the rotational freedom of the phenyl rings
ꢀ
around the C C bonds that link the fluorene units (Figure 7,
inset b).[66] The same behavior was also observed with tet-
raoctylindenofluorene dimers and trimers with maximum
absorptions in solution recorded at 371 and 383 nm, respec-
tively.[70] These features clearly indicate the effectiveness of
using fused architectures to increase the planarity and thus
electronic delocalisation throughout the structure.
Figure 8. Absorption (top) and photoluminescence (bottom, lexc
275 nm) of 1 in solution in cyclohexane (a) and in thin film (c).
=
The photoluminescence of 2 was studied in cyclohexane
(Figure 10). The emission maxima recorded at 397 and
420 nm are also redshifted compared with those of 1, that is,
347 and 366 nm,[29] which indicates again the extension of
the conjugation in p-system-1. These transitions have been
assigned to 0–0 and 0–1 singlet transitions.[25] As expected,
the Stokes shift is small and consistent with a highly rigid
structure. The fluorescence quantum yield of 2 in solution[41]
is very high (90% for 2 vs. 62% for 1), which indicates that
non-radiative decay pathways are nearly suppressed in such
structures. Before discussing the behavior of 2 in OLEDs
and to determine its stability against the formation of oxida-
tive keto-type defects, spin-coated films of 3p-2spiro com-
pounds 1 and 2 on quartz substrates were exposed to ther-
mal stress conditions under an ambient atmosphere.
A comparison of the thin-film and solution fluorescence
spectra of 1 reveals an 8 nm redshift between the two first
emission peaks and a broader spectrum (Figure 8). This may
be explained by the different dielectric constants of the two
different environments.[18,71] Note that the intensity of the
relative maxima in the solid state is highly dependent on the
preparation conditions (vacuum deposition vs. spin-
coated).[19,71] In our case the second and third bands (374/
386 nm) in the solid state are higher in intensity than those
in solution (355/365 nm). The UV/Vis thin-film spectrum of
1 (Figure 8) is almost identical to the solution spectrum[29]
with only a slight bathochromic shift in the maximum from
344 to 349 nm.
Figure 9. Normalized thin-film photoluminescent spectra of 1 after an-
nealing for 1 h at different temperatures in air (lexc =275 nm) and for
60 h at 2008C.
The evolution of the fluorescence spectrum of 1 under
thermal stress conditions is presented in Figure 9. Gradual
heating of the spin-coated film in air from room tempera-
ture to 2008C (1 h for each stage) did not lead to clear pho-
toluminescent spectral changes. Even after 60 h at 2008C, no
signs of ketonic defect formations or aggregates (low-energy
emission band beyond 500 nm) were detected.
The UV/Vis and thin-film fluorescence spectra of 2 pres-
ent a fine vibronic structure and are almost identical to their
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11335

C. Poriel, J. Rault-Berthelot et al.
Figure 11. Normalized thin-film photoluminescent spectra of 2 after an-
nealing for 1 h at different temperatures in air (lexc =320 nm) and for
60 h at 2008C.
methyl group at the bridges is chemically degraded, which
leads to the well-known low-energy emission. With all these
recent findings, and as 1 did not present any low-energy
emission, we assume that the dialkyl bridgeheads in 2 are re-
sponsible for the appearance of this long-wavelength emis-
sion band. It is then likely that the introduction of aryl
groups instead of alkyl groups at the bridgeheads will re-
solve this issue.
Figure 10. Absorption (top) and photoluminescence (bottom, lexc
=
320 nm) spectra of 2 in solution (cyclohexane, a) and as a thin film
(c).
Thermal properties: The thermal properties were evaluated
by thermogravimetric analysis (TGA) under a nitrogen at-
mosphere at a rate of 58Cminꢀ1 (Table 2). The decomposi-
tion temperature (Td, corresponding to 5% loss) increases
significantly in the ketone series as the chain length increas-
es (1358C for 9-fluorenone, 2658C for 3, 3858C for 10; see
the Supporting Information). Dispiro 2 also exhibits good
thermal stability with a high Td (3658C), which is highly im-
portant for improving the lifetime of OLEDs.
solution spectra, being redshifted by only around 5 and
9 nm, respectively (Figure 10). However, and as observed
for 1, compared with the solution spectrum, the thin-film
emission spectrum of 2 is broader and the intensity of the
second fluorescence band appears to have a higher intensity.
These results are consistent with those recently reported by
Mꢀllen and co-workers on a similar compound.[25] However,
in the thin-film fluorescence spectrum a broad band appears
between 500 and 600 nm. The intensity of this band increas-
es slightly under thermal stress (RT to 2008C, Figure 11) in
an ambient atmosphere. Mꢀllen and co-workers have dem-
onstrated that fully alkylated poly(indenofluorene) and
ladder-type poly(pentaphenylene) exhibit less stability to-
wards oxidative degradation than their aryl counterparts as
a result of the dialkyl bridgeheads, which are sensitive to ox-
idation.[2,22,24,72] Moreover, Bo and co-workers recently re-
ported an identical thermal stress study of an analogous
pentaphenylene derivative with four spiro-linked fluorenes
at the bridges (no dialkyl bridgeheads). By heating a thin
film up to 2008C they reported no evident photoluminescent
spectral change.[26] In addition, Bredas and co-workers have
also investigated in depth the oxidative degradation process
that occurs in ladder-type poly-p-phenylene bearing either a
hydrogen atom or a methyl group (in addition to an aryl
group) at the bridge.[73] Despite a significant decrease in the
degree of degradation, ladder-type poly-p-phenylene with a
OLEDs using 2 as the emitting layer: Various OLED struc-
tures were fabricated by using 2 as the emitting layer. At
first, basic structures using ITO as the anode, PEDOT
doped with poly(styrene sulfonate) (PSS) as the hole inject-
ing layer (HIL), 2 as the emitting layer (EML) and calcium
as the cathode were elaborated. As the LUMO of 2 is very
low (ꢀ2.1 eV), a cathode with a low work function such as
calcium (ꢀ2.9 eV) was used to enable charge injection into
the emitting layer 2 (Figure 12a).
A comparison of the basic devices made from both spin-
coated and evaporated films of 2 was carried out. Different
solvents were used for the spin-coating: xylene, chloroform,
1,2-dichlorobenzene and cyclohexanone. Cyclohexanone ap-
peared to be the best solvent for 2 when heated above 458C
and allowed uniform films to be made, whereas the use of
other solvents resulted in poor quality films. Thus, the devi-
ces made by spin-coating used cyclohexanone as the solvent.
OLEDs were made by using different layer thicknesses for
11336
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

3p-2Spiro Ladder-Type Phenylene Materials
FULL PAPER
Figure 12. Schematic representation of the energy band diagram associated with the OLED devices investigated (before contact). HBL and ETL refer to
the hole-blocking layer and the electron-transporting layer, respectively.
the emitting layer. Luminance–voltage characteristics for 60
and 120 nm thick emitting layers deposited by evaporation
and spin-coating are displayed in Figure 13. Irrespective of
(ETL) of tris(8-hydroxyquinolinolato)aluminiumACTHNGUTRENNU(G III) (Alq3)
was thermally evaporated on top of the BCP, which resulted
in devices with the following structure ITO/PEDOT/2/BCP/
Alq3/Ca (Figure 12c).
The luminous efficiencies of the ITO/PEDOT/2/Ca, ITO/
PEDOT/2/BCP/Ca and ITO/PEDOT/2/BCP/Alq3/Ca devi-
ces are compared in Figure 14. The BCP-based OLED ex-
Figure 13. Luminance–voltage characteristics of ITO/PEDOT/2/Ca devi-
ces in which 2 was deposited by spin-coating and evaporation.
the deposition process, the turn-on voltage remained un-
changed for an equivalent thickness (see the semi-logarith-
mic graph displayed in the insert of Figure 13). However,
spin-coated and evaporated films exhibited very different
behavior. Film prepared by the evaporation process led to
luminance ten times higher than that observed for films
formed by the spin-coating process. This is due to the quali-
ty of the film (see images of the surface in the Supporting
Information). Indeed, the evaporation process leads to an
almost defect-free film. In contrast, the need to use a warm
solvent to perform the spin-coating makes it difficult to pre-
vent the formation of aggregates when the solution is depos-
ited onto the cold substrate. This results in poorer quality
films and as a consequence in poorer performances.
To improve the performances of the devices, a hole-block-
ing layer (HBL) of 2,9-dimethyl-4,7-diphenyl-1,10-phenan-
throline (bathocuproin, BCP), was evaporated between the
emissive layer and the calcium cathode (Figure 12b). The
low HOMO level of BCP (ꢀ6.4 eV) was intended to pre-
vent holes from crossing the device and reaching the calci-
um cathode without recombination. Furthermore, to im-
prove the transport of electrons, an electron-transport layer
Figure 14. Luminous efficiency versus current density for the ITO/
PEDOT/2/Ca, ITO/PEDOT/2/BCP/Ca and ITO/PEDOT/2/BCP/Alq3/Ca
OLEDs.
hibits the best performance. The energy-band diagram in
Figure 12b shows that the energy barrier between the
HOMO levels of 2 and BCP is 1 eV, which is sufficient to
block of holes and confine them to the emitting layer. This
results in an increase in radiative recombination and there-
fore an enhancement in the performance of the BCP-based
OLED compared with the single layer device. Indeed, with-
out BCP, holes can cross the emitting layer to the calcium
electrode without radiative recombination, which would
reduce the efficiency.
The insertion of Alq3 above BCP slightly reduces the lu-
minous efficiency from 0.16 to 0.1 CdAꢀ1. This shows that 2
enables the transport of electrons without the need for an
additional ETL. Alq3 does not favor the transport of elec-
trons in the emissive layer from the calcium cathode as the
LUMO level of BCP (ꢀ2.8 eV) perfectly fits the work func-
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11337

C. Poriel, J. Rault-Berthelot et al.
tion of calcium (ꢀ2.9 eV). Thus the optimum OLED struc-
ture is ITO/PEDOT/2/BCP/Ca. The associated current den-
sity–luminance–voltage curves are displayed in Figure 15
with a turn-on voltage of 8 V (measured precisely by using
the semi-logarithmic plot) and a maximum luminance of
470 Cdmꢀ2.
Figure 15. Current density–luminance–voltage characteristics of an ITO/
PEDOT/2 (100 nm)/BCP/Ca OLED.
The normalized electroluminescence spectrum recorded
for the ITO/PEDOT/2/BCP/Ca device (Figure 16a) reveals
three main emission peaks, two very sharp peaks at 404 and
427 nm and one much broader centered around 560 nm. A
fourth minor contribution can be observed at 452 nm. At
low wavelengths the emission spectrum of the OLED is
very similar to the thin-film fluorescence spectrum of 2. The
contribution at 560 nm seems to correspond to the broad
fluorescence band that appeared between 500 and 600 nm
when the sample was heated at 2008C for 60 h. However, it
seems difficult to correlate these two bands (ambient atmos-
phere vs. inert atmosphere) and further studies on the exact
nature of this emission are required. The chromatic coordi-
nates calculated from the electroluminescence spectrum in
the CIE 1964 chromaticity diagram are (0.30, 0.27), as can
be seen in Figure 16b.
Figure 16. a) Normalized electroluminescence spectrum of the ITO/
PEDOT/2/BCP/Ca OLED; b) corresponding color on a CIE 1964 chro-
maticity diagram.
devices displayed moderate efficiencies, the use of a hole-
transporting layer could be avoided because the HOMO
level of 2 fits well with the work function of the anode ITO/
PEDOT. The electroluminescence spectrum of 2 showed
two blue contributions (404 and 427 nm) and an additional
low-energy emission band (560 nm), which leads to chromat-
ic coordinates of (0.30, 0.27). After the application of ther-
mal stress in the presence of oxygen, a low-energy emission
band was also observed in solid-state fluorescence studies.
However, the relationship between these low-energy bands,
obtained under either an ambient or inert atmosphere, is
not evident as they may arise from different causes. Further
studies are currently underway in our laboratory to deter-
mine the exact nature of these low-energy emissions.
Conclusions
We have developed new efficient routes to the synthesis of
3p-2spiro ladder-type oligomers 1 and 2 by using extended
diketone derivatives as key intermediates. The structural,
electrochemical, thermal and optical properties of 2 have
been studied in detail and compared with the less extended
oligomer 1. Because the p-system-1 conjugation increases
on going from 1 to 2, we logically observed a band-gap con-
traction from 3.59 eV for 1 to 3.26 eV for 2. Compound 2
exhibited intense fluorescence and a very small Stokes shift
in solution as well as in the spin-coated thin film due to the
rigidity of the backbone. The OLED application studies
were carried out with 2 as the emissive layer. Although the
Experimental Section
Synthesis: Commercially available reagents and solvents were used with-
out further purification except for those detailed below. Dichloromethane
and acetonitrile were distilled from the P2O5 drying agent Sicapent
(Merck). THF and diethyl ether, were distilled from sodium/benzophe-
none prior to use. Light petroleum refers to the fraction with b.p. 40–
608C. Reactions were stirred magnetically, unless otherwise indicated.
Analytical thin-layer chromatography was carried out by using alumini-
11338
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

3p-2Spiro Ladder-Type Phenylene Materials
FULL PAPER
um-backed plates coated with Merck Kieselgel 60 GF254 and visualisation
under UV light (at 254 and/or 365 nm). Chromatography was carried out
by using silica 60A CC 40–63 mm (SDS). 1H and 13C NMR spectra were
recorded by using Bruker 300 MHz instruments (1H NMR frequency; the
corresponding 13C NMR frequency is 75 MHz). The chemical shifts are
given in ppm and the J values in Hz. In the 13C NMR spectra, the signals
corresponding to CH, CH2 and Me groups, assigned from DEPT experi-
ments, are noted; all others are C. The residual signals for the NMR sol-
vents are as follows: CDCl3: d=7.26 ppm for the proton and d=
77.00 ppm for the carbon; CD2Cl2: d=5.32 ppm for the proton and d=
53.80 ppm for the carbon; (CD3)2SO: d=2.50 ppm for the proton and d=
39.43 ppm for the carbon. The following abbreviations have been used
for the NMR assignments: s for singlet, d for doublet, t for triplet, q for
quartet and m for multiplet; br refers to a broad signal. High-resolution
mass spectra were recorded either at the Centre Rꢂgional de Mesure
Physique de lꢆOuest (Rennes) or with a Bruker Microflex LT spectrome-
ter. 2-Iodobiphenyl was prepared according to a literature procedure.[42]
Diketone 3 was prepared in a three-step synthesis as previously report-
ed.[29,41] Full characterisation of DSF-IF 1 may be found in our previous
work.[28] 9,9-Dioctylfluorene-2-boronate ester 7 was prepared according
to literature procedures starting from the commercially available 2-bro-
mofluorene.[44–46] Diester 6 was prepared by a modification of the Lamba
and Tour procedure starting from 1,4-dibromo-2,5-dimethylbenzene in a
two-step oxidation reaction followed by esterification (see the Supporting
Information).[47] The names of chemicals have been generated with the
naming service of ACD-I lab, which determines the chemical name ac-
cording to systematic application of the nomenclature rules agreed upon
by the International Union of Pure and Applied Chemistry and Interna-
tional Union of Biochemistry and Molecular Biology.
1383, 1227, 1131, 1109, 1020 cmꢀ1
; HRMS (ESI): m/z: calcd for
C70H94O4+Na: 1021.70498 [M+Na]+; found: 1021.7033.
2,5-Bis(9,9-dioctyl-9H-fluoren-2-yl)terephthalic acid (9): A solution of
sodium hydroxide (10n, 2.1 mL, 21.0 mmol) was added to a suspension
of 8 (500 mg, 0.5 mmol) in ethanol (20 mL). The mixture was heated at
reflux for 12 h. A solution of concentrated hydrochloric acid (37%) was
then carefully added dropwise until pH 1. Filtration gave 9 as a yellow
solid (350 mg, 74%). M.p. 2038C (ethanol); 1H NMR (300 MHz,
[D6]DMSO): d=13.06 (brs, 2H; exch D2O, OH), 7.88–7.81 (m, 4H;
ArH), 7.75 (s, 2H; ArH), 7.45–7.30 (m, 10H; ArH), 1.99 (t, J=7.2 Hz,
8H; CH2), 1.17–1.01 (m, 40H; CH2), 0.76 (t, J=7.2 Hz, 12H; Me), 0.63–
0.52 ppm (m, 8H; CH2); 13C NMR (75 MHz, [D6]DMSO): d=168.8 (C),
150.4 (C), 150.1 (C), 140.1 (C), 139.9 (C), 139.7 (C), 138.4 (C), 134.3 (C),
130.8 (CH), 127.2 (CH), 127.0 (CH), 126.8 (CH), 122.9 (CH), 122.7 (C),
119.8 (CH), 119.6 (CH), 54.6 (C), 31.0 (CH2), 29.1 (CH2), 28.4 (CH2),
28.4 (CH2), 23.2 (CH2), 21.9 (CH2), 13.8 ppm (Me), one CH2 signal not
observed; IR (KBr): n˜ =3070, 2963, 2927, 2850, 2650–2540, 1694 (C=O),
1460, 1430, 1407, 1287, 1243, 1128 cmꢀ1; HRMS (ESI): m/z: calcd for
C66H86O4+Na: 965.64238; [M+Na]+; found: 965.6418.
9,9,18,18-Tetraoctyl-9,18-dihydrobenzoACHTGNUTREN[NNUG 5,6]-s-indacenoACHUTNRTGEG[NNUN 1,2-b]indenoACHTUNGTRENNUNG[2,1-
h]fluorene-6,15-dione (10): Oxalyl chloride (201 mg, 134 mL, 1.59 mmol)
and DMF (116 mg, 123 mL, 1.59 mmol) were added dropwise to a suspen-
sion of 9 (100 mg, 0.11 mmol) in CH2Cl2 (7 mL) and the solution was
stirred at room temperature for 3 h. After removing the solvent in vacuo,
CH2Cl2 (5 mL) was added and the solution was cooled to 08C before the
addition of TiCl4 (80 mg, 46 mL, 0.42 mmol). The dark mixture was stirred
for 4 h at room temperature. The reaction was quenched with iced water
(50 mL). The mixture was extracted with CH2Cl2 (3ꢅ20 mL) and the ex-
tracts were dried (MgSO4). The solvent was removed in vacuo and the
residue was purified by column chromatography on silica gel eluting with
CH2Cl2/light petroleum (2:3) to give 10 as a grey solid (88 mg, 91%).
M.p. 2268C (acetonitrile); 1H NMR (300 MHz, CDCl3): d=8.00 (s, 2H;
ArH), 7.86 (s, 2H; ArH), 7.72 (m, 2H; ArH), 7.52 (s, 2H; ArH), 7.39–
7.36 (m, 6H; ArH), 2.03 (t, J=7.2 Hz, 8H; CH2), 1.26–1.07 (m, 40H;
CH2), 0.80 (t, J=7.2 Hz, 12H; Me), 0.68–0.66 ppm (m, 8H; CH2);
13C NMR (75 MHz, CDCl3): d=192.7 (C), 159.5 (C), 150.6 (C), 146.0
(C), 143.1 (C), 142.8 (C), 140.0 (C), 139.7 (C), 133.7 (C), 127.9 (CH),
127.2 (CH), 122.9 (CH), 120.0 (CH), 115.9 (CH), 115.6 (CH), 115.2
(CH), 55.8 (C), 40.2 (CH2), 31.7 (CH2), 30.0 (CH2), 29.2 (2ꢅCH2; deter-
mined by HMQC), 23.8 (CH2), 22.6 (CH2), 14.0 ppm (Me); IR (KBr):
n˜ =3052, 2954, 2924, 2851, 1704 (C=O), 1605, 1421, 1273, 1143 cmꢀ1; UV/
Dispiro[fluorene-9,6’-indenoACTHNUTRGNE[NUG 1,2-b]fluorene-12’,9“-fluorene] (1): 2-Iodo-
biphenyl (0.65 g, 2.34 mmol) was dissolved in dry THF (40 mL) in a
Schlenk tube under an argon atmosphere. The solution was degassed and
cooled to ꢀ788C before the dropwise addition, over 35 min, of nBuLi
(1.6m in hexane, 1.6 mL, 2.56 mmol). The resulting yellow solution was
stirred for a further 55 min and diketone 3 (0.30 g, 1.06 mmol) partially
dissolved in dry and degassed THF (100 mL) was added dropwise
through a cannula. The reaction was stirred for 2 d (from ꢀ788C to room
temperature) and the resulting mixture was poured into a saturated solu-
tion of ammonium chloride (15 mL) and extracted with CH2Cl2 (3ꢅ
50 mL). The extracts were dried (MgSO4), the solvent was removed in
vacuo and the residue was used in the next step without any further pu-
rification. Hydrochloric acid (5 drops) was added to a suspension of the
crude mixture of difluorenol 5 in acetic acid (50 mL) and the resulting
mixture was stirred for 2 h at 1008C. After cooling to room temperature,
water was added (20 mL) and a colorless solid was filtered and purified
by column chromatography on silica gel eluting with CH2Cl2/light petro-
leum (2:8) to give 1 as a colorless solid (0.15 g, 26%). The spectroscopic
analyses and purity of 1 was in accordance with our previous work.[28]
Vis
(CH2Cl2):
lmax =381
(e=65589),
363
(44657),
326 nm
(92069 mꢀ1 cmꢀ1); HRMS (ESI): m/z: calcd for C66H82O2: 906.63148;
[M]+; found: 906.6323; elemental analysis calcd (%) for C66H82O2: C
87.36, H 9.11; found: C 87.42, H 9.32.
15-Biphenyl-2-yl-15-hydroxy-9,9,18,18-tetraoctyl-15,18-dihydrobenzo-
AHCTUNGTREGNU[NN 5,6]-s-indacenoACHTUNRTGENGN[UN 1,2-b]indenoACHTNUGTRENN[UGN 2,1-h]fluoren-6(9H)-one (12): 2-Bromobi-
phenyl (77 mg, 57 mL, 330 mmol) was dissolved in dry THF (20 mL) in a
Schlenk tube under an argon atmosphere and the solution was degassed.
The solution was cooled to ꢀ788C. nBuLi (1.6m in hexane, 315 mL,
495 mmol) was added dropwise over 45 min. The resulting yellow solution
was stirred for a further 45 min and diketone 10 (100 mg, 110 mmol), dis-
solved in dry, degassed and cooled (ꢀ788C) THF, was added dropwise
through a cannula. The reaction was stirred overnight (from ꢀ788C to
room temperature) and the resulting mixture was poured into a saturated
solution of ammonium chloride (15 mL) and extracted with CH2Cl2 (3ꢅ
20 mL). The extracts were dried (MgSO4), the solvent was removed in
vacuo and the residue was purified by column chromatography on silica
gel eluting with CH2Cl2/light petroleum (1:2) to give 12 as a red solid
(47 mg, 40%). 1H NMR (300 MHz, CDCl3): d=8.65 (d, J=7.5 Hz, 1H;
ArH), 7.97 (s, 1H; ArH), 7.71–7.61 (m, 3H; ArH), 7.53–7.29 (m, 11H;
ArH), 7.21 (s, 1H; ArH), 7.01 (d, J=7.5 Hz, 1H; ArH), 6.87 (t, J=
7.5 Hz, 1H; ArH), 6.71 (br m, 1H; ArH), 6.53 (br m, 1H; ArH), 6.34
(br m, 1H; ArH), 6.08 (br m, 1H; ArH), 2.48 (s, 1H; exch D2O, OH),
1.99 (m, 8H; CH2), 1.34–1.08 (m, 40H; CH2), 0.80 (t, J=7.2 Hz, 12H;
Me), 0.71–0.53 ppm (m, 8H; CH2); 13C NMR (300 MHz, CDCl3): d=
193.3 (C), 158.9 (C), 157.9 (C), 152.6 (C), 151.2 (C), 150.5 (C), 149.5 (C),
144.9 (C), 143.8 (C), 142.2 (C), 141.8 (C), 141.4 (C), 141.2 (C), 140.4 (C),
Diethyl 2,5-bis(9,9-dioctyl-9H-fluoren-2-yl)terephthalate (8): Fluorene-2-
boronate 7 (2.88 g, 5.56 mmol), potassium carbonate (0.84 g, 6.07 mmol),
tetrakis(triphenylphosphine)palladium(0) (0.91 g, 0.1 mmol) and water
(15 mL) were added to a solution of 6 (1.0 g, 2.6 mmol) in THF (30 mL)
in a Schlenk tube. The Schlenk tube was degassed and the mixture was
stirred at reflux under an argon atmosphere for 12 h. The cooled mixture
was then extracted with diethyl ether (5ꢅ50 mL) and the extracts were
washed with brine and dried (MgSO4). The solvent was removed in
vacuo and the residue was purified by column chromatography on silica
gel eluting with CH2Cl2/light petroleum (1:9) to give 8 as a colorless solid
(1.90 g, 72%). 1H NMR (300 MHz, CDCl3): d=7.87 (s, 2H; ArH), 7.77–
7.74 (m, 4H; ArH), 7.39–7.32 (m, 10H; ArH), 4.10 (q, J=7.2 Hz, 4H;
OCH2CH3), 1.99 (t, J=7.2 Hz, 8H; CH2), 1.24–1.08 (m, 40H; CH2), 1.00
(t, J=7.2 Hz, 6H; OCH2CH3), 0.82 (t, J=7.2 Hz, 12H; Me), 0.70–
0.61 ppm (m, 8H; CH2); 13C NMR (75 MHz, CDCl3): d=168.6 (C),
150.93 (C), 150.89 (C), 141.2 (C), 140.7 (C), 140.6 (C), 138.9 (C), 133.9
(C), 131.5 (CH), 127.21 (CH), 127.17 (CH), 126.8 (C), 122.9 (CH), 122.8
(CH), 119.8 (CH), 119.4 (CH), 61.2 (CH2), 55.2 (C), 40.4 (CH2), 31.8
(CH2), 30.1 (CH2), 29.3 (CH2), 23.8 (CH2), 22.6 (CH2), 14.0 (Me),
13.8 ppm (Me); IR (KBr): n˜ =3056, 2960, 2925, 2850, 1723 (C=O), 1466,
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11339

C. Poriel, J. Rault-Berthelot et al.
140.2 (C), 140.0 (C), 139.2 (C), 138.4 (C), 135.7 (C), 134.2 (C), 131.5
(CH), 129.3 (CH), 129.1 (CH), 127.6 (CH), 127.4 (CH), 127.3 (CH),
127.2 (CH), 127.1 (CH), 126.8 (CH), 126.5 (CH), 126.4 (CH), 125.8
(CH), 122.9 (CH), 122.8 (CH), 119.9 (CH), 119.8 (CH), 116.3 (CH),
115.7 (CH), 115.6 (CH), 114.9 (CH), 114.8 (CH), 81.9 (COH), 55.7 (C),
54.8 (C), 40.5 (CH2), 40.3 (CH2), 40.2 (CH2), 40.0 (CH2), 31.8 (CH2), 31.7
(CH2), 30.3 (CH2), 30.0 (CH2), 29.4 (CH2), 29.2 (CH2), 24.1 (CH2), 23.8
(CH2), 22.6 (CH2), 22.5 (CH2), 14.1 (Me), 14.0 ppm (Me); IR (KBr): n˜ =
3056, 2925, 2853, 1715 (C=O), 1597, 1470, 1420, 1281, 1155 cmꢀ1; HRMS
(ESI): m/z: calcd for C78H92O2+Na: 1083.69950 [M+Na]+; found:
1083.7001.
CH2), 1.30–1.08 (m, 40H; CH2), 0.81 (t, J=7.2 Hz, 12H; Me), 0.73–
0.54 ppm (m, 8H; CH2); 13C NMR (75 MHz, CDCl3): d=52.3 (C), 151.9
(C), 151.0 (C), 150.0 (C), 141.2 (C), 141.0 (C), 140.7 (C), 140.6 (C), 140.4
(C), 139.9 (C), 139.3 (C), 131.5 (CH), 129.0 (CH), 127.3 (CH), 127.0
(CH), 126.9 (CH), 126.7 (CH), 126.5 (CH), 125.4 (CH), 122.8 (CH),
122.7 (CH), 119.7 (CH), 115.8 (CH), 115.7 (CH), 114.6 (CH), 81.7
(COH), 54.8 (C), 40.7 (CH2), 31.7 (CH2), 30.3 (CH2), 30.0 (CH2), 29.2
(CH2), 23.8 (CH2), 22.6 (CH2), 14.0 ppm (Me); IR (KBr): n˜ =3056, 2925,
2852, 1583, 1477, 1435, 1418, 1271, 1180 cmꢀ1; HRMS (ESI): m/z: calcd
for C90H102O2+Na: 1237.77775 [M+Na]+; found: 1237.7775.
9’,9’,18’,18’-Tetraoctyl-9’,18’-dihydrodispiro[fluorene-9,6’-benzo
ACHTUNGTRENNUNG[5,6]-s-
indaceno[1,2-b]indeno[2,1-h]fluorene-15’,9“-fluorene] (2): Diol 11
G
ACHTUNGTRENNUNG
9,9,18,18-Tetraoctyl-9,18-dihydro-15H-spiro
ACHTUNGERTN[UNGN benzoACHTUNTRGEG[NNUN 5,6]-s-indacenoACHTUNGTREN[NNGU 1,2-b]-
(342 mg, 0.28 mmol) was dissolved in CH2Cl2 (100 mL) and a solution of
boron trifluoride diethyl etherate (145 mL, 166 mg, 1.12 mmol) was
added. The solution was stirred at room temperature for 10 min. The sol-
vent was evaporated in vacuo. The residue was purified by column chro-
matography on silica gel eluting with CH2Cl2/light petroleum (1:2) to
give 2 as a colorless solid (210 mg, 64%). M.p. 2468C (CH2Cl2/acetoni-
trile); 1H NMR (300 MHz, CD2Cl2): d=8.01 (d, J=7.5 Hz, 4H; ArH),
7.56 (s, 2H; ArH), 7.48 (t, J=7.5 Hz, 4H; ArH), 7.35 (d, J=7.5 Hz, 2H;
ArH), 7.28–7.12 (m, 12H; ArH), 6.98 (s, 2H; ArH), 6.88 (d, J=7.5 Hz,
4H; ArH), 1.93 (t, J=7.2 Hz, 8H; CH2), 1.27–1.00 (m, 40H; CH2), 0.79
(t, J=7.2 Hz, 12H; Me), 0.65–0.40 ppm (m, 8H; CH2); 13C NMR
(75 MHz, CDCl3): d=151.0 (C), 150.8 (C), 149.6 (C), 149.1 (C), 147.8
(C), 141.9 (C), 141.8 (C), 141.0 (C), 140.9 (C), 140.8 (C), 127.9 (CH),
127.7 (CH), 126.6 (CH), 126.4 (CH), 124.4 (CH), 122.6 (CH), 120.1
(CH), 119.6 (CH), 115.2 (CH), 115.0 (CH), 114.4 (CH), 65.6 (C), 54.7
(C), 40.6 (CH2), 31.7 (CH2), 30.0 (CH2), 29.7 (CH2), 29.2 (CH2), 23.7
(CH2), 22.6 (CH2), 14.0 ppm (Me); IR (KBr): n˜ =3059, 2924, 2850, 1461,
1373, 1267 cmꢀ1; UV/Vis (CH2Cl2): lmax =394 (e=97366), 373 (64400),
indeno[2,1-h]fluorene-6,9’-fluoren]-15-one (13): Compound 12 (160 mg,
ACHTUNGTRENNUNG
0.15 mmol) was dissolved in CH2Cl2 (25 mL) and a solution of boron tri-
fluoride diethyl etherate was added (76 mL, 86 mg, 0.60 mmol). The solu-
tion was stirred at room temperature for 10 min and the solvent was
evaporated in vacuo. The residue was purified by column chromatogra-
phy on silica gel eluting with CH2Cl2/light petroleum (1:2) to give 13 as a
red solid (127 mg, 81%). 1H NMR (300 MHz, CDCl3): d=8.19 (s, 1H;
ArH), 7.99 (d, J=7.5 Hz, 2H; ArH), 7.94 (s, 1H; ArH), 7.82 (s, 1H;
ArH), 7.68 (d, J=7.5 Hz, 1H; ArH), 7.49 (t, J=7.5 Hz, 2H; ArH), 7.41–
7.17 (m, 10H; ArH), 7.02 (s, 1H; ArH), 6.91 (d, J=7.5 Hz, 2H; ArH),
6.86 (s, 1H; ArH), 2.05 (t, J=7.2 Hz, 4H; CH2), 1.90 (t, J=7.2 Hz, 4H;
CH2), 1.27–0.99 (m, 40H; CH2), 0.85–0.71 (m, 16H; Me/CH2), 0.61–
0.42 ppm (m, 4H; CH2); 13C NMR (75 MHz, CDCl3): d=193.3 (C), 158.7
(C), 156.4 (C), 151.6 (C), 151.1 (C), 150.4 (C), 148.4 (C), 147.5 (C), 144.6
(C), 144.0 (C), 143.0 (C), 142.6 (C), 142.0 (C), 141.1 (C), 140.5 (C), 140.0
(C), 139.8 (C), 135.4 (C), 134.3 (C), 133.1 (CH), 131.3 (CH), 129.4 (CH),
128.7 (CH), 128.14 (CH), 128.06 (CH), 127.9 (CH), 127.6 (CH), 127.3
(CH), 124.4 (CH), 122.6 (CH), 120.2 (CH), 116.0 (CH), 115.7 (CH),
115.5 (CH), 115.2 (CH), 115.1 (CH), 114.5 (CH), 66.0 (C), 55.6 (C), 54.8
(C), 40.5 (CH2), 40.3 (CH2), 31.8 (CH2), 31.7 (CH2), 30.0 (CH2), 29.9
(CH2), 29.7 (CH2), 29.18 (CH2), 29.15 (CH2), 29.1 (CH2), 23.8 (CH2), 23.7
(CH2), 22.6 (CH2), 22.5 (CH2), 14.1 (Me), 14.0 ppm (Me); IR (KBr): n˜ =
3060, 2923, 2852, 1714 (C=O), 1597, 1419, 1283 cmꢀ1; UV/Vis (CH2Cl2):
339 nm (18806mꢀ1 cmꢀ1); HRMS (LSIMS): m/z: calcd for C90H98
:
1178.76685 [M]+; found: 1178.7662; elemental analysis calcd (%) for
C90H98: C 91.63, H 8.37; found: C 91.13, H 8.57.
Spectroscopic studies: Cyclohexane (ACS grade) and toluene (semicon-
ductor grade) were purchased from Alfa Aesar. Oligomers 1 and 2
(15 mgmLꢀ1 in toluene, 90 mL) were deposited on quartz substrate by
using a home-made spin-coater and the UV/Vis and photoluminescence
spectra were immediately recorded. UV/Vis spectra of the thin films
were recorded by using a UV/Vis SHIMADZU UV-1605 spectrophotom-
eter. UV/Vis spectra of solutions were recorded by using either a UV/
Vis/NIR CARY 5000-Varian (for quantum yield determination) or a UV/
Vis UVIKON XL Biotech spectrophotometer. The optical band gap was
calculated from the absorption edge of the UV/Vis absorption spectra by
using the formula DEopt [eV]=hclꢀ1, in which l is the absorption edge
(in metres). With h=6.626ꢅ10ꢀ34 Js (1 eV=1.602ꢅ10ꢀ19 J) and c=
2.997ꢅ108 msꢀ1. Photoluminescence spectra of the solutions (cyclohex-
ane) or thin films were recorded with a PTI spectrofluorimeter (PTI-814
PDS, MD 5020, LPS 220B) by using a xenon lamp. Quantum yields in so-
lution (fsol) were calculated relative to quinine sulfate (fsol =0.546 in 1n
H2SO4) using standard procedures.[41] fsol was determined according to
Equation (1), in which subscripts s and r refer to the sample and the ref-
erence, respectively. The integrated area of the emission peak in arbitrary
units is given as T, n is the refracting index of the solvent (ns =1.42662
for cyclohexane) and A is the absorbance. IR spectra were recorded on a
BIORAD IRFTS175C spectrometer.
lmax =385
(e=21419),
365
(14851),
322
(13367),
309 nm
(12889mꢀ1 cmꢀ1); HRMS (ESI): m/z: calcd for C78H90O+Na: 1065.68894
[M+Na]+; found: 1065.6883.
6,15-Dibiphenyl-2-yl-9,9,18,18-tetraoctyl-6,9,15,18-tetrahydrobenzo
indaceno[1,2-b]indeno[2,1-h]fluorene-6,15-diol (11)
ACHTUNGTRENN[UNG 5,6]-s-
N
ACHTUNGTRENNUNG
Route 1: 2-Bromobiphenyl (1.85 g, 1.37 mL, 7.93 mmol) was dissolved in
dry THF (60 mL) in a Schlenk tube under an argon atmosphere and the
solution was degassed and cooled to ꢀ788C. nBuLi (1.6m in hexane,
5.46 mL, 8.73 mmol) was added dropwise over 45 min. The resulting
yellow solution was stirred for
a further 45 min and 10 (0.80 g,
0.88 mmol) dissolved in dry, degassed and cooled (ꢀ788C) THF was
added dropwise through a cannula. The reaction was stirred overnight
(from ꢀ788C to room temperature) and the resulting mixture was
poured into a saturated solution of ammonium chloride (40 mL) and ex-
tracted with CH2Cl2 (5ꢅ20 mL). The extracts were dried (MgSO4), the
solvent was removed in vacuo and the residue was purified by column
chromatography on silica gel eluting with CH2Cl2/light petroleum (1:2) to
give 11 as a red solid (0.34 g, 31%).
Route 2: Diketone 10 (100 mg, 0.11 mmol) was dissolved in dry and de-
gassed THF (10 mL) in a Schlenk tube under an argon atmosphere. A so-
lution of biphenyl-2-ylmagnesium bromide (0.5m in diethyl ether, 2.1 mL,
0.99 mmol) was added. The reaction was stirred overnight at room tem-
perature and the resulting mixture was poured into a saturated solution
of ammonium chloride (10 mL) and extracted with CH2Cl2 (5ꢅ10 mL).
The extracts were dried (MgSO4), the solvent was removed in vacuo and
the residue was purified by column chromatography on silica gel eluting
with CH2Cl2/light petroleum (1:2) to give 11 as a red solid (29 mg, 22%).
M.p. 1718C; 1H NMR (300 MHz, CDCl3): d=8.64 (d, J=7.5 Hz, 2H;
ArH), 7.69–7.61 (m, 4H; ArH), 7.53 (s, 2H; ArH), 7.40 (t, J=7.5 Hz,
2H; ArH), 7.34–7.28 (m, 6H; ArH), 7.12 (s, 2H; ArH), 7.00 (s, 2H;
ArH), 6.94 (d, J=7.5 Hz, 2H; ArH), 6.83 (t, J=7.5 Hz, 2H; ArH), 6.62
(br m, 2H; ArH), 6.51 (br m, 2H; ArH), 6.20 (br m, 2H; ArH), 6.02
(br m, 2H; ArH), 2.32 (s, 2H; exch D2O, OH), 1.95 (t, J=7.2 Hz, 8H;
ꢀ
ꢁ
2
ðTsArÞ ns
ð1Þ
ꢀsol ¼ 100ꢀref
nr
ðTrAsÞ
Thermal analysis: Thermogravimetric analyses (TGA) were carried out
with a Rigaku Thermoflex instrument under a nitrogen atmosphere be-
tween room temperature and 10008C at a heating rate of 58Cminꢀ1
.
Melting points were determined by using an Electrothermalꢇ melting
point apparatus.
Electrochemical studies: All electrochemical experiments were per-
formed under an argon atmosphere using a platinum disk electrode (di-
ameter 1 mm). The counter electrode was a vitreous carbon rod and the
reference electrode was a silver wire in a 0.1m solution of AgNO3 in
11340
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342

3p-2Spiro Ladder-Type Phenylene Materials
FULL PAPER
CH3CN. Ferrocene was added to the electrolyte solution at the end of a
series of experiments. The ferrocene/ferrocenium (Fc/Fc+) couple served
as internal standard. The three-electrode cell was connected to a PAR
Model 173 potentiostat monitored with a PAR Model 175 signal genera-
tor and a PAR Model 179 signal coulometer. CV traces were recorded
on an XY SEFRAM-type TGM 164 instrument. Acetonitrile with less
than 5% of water (ref. SDS 00610S21) and CH2Cl2 with less than
100 ppm of water (ref. SDS 02910E21) were used without purification.
Activated Al2O3 was added to the electrolytic solution to remove excess
moisture. For further comparison of the electrochemical and optical
properties, all potentials are referred to the SCE electrode, which was
calibrated at ꢀ0.405 V versus Fc/Fc+. Following the work of Jenekhe and
co-workers,[57] we estimated the electron affinity (EA), or LUMO, and
under vacuum (ca. 10ꢀ6 mbar) at a low deposition rate of 0.1 nmsꢀ1. The
layer thickness was monitored in situ by a piezoelectric quartz balance
during the evaporation. Calcium cathodes were finally evaporated
through a shadow mask. The OLEDs were then stored and characterised
under an inert atmosphere in a nitrogen glove box. Current–voltage–lu-
minance (I–V–L) curves were recorded by using a Keithley 4200 SCS in-
strument. Light emission was collected by using a calibrated photodiode
coupled to a Hamamatsu Photosensor amplifier C9329. Electrolumines-
cence spectra were measured with a CCD spectrometer (Ocean Optics
HR 2000). The microscope images were taken with a Zeiss Axio Imager
A1/F1 coupled to a digital camera using a ꢅ20 lens.
the ionisation potential (IP), or HOMO, from the redox data. The
onset
red
LUMO level was calculated from LUMO [eV]=ꢀ[E
(vs. SCE)+4.4]
onset
ox
and the HOMO level from HOMO [eV]=ꢀ[E
(vs. SCE)+4.4], based
Acknowledgements
on an SCE energy level of 4.4 eV relative to the vacuum. The electro-
chemical gap was calculated from DEEl =jHOMOꢀLUMOj (in eV).
N.C. thanks le ministꢁre de lꢆꢂducation nationale, de la recherche et de la
technologie (MENRT) for a studentship. The authors would like to
thank the CINES (Centre Informatique National de lꢆEnseignement Su-
pꢂrieur, Montpellier) for awarding computing time, the CRMPO (Centre
Rꢂgional de Mesure Physique de lꢆOuest) for high-resolution mass meas-
urements and CHN analyses. Prof. Muriel Hissler and Dr. Olivier Jean-
nin are strongly acknowledged for helpful discussions in thin-film fluores-
cence studies and in X-ray diffraction, respectively. We also thank
Damien Thirion, Ali Yassin, Sꢂbastien Bivaud and Stephanie Fryars for
invaluable technical assistance.
Computational details: Calculations were performed on a simplified
model of 2 in which the octyl chains were omitted and replaced by
methyl groups. Full geometry optimisation by DFT[74, 75] calculations were
performed with the hybrid Becke-3 parameter exchange function-
[60,76,77]
al
and the Lee–Yang–Parr non-local correlation functional[78]
(B3LYP or UB3LYP) implemented in the Gaussian 03 (Revision D.02)
suite of programs[56] using the 6-31G* basis set[79] and the default conver-
gence criterion implemented in the program. The figures were generated
with MOLEKEL 4.3.[59]
X-Ray determination: Data were collected by using a Rigaku MM007
RA/confocal optics generator and Saturn70 CCD diffractometer. In both
2 and 8 the crystal quality/diffraction was poor and several different crys-
tals were examined with consistent results. The structures were solved by
direct methods and refined by full-matrix least-squares on F2 values of all
data.[80]
[1] K. Mꢀllen, U. Scherf, Organic Light-Emitting Devices: Synthesis
Properties and Applications, Wiley-VCH, Weinheim, 2006.
[2] A. C. Grimsdale, K. Mꢀllen, Macromol. Rapid Commun. 2007, 28,
1676–1702.
[3] U. Scherf, J. Mater. Chem. 1999, 9, 1853–1864.
[4] V. N. Bliznyuk, S. A. Carter, J. C. Scott, G. Klꢈrner, R. D. Miller,
D. C. Miller, Macromolecules 1999, 32, 361–369.
[5] J.-I. Lee, G. Klaerner, R. D. Miller, Chem. Mater. 1999, 11, 1083–
1086.
[6] L. Romaner, A. Pogantsch, P. Scandiucci de Freitas, U. Scherf, M.
Gaal, E. Zojer, E. J. W. List, Adv. Funct. Mater. 2003, 13, 597–601.
[7] M. Gaal, E. J. W. List, U. Scherf, Macromolecules 2003, 36, 4236–
4237.
[8] T. P. I. Saragi, T. Spehr, A. Siebert, T. Fuhrmann-Lieker, J. Salbeck,
Chem. Rev. 2007, 107, 1011–1065.
[9] R. Pudzich, T. Fuhrmann-Lieker, J. Salbeck, Adv. Polym. Sci. 2006,
199, 83–142.
[10] R. Wu, J. S. Schumm, D. L. Pearson, J. M. Tour, J. Org. Chem. 1996,
61, 6906–6921.
[11] J. M. Tour, R. Wu, J. S. Schumm, J. Am. Chem. Soc. 1991, 113,
7064–7066.
[12] J. M. Tour, R. Wu, J. S. Schumm, J. Am. Chem. Soc. 1990, 112,
5662–5663.
Oligomer 2: C98H115; M=1292.90; yellow platelet; crystal size 0.35ꢅ
0.08ꢅ0.01 mm; triclinic; P1; a=12.676(7), b=13.014(8), c=13.240(7) ꢄ;
¯
a=84.53(5), b=76.25(4), g=83.29(5)8; V=2101.9(19) ꢄ3; Z=1; 1calcd
=
1.021 Mgmꢀ3
; MoKa radiation (confocal optic, l=0.71073 ꢄ); m=
0.057 mmꢀ1; T=93(2) K; 15920 data (6963 unique, Rint =0.0792, 2.030<
1
2
2
q<25.428); wR={ꢀ[w(Fo ꢀFc2)2]/ꢀ[w(Fo )2]} = =0.3413; conventional R=
2
0.1218 for F values of reflections with Fo2 >2s(Fo ) (4932 observed reflec-
2
tions); S=1.082 for 480 parameters. Residual electron density extremes
were 0.547 and ꢀ0.299 eꢄꢀ3
.
Compound 8: C70H94O4; M=999.45; colorless platelet; crystal size 0.1ꢅ
¯
0.1ꢅ0.01 mm; triclinic; P1; a=8.963(3), b=18.085(6), c=19.344(6) ꢄ;
a=97.389(6), b=100.353(5), g=94.091(5)8; V=3044.6(17) ꢄ3; Z=2;
1calcd =1.090 Mgmꢀ3; MoKa radiation (confocal optic, l=0.71073 ꢄ); m=
0.065 mmꢀ1; T=93(2) K; 30488 data (11031 unique, Rint =0.0887, 2.16<
1
2
2
q<25.368); wR={ꢀ[w(Fo ꢀFc2)2]/ꢀ[w(Fo )2]} = =0.1838; conventional R=
2
0.0958 for F values of reflections with Fo2 >2s(Fo ) (7989 observed reflec-
2
tions); S=1.053 for 669 parameters. Residual electron density extremes
were 0.506 and ꢀ0.268 eꢄꢀ3
.
CCDC-689480 and 689481 contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif
[13] J. Guay, A. Diaz, R. Wu, J. M. Tour, J. Am. Chem. Soc. 1993, 115,
1869–1874.
[14] N. Johansson, D. A. Dos Santos, S. Guo, J. Cornil, M. Fahlman, J.
Salbeck, H. Schenk, H. Arwin, J.-L. Bredas, W. R. Salaneck, J.
Chem. Phys. 1997, 107, 2542–2549.
[15] N. Johansson, J. Salbeck, J. Bauer, F. Weissçrtel, P. Brçms, A. An-
dersson, W. R. Salaneck, Adv. Mater. 1998, 10, 1136–1141.
[16] R. Pudzich, J. Salbeck, Synth. Met. 2003, 138, 21–31.
[17] J. Salbeck, F. Weissçrtel, J. Bauer, Macromol. Symp. 1997, 125, 121–
132.
[18] J. Salbeck, N. Yu, J. Bauer, F. Weissçrtel, H. Besten, Synth. Met.
1997, 91, 209–215.
[19] T. Spehr, R. Pudzich, P. Fuhrmann, J. Salbeck, Org. Electron. 2003,
4, 61–69.
[20] U. Scherf, K. Mꢀllen, Macromolecules 1992, 25, 3546–3548.
[21] J. Jacob, S. Sax, T. Piok, E. J. W. List, A. C. Grimsdale, K. Mꢀllen, J.
Am. Chem. Soc. 2004, 126, 6987–6995.
EL fabrication and testing: OLEDs were fabricated by using the follow-
ing procedure. Indium–tin oxide (ITO) substrates on glass from Merck
were submitted to solvent ultrasonic cleansing with acetone and isopro-
panol followed by UV/ozone treatment for 20 min. A layer of poly(3,4-
ethylenedioxythiophene) doped with poly(styrene sulfonate) (PEDOT/
PSS from HC Starck) was then deposited onto the ITO by spin-coating
at 4000 rpm from a 3 wt% water dispersion to form a layer 40 nm thick.
This layer improves hole injection from the ITO to the HOMO level of
the organic material and increases the performances and the lifetime of
the device.[81] PEDOT/PSS was subsequently annealed at 808C under
vacuum for 30 min. Then the emissive layer was either spin-coated or
thermally evaporated. The spin-coating was carried out at 1000 rpm from
a 20 mgmLꢀ1 solution in cyclohexanone. The evaporation was performed
Chem. Eur. J. 2008, 14, 11328 – 11342
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11341

C. Poriel, J. Rault-Berthelot et al.
[22] J. Jacob, S. Sax, M. Gaal, E. J. W. List, A. C. Grimsdale, K. Mꢀllen,
Macromolecules 2005, 38, 9933–9938.
[23] F. Schindler, J. Jacob, A. C. Grimsdale, U. Scherf, K. Mꢀllen, J. M.
Lupton, J. Feldmann, Angew. Chem. 2005, 117, 1544–1549; Angew.
Chem. Int. Ed. 2005, 44, 1520–1525.
[24] A. K. Mishra, M. Graf, F. Grasse, J. Jacob, E. J. W. List, K. Mꢀllen,
Chem. Mater. 2006, 18, 2879–2885.
[25] G. Zhou, M. Baumgarten, K. Mꢀllen, J. Am. Chem. Soc. 2007, 129,
12211–12221.
[26] Y. Wu, J. Zhang, Z. Bo, Org. Lett. 2007, 9, 4435–4438.
[27] D. Vak, B. Lim, S.-H. Lee, D.-Y. Kim, Org. Lett. 2005, 7, 4229–4232.
[28] D. Horhant, J.-J. Liang, M. Virboul, C. Poriel, G. Alcaraz, Rault- J.
Berthelot, Org. Lett. 2006, 8, 257–260.
[29] C. Poriel, J.-J. Liang, J. Rault-Berthelot, F. Barriꢁre, N. Cocherel,
A. M. Z. Slawin, D. Horhant, M. Virboul, G. Alcaraz, N. Audebrand,
L. Vignau, N. Huby, L. Hirsch, G. Wantz, Chem. Eur. J. 2007, 13,
10055–10069.
[54] M. Zhang, C. Yang, A. K. Mishra, W. Pisula, G. Zhou, B. Schmaltz,
M. Baumgarten, K. Mꢀllen, Chem. Commun. 2007, 1704–1706.
[55] In the simplified model of 2, octyl chains have been replaced by
methyl groups.
[56] Gaussian 03, Revision D.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomer-
y, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Lyen-
gar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. C. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannen-
berg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford,
CT, 2004.
[30] C. Poriel, J. Rault-Berthelot, F. Barriꢁre, A. M. Z. Slawin, Org. Lett.
2008, 10, 373–376.
[31] M. Kimura, S. Kuwano, Y. Sawaki, H. Fujikawa, K. Noda, Y. Taga,
K. Takagi, J. Mater. Chem. 2005, 15, 2393–2398.
[32] L.-H. Xie, X.-Y. Hou, C. Tang, Y.-R. Hua, R.-J. Wang, R.-F. Chen,
Q.-L. Fan, L.-H. Wang, W. Wei, B. Peng, W. Huang, Org. Lett. 2006,
8, 1363–1366.
[57] A. P. Kulkarni, C. J. Tonzola, A. Babel, S. A. Jenekhe, Chem. Mater.
2004, 16, 4556–4573.
[33] L. Oldridge, M. Kastler, K. Mꢀllen, Chem. Commun. 2006, 885–
887.
[58] Y. Zhu, R. D. Champio, S. A. Jenekhe, Macromolecules 2006, 39,
8712–8719.
[34] F. Uckert, S. Setayesh, K. Mꢀllen, Macromolecules 1999, 32, 4519–
4524.
[35] T. Nakagawa, D. Kumaki, J.-I. Nishida, S. TOkito, Y. Yamashita,
Chem. Mater. 2008, 20, 2615–2617.
[36] H. Usta, A. Facchetti, T. J. Marks, Org. Lett. 2008, 10, 1385–1388.
[37] R. G. Clarkson, M. Gomberg, J. Am. Chem. Soc. 1930, 52, 2881–
2891.
[59] MOLEKEL 4.3, H. P. Flꢀkiger, S. Lꢀthi, S. Portmann, J. Weber,
Swiss National Supercomputing Centre CSCS, Manno (Switzerland),
2000.
[60] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[61] C. Zhao, Y. Zhang, M.-K. Ng, J. Org. Chem. 2007, 72, 6364–6371.
[62] J. Rault-Berthelot, C. Poriel, F. Justaud, F. Barriꢁre, New J. Chem.
2008, 32, 1259–1266.
[38] J. H. Weisburger, E. K. Weisburger, F. E. Ray, J. Am. Chem. Soc.
1950, 72, 4250–4253.
[39] J. H. Weisburger, E. K. Weisburger, F. E. Ray, J. Am. Chem. Soc.
1950, 72, 4253–4255.
[63] A. Kuboyama, Bull. Chem. Soc. Japan 1964, 37, 1540–1544.
[64] F. Jaramillo-Isaza, M. L. Turner, J. Mater. Chem. 2006, 16, 83–89.
[65] Optical band gap estimated from the low-energy edge of the UV/
Vis spectrum in CH2Cl2.
[40] B. Winter-Werner, F. Diederich, V. Gramlich, Helv. Chim. Acta
1996, 79, 1338–1360.
[66] M. BelletÞte, M. Ranger, S. Beauprꢂ, M. Leclerc, G. Durocher,
Chem. Phys. Lett. 2000, 316, 101–107.
[41] S. Merlet, M. Birau, Z. Y. Wang, Org. Lett. 2002, 4, 2157–2159.
[42] C. Poriel, Y. Ferrand, S. Juillard, P. Le Maux, G. Simonneaux, Tetra-
hedron 2004, 60, 145–158.
[67] F.-I. Wu, R. Dodda, D. Sahadeva Reddy, C.-F. Shu, J. Mater. Chem.
2002, 12, 2893–2897.
[68] F. Le Floch, Ph.D. Thesis, University of Rennes I (France), 2003.
[69] K.-T. Wong, Y.-Y. Chien, R.-T. Chen, C.-F. Wang, Y.-T. Lin, H.-H.
Chiang, P.-Y. Hsieh, C.-C. Wu, C.-H. Chou, Y. O. Su, G.-H. Lee, S.-
M. Peng, J. Am. Chem. Soc. 2002, 124, 11576–11577.
[70] M. M. Elmahdy, G. Floudas, L. Oldridge, A. C. Grimsdale, K.
Mꢀllen, ChemPhysChem 2006, 7, 1431–1441.
[71] H. Etori, X. L. Jin, T. Yasuda, S. Mataka, T. Tsutsui, Synth. Met.
2006, 156, 1090–1096.
[72] J. Jacob, J. Zhang, A. C. Grimsdale, K. Mꢀllen, M. Gaal, E. J. W.
List, Macromolecules 2003, 36, 8240–8245.
[43] Our previously reported approach to 1 led to an overall yield of
around 20%.[29]
[44] M. Sonntag, K. Kreger, D. Hanft, P. Strohriegl, S. Setayesh, D. de
Leeuw, Chem. Mater. 2005, 17, 3031–3039.
[45] M. BelletÞte, S. Beauprꢂ, J. Bouchard, P. Blondin, M. Leclerc, G.
Durocher, J. Phys. Chem. B 2000, 104, 9118–9125.
[46] T. Li, T. Yamamoto, H.-L. Lan, J. Kido, Polym. Adv. Technol. 2004,
15, 266–269.
[47] J. J. S. Lamba, J. M. Tour, J. Am. Chem. Soc. 1994, 116, 11723–
11736.
[48] R. Laufer, G. I. Dmitrienko, J. Am. Chem. Soc. 2002, 124, 1854–
1855.
[73] L. Romaner, G. Heimel, H. Wiesenhofer, P. Scandiucci de Freitas, U.
Scherf, J. L. Brꢂdas, E. Zojer, E. J. W. List, Chem. Mater. 2004, 16,
4667–4674.
[49] H. Koyama, T. Kamikawa, J. Chem. Soc. Perkin Trans. 1 1998, 203–
209.
[50] Note that several other groups have also reported low-to-moderate
yields for similar coupling reactions, see references. [27], [31] and
[52].
[51] When the cyclisation was performed by heating at reflux in AcOH/
HCl, we always observed the presence of an impurity with a strong
yellow emission under a UV lamp (365 nm), which, despite several
attempts, could not be separated from 2.
[74] P. Hohenberg, W. Kohn, Phys. Rev. 1964, 136, B864–B871.
[75] R. G. Parr, W. Yang, Density-Functional Theory of Atoms and Mole-
cules, Oxford University, Oxford, 1989.
[76] A. D. Becke, Phys. Rev. 1988, 38, 3098–3100.
[77] A. D. Becke, J. Chem. Phys. 1993, 98, 1372–1377.
[78] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[79] P. C. Hariharan, J. A. Pople, Chem. Phys. Lett. 1972, 16, 217–219.
[80] SHELXTL, version 6.1, G. M. Sheldrick, Bruker AXS, Madison,
WI, 2001
[52] J. Luo, Y. Zhou, Z.-Q. Niu, Q.-F. Zhou, Y. Ma, J. Pei, J. Am. Chem.
Soc. 2007, 129, 11314–11315.
[81] A. van Dijken, A. Perro, E. A. Meulenkamp, K. Brunner, Org. Elec.
2003, 4, 131–141.
[53] K.-T. Wong, L.-C. Chi, S.-C. Huang, Y.-L. Liao, Y.-H. Liu, Y. Wang,
Org. Lett. 2006, 8, 5029–5032.
Received: July 15, 2008
Published online: November 13, 2008
11342
www.chemeurj.org
ꢃ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11328 – 11342